JP5036620B2 - Operating state determination device - Google Patents

Description

  The present invention relates to an operation state determination device that determines an operation state of a driver.

  Conventionally, as one of the research and development fields of an intelligent transport system (ITS), there is a field of a safe driving support system. In the field of this safe driving support system, research and development of accident reduction by driving support such as providing information, alerting, warning, and control to a driver of a running vehicle (own vehicle) is being promoted.

Focusing on the fact that the driver's driving operation changes gradually from normal operation (normal operation) due to physical and mental fatigue, etc., the driver's steering operation is one of the driving operations. It has been proposed to detect the change in steering angle and to determine that the driving state is a doze if the low-frequency component of the detected change is greater than or equal to the normal sample value (see, for example, Patent Document 1). . It has also been proposed to determine whether or not the driver is in an awake driving state based on whether or not the integral value of the high-frequency component of the steering angle change exceeds a reference value (for example, see Patent Document 2).
Japanese Patent Laid-Open No. 5-58192 (for example, abstract, paragraphs [0017], [0018], [0024]-[0025], FIG. 1, etc.) JP-A-60-157927 (for example, page 5, upper left column, line 5 to upper right column, line 18, line 3, FIG. 5, etc.)

  As described in Patent Documents 1 and 2, whether or not the driver is in a drowsy driving state (or whether or not he is in a driving state of awakening) from a change in a low frequency component or a high frequency component of a steering angle change of the steering wheel operation. Therefore, the driving state of the driver is simply determined from the increase / decrease change of the driving operation (handle operation) of the driver.

  In this case, there is a problem that an erroneous determination of the driving state may occur depending on the driving environment such as the state of the driving path or the driver's daily habit. For example, when driving on a long straight road or on an expressway, even if it is a normal steering wheel operation, the operation is gentle and may be erroneously determined as an abnormal driving state such as a snooze driving. Also, depending on the driver's daily driving habits, even a normal steering wheel operation may be erroneously determined as an abnormal driving state for an agile operation. Such an erroneous determination may occur not only in the steering wheel operation but also in other driving operations such as a pedal operation.

  SUMMARY OF THE INVENTION An object of the present invention is to enable a highly accurate determination of a driving state in consideration of a driving environment such as a state of a driving path and a driver's daily habit.

In order to achieve the above-described object, the driving state determination device of the present invention is a driving state determination device that determines the driving state of a driver of a traveling vehicle, and detects the position of the vehicle. Based on the means, the driving operation detecting means for detecting the driving operation of the driver, the detected position of the own vehicle position detecting means, and the detecting operation of the driving operation detecting means, the position of the own vehicle is corrected with respect to the travel path. A characteristic calculation unit that repeatedly calculates the driving operation characteristic of the driver from the control characteristic of the closed loop model of the driving operation of the driver, and a characteristic having a high appearance frequency of the driving operation characteristic that is repeatedly calculated by the characteristic calculation unit a data storage unit for holding freely rewritten, the difference between the driving operation characteristics calculated by the data storage unit oPERATION operating characteristics when held normally by said characteristic calculation means Is characterized by comprising determination means for determining operating conditions of Luo driver (claim 1).

The characteristic calculation means repeatedly calculates the driving operation characteristic of the driver by identifying the transfer function by either the SSARX method which is one of the closed-loop subspace identification methods or the prediction error method effective for the identification of the closed-loop system. The determination means preferably determines the driving state of the driver from the resonance frequency of the latest driving operation characteristic calculated by the characteristic calculation means with respect to the normal driving operation characteristic. .

  In the case of the invention of claim 1, paying attention to the fact that the driving operation of the driver for correcting the position of the vehicle with respect to the travel path can be regarded as a closed loop model (feedback control model), the vehicle position detection means detects the vehicle Based on the position and the driving operation of the driver detected by the driving operation detection means, the characteristic calculation means calculates the control characteristic of the closed loop model as the driving operation characteristic of the driver.

The normal driving operation characteristics and the latest driving operation characteristics from time to time obtained based on this calculation are characteristics including the driving environment such as the condition of the driving path, the driver's habit, and the like. For this reason, when a characteristic with a high frequency of appearance of the driving operation characteristic repeatedly calculated by the characteristic calculating means is retained as a normal driving operation characteristic, it is calculated by the normal driving operation characteristic and the characteristic calculating means. The difference from the latest driving operation characteristics excludes the state of the driving environment and the driver's daily habits.

  Therefore, based on the difference between the normal driving operation characteristic and the latest driving operation characteristic, the driving state of the driver can be accurately determined by the determining means.

  By the way, it has been found that there is a resonance frequency in the driving operation characteristics of the driver calculated from the control characteristics of the closed loop model, and this resonance frequency best represents the characteristics of the driving operation of the driver.

  In the case of the invention of claim 2, the determination means determines the driving state of the driver from the resonance frequency level of the latest driving operation characteristic with respect to the normal driving operation characteristic, thereby further increasing the accuracy of the driving state of the driver. Judgment can be made.

  Next, in order to describe the present invention in more detail, an embodiment will be described in detail with reference to FIGS.

  FIG. 1 shows a block configuration of an operation state determination device 2 mounted on the host vehicle 1, and FIG. 2 is a schematic diagram of a closed loop model of the determination processing unit. FIG. 3 shows frequency response characteristics of the driver model of FIG. 2, and FIG. 4 shows changes in frequency response characteristics of the driver model of FIG. 2 according to the driving state of the driver.

  1 includes, for example, a lane recognition system formed by the photographing means 3 and the lane recognition processing unit 4, and is based on a steering operation that corrects a displacement of the position of the own vehicle 1 with respect to a lane (white line) on the travel path. Determine the operating status of the driver.

  The photographing means 3 is composed of a monochrome or color CCD monocular camera that photographs at least the front road surface of the traveling vehicle 1, and outputs a captured image (a frame image or a field image) every moment to the lane recognition processing unit 4.

  The lane recognition processing unit 4 executes image processing by a microcomputer, differentiates the captured image of the photographing unit 3 every time and performs binarization processing, and by image recognition of the binary image of the processing result, for example, the above-mentioned A feature portion of a road white line defined in advance is cut out from the binary image, and the road white line of the own vehicle traveling lane that changes from time to time is detected from pattern matching processing or the like to recognize the traveling lane of the own vehicle 1.

  The travel lane recognition information of the lane recognition processing unit 4 is input to the determination processing unit 7 together with detection information of various sensors of the traveling state of the host vehicle 1 such as the steering angle sensor 5 and the vehicle speed sensor 6.

  The determination processing unit 7 executes software processing of the microcomputer to form the vehicle position detection means, driving operation detection means, characteristic calculation means, and determination means of the present invention.

  The own vehicle position detecting means detects the position of the own vehicle 1 from moment to moment based on the detected steering angle of the steering angle sensor 5, the detected vehicle speed of the vehicle speed sensor 6, and the like.

  The driving operation detection means detects the driving operation (handle operation) of the driver from moment to moment based on the detected steering angle of the steering angle sensor 5.

  The characteristic calculating means is a closed-loop model in which the driver corrects the positional deviation of the own vehicle 1 with respect to the traveling road by a steering operation based on the detection position of the own vehicle position detecting means with respect to the traveling road and the detection operation of the driving operation detecting means. The control characteristic of the feedback control model) is obtained, and the driving operation characteristic of the driver is repeatedly calculated (identified).

  That is, the travel position of the host vehicle 1 relative to the travel lane is detected from the travel lane of the host vehicle 1 recognized by the lane recognition processing unit 4 and the detection position of the host vehicle position detection means, and collected in the data storage unit 8. For example, the normal position (target position) of the vehicle 1 is repeatedly updated from the initial set position to the travel position having the highest appearance frequency.

  Furthermore, the deviation (error) of the running position from time to time with respect to the target position is input u, the steering operation of the driver for correcting it, that is, the detected steering angle of the driving operation detecting means is output y, and the so-called "person-" in FIG. A transfer function of the closed-loop model of “automobile” is obtained to calculate (identify) the driving operation characteristics (steering characteristics) of the driver.

  In FIG. 2, G is a transfer function of the driving state of the driver, C is a transfer function of the own vehicle 1, r1 is a target value, r2 is noise such as soot applied to an output signal of steering operation of the steering angle y, and v is Disturbance, H, is its transfer function.

  In the identification for a closed loop system such as the closed loop model of FIG. 2, since the external noise and the control input have a correlation, the identification method of the open loop system cannot generally be applied to the closed loop system as it is.

  Accordingly, the characteristic calculation means identifies the transfer function G by adopting either the SSARX method which is one of the closed-loop subspace identification methods or the prediction error method (PEM method) which is effective for the identification of the closed-loop system. .

  Then, when r1 is 0 and r2 is white noise and the Bode diagram is obtained by identifying the transfer function G by simulation of the SSARX method and the prediction error method, the frequency response characteristic of FIG. 3 is obtained.

  In FIG. 3, the solid line is the characteristic of the SSARX method, and the broken line is the characteristic of the prediction error method. 3 that the transfer function G, that is, the driving state of the driver has a resonance frequency.

  Furthermore, in order to verify the change of the transfer function G depending on the driving state of the driver, r1 is set to 0, r2 is set to white noise, and the driver operates the steering wheel so that the vehicle 1 travels straight along the traveling path. When the driver is concentrating on driving (equivalent to normal driving conditions) or when driving while solving a given calculation problem (equivalent to mild abnormal driving conditions such as mild sideways driving or dozing) On the other hand, when the Bode diagram was obtained by identifying the transfer function G by simulation of the SSARX method and the prediction error method, the frequency response characteristics of FIG. 4 were obtained. 4 are normal and abnormal characteristics identified by the SSARX method, and β1 and β2 in FIG. 4 are normal and abnormal characteristics identified by the prediction error method.

  As is apparent from FIG. 4, in the case of the mild abnormal operating state, since the correction operation is performed in a hurry, the resonance frequency shifts to the higher side than in the normal operating state at normal time. There was found. In addition, when it becomes a severe abnormal driving state such as complete doze driving, the correction operation becomes slow and delayed, and the resonance frequency shifts to the lower side.

  Therefore, the determining means determines the driving state of the driver from the difference between the normal driving operation characteristic and the latest driving operation characteristic based on the characteristic of the transfer function G calculated by the characteristic calculating means.

Incidentally, the driving operation characteristics of the normal, the first is the initial setting characteristics or previous characteristics held in the data storage unit 8, according to traveling, based on the characteristics of the transfer function G characteristic calculation means for calculating weighting Is corrected to a characteristic having a high appearance frequency by the correction process or the majority process, and is stored in the data storage unit 8 in a rewritable manner.

  Then, the determination means determines the characteristic of the latest transfer function G calculated by the characteristic calculation means for the resonance frequency of the normal driving operation characteristic of the data storage unit 8 as the difference between the normal driving operation characteristic and the latest driving operation characteristic. The resonance frequency of (the latest driving operation characteristics) is detected, and the driving state of the driver is determined from the result. In other words, if the resonance frequency of the latest driving operation characteristic is higher than the resonance frequency of the normal driving operation characteristic, it is determined that the driving operation characteristic is normal and the resonance frequency of the latest driving operation characteristic is normal driving operation. When the resonance frequency is lower than the characteristic resonance frequency, it is determined that the operation state is severely abnormal.

  At this time, the normal driving operation characteristic and the latest driving operation characteristic are both the characteristics of the transfer function G of the closed loop model including the driving environment such as the condition of the driving path and the driver's habit. Therefore, the determination means can accurately determine the driving state of the driver without being affected by the driving environment, the driver's habit, and the like.

  And the determination output of the determination processing unit 7 according to the determination result of the determination means is supplied to the notification processing unit 9 and the driving regulation unit 10 at the subsequent stage of the driving recognition state determination device 2, and if it is a mild abnormal driving state, The notification processor 9 alerts the driver by sound, vibration, or the like. If the driving state is severe and abnormal, the driving restricting unit 10 further reduces the throttle valve opening relative to the depression of the accelerator pedal or applies a brake.

  Therefore, in the case of the present embodiment, it is possible to perform a highly accurate determination of the driving state in consideration of the driving environment and the driver's daily habit while the host vehicle 1 is traveling, and based on this determination, the driving state of the driver is determined. It is possible to dramatically improve safety by performing appropriate notification and driving regulation.

  The present invention is not limited to the above-described embodiment, and various modifications other than those described above can be made without departing from the spirit of the present invention. When equipped, the lane recognition system of the photographing means 3 and the lane recognition processing unit 4 is omitted, the vehicle position detection means is formed by the GPS of the car navigation system, and from the map information of the car navigation system and the GPS position information, You may make it recognize the position of the own vehicle 1 with respect to the travel lane of the vehicle 1, and a travel lane.

  Further, the driving function characteristic of the driver may be calculated by identifying the transfer function G by a method other than the SSARX method and the prediction error method by the characteristic calculating means.

  Further, any alerting notification method or driving regulation method based on the determination result may be used, and the determination result may be used for other control of the vehicle 1 or the like.

  Next, the driving state of the driver may be a brake pedal operation state or an accelerator pedal operation state. In this case, the driving operation detection means detects the operation amount of the brake pedal or the accelerator pedal, and the vehicle position detection means What is necessary is just to detect the position of the traveling direction (front-back direction) of the own vehicle 1 with respect to the traveling path.

  And this invention is applicable to determination of the driving | running state of the driver of various vehicles.

It is a block diagram of one embodiment of the present invention. It is a schematic diagram of the closed loop model of the determination processing unit of FIG. FIG. 3 is an explanatory diagram of an example of frequency response characteristics of the driver model of FIG. 2. It is explanatory drawing of the change of the frequency response characteristic of the driver model of FIG. 2 according to the driving | running state of a driver.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Own vehicle 2 Driving | running state determination apparatus 7 Determination processing part

Claims (2)

A driving state determination device that determines a driving state of a driver of a traveling vehicle,
Own vehicle position detecting means for detecting the position of the own vehicle;
Driving operation detection means for detecting the driving operation of the driver;
Based on the detection position of the own vehicle position detection means and the detection operation of the driving operation detection means, the driving operation characteristics of the driver from the closed loop model control characteristics of the driving operation of the driver that corrects the position of the own vehicle with respect to the travel path. Characteristic calculation means for repeatedly calculating,
A data accumulating unit that holds a high-frequency appearance characteristic of the driving operation characteristic repeatedly calculated by the characteristic calculating unit as a normal driving operation characteristic,
Operation, characterized in that a determination means for determining the difference between the driver's driving state of the driving operation characteristics calculated by the OPERATION Operation characteristics when held normally the characteristic calculating means by said data storage unit State determination device. In the driving | running state determination apparatus of Claim 1,
The characteristic calculation means repeatedly calculates a driving operation characteristic of a driver by identifying a transfer function by an identification method such as a SSARX method which is one of closed-loop subspace identification methods and a prediction error method effective for identification of a closed-loop system. ,
The determination means determines an operation state of a driver from a resonance frequency of the latest driving operation characteristic calculated by the characteristic calculation means with respect to the normal driving operation characteristic.

Images