JP7632231B2 - Vehicle suspension control device - Google Patents

Description

This disclosure relates to a vehicle suspension control device that includes an actuator that applies a vertical control force between the unsprung structure and the sprung structure of the vehicle.

Patent Document 1 discloses an active suspension system for a vehicle. In this active suspension system, when a vertical acceleration sensor fails, a suspension characteristic control device controls the control gain of the suspension control based on the detection signal of the vertical acceleration sensor so that it becomes zero. In other words, the suspension control is stopped. More specifically, failure determination in the active suspension system is performed as follows. That is, when vertical acceleration signals are input from all three vertical acceleration sensors, these vertical acceleration sensors are determined to be normal. On the other hand, when a vertical acceleration signal is not input from any of the three vertical acceleration sensors, the vertical acceleration sensor to which the vertical acceleration signal is not input is determined to have failed.

Japanese Patent Application Publication No. 04-039108

Causes of abnormalities in the sprung acceleration sensor (vertical acceleration sensor) include not only a lack of signal input due to a broken wire or other reasons, but also, for example, an abnormal signal output from the sprung acceleration sensor. If the output from an abnormal sprung acceleration sensor is used, it will be impossible to perform proper suspension control. For this reason, it is desirable to be able to quickly detect abnormalities in the sprung acceleration sensor.

This disclosure was made in consideration of the above-mentioned problems, and aims to provide a vehicle suspension control device that can quickly detect an abnormality in a sprung acceleration sensor used to control a suspension actuator.

A vehicle suspension control device according to one aspect of the present disclosure includes an actuator, at least four sprung acceleration sensors, and an electronic control unit. The actuator applies a vertical control force between an unsprung structure and a sprung structure of a vehicle. The at least four sprung acceleration sensors detect vertical acceleration of the sprung structure. The electronic control unit controls the actuator based on the detected accelerations of the at least four sprung acceleration sensors. The electronic control unit executes a sensor identification process. In the sensor identification process, the electronic control unit calculates a specific value related to a vertical displacement, a roll angle, or a pitch angle at the same position of the sprung structure based on the detected accelerations of the sprung acceleration sensors that are one less than all of the at least four sprung acceleration sensors, for each of a plurality of different combinations of the sprung acceleration sensors that are one less than all of the at least four sprung acceleration sensors, and identifies an abnormal sprung acceleration sensor from among the at least four sprung acceleration sensors based on a comparison of the specific values between the plurality of different combinations.

A vehicle suspension control device according to another aspect of the present disclosure includes an actuator, at least four sprung acceleration sensors, a gyro sensor, and an electronic control unit. The actuator applies a vertical control force between an unsprung structure and a sprung structure of the vehicle. The at least four sprung acceleration sensors detect vertical acceleration of the sprung structure. The gyro sensor detects angular velocity of the sprung structure. The electronic control unit controls the actuator based on the detected accelerations of the at least four sprung acceleration sensors. The electronic control unit executes an abnormality detection process. In the abnormality detection process, the electronic control unit calculates a first estimated angular velocity related value of the sprung structure at the position of the gyro sensor based on the detected accelerations of all of the at least four sprung acceleration sensors, and determines whether or not an abnormal sprung acceleration sensor is included among the at least four sprung acceleration sensors based on a comparison between the calculated first estimated angular velocity related value and the detected angular velocity related value of the gyro sensor.

The vehicle suspension control device according to each aspect of the present disclosure can quickly detect an abnormality in the sprung acceleration sensor, compared to an example in which a process for detecting a sensor abnormality based on the presence or absence of abnormal behavior in the output signal of the sprung acceleration sensor is performed alone using a long determination time.

1 is a diagram illustrating an example of a configuration of a vehicle according to a first embodiment; 1 is a diagram illustrating an example of a configuration of a suspension according to a first embodiment. FIG. 4 is a flowchart showing a process related to suspension control according to the first embodiment. FIG. 11 is a supplementary diagram for explaining the abnormality detection process (step S106). 13 is a supplementary diagram for explaining the sensor identification process according to the second embodiment. FIG. 10 is a flowchart showing a process related to suspension control according to the second embodiment; FIG. 11 is a diagram illustrating an example of a configuration of a vehicle according to a third embodiment. 13 is a flowchart showing a process related to suspension control according to the third embodiment.

When the number, quantity, amount, range, etc. of each element is mentioned in the embodiments described below, the technical ideas of this disclosure are not limited to the mentioned numbers, unless otherwise specified or clearly specified in principle.

1. First embodiment
1-1. Vehicle Configuration FIG. 1 is a diagram that shows a schematic example of the configuration of a vehicle 1 according to the first embodiment. The vehicle 1 includes wheels 2 and suspensions 3 that suspend the wheels 2 from a vehicle body 6 (see FIG. 2). The wheels 2 include a left front wheel 2FL, a right front wheel 2FR, a left rear wheel 2RL, and a right rear wheel 2RR. Suspensions 3FL, 3FR, 3RL, and 3RR are provided for the left front wheel 2FL, the right front wheel 2FR, the left rear wheel 2RL, and the right rear wheel 2RR, respectively. In the following description, unless there is a particular need to distinguish between them, each wheel will be referred to as a wheel 2 and each suspension will be referred to as a suspension 3.

Figure 2 is a diagram showing an example of the configuration of the suspension 3 according to the first embodiment. The suspension 3 is provided to connect between the unsprung structure 4 and the sprung structure 5 of the vehicle 1. The unsprung structure 4 includes the wheels 2. The sprung structure 5 includes the vehicle body 6. The suspension 3 includes a spring 3S, a damper (shock absorber) 3D, and an actuator 3A. The spring 3S, the damper 3D, and the actuator 3A are provided in parallel between the unsprung structure 4 and the sprung structure 5. The spring 3S has a spring constant K. The damper 3D has a damping coefficient C. The actuator 3A applies a control force Fc in the vertical direction between the unsprung structure 4 and the sprung structure 5. This controls the stroke of the suspension 3.

More specifically, the actuator 3A is, for example, an electric or hydraulic active actuator (an actuator constituting a so-called full active suspension). However, the "actuator" according to the present disclosure is not particularly limited as long as it can apply a vertical control force Fc between the unsprung structure 4 and the sprung structure 5. Specifically, the actuator may be, for example, an actuator that varies the damping force generated by the damper 3D, or an actuator of an active stabilizer device. Furthermore, the actuator may be, for example, an actuator (for example, an electric motor) that generates a vehicle longitudinal force (driving force and braking force) acting on the wheel in a vehicle equipped with a suspension configured to convert the vehicle longitudinal force (driving force and braking force) acting on the wheel into a control force Fc by utilizing the suspension geometry. The electric motor may be, for example, an in-wheel motor (IWM) provided on the wheel, or an electric motor that can drive the wheel via the vehicle drive shaft.

The vehicle 1 further includes an electronic control unit (ECU) 10. The ECU 10 includes a processor, a storage device, and an input/output interface. The input/output interface receives sensor signals from sensors 12 attached to the vehicle 1, and outputs operation signals to the actuator 3A. The storage device stores various control programs for controlling the actuator 3A. The processor reads and executes the control programs from the storage device. This realizes suspension control using the actuator 3A. There may be multiple ECUs 10.

The sensors 12 include, for example, a sprung acceleration sensor that detects the vertical acceleration of the sprung structure 5. In the example shown in FIG. 1, four sprung acceleration sensors 14-i (i=1 to 4) are provided. More specifically, the sprung acceleration sensor 14-1 detects the sprung acceleration at a first position toward the right front wheel 2FR as viewed from the sprung center of gravity position (center of gravity position GC of the sprung structure 5). The sprung acceleration sensor 14-2 detects the sprung acceleration at a second position toward the left front wheel 2FL as viewed from the sprung center of gravity position GC. The sprung acceleration sensor 14-3 detects the sprung acceleration at a third position toward the right rear wheel 2RR as viewed from the sprung center of gravity position GC. The sprung acceleration sensor 14-4 detects the sprung acceleration at a fourth position toward the left rear wheel 2RL as viewed from the sprung center of gravity position GC. That is, the numbers 1 to 4 of the symbol "i" correspond to the right front wheel 2FR, the left front wheel 2FL, the right rear wheel 2RR, and the left rear wheel 2RL, respectively. The distance from the sprung center of gravity position GC to each of the first to fourth positions is arbitrary. The sensors 12 include, for example, a lateral acceleration sensor, a suspension stroke sensor, a vehicle height sensor, an unsprung acceleration sensor, and a wheel speed sensor.

1-2. Suspension control The ECU 10 executes suspension control to control the actuator 3A based on the detection signals of the four sprung acceleration sensors 14-i described above. This suspension control includes, as an example, feedback control based on a skyhook control law for reducing vibration of the sprung structure 5 (hereinafter referred to as "FB control of sprung state quantity" or simply "FB control"). Note that, as an example, this FB control is executed for each of the four wheels 2, but instead, for example, it may be executed for only the left and right front wheels 2F or only the left and right rear wheels 2R.

1-2-1. Example of basic configuration of FB control of sprung state quantity In this FB control, the ECU 10 controls the actuator 3A to generate a control force Fc corresponding to a required control amount X required to reduce vibration of the sprung structure 5. In one example described below, the ECU 10 calculates, as the required control amount X, a vertical required control amount Fz, a roll required control amount Mr, and a pitch required control amount Mp, which are required control amounts required to suppress each mode vibration (vertical vibration (heave vibration), roll vibration, and pitch vibration) of the sprung center of gravity position GC.

The detection value of the sprung acceleration detected by each sprung acceleration sensor 14-i (i=1 to 4) is hereinafter referred to as "detected acceleration Z _i ''". The ECU 10 calculates each mode acceleration of the sprung center of gravity position GC (i.e., vertical acceleration Z _g '', roll acceleration Φ _{g '', and pitch acceleration Θ g '') from the detection accelerations Z 1 '' to Z 4 '' detected by the four sprung acceleration sensors 14-1 to 14-4, respectively. For example, the ECU 10 calculates the vertical acceleration Z g '', roll acceleration Φ g ' ' , and pitch} acceleration Θ _g '' according to the following equations (1) to (4). The positions of each sprung acceleration sensor 14-i (i=1 to 4) in the X direction (traveling direction of the vehicle 1) and the Y direction (lateral direction of the vehicle 1) are denoted by L _i and W _i , respectively (see FIG. 1). The X-direction position and the Y-direction position of the sprung center of gravity position GC are _Lg and _Wg , respectively. These parameters (L _i , W _i , _Lg , and _Wg ) are obtained in advance and stored in the storage device of the ECU 10.

By using the accelerations Z ₁ '' to Z ₄ '' detected at the four positions, the vertical acceleration Z _g '', roll acceleration Φ _g '', and pitch acceleration Θ _g '' of the sprung center of gravity position GC can be calculated with high accuracy.

Next, the ECU 10 calculates each modal velocity (vertical velocity Z _g ', roll velocity Φ _g ', and pitch velocity Θ _g ') of the sprung center of gravity position GC by integrating each modal acceleration (vertical acceleration Z _{g '} ', roll acceleration Φ _g '', and pitch acceleration Θ _g ''). The vertical velocity Z _g ', roll velocity Φ _g ', and pitch velocity Θ _g ' of the sprung center of gravity position GC are respectively expressed by the following equations (5) to (7).

Similarly, the ECU 10 calculates each modal displacement (vertical displacement Z _g , roll angle (roll angular displacement) Φ _g , and pitch angle (pitch angular displacement) Θ _g ) of the sprung center of gravity position GC by integrating each modal speed (vertical speed Z _{g '} , roll speed Φ _{g '} , and pitch speed Θ _{g '} ). The vertical displacement Z _g (see FIG. 1), roll angle Φ _g , and pitch angle Θ _g of the sprung center of gravity position GC are respectively expressed by the following equations (8) to (10). Note that the sign of the vertical displacement Z _g is positive when directed upward and negative when directed downward.

Next, the ECU 10 calculates a vertical required control amount _Fz , a roll required control amount _Mr , and a pitch required control amount _Mp for suppressing each mode vibration (vertical vibration, roll vibration, pitch vibration) of the sprung center of gravity position GC. Here, the vertical required control amount _Fz is assumed to be positive when an upward control force is required. The roll required control amount _Mr is assumed to be positive when a right-downward and left-upward control moment is required. The pitch required control amount _Mp is assumed to be positive when a front-wheel-downward and rear-wheel-upward control moment is required.

The ECU 10 calculates these vertical required control amount _Fz , roll required control amount _Mr , and pitch required control amount _Mp using the respective mode accelerations (vertical acceleration _Zg '', roll acceleration _Φg '', and pitch acceleration _Θg '', the respective mode velocities (vertical velocity _Zg ', roll velocity _Φg ', and pitch velocity _Θg '), and the respective mode displacements (vertical displacement _Zg , roll angle _Φg , and pitch angle _Θg ) acquired as described above. For example, the vertical required control amount _Fz , roll required control amount _Mr , and pitch required control amount _Mp are given by the following equations (11) to (13), respectively.

In equations (11) to (13), _G1z , _G2z , _G3z , _G1r , _G2r , _G3r , _G1p , _G2p , and _G3p are control gains. In the following description, unless there is a particular need to distinguish between them, each control gain will be referred to as control gain G.

As shown by equations (11) to (13), the vertical required control amount F _z , the roll required control amount M _r , and the pitch required control amount M _p each have a displacement term, a velocity term, and an acceleration term related to the displacement, velocity, and acceleration of the sprung structure 5.

Specifically, the vertical required control amount _Fz has a displacement term which is the product of the vertical displacement _Zg and the control gain _G1z , a velocity term which is the product of the vertical velocity _Zg ' and the control gain _G2z , and an acceleration term which is the product of the vertical acceleration _Zg '' and the control gain _G3z .

Similarly, the roll required control amount M _r has a displacement term which is the product of the roll angle Φ _g and the control gain G1 _r , a velocity term which is the product of the roll velocity Φ _g ' and the control gain G2 _r , and an acceleration term which is the product of the roll acceleration Φ _g '' and the control gain G3 _r .

Similarly, the pitch request control amount _Mp has a displacement term which is the product of the pitch angle Θ _g and the control gain G1 _p , a velocity term which is the product of the pitch speed Θ _g ' and the control gain G2 _p , and an acceleration term which is the product of the pitch acceleration Θ _g '' and the control gain G3 _p .

Next, the ECU 10 converts (transforms) the required control amounts _Fz , _Mr , and _Mp at the sprung center of gravity position GC into required control amounts for the positions of the wheels 2 (2FR, 2FL, 2RR, and 2RL). The required control amounts for the positions of the wheels 2 correspond to the target control forces Fct of the actuators 3A corresponding to the wheels 2.

The tread width of the front wheels 2F is _Tf , and the tread width of the rear wheels 2R is _Tr . The distance between the front wheel axle and the sprung center of gravity GC is _lf , and the distance between the rear wheel axle and the sprung center of gravity GC is _lr . In this case, the target control forces Fct ( _Fctfr , _Fctfl , _Fctrr , and _Fctrl ) of the actuators 3A are expressed by the following equation (14).

The ECU 10 can convert the required control amounts _Fz , _Mr , and _Mp at the sprung center of gravity position GC into target control forces Fct of the actuators 3A according to the formula (14). Alternatively, the ECU 10 may calculate the target control forces Fct of the wheels 2 by referring to a map based on the required control amounts _Fz , _Mr , and _Mp .

The ECU 10 controls the actuator 3A to generate a control force Fc according to the required control amount X (here, _Fz , _Mr , and _Mp ) calculated as described above. Specifically, the ECU 10 commands the actuator 3A corresponding to each wheel 2 to generate target control forces Fct ( _Fctfr , _Fctfl , Fctrr, and _Fctrl ) for each wheel 2 after conversion from the required control amount X ( _Fz , _Mr , _{and Mp} ) according to equation (14).

In the example shown in the above formulas (11) to (13), the required control amount X has three terms: a displacement term, a speed term, and an acceleration term. When executing this FB control, instead of such an example, the required control amount X may have only one or two control terms: a displacement term, a speed term, and an acceleration term. Also, as shown in formulas (11) to (13), each control term of the required control amount X is expressed by the product of the sprung state amount Y and the control gain G. Therefore, when the control gain G increases, the value (control amount) of the control term having the control gain G, in other words, the control force Fc, increases, and as a result, the control effect (vibration suppression effect or vibration excitation effect) increases.

In addition, the required control amount X in the FB control of the sprung state amount Y may be, for example, a product of the sprung state amount Y (sprung displacement _Zs (see FIG. 2), sprung velocity Zs _' , _sprung acceleration Zs _' ') at the position of each wheel 2 and a control gain, instead of the required control amounts _{Fz ,} Mr, and Mp for suppressing each mode vibration of the sprung center of gravity position GC. The sign of the sprung displacement _Zs is positive when it is upward and negative when it is downward. Also, the sprung state amount Y used in calculating the required control amount X may be the in-phase or out-of-phase component of the sprung state amount Y at the position of the left and right front wheels 2F or the left and right rear wheels 2R.

1-2-2. Issues and Solutions for Suspension Control Based on the Spring Acceleration Sensor In the following description, when there is no need to distinguish between the sprung acceleration sensors 14-i (i = 1 to 4), each sprung acceleration sensor will be referred to as the sprung acceleration sensor (or simply the "acceleration sensor") 14.

1-2-2-1. Sensor Abnormality Detection If an abnormality such as a breakdown occurs in the acceleration sensor 14, suspension control will be performed based on erroneous sensor information. As a result, the vehicle body 6 may be significantly vibrated or the vibration damping effect may be reduced. For this reason, it is necessary to quickly detect an abnormality in the acceleration sensor 14.

Examples of causes of an abnormality in the acceleration sensor 14 include sticking of movable parts inside the sensor, and disconnection (including intermittent disconnection) or short circuit of the sensor circuit or signal line. Here, the output value of the acceleration sensor 14 usually fluctuates all the time while the vehicle is traveling. More specifically, the sensor output signal changes significantly due to various road surface inputs. For this reason, for example, based on the waveform of the sensor output signal (for example, the magnitude of the output signal), it is difficult to accurately determine whether the sensor output signal is abnormal or whether the sprung acceleration Z _s ″, which is actually changing significantly, is being detected normally. In addition, if an abnormality in the acceleration sensor 14, which is difficult to accurately determine in this way, is quickly determined from the waveform of the sensor output signal by simply shortening the determination time, there is a risk of erroneous determination. For example, if the acceleration sensor 14 is erroneously determined to be abnormal when it is normal, the suspension control will be stopped excessively, and the vibration damping effect that should be obtained will not be obtained.

In view of the above-mentioned problems, in this embodiment, the processes executed by the ECU 10 in association with the suspension control (for example, the above-mentioned FB control) based on the acceleration Z _i ″ detected by the sprung acceleration sensor 14 include the following “abnormality detection process” and “sensor identification process”. Specific examples of these processes will be described using the flowchart shown in FIG. 3, but the outline of these processes is as follows.

In the abnormality detection process, the ECU 10 calculates an estimated acceleration Z _iest '' of the sprung acceleration at the position of the remaining acceleration sensor 14 based on the detected acceleration Z i '' of the acceleration sensors 14 that is one less than all of the four acceleration sensors 14 (i.e., _three ). Then, the ECU 10 determines whether or not an abnormal acceleration sensor 14 is included among the four acceleration sensors 14 based on a comparison between the calculated estimated acceleration Z _iest '' and the detected acceleration Z _i '' of the remaining acceleration sensor 14.

The sensor identification process is executed in response to the abnormality detection process detecting the presence of an abnormal acceleration sensor 14. In the sensor identification process, the ECU 10 identifies (determines) as the abnormal acceleration sensor 14 the acceleration sensor whose output signal exhibits abnormal behavior over a predetermined determination time T0 among the four acceleration sensors 14.

1-2-2-2. Suspension control when sensor abnormality is detected Of the abnormality detection process and the sensor identification process according to this embodiment, for example, only the abnormality detection process may be executed in association with the suspension control. Then, when it is determined by the abnormality detection process that an abnormal acceleration sensor 14 is included, for example, the suspension control may be stopped. Also, both the abnormality detection process and the sensor identification process may be executed in association with the suspension control. Then, for example, the suspension control may be stopped after the abnormal acceleration sensor 14 has been identified by the sensor identification process.

If suspension control is then halted as described above when a sensor abnormality is detected by executing at least the abnormality detection process out of the abnormality detection process and the sensor identification process, the state of the sprung structure 5 may suddenly change. Therefore, in this embodiment, the processes executed by the ECU 10 include the following "calculation process" and "abnormality control process." Specific examples of these processes are also described using the flowchart shown in FIG. 3, but the outline of these processes is as follows.

For ease of explanation, the sprung state quantity calculated using the remaining acceleration sensors 14 excluding the abnormal acceleration sensor 14 under a situation in which a sensor abnormality occurs and the required control quantity based thereon will be referred to as "Ya" and "Xa", respectively. The calculation process is a process for calculating the sprung state quantity Ya based on the detected acceleration Z _i '' of the remaining acceleration sensors 14 excluding the identified abnormal acceleration sensor 14 after the abnormal acceleration sensor 14 is identified by the sensor identification process. The abnormality control process is a process for controlling the actuator 3A to generate a control force Fc corresponding to the required control quantity Xa, which is the product of the sprung state quantity Ya calculated by the calculation process and the control gain G.

3 is a flowchart showing a process related to the suspension control according to the embodiment 1. The process of this flowchart is executed repeatedly while the vehicle 1 is traveling, for example.

<Step S100>
In step S100, the ECU 10 acquires acceleration sensor information of the sprung structure 5. Specifically, as an example of the acceleration sensor information referred to here, the ECU 10 acquires the detected acceleration Z _i ″ of each of the four sprung acceleration sensors 14-i.

In addition, the "sprung acceleration sensor" according to the present disclosure does not necessarily have to be installed on the sprung structure 5 as in the example of the sprung acceleration sensor 14, as long as it can acquire the detected acceleration Z _i ''. Specifically, the "sprung acceleration sensor" according to the present disclosure may be configured, for example, by a combination of a vehicle height sensor that detects the vertical displacement of the vehicle body 6 (sprung structure 5) and an unsprung acceleration sensor. More specifically, the detected acceleration Z _i '' may be acquired by calculation based on the detection values of the vehicle height sensor and the unsprung acceleration sensor, or may be acquired by estimation by an observer that uses the detection values of these two sensors. In an example in which a combination of a vehicle height sensor and an unsprung acceleration sensor is used as the sprung acceleration sensor, an abnormality in either one or both of the vehicle height sensor and the unsprung acceleration sensor corresponds to an abnormality in the above-mentioned "sprung acceleration sensor".

<Step S102>
Next, in step S102, the ECU 10 calculates an estimated acceleration Z _iest '' of the sprung acceleration at the position of the remaining acceleration sensor 14 from the detected accelerations Z _i '' of the three acceleration sensors 14. In other words, this estimated acceleration Z _iest '' corresponds to an estimated sensor value of the remaining acceleration sensor 14. It may be arbitrarily determined which of the four sprung acceleration sensors 14-i is to be used as the "remaining acceleration sensor 14". In addition, in the calculation example described below, the acceleration sensor 14-4 (i.e., the acceleration sensor corresponding to the left rear wheel 2RL) is used as the "remaining acceleration sensor 14".

Specifically, under the condition that the sprung structure 5 is a rigid body, the ECU 10 calculates an estimated acceleration Z 4est '' of the remaining acceleration sensor 14-4 from the detected accelerations Z ₁ '', Z ₂ '', and Z ₃ '' of the three acceleration sensors 14-1, 14-2, and 14-3 acquired in step S100. At this time, for example, the ECU 10 first converts the detected accelerations Z ₁ '', Z ₂ '', and Z ₃ '' into a vertical acceleration Z _g '', roll acceleration Φ _g '', and pitch acceleration Θ _g '' of the sprung center of gravity position GC in accordance with the following equations (15) to ( ₁₈ ). Next, the ECU 10 calculates the estimated acceleration Z _4est '' from the vertical acceleration Z _g '', roll acceleration Φ _g '', and pitch acceleration Θ _g '' in accordance with the following equation (19).

<Step S104>
Next, in step S104, the ECU 10 executes a filtering process of applying a band pass filter (BPF) to each of the estimated acceleration Z _4est '' and the detected acceleration (actual sensor value) Z ₄ '' of the acceleration sensor 14-4 acquired in step S102 and step S100, respectively.

According to the processing of step S102 described above, the sensor value of the remaining one of the three acceleration sensors 14 is estimated as the estimated acceleration Z _4est ''. For this reason, due to the influence of twisting and bending of the vehicle body 6, etc., the estimated acceleration Z _4est '' may deviate from the detected acceleration Z ₄ '' with regard to high frequency components. In addition, due to the influence of sensor drift, etc., the estimated acceleration Z _4est '' may deviate from the detected acceleration Z ₄ '' with regard to low frequency components as well.

Therefore, in step S104, the ECU 10 applies a BPF, which has a pass band that is less susceptible to these influences (e.g., 1 to 5 Hz), to each of the estimated acceleration Z _4est '' and the detected acceleration Z ₄ ''. This makes it possible to extract frequency band components in which deviation between the two is less likely to occur, making it possible to perform the next abnormality detection process more accurately.

<Step S106>
The process of step S106 corresponds to an example of the above-mentioned "abnormality detection process." In step S106, to compare the estimated acceleration Z _4est "with the detected acceleration Z ₄ ", the ECU 10 uses, as an example, the absolute value of the difference ΔZ between the estimated acceleration Z _4est "and the detected acceleration Z ₄ " (= Z _4est "- _{Z 4} "). Then, it is determined whether the absolute value of the difference ΔZ is less than a predetermined threshold value TH1.

If there is no abnormality in the three acceleration sensors 14-1, 14-2, and 14-3 used to calculate the estimated acceleration Z _4est '', the estimated acceleration Z _4est '' and the detected acceleration Z ₄ '' will be equal or substantially equal. Therefore, the absolute value of the difference ΔZ will be small. Therefore, if the absolute value of ΔZ is less than the threshold value TH1 (step S106; Yes), the ECU 10 determines that the four acceleration sensors 14 are normal (in other words, none of the acceleration sensors 14 is abnormal).

On the other hand, if the absolute value of ΔZ is equal to or greater than the threshold value TH1 (step S106; No), the ECU 10 determines that an abnormal acceleration sensor 14 is included among the four acceleration sensors 14.

Now, with reference to FIG. 4 , the reason why the processing in step S106 makes it possible to quickly determine that an abnormal acceleration sensor 14 is included among the four acceleration sensors 14, regardless of which of the four acceleration sensors 14 is to _{be used for calculating the estimated acceleration Z iest '} ', and by simply comparing the estimated acceleration Z _iest '' with the detected acceleration Z i '', will be described.

Figure 4 is a supplementary diagram for explaining the abnormality detection process (step S106). More specifically, Figures 4(A) to 4(D) all show an example in which an abnormality (x) has occurred in acceleration sensor 14-3, and the remaining three acceleration sensors 14-1, 14-2, and 14-4 are normal (o).

FIG. 4(A) shows an example in which the estimated acceleration Z _iest '' is calculated for the acceleration sensor 14-3. In this example, the estimated acceleration Z _3est '' calculated using the three normal acceleration sensors 14-1, 14-2, and 14-4 deviates from the detected acceleration Z ₃ '' of the abnormal acceleration sensor 14-3. In each of the remaining examples shown in FIGS. 4(B) to 4(D), the estimated acceleration Z _iest '' is calculated using the detected acceleration Z _i '' of the abnormal acceleration sensor 14, and therefore deviates from the detected acceleration Z _i '' of the normal acceleration sensor 14.

In this way, according to the process of step S106, it is possible to quickly determine whether an abnormal acceleration sensor 14 is included by simply comparing the estimated acceleration _{Z iest '} ' with the detected acceleration Z _i '' once for any acceleration sensor 14 as the calculation target of the estimated acceleration Z iest ''. More specifically, an abnormality can be detected more quickly than when the process of detecting a sensor abnormality based on the presence or absence of abnormal behavior of the output signal of the sprung acceleration sensor is performed alone using a long determination time (e.g., 30 seconds).

In addition, the value used in the comparison between the estimated acceleration Z _iest '' and the detected acceleration Z _i '' in step S106 is not limited to the above-mentioned simple difference ΔZ as long as it is a method capable of determining (evaluating) the difference between the two values. Specifically, for example, in order to compare the estimated acceleration Z _iest '' and the detected acceleration Z _i '' after more sufficiently removing high-frequency noise that may be superimposed on each of them, a low-pass filter (LPF) may be applied to both of them. Then, the absolute value of the difference between the two values after the LPF may be used for the comparison. Alternatively, the absolute value of the difference between the moving average values of the estimated acceleration Z _iest '' and the detected acceleration Z _i '' may be used for the comparison. Furthermore, instead of using these differences, for example, a correlation coefficient related to the estimated acceleration Z _iest '' and the detected acceleration Z _i '' may be used for the comparison.

<Step S108>
As described above, if the determination result in step S106 is Yes (i.e., all four acceleration sensors 14 are normal), the process proceeds to step S108. In step S108, the ECU 10 executes this FB control, which is an example of suspension control based on the acceleration Z _i ″ detected by the four sprung acceleration sensors 14, while using a normal control gain G (hereinafter, for convenience, referred to as a "normal gain Gn"). This allows the FB control to continue as normal.

In addition, the normal gain Gn here refers to a basic value of the control gain G used in suspension control based on the acceleration Z _i ″ detected by the sprung acceleration sensor 14. In this example of FB control that utilizes a required control amount X having three control terms, a proportional term, a speed term, and an acceleration term, the basic values of the control gain G of the three control terms that are set in advance based on the balance between the control terms in accordance with the skyhook control law correspond to the normal gain Gn.

<Step S110>
If the determination result in step 106 is No (i.e., if an abnormal acceleration sensor 14 is included), the process proceeds to step S110. The process in step S110 corresponds to an example of the above-mentioned "sensor identification process."

The sensor identification process is performed for each of the four acceleration sensors 14. In the sensor identification process, the ECU 10 identifies (determines) as an abnormal acceleration sensor 14 an acceleration sensor whose output signal shows "abnormal behavior" over a predetermined determination time T0. This determination time T0 is, for example, a period of about 10 seconds or 5 seconds, with the current time as the end point.

An example of the "abnormal behavior" of the sensor signal to be determined in the sensor identification process is caused by sticking of the acceleration sensor 14. Specifically, when sticking occurs, the fluctuation range of the output signal of the acceleration sensor 14 (i.e., the detected acceleration Z _i '') becomes small. Therefore, in the sensor identification process, the ECU 10 determines whether the time during which the fluctuation range of the output signal of the acceleration sensor 14 is maintained at or below a predetermined threshold TH2 exceeds a determination time T0. This threshold TH2 is determined in advance as a value equivalent to the upper limit of the fluctuation range obtained when sticking occurs, for example.

Another example of "abnormal behavior" of the sensor signal occurs due to a disconnection or short circuit in the acceleration sensor 14. Specifically, if there is a disconnection in the acceleration sensor 14, the output signal of the acceleration sensor 14 will continuously show zero. If there is a short circuit in the acceleration sensor 14, the output signal of the acceleration sensor 14 will continuously show the same value as the power supply voltage. In this way, if there is a disconnection or short circuit in the acceleration sensor 14, the acceleration sensor 14 will continuously output an output signal outside the normal sensor output range. Therefore, in the sensor identification process, the ECU 10 determines whether the time during which the acceleration sensor 14 continues to output an output signal outside the normal sensor output range exceeds the determination time T0. As a result, an acceleration sensor 14 that continues to output an output signal outside the sensor output range for a time exceeding the determination time T0 is determined (identified) to be abnormal.

If an abnormal acceleration sensor 14 is identified in step S110 (step S110; Yes), the process proceeds to step S112. On the other hand, if an abnormal acceleration sensor 14 is not identified (step S110; No), the process proceeds to step S108, which has already been described.

In addition, if the absolute value of ΔZ is equal to or greater than threshold value TH1 in step S106 and processing proceeds to step S110, an abnormal acceleration sensor 14 is basically identified. However, there is a possibility that an erroneous determination will be made in step S106, and as a result, although it is exceptional, processing may proceed to step S108 after proceeding to step S110. The processing of step S110 also serves as a double-check function for detecting sensor anomalies in such exceptional cases. In other words, the sensor identification processing can be used to suppress erroneous determinations in the anomaly detection processing.

<Step S112>
The process of step S112 corresponds to an example of the above-mentioned "calculation process." In step S112, the ECU 10 calculates the sprung state quantity Ya based on the detected accelerations Z _i ″ of the remaining three acceleration sensors 14 excluding the acceleration sensor 14 identified by the sensor identification process as having an abnormality.

Specifically, for example, when the FB control is executed as the suspension control based on the sprung state quantity Ya, the ECU 10 calculates the sprung state quantity Ya of the acceleration term, the velocity term, and the displacement term. This calculation can be performed, for example, as follows. More specifically, the ECU 10 uses the above equations (16) to (18) to calculate each mode acceleration (vertical acceleration Z _{g '', roll acceleration Φ g ''} , and pitch acceleration Θ _g '') which is the sprung state quantity Ya of the acceleration term of each of the above equations (11) to ( ₁₃ ) based on the acceleration Z _i '' detected by the three normal acceleration sensors 14. Next, the ECU 10 calculates each mode velocity (vertical velocity Z _g ', roll velocity Φ _g ', and pitch velocity Θ _g ') which is the sprung state quantity Ya of the velocity term by integrating each calculated mode acceleration according to the equations (5) to (7). Similarly, the ECU 10 further integrates each calculated modal speed according to equations (8) to (10) to calculate each modal displacement (vertical displacement Z _g , roll angle Φ _g , and pitch angle Θ _g ), which are the sprung state quantities Ya of the displacement terms.

<Step S114>
The process of step S114 and the subsequent step S116 corresponds to one example of the above-mentioned "abnormality control process". First, in step S114, the ECU 10 sets an abnormality gain Ga that is smaller than the normal gain Gn. Note that the amount of reduction of the abnormality gain Ga with respect to the normal gain Gn is determined in advance.

Here, in the example of FB control having a displacement term, a velocity term, and an acceleration term, it is the acceleration term that has a particularly large effect on the stability of the FB control. Therefore, in step S114, the change to the abnormality gain Ga is performed by focusing on the acceleration term, which has a particularly large effect on the stability of the FB control. Specifically, of the three control terms, at least the control gain G of the acceleration term is changed to the abnormality gain Ga. For example, only the control gain G of the acceleration term is changed to the abnormality gain Ga. Alternatively, along with the acceleration term, the control gain G of one or both of the displacement term and the velocity term may also be changed to the abnormality gain Ga. Note that, when the control gain G of multiple control terms is changed, the amount of reduction of the abnormality gain Ga relative to the normal gain Gn may be determined in advance for each control term. For example, when all the control gains G of the three control terms are reduced, the amount of reduction of the control gain G of the acceleration term may be increased more than the amount of reduction of at least one of the control gains G of the displacement term and the velocity term.

<Step S116>
Next, in step S116, the ECU 10 executes this FB control, which is an example of suspension control based on the acceleration Z _i ″ detected by the three normal sprung acceleration sensors 14, while using the abnormality gain Ga set in step S114. More specifically, the ECU 10 controls the actuator 3A to generate the control force Fc corresponding to the required control amount Xa, which is the product of the sprung state amount Ya calculated in step S112 and the abnormality gain Ga.

1-3. Effects As described above, according to the first embodiment, the above-mentioned abnormality detection process is executed as a process related to suspension control based on the detected acceleration Z _i ″ of the sprung acceleration sensor 14. According to this abnormality detection process, as already described with reference to FIG. 4, it is possible to quickly determine that an abnormal acceleration sensor 14 is included among the four acceleration sensors 14. As a result, when the abnormality detection process is executed prior to the start of suspension control, it is possible to quickly determine that all four acceleration sensors 14 are normal, and the control start timing can be advanced. Furthermore, when the abnormality detection process is executed constantly or periodically during the execution of suspension control, it is possible to quickly start measures (such as continuing control by abnormality control process or stopping control) in response to the detection of a sensor abnormality.

In addition, in the sensor abnormality detection according to the first embodiment, the above-mentioned sensor identification process is executed together with the abnormality detection process. That is, the sensor identification process is executed under a situation in which it is determined by the abnormality detection process that an abnormal acceleration sensor 14 is included among the four acceleration sensors 14. By performing the determination in two stages in this way, it becomes possible to identify the abnormal acceleration sensor 14 while shortening the determination time T0 used in the sensor identification process compared to the case in which only the sensor identification process is executed without the abnormality detection process. Specifically, for example, in the case in which a determination time T0 of, for example, 30 seconds is required for accurate determination in the example of only the sensor identification process, the determination time T0 can be shortened to, for example, 10 seconds or 5 seconds by this two-stage determination. As described above, the combination of the abnormality detection process and the sensor identification process makes it possible to quickly identify the abnormal acceleration sensor 14.

Furthermore, according to the first embodiment, after the abnormal acceleration sensor 14 is identified by the sensor identification process, the above-mentioned calculation process and abnormality control process are executed. As a result, the sprung state quantity Ya can be calculated (estimated) based on the acceleration detected by the acceleration sensors 14 excluding the abnormal acceleration sensor 14, and _suspension control can be continued. More specifically, compared to the case where the acceleration detected by the abnormal acceleration sensor 14 is continued after the occurrence of an abnormality, vibration of the sprung structure 5 (vehicle body 6) caused by the sensor abnormality can be prevented. Furthermore, by continuing the suspension control using the remaining normal acceleration sensor 14, the comfort of the vehicle 1 and the stability of the suspension control can be ensured even when an abnormality is detected.

Furthermore, according to the abnormality control process, an abnormality gain Ga is used that is smaller than the normal gain Gn that is used when the four acceleration sensors 14 are normal. In this way, by reducing the required control amount Xa using the abnormality gain Ga when a sensor abnormality is detected, it is possible to suppress the discomfort in control that may occur when suspension control is continued using a smaller number of acceleration sensors 14 than normal. More specifically, it is possible to more appropriately maintain the stability of suspension control in a situation where a smaller number of acceleration sensors 14 than normal is used.

As already explained, in the example of FB control having a displacement term, a velocity term, and an acceleration term, the change to the abnormality gain Ga may be made with a focus on the acceleration term, which has a particularly large effect on the stability of the FB control. Here, since reducing the control gain G may lead to a decrease in the vibration damping effect of this FB control, it is desirable to keep the change to the abnormality gain Ga to the minimum necessary to ensure the stability of the FB control. Therefore, by changing only the control gain G of the acceleration term to the abnormality gain Ga, control stability may be ensured while suppressing the decrease in the vibration damping effect. Alternatively, for the same purpose, the change to the abnormality gain Ga may be made to only the acceleration term and either the displacement term or the velocity term, rather than all three control terms.

1-4. Modifications In the above-described first embodiment, the vehicle 1 including four sprung acceleration sensors 14 has been exemplified. However, the "abnormality detection process" described in the first embodiment may be applied to a vehicle including five or more sprung acceleration sensors 14. In an example including five or more sprung acceleration sensors 14, the location of the one or more sprung acceleration sensors 14 added to the four sprung acceleration sensors 14-1, 14-2, 14-3, and 14-4 is not particularly limited.

For example, the "abnormality detection process" in an example having five sprung acceleration sensors 14 is executed as follows. That is, according to the above equations (1) to (4), the vertical acceleration Z _g '', the roll acceleration Φ _g '', and the pitch acceleration Θ _g '' are calculated from the detected accelerations of any four sprung acceleration sensors 14. Next, an estimated acceleration at the position of the remaining acceleration sensor 14 is calculated using an equation similar to the above equation (19). The calculated estimated acceleration is then compared with the detected acceleration of the remaining acceleration sensor 14. As a result, it can be determined that an abnormal acceleration sensor 14 is included, similar to the example having four acceleration sensors 14.

In addition, the "sensor identification process" in a vehicle equipped with, for example, five sprung acceleration sensors 14 can be performed using a method similar to the process of step S110 described above (first method). In addition, the sensor identification process in an example equipped with five or more sprung acceleration sensors 14 may be performed using, for example, a second method as follows.

In the second method, for example, in the case of a vehicle equipped with five sprung acceleration sensors 14, the above formulas (15) to (19) are used to calculate the estimated acceleration of one of the remaining two acceleration sensors 14 from the detected acceleration of the three acceleration sensors 14, and the process is executed multiple times while sequentially changing the combination of the three acceleration sensors 14 used for the estimated acceleration and the acceleration sensor 14 that is the subject of the calculation of the estimated acceleration. As a result, if the estimated acceleration and the detected acceleration of the acceleration sensor 14 that is the subject of the calculation of the estimated acceleration are equal or substantially equal, it is determined that the four acceleration sensors 14 that were the subject of the comparison are normal. Therefore, by executing such a process multiple times, it is possible to identify an abnormal acceleration sensor 14. More specifically, in a configuration example including the four acceleration sensors 14-1 to 14-4 shown in FIG. 1 and an acceleration sensor 14-g arranged at the sprung center of gravity position GC, if the estimated acceleration at the sprung center of gravity position GC based on the three acceleration sensors 14-1, 14-2, and 14-3 is equal or substantially equal to the detected acceleration of the acceleration sensor 14-g, it can be determined that the remaining acceleration sensor 14-4 is abnormal.

In addition, in an example having five or more sprung acceleration sensors 14, if there are four or more wheels 2 corresponding to acceleration sensors 14 that are not abnormal, the following technology can be used to calculate the sprung state quantity Ya used for suspension control while reducing the effect of torsion of the sprung structure 5 (vehicle body 6). Details of this technology are described in JP 2019-010912 A. The technology is summarized as follows.

That is, according to the above technique, first, the respective mode accelerations (vertical acceleration Z g '', roll acceleration Φ _g '', and pitch acceleration Θ _g '') of the spring center of gravity position GC are calculated from the detected accelerations of the four sprung acceleration sensors 14 according to the same four equations as the above equations ( ₁ ) to (4). Then, according to the technique, under the condition that the sprung structure 5 is a rigid body, the respective mode accelerations of the sprung center of gravity position GC are converted (transformed) into sprung accelerations (vertical accelerations) Z _s1 '' to Z _s4 '' at the positions of the respective wheels 2. This conversion can be performed, for example, using the following equation (20).

By utilizing the above-described technology, a vertical acceleration Z _si ″ in which the influence of torsional vibration is reduced can be obtained. By using the vertical acceleration Z _si ″ in which the influence of torsional vibration is reduced as the sprung state quantity Ya for suspension control, it is possible to improve the stability of suspension control and to improve the vibration damping effect by using a large control gain G, compared to a case in which the influence of torsion is not eliminated. This is convenient for favorably continuing suspension control under a situation in which a smaller number of acceleration sensors 14 are used compared to normal times due to detection of a sensor abnormality. In addition, if a sufficient number of acceleration sensors 14 remain at the time of detection of a sensor abnormality to calculate the vertical acceleration Z _si ″ taking into account the influence of bending of the sprung structure 5 (vehicle body 6), the vertical acceleration Z _si ″ may be calculated using a known method so as to reduce the influence of the bending.

2. Second embodiment
The second embodiment differs from the first embodiment in the method of detecting an abnormality in a sensor. Specifically, in the second embodiment, the following "sensor identification process" is executed to detect an abnormality in a sensor.

FIG. 5 is a supplementary diagram for explaining the sensor identification process according to the second embodiment. In the sensor identification process of this embodiment, the ECU 10 calculates a "specific value SV" which is a related value of the vertical displacement (heave) Z, roll angle Φ, or pitch angle Θ at the same position of the sprung structure 5 based on the detected acceleration Z _i '' of three acceleration sensors 14 out of the four sprung acceleration sensors 14 equipped in the vehicle 1. The "same position" here is, for example, the sprung center of gravity position GC, but may be any position on the sprung structure 5 other than the sprung center of gravity position GC. In addition, the related value of the vertical displacement Z, roll angle Φ, or pitch angle Θ here may be not only the vertical displacement Z, roll angle Φ, or pitch angle Θ itself, but also, for example, the vertical velocity Z', roll velocity Φ', pitch velocity Θ', vertical acceleration Z'', roll acceleration Φ'', or pitch acceleration Θ''.

The ECU 10 calculates the specific value SV for each of the four different combinations of the three acceleration sensors 14. Specifically, in the example of a vehicle 1 equipped with four acceleration sensors 14, as shown in FIG. 5, the different combinations of the three acceleration sensors 14 are combinations A to D. The table shown in FIG. 5(B) shows the breakdown of the four combinations A to D of the three acceleration sensors 14 shown in FIG. 5(A) using the wheel 2 number i (i=1 to 4).

In the sensor identification process, the ECU 10 identifies an abnormal acceleration sensor 14 from among the four acceleration sensors 14 based on a comparison of the specific values SV between the four different combinations A to D.

Specifically, if all four acceleration sensors 14 are normal, the specific values SV calculated using each of combinations A to D will be equal or substantially equal. Therefore, if the specific values SV calculated using each of combinations A to D are all within the specified range Rth, it can be determined that all four acceleration sensors 14 are normal.

On the other hand, if an abnormality occurs in any one of the four acceleration sensors 14, a difference exceeding the predetermined range Rth will be observed in the specific value SV among the combinations A to D. Specifically, for example, if an abnormality occurs in the acceleration sensor 14-1 with the sensor number i=1, the specific value SV D of the combination D not including the acceleration sensor 14-1 will deviate from the specific values SV _A , SV _B , and SV _C of the three combinations A to C including the acceleration sensor 14-1. Therefore, for example, if the absolute value of the difference ΔSV between the specific value SV _{A-C and} the specific value SV _D is equal to or greater than a predetermined threshold value TH2, it can be determined that an abnormality occurs in the acceleration sensor 14-1. Note that the value used to compare the specific value SV _A-C with the specific value SV _D is not limited to the above-mentioned simple difference ΔSV, as long as it is a method capable of determining (evaluating) the difference between the two values. That is, similar to the example of the processing of step S106 described above, for example, the absolute value of the difference between the two values after the LPF, the absolute value of the difference between their respective moving average values, or the correlation coefficient between the two may be used for the above comparison.

Based on a similar concept, an abnormality in the acceleration sensor 14-2 with sensor number i=2 can be determined by noting that the specific value SV _C of combination C not including the acceleration sensor 14-2 deviates from the specific values SV _A , SV _B , and SV _D of the remaining combinations A, B, and D. An abnormality in the acceleration sensor 14-3 with sensor number i=3 can be determined by noting that the specific value SV _B of combination B deviates from the others. Furthermore, an abnormality in the acceleration sensor 14-4 with sensor number i=4 can be determined by noting that the specific value SV _A of combination A deviates from the others.

As described above, according to the sensor identification process of the second embodiment, by calculating a specific value SV for each of the four combinations A to D and comparing the specific values SV between the combinations A to D, it is possible to quickly determine whether the four acceleration sensors 14 are normal or not, and to quickly identify one abnormal acceleration sensor 14 that may be included among the four acceleration sensors 14. More specifically, an abnormality can be detected more quickly than when the process of detecting a sensor abnormality based on the presence or absence of abnormal behavior of the output signal of the sprung acceleration sensor is performed alone using a long determination time (e.g., 30 seconds).

Figure 6 is a flowchart showing processing related to suspension control according to the second embodiment. The processing of this flowchart is executed repeatedly, for example, while the vehicle 1 is traveling. Note that explanations of the processing of steps that are the same as the steps shown in the flowchart shown in Figure 3 are omitted or simplified.

6, in step S200 following step S100, the ECU 10 calculates specific values SV (SV _A to SV _D ) for each of the four combinations A to D. The specific values SV can be calculated, for example, by using the above formula (15) and any one of formulas (16) to (18).

Next, in order to perform the sensor identification process more accurately, in step S202, ECU 10 applies a BPF to each specific value SV calculated in step S200 based on the same concept as in step S104 described above.

The process of the next step S204 corresponds to an example of the "sensor identification process" described above. In step S204, the ECU 10 compares the specific values SV between the four combinations A to D as already described, thereby determining whether or not there is an abnormal acceleration sensor 14, and identifying the acceleration sensor 14 if there is an abnormal acceleration sensor 14.

As a result, if all four acceleration sensors 14 are normal (step S204; No), the process proceeds to step S108. On the other hand, if an abnormal acceleration sensor 14 is present and the abnormal acceleration sensor 14 has been identified (step S204; Yes), the process proceeds to step S112.

In addition, the "sensor identification process" described in the second embodiment as above may be applied to a vehicle equipped with five or more acceleration sensors 14, instead of the example of four acceleration sensors 14. Specifically, in the example of a vehicle equipped with five acceleration sensors 14, for example, the calculation of the "specific value SV" based on the detected acceleration of four of the five acceleration sensors 14 may be performed for each of five different combinations of the four acceleration sensors 14. Then, based on a similar idea to the example of vehicle 1 equipped with four acceleration sensors 14 described above, an abnormal acceleration sensor 14 may be identified from among the five acceleration sensors 14 based on a comparison of the specific values SV between the five different combinations.

3. Third embodiment
The method of detecting an abnormality in a sensor in the third embodiment is different from that in the first and second embodiments. Specifically, in the third embodiment, in order to detect an abnormality in a sensor, the following "abnormality detection process" and "sensor identification process" are executed.

Figure 7 is a diagram that shows an example of the configuration of a vehicle 7 according to embodiment 3. The vehicle 7 is configured similarly to the vehicle 1 shown in Figure 1, except that the vehicle 7 additionally includes a gyro sensor 8.

The gyro sensor 8 is attached to the sprung structure 5 (vehicle body 6) and detects the angular velocity of the sprung structure 5. Specifically, this angular velocity is, for example, the roll velocity Φ' or pitch velocity Θ' of the sprung structure 5.

In the "abnormality detection process" according to the present embodiment, the ECU 10 calculates a first estimated angular velocity ω gyest 1 which is an estimated value of the angular velocity of the sprung structure 5 at the position of the gyro sensor 8, based on the detected accelerations _{Z i} ″ of all four acceleration sensors 14 provided on the vehicle 7. Then, the ECU 10 determines whether or not an abnormal acceleration sensor 14 is included among the four acceleration sensors 14, based on a comparison between the calculated first estimated angular velocity ω _gyest 1 and the detected angular velocity ω _gy of the gyro sensor 8. Note that the first estimated angular velocity ω _gyest 1 is an example of a "first estimated angular velocity related value of the sprung structure at the position of the gyro sensor" according to the present disclosure. The "first estimated angular velocity related value" may be, for example, a first estimated acceleration or a first estimated angle which is an estimated value of the angular acceleration or angle, respectively, of the sprung structure 5 at the position of the gyro sensor 8. Similarly, the detected angular velocity ω _gy is an example of a "detected angular velocity related value of the gyro sensor" according to the present disclosure. The "detected angular velocity related value" may be, for example, a detected angular acceleration which is a time differential value of the detected angular velocity ω _gy , or a detected angle which is a time integral value of the detected angular velocity ω _gy .

Specifically, if all four acceleration sensors 14 are normal, the first estimated angular velocity ω _gyest 1 based on the detected accelerations Z _i ″ of these four acceleration sensors 14 is equal or substantially equal to the detected angular velocity ω _gy of the gyro sensor 8. On the other hand, if an abnormality occurs in any of the four acceleration sensors 14, the first estimated angular velocity ω _gyest 1 deviates from the detected angular velocity ω _gy . Therefore, according to this abnormality detection process, it is possible to quickly determine whether or not an abnormal acceleration sensor 14 is included based on a comparison between the first estimated angular velocity ω _gyest 1 and the detected angular velocity ω _gy . More specifically, an abnormality can be detected more quickly than when a process for detecting a sensor abnormality based on the presence or absence of abnormal behavior of the output signal of the sprung acceleration sensor is performed alone using a long determination time (e.g., 30 seconds).

The "sensor identification process" according to this embodiment also includes the following first and second examples. Specifically, the first example of this sensor identification process is the same as the sensor identification process described in embodiment 1 (see step S110).

Next, a second example of the sensor identification process will be described. In the second example, the ECU 10 executes the following process after detecting that an abnormal acceleration sensor 14 is included by the abnormality detection process. That is, the ECU 10 calculates a second estimated angular velocity ω gyest 2 which is an estimated value of the angular velocity of the sprung structure 5 at the position of the gyro sensor 8 based on the detected accelerations Z _i ″ of three of the four acceleration sensors 14 equipped in the vehicle 7. The second estimated angular velocity ω _gyest 2 is an example of _a “second estimated angular velocity related value of the sprung structure at the position of the gyro sensor” according to the present disclosure. The “second estimated angular velocity related value” may be, for example, a second estimated acceleration or a second estimated angle which is an estimated value of the angular acceleration or angle of the sprung structure 5 at the position of the gyro sensor 8, respectively.

Then, the ECU 10 repeatedly calculates the second estimated angular velocity ω _gyest 2 a required number of times while changing the combination of the three acceleration sensors 14 until the calculated second estimated angular velocity ω _gyest 2 and the detected angular velocity ω _gy are equal or substantially equal to each other. As a result, when the calculated second estimated angular velocity ω _gyest 2 and the detected angular velocity ω _gy are equal or substantially equal to each other, the ECU 10 identifies (confirms) that the remaining sprung acceleration sensor 14 that was not used in the latest (i.e., current) calculation of the second estimated angular velocity ω _gyest 2 is the abnormal acceleration sensor 14.

Specifically, in the second example, when the first second estimated angular velocity ω _gyest 2 is calculated using, for example, the three acceleration sensors 14-1, 14-2, and 14-3, if this second estimated angular velocity ω _gyest 2 and the detected angular velocity ω _gy are equal or substantially equal, the remaining acceleration sensor 14-4 is determined to be abnormal. On the other hand, if this first second estimated angular velocity ω _gyest 2 deviates from the detected angular velocity ω _gy , it can be determined that one of the three acceleration sensors 14-1, 14-2, and 14-3 is abnormal. Therefore, the combination of the three acceleration sensors 14 is changed from that of the first calculation, and the second second estimated angular velocity ω _gyest 2 is calculated. By repeating the calculation of the second estimated angular velocity ω _gyest 2 and the comparison of the second estimated angular velocity ω _gyest 2 with the detected angular velocity ω _gy a required number of times (that is, as necessary), an abnormal acceleration sensor 14 can be identified.

Figure 8 is a flowchart showing processing related to suspension control according to the third embodiment. The processing of this flowchart is executed repeatedly, for example, while the vehicle 1 is traveling. Note that explanations of the processing of steps that are the same as the steps shown in the flowchart shown in Figure 3 are omitted or simplified.

In FIG. 8, in step S300 following step S100, the ECU 10 acquires the detected angular velocity ω _gy of the gyro sensor 8.

Next, in step S302, the ECU 10 calculates a first estimated angular velocity ω _gyest 1 based on the accelerations Z _i ″ detected by all of the four acceleration sensors 14. In an example in which the angular velocity of the gyro sensor 8 is the roll velocity Φ', the first estimated angular velocity ω _gyest 1 can be calculated using the above equations (1), (3), and (6) while changing the position of the calculation target from the sprung center of gravity position GC to the position of the gyro sensor 8. The first estimated angular velocity ω _gyest 1 in the example in which the pitch velocity Θ' is used can be calculated in a similar manner using the above equations (1), (4), and (7).

Note that the BPF may be applied to the detected angular velocity ω _gy and the first estimated angular velocity ω _gyest 1 acquired in steps S300 and S302, similarly to the process in step S104.

The process of the next step S304 corresponds to an example of the above-mentioned "abnormality detection process." In step S304, the ECU 10 determines whether or not an abnormal acceleration sensor 14 is included among the four acceleration sensors 14 based on a comparison between the first estimated angular velocity ω _gyest 1 and the detected angular velocity ω _gy , as already described. If the result of this determination is No, the process proceeds to step S108, and if the result of this determination is Yes, the process proceeds to step S306.

The process of step S306 corresponds to an example of the "sensor identification process" described above. As already explained, either the first example or the second example described above may be used as the sensor identification process. As a result, if an abnormal acceleration sensor 14 is identified (step S306; Yes), the process proceeds to step S112. On the other hand, if an abnormal acceleration sensor 14 is not identified (step S306; No), the process proceeds to step S108.

Additionally, the "abnormality detection process" described above in the third embodiment may be applied to a vehicle equipped with five or more acceleration sensors 14, instead of the example of four acceleration sensors 14. Specifically, based on the same idea as the above example, in the example of a vehicle equipped with, for example, five acceleration sensors 14, the "first estimated angular velocity ω _gyest 1" calculated based on the detected accelerations Z _i '' of four of the five acceleration sensors 14 may be compared with the detected angular velocity ω _gy to detect whether an abnormal acceleration sensor 14 is included.

Furthermore, the second example of the "sensor identification process" described in the third embodiment may be applied to a vehicle equipped with five or more acceleration sensors 14, instead of the example of four acceleration sensors 14. Specifically, based on the same idea as the above example, in the example of a vehicle equipped with, for example, five acceleration sensors 14, the second estimated angular velocity ω _gyest 2 based on the detected accelerations Z _i ″ of four of the five acceleration sensors 14 may be compared with the detected angular velocity ω _gy to identify an abnormal acceleration sensor 14.

(Reference example)
In a vehicle having at least four sprung acceleration sensors, there is a possibility that an abnormality occurs in a plurality of the sprung acceleration sensors. In an example in which the number of remaining normal sprung acceleration sensors is four, the sprung state quantity Ya used in the suspension control (for example, the above-mentioned FB control) can be calculated while eliminating the influence of the torsion of the sprung structure 5 (vehicle body 6) by using, for example, the technique described with reference to the formula (20). In an example in which the number of remaining normal sprung acceleration sensors is three, the sprung state quantity Ya can be calculated by the process of step S112 (for example, see FIG. 3). On the other hand, it is also assumed that the number of remaining normal sprung acceleration sensors will be two or one.

Whether the number of remaining normal sprung acceleration sensors is two or one can be determined, for example, by executing the sensor identification process described in step S100 alone. If the number of remaining normal sprung acceleration sensors is two or one, suspension control may be continued, for example, by the following method. That is, for example, in a configuration example in which the two or one remaining normal sprung acceleration sensor is attached to a position sufficiently close to the wheel (for example, a suspension tower), suspension control based on the sprung state quantity based on the acceleration detected by the remaining normal acceleration sensor (for example, the above-mentioned FB control) may be performed on a wheel-by-wheel basis for only the wheel corresponding to that acceleration sensor.

On the other hand, even if it is difficult to calculate the sprung state quantity in an example where there are only two or one normal sprung acceleration sensors remaining, if the vehicle is equipped with a height sensor or a stroke sensor, suspension control may be continued for each wheel as follows. That is, instead of FB control based on the sprung state quantity, the suspension actuator may be controlled to generate a control force Fc corresponding to a required control quantity obtained by multiplying the suspension stroke speed obtained from the height sensor or stroke sensor by a control gain. This enables suspension control simulating a damper (shock absorber) in a situation where the sprung acceleration sensor cannot be used satisfactorily due to the occurrence of an abnormality, and makes it possible to ensure the stability of vehicle behavior.

Reference Signs List 1, 7 Vehicle 2 Wheel 3 Suspension 3A Suspension actuator 4 Unsprung structure 5 Spring upper structure 6 Vehicle body 8 Gyro sensor 10 Electronic control unit (ECU)
12 Sensors 14 Spring acceleration sensor

Claims (7)

an actuator that applies a vertical control force between an unsprung structure and a sprung structure of the vehicle;
At least four sprung acceleration sensors for detecting vertical acceleration of the sprung structure;
an electronic control unit that controls the actuator based on the acceleration detected by the at least four sprung acceleration sensors;
A vehicle suspension control device comprising:
The electronic control unit executes a sensor identification process,
In the sensor identification process, the electronic control unit:
Calculating a specific value that is a related value of a vertical displacement, a roll angle, or a pitch angle at the same position of the sprung structure based on the detected accelerations of the sprung acceleration sensors that are one less than all of the at least four sprung acceleration sensors, for each of a plurality of different combinations of the sprung acceleration sensors that are one less than all of the at least four sprung acceleration sensors;
a control device for controlling a suspension of a vehicle, the control device comprising: a step of: identifying an abnormal sprung acceleration sensor from among the at least four sprung acceleration sensors based on a comparison of the specific values among the plurality of different combinations. an actuator that applies a vertical control force between an unsprung structure and a sprung structure of the vehicle;
At least four sprung acceleration sensors for detecting vertical acceleration of the sprung structure;
A gyro sensor for detecting an angular velocity of the sprung structure;
an electronic control unit that controls the actuator based on the acceleration detected by the at least four sprung acceleration sensors;
A vehicle suspension control device comprising:
The electronic control unit executes an abnormality detection process,
In the abnormality detection process, the electronic control unit
Calculating a first estimated angular velocity related value of the sprung structure relative to a position of the gyro sensor based on the accelerations detected by all of the at least four sprung acceleration sensors;
and determining whether or not an abnormal sprung acceleration sensor is included among the at least four sprung acceleration sensors based on a comparison between the calculated first estimated angular velocity-related value and the detected angular velocity-related value of the gyro sensor. 3. The vehicle suspension control device according to claim 2, wherein the electronic control unit executes a sensor identification process to identify, as the abnormal sprung acceleration sensor, one of the at least four sprung acceleration sensors whose output signal exhibits abnormal behavior for a predetermined determination time, after determining through the abnormality detection process that an abnormal sprung acceleration sensor is included. The electronic control unit executes a sensor identification process after determining that an abnormal sprung acceleration sensor is included by the abnormality detection process,
In the sensor identification process, the electronic control unit:
calculating a second estimated angular velocity related value of the sprung structure relative to the position of the gyro sensor based on accelerations detected by the sprung acceleration sensors, the number of which is one less than all of the at least four sprung acceleration sensors;
Repeating the calculation of the second estimated angular velocity related value a necessary number of times while changing the combination of the sprung acceleration sensors, the number of which is one less than all of the sprung acceleration sensors, until the calculated second estimated angular velocity related value and the detected angular velocity related value of the gyro sensor become equal or substantially equal to each other;
3. The vehicle suspension control device according to claim 2, wherein when the calculated second estimated angular velocity related value and the detected angular velocity related value of the gyro sensor become equal or substantially equal, one remaining sprung acceleration sensor not used in the latest calculation of the second estimated angular velocity related value is identified as the abnormal sprung acceleration sensor. The electronic control unit includes:
a calculation process for calculating a sprung state quantity based on the detected accelerations of the remaining sprung acceleration sensors excluding the identified abnormal sprung acceleration sensor after the abnormal sprung acceleration sensor is identified by the sensor identification process;
an abnormality control process for controlling the actuator so as to generate the control force corresponding to a required control amount, which is a product of the calculated sprung state amount and a control gain;
5. The vehicle suspension control device according to claim 1 , 3 or 4 , further comprising: 6. The vehicle suspension control device according to claim 5 , wherein in the abnormality control process, the electronic control unit reduces the control gain compared to a case in which all of the at least four sprung acceleration sensors are normal. The required control amount is required to reduce vibration of the sprung structure, and includes at least one of the displacement term and the velocity term and the acceleration term among a displacement term, a velocity term, and an acceleration term related to a displacement, a velocity, and an acceleration term of the sprung structure,
7. The vehicle suspension control device according to claim 6 , wherein in the abnormality control process, the electronic control unit reduces only the control gain of the acceleration term.

Images