JP7695845B2 - Vehicle control device and vehicle control method - Google Patents

Description

The present invention relates to the configuration of a vehicle control device and its control, and in particular to technology that is effective in improving ride comfort.

Traditionally, the focus of autonomous driving technology development has been on developing essential technologies for autonomous driving, such as route following. However, as autonomous driving technology advances, the value people place on moving spaces has become more diverse, and there is an increasing demand for improved ride comfort.

As a result, in recent years, there has been progress in the development of vehicle control technologies that improve the comfort of drivers and passengers by optimally integrating and controlling multiple, heterogeneous actuators.

As background technology in this technical field, for example, there is technology such as that in Patent Document 1. Patent Document 1 discloses "a roll vibration damping control device for a vehicle that has a roll angular acceleration detection device that detects the roll angular acceleration of the vehicle body, an actuator that generates a roll moment to be applied to the vehicle body, and a control unit that controls the actuator, and the control unit stores the roll moment of inertia, roll damping coefficient, and equivalent roll stiffness of the vehicle."

Patent Document 2 also discloses a driving support method that "predicts the behavior of a vehicle when driving according to a driving plan of the vehicle, calculates an evaluation value for the predicted vehicle behavior based on the driving plan, and calculates the control amount of an actuator that controls the vehicle based on the evaluation value."

JP 2020-59477 A JP 2020-26189 A

According to the above-mentioned patent document 1, it is possible to effectively suppress the roll vibration of the vehicle body without causing any change in the dynamic characteristics of the roll motion of the vehicle.

However, in Patent Document 1, as explained in each embodiment, the roll vibration is suppressed by controlling the same type of actuator, and there is no specific explanation of how the various actuators are controlled to suppress roll vibration when different types of actuators are used in combination.

In addition, according to the above-mentioned Patent Document 2, it is possible to reduce delays in vehicle control even when the vehicle's behavior changes dynamically.

However, although Patent Document 2 mentions a control technology that improves path tracking by optimally controlling multiple actuators by predicting vehicle motion using a state predictor, it does not explain any specific control techniques related to acceleration or jerk, which is a change in acceleration that is known to affect ride comfort.

The object of the present invention is to provide a high-performance vehicle control device and vehicle control method that can control the acceleration and jerk that affect ride comfort to the desired movement while improving path tracking in a vehicle control device that controls different types of actuators in an integrated manner.

In order to achieve the above object, the present invention provides a vehicle control device that controls vehicle motion by operating a plurality of actuators, the vehicle control device comprising: a predictor that predicts future vehicle motion based on a vehicle motion prediction model and an actuator command predicted value; and an optimizer that calculates, by iterative calculation, the actuator command predicted value that minimizes a difference between an arbitrary motion target value and the future vehicle motion that is a predicted value by the predictor, wherein the predictor has a first motion predictor that calculates a first motion predicted state obtained by differentiating vehicle dynamics one or more times, and the optimizer calculates, based on the first motion predicted state calculated by the first motion predictor, the actuator command predicted value that minimizes the difference between the motion target value and the future vehicle motion .

Moreover, the present invention is a vehicle control method for controlling vehicle motion by operating a plurality of actuators, comprising: (a) a step of predicting future vehicle motion based on a vehicle motion prediction model and an actuator command prediction value; and (b) a step of calculating, by iterative calculation, an actuator command prediction value that minimizes a difference between an arbitrary motion target value and the future vehicle motion that is the prediction value predicted in the step (a), wherein in the step (a), a first motion prediction state obtained by differentiating vehicle dynamics one or more times is calculated, and in the step (b), an actuator command prediction value that minimizes the difference between the motion target value and the future vehicle motion is calculated based on the first motion prediction state calculated in the step (a ).

The present invention provides a vehicle control device that controls different types of actuators in an integrated manner, and realizes a high-performance vehicle control device and vehicle control method that can improve path tracking while controlling the acceleration and jerk that affect ride comfort to the desired movement.

This will ensure the route tracking capabilities necessary for autonomous driving while improving the ride comfort for drivers and passengers.

Problems, configurations, and advantages other than those described above will become clear from the description of the embodiments below.

1 is a diagram showing a schematic configuration of a vehicle according to a first embodiment of the present invention; 2 is a block diagram showing a schematic configuration of a control system for controlling the vehicle shown in FIG. 1 . FIG. 1 is a block diagram showing a general model predictive control. 1 is a block diagram showing a control configuration of a vehicle control device according to a first embodiment of the present invention; 5 is a waveform diagram showing the effect of the vehicle control device of FIG. 4. FIG. 11 is a block diagram showing a control configuration of a vehicle control device according to a second embodiment of the present invention. FIG. 11 is a block diagram showing a control configuration of a vehicle control device according to a third embodiment of the present invention.

The following describes an embodiment of the present invention with reference to the accompanying drawings. Similar components are given similar reference numerals and similar descriptions will not be repeated.

A vehicle control device and a vehicle control method according to a first embodiment of the present invention will be described with reference to Figures 1 to 5. Figure 1 is a diagram showing a schematic configuration of a vehicle according to this embodiment.

As shown in FIG. 1, the vehicle 1 of this embodiment is equipped with wheels 11, a motor 12, a suspension 13, a steering wheel 14, a brake 15, and a stabilizer 16 on a vehicle body 10.

In the following, the fore-and-aft direction of vehicle 1 is the x-axis (forward direction is positive), the left-right direction is the y-axis (left direction is positive), and the up-down direction is the z-axis (upward direction is positive).

The wheels 11 support the vehicle body 10 and exert grip on the road surface. In this embodiment, there are four wheels: a left front wheel 11FL, a right front wheel 11FR, a left rear wheel 11RL, and a right rear wheel 11RR.

In the following, the reference numerals corresponding to the left front wheel 11FL will be marked with FL, the reference numerals corresponding to the right front wheel 11FR will be marked with FR, the reference numerals corresponding to the left rear wheel 11RL will be marked with RL, and the reference numerals corresponding to the right rear wheel 11RR will be marked with RR. Furthermore, the reference numerals corresponding to both the left front wheel 11FL and the right front wheel 11FR will be marked with F, and the reference numerals corresponding to both the left rear wheel 11RL and the right rear wheel 11RR will be marked with R.

Each of the wheels 11 is fitted with an in-wheel motor 12 (12FL, 12FR, 12RL, 12RR), and these motors 12 allow each of the wheels 11 to rotate (forward and reverse) independently.

Suspensions 13 (13FL, 13FR, 13RL, 13RR) are provided between each of the motors 12 and the vehicle body 10. These suspensions 13 absorb vibrations and shocks generated in each of the wheels 11, improving the stability and ride comfort of the vehicle body 10.

These suspensions 13 include, for example, semi-active suspensions that combine a damper with adjustable viscosity and a coil spring, fully active suspensions that combine an actuator with adjustable stroke, a damper and a coil spring, and electric suspensions that use a combination of a linear motor or a rotary motor and a rotary-linear mechanism.

The steering wheel 14 is a device for steering the wheels 11 and determining the direction of travel of the vehicle 1. In this embodiment, the steering wheel 14 is provided with three wheels: a steering wheel 14FL for steering the left front wheel 11FL, a steering wheel 14FR for steering the right front wheel 11FR, and a steering wheel 14R for steering the left rear wheel 11RL and the right rear wheel 11RR.

The brakes 15 are devices for braking the rotation of the wheels 11, and in this embodiment, there are four brakes: a brake 15FL for the left front wheel 11FL, a brake 15FR for the right front wheel 11FR, a brake 15RL for the left rear wheel 11RL, and a brake 15RR for the right rear wheel 11RR.

The stabilizer 16 is a device that moves in conjunction with the up and down movement of the left and right wheels to suppress the amount of roll of the vehicle. In this embodiment, the stabilizer is a controlled stabilizer whose torsion angle can be adjusted electrically. It is a device that tilts the body 10 in the roll direction, which is the rotational movement around the x-axis, when the vehicle 1 turns, etc., to adjust the amount of roll. In this embodiment, there are two stabilizers: a front stabilizer 16F and a rear stabilizer 16R.

Figure 2 is a block diagram showing the schematic configuration of a control system that controls the vehicle 1 in Figure 1.

The vehicle control device 2 mounted on the vehicle 1 receives vehicle motion command values calculated by a higher-level computer based on map information, self-position information, and a target ride comfort index. The vehicle control device 2 also has a function of calculating multiple actuator commands to reduce the error between the received motion command values and current motion information obtained from an acceleration sensor or the like, and transmitting operation command values to each actuator.

Figure 2 shows an example in which forward/backward command values, left/right command values, up/down command values, roll command values, pitch command values, and yaw command values are input to the vehicle control device 2 as motion command values, and operation command values for the drive actuator, brake actuator, active suspension actuator, steering actuator, etc. are output.

There is a technology called model predictive control that can optimally control the error between the target motion and the current value shown in Figure 2.

Figure 3 shows a schematic diagram of a general model predictive control technology that is incorporated into the vehicle control device 2 and performs optimal actuator control based on vehicle motion prediction.

In model predictive control, the motion predictor 21 is equipped with a vehicle dynamics model to simulate vehicle motion, and predicts vehicle motion based on operation commands for multiple actuators. The optimizer 22 compares the motion command value received from the host computer 41 with the motion prediction value obtained from the motion predictor 21, and optimizes the operation command value for the actuator 42 by iterative calculation so that the error of the future motion prediction value for a certain interval is minimized relative to the motion command value.

At this time, the weight of the evaluation function used for optimization by the evaluation weight setting unit 23 can be adjusted so that any of a number of vehicle movements can be controlled. For example, it is possible to adjust the control so that suppression of changes in the attitude angle of the vehicle 1 is given priority over the path following error of the vehicle 1.

The constraint condition setting unit 24 also enables the optimization calculation by the optimizer 22 after constraining any motion state or actuator operation command so that it does not exceed a preset upper or lower limit. For example, it is possible to run the vehicle while constraining the path tracking error to always be 10 m or less.

In this way, typical model predictive control calculates operation command values that optimally operate the actuators based on predictions of vehicle dynamics, optimizing to minimize the error with respect to the target value and not exceeding various constraints, and transmits these values to the actuators 42.

Figure 4 shows a control block diagram of the vehicle control device 2 in this embodiment. As in Figure 3, this embodiment also includes an optimizer 22, an evaluation weight setting unit 23, and a constraint condition setting unit 24. The motion predictor 21 includes a first motion predictor 31 and a second motion predictor 32. The first motion predictor 31 includes a vehicle motion model that is a second-order derivative of a general vehicle motion model that calculates the acceleration change of the vehicle motion.

Here, the motion model of a general vehicle motion model that calculates the change in acceleration of vehicle motion, i.e., the forward/backward x, upward/downward z, and pitch θ, which is the rotational motion around the y-axis, is, for example, the motion model shown in equations (1) to (3).

Here, m is the mass of the entire vehicle, ms is the sprung mass, hp is the deviation between the pitch center and the vehicle's center of gravity, ρ is a coefficient expressing the influence of various motions, and f is the input to various actuators. In order to achieve optimal control, these equations of motion are followed and expressed as the linear state equation shown in equation (4).

where X and U are the state and input, respectively, and A and B are the coefficient matrix of the state vector and the coefficient matrix of the input vector, respectively. In equation (4), for example, the state includes the position and velocity of the vehicle or the angle and angular velocity of the vehicle attitude.

Here, the first motion predictor 31 in this embodiment has equation (5), which is the second derivative of equation (4), as the vehicle motion prediction model.

At this time, because equation (5) is the second derivative of the equations of motion shown in equations (1) to (3), it includes the acceleration and jerk, as well as the angular acceleration and angular jerk of the attitude.

In addition, by using second-order differentiation, the input handled by the state equation is given as the rate of change of the actuator command change (the second-order differential value of U).

Then, we obtain a second predicted motion state by integrating the state predicted by equation (5) twice. The state obtained here is the same as state X in equation (4).

Furthermore, the predicted state combiner 33 converts the first and second motion prediction models, and the actuator input U, into a form that can be simultaneously optimized. Specifically, an extended system of equation (6) is constructed.

However, in reality, U included in the input vector is uniquely determined by the integral of the input of the second derivative of U in equation (5), so it is not appropriate to treat it as an independent control input. Therefore, we define equation (7), which includes U in the state vector, as a new extended system.

Since the equation (8) obtained in this way simultaneously includes the state of position, velocity, acceleration, jerk, angle, angular velocity, angular acceleration, angular jerk, various actuator commands, etc., it is possible to predict these changes and simultaneously optimally control them using the optimizer 22.

At this time, the value calculated as the result of optimization is the second-order differential value of the actuator command, so the signal actually sent to the actuator 42 is the second-order integrated value (actuator command value).

In the optimizer 22, the weight setting value Q by the evaluation weight setting unit 23 and the equation (10) which is an optimization function based on the error E between the target motion and the motion prediction value shown in equation (9) are defined, and a constrained optimization calculation is performed based on equations (11) and (12) which formulate the constraint conditions on the state and input of the augmented system represented in equation (8) as inequality constraint conditions.

Here, Hp is the prediction interval, and Gx and Gu are the coefficient matrices for the state and input constraints, respectively.

With the above configuration, in this embodiment, it is possible to simultaneously control the path following control of the vehicle 1 and the acceleration and jerk motion states that affect the ride comfort of the driver and passengers so as not to exceed the actuator operating limits.

The effect of this embodiment will be explained using Figure 5. Figure 5 shows, from the top, the waveforms of vehicle speed, in-wheel motor drive torque, longitudinal jerk, pitch angular acceleration, and pitch angular velocity. In the vehicle speed waveform, the dashed line shows the speed command value, the dotted line shows the control waveform of the conventional technology, and the solid line shows the control waveform of the present invention.

Conventional technology cannot simultaneously perform speed tracking control and suppress longitudinal jerk and pitch angular acceleration, resulting in the longitudinal jerk and pitch angular acceleration exceeding the allowable values (dashed lines) set from the perspective of ride comfort. On the other hand, the present invention can achieve driving that simultaneously suppresses longitudinal jerk and pitch angular acceleration while performing speed tracking control and taking into account the limits of the drive torque, thereby enabling vehicle driving control with improved ride comfort.

As described above, the vehicle control device 2 of this embodiment includes a motion predictor 21 that predicts future vehicle motion based on a motion prediction model of the vehicle 1 and actuator command values, and an optimizer 22 that calculates actuator command values that minimize the difference between an arbitrary motion target value and a value predicted by the motion predictor 21 through iterative calculations. The motion predictor 21 has a first motion predictor 31 that calculates a first motion prediction state obtained by differentiating the dynamics of the vehicle 1 by one or more orders of magnitude.

The motion predictor 21 also has a second motion predictor 32 that calculates a second motion predicted state by integrating the first motion predicted state to one or more orders.

The vehicle control device 2 also includes a predicted state combiner 33 that combines the first motion prediction state and the second motion prediction state.

The predicted state combiner 33 further combines the actuator command predicted values, which are predicted based on the integral of the change in the actuator command values.

The optimizer 22 calculates the optimization result based on the amount of change in the actuator command value obtained by differentiating the dynamics of the vehicle 1 by one or more orders of magnitude.

The first motion predictor 31 and the second motion predictor 32 predict one or more of the six degrees of freedom of motion of the vehicle 1, including the translational motion in the forward/backward, up/down, and left/right directions, and the rotational motions around each axis, namely roll, pitch, and yaw.

The first motion predictor 31 also predicts one or more of the acceleration or jerk of the translational motion of the vehicle 1, and the angular acceleration or angular jerk of the rotational motion of the vehicle 1.

The second motion predictor 32 also predicts one or more of the position or velocity of the translational motion of the vehicle 1, and the angle or angular velocity of the rotational motion.

The optimizer 22 also limits one or more predicted values consisting of the first and second motion predicted states and the actuator command predicted value predicted based on the integral of the change in the actuator command value so that they do not exceed an arbitrary limit value.

This allows for optimal control of the path following control required for autonomous driving of automobiles, while at the same time limiting the acceleration and jerk motion states that affect the ride comfort of the driver and passengers to within allowable values.

A vehicle control device and a vehicle control method according to a second embodiment of the present invention will be described with reference to FIG. 6. FIG. 6 is a block diagram showing the control configuration of the vehicle control device 2 of this embodiment.

As shown in FIG. 6, the vehicle control device 2 of this embodiment differs from the first embodiment (FIG. 4) in that the predicted state combiner 33 does not use actuator command values when converting the first and second motion prediction models into an optimizable format. The other configurations are the same as those of the first embodiment (FIG. 4).

In model predictive control, computational cost is often an issue. It can be reduced by reducing the dimensionality of the state vector.

Therefore, for the control purpose of improving ride comfort, it is assumed that the actuator performance limit will not be used, and a configuration is considered in which the actuator command value is not included in the predicted state combiner 33.

By doing this, the number of dimensions required for the optimization calculation can be reduced by the number of actuators being handled, which is expected to reduce calculation costs and the cost of computers to be installed in vehicles.

A vehicle control device and a vehicle control method according to a third embodiment of the present invention will be described with reference to FIG. 7. FIG. 7 is a block diagram showing the control configuration of the vehicle control device 2 of this embodiment.

As shown in FIG. 7, the vehicle control device 2 of this embodiment differs from the first embodiment (FIG. 4) in that when the motion predictor 21 predicts the vehicle motion, the second motion predictor 32 is not used, and further, the predicted state combiner 33 does not convert the format of the first motion prediction model. The other configurations are the same as those of the first embodiment (FIG. 4).

This embodiment is a control configuration that focuses only on controlling the ride comfort. In this embodiment, only the acceleration, jerk, and angular acceleration and angular jerk of the attitude can be controlled as states.

By doing this, it is possible to further reduce computational costs.

The present invention is not limited to the above-described embodiments, but includes various modified examples. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and are not necessarily limited to those having all of the configurations described. It is also possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. It is also possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

1...vehicle, 2...vehicle control device, 10...vehicle body, 11...wheels, 12...motor, 13...suspension, 14...steering, 15...brakes, 16...stabilizer, 21...motion predictor, 22...optimizer, 23...evaluation weight setting unit, 24...constraint condition setting unit, 31...first motion predictor, 32...second motion predictor, 33...predicted state combiner, 41...host computer, 42...actuator.

Claims (15)

A vehicle control device that controls a vehicle's motion by operating a plurality of actuators,
a predictor for predicting future vehicle motion based on a vehicle motion prediction model and an actuator command prediction value;
an optimizer that calculates, by iterative calculation, the actuator command predicted value that minimizes a difference between an arbitrary motion target value and the future vehicle motion that is a predicted value by the predictor;
The predictor includes a first motion predictor that calculates a first motion predicted state obtained by differentiating the dynamics of the vehicle by one or more orders of magnitude;
The optimizer calculates an actuator command predicted value that minimizes a difference between the motion target value and the future vehicle motion based on the first motion predicted state calculated by the first motion predictor . The vehicle control device according to claim 1,
The predictor has a second motion predictor that calculates a second motion predicted state obtained by integrating the first motion predicted state by one or more orders,
The optimizer calculates an actuator command predicted value that minimizes a difference between the motion target value and the future vehicle motion based on the first motion predicted state and the second motion predicted state calculated by the second motion predictor . The vehicle control device according to claim 2,
a prediction state combiner for combining the first motion prediction state and the second motion prediction state ;
The optimizer calculates an actuator command predicted value that minimizes a difference between the motion target value and the future vehicle motion based on a motion predicted value obtained by combining the first motion predicted state and the second motion predicted state by the predicted state combiner . The vehicle control device according to claim 3,
The predicted state combiner further combines an integral value of the change amount of the actuator command predicted value ,
the optimizer calculates an actuator command predicted value that minimizes a difference between the motion target value and the future vehicle motion, based on a motion predicted value obtained by combining the first motion predicted state, the second motion predicted state, and an integral value of an amount of change in the actuator command predicted value by the predicted state combiner. The vehicle control device according to claim 3,
the optimizer calculates an optimization result based on a change amount of an actuator command prediction value obtained by differentiating the vehicle dynamics by one or more orders of magnitude;
A vehicle control device that generates and outputs actuator command values based on the calculated optimization results . The vehicle control device according to claim 3,
A vehicle control device in which the first motion predictor and the second motion predictor predict one or more of the vehicle's forward/backward, up/down, and left/right translational motions, and roll, pitch, and yaw, which are rotational motions around each axis, for a total of six degrees of freedom. The vehicle control device according to claim 3,
A vehicle control device wherein the first motion predictor predicts one or more of an acceleration or jerk of a translational motion of the vehicle, and an angular acceleration or angular jerk of a rotational motion of the vehicle. The vehicle control device according to claim 3,
A vehicle control device, wherein the second motion predictor predicts one or more of a position or a velocity of a translational motion of a vehicle, and an angle or an angular velocity of a rotational motion. The vehicle control device according to claim 8,
The optimizer limits one or more predicted values, consisting of the first motion predicted state, the second motion predicted state, and an integral value of the amount of change in the actuator command predicted value, so as not to exceed an arbitrary limit value. A vehicle control method for controlling a motion of a vehicle by operating a plurality of actuators, comprising:
(a) predicting future vehicle motion based on a vehicle motion prediction model and an actuator command prediction value;
(b) calculating, by iterative calculation, an actuator command prediction value that minimizes a difference between an arbitrary motion target value and the future vehicle motion that is the prediction value predicted in the (a) step;
having
In the step (a), a first motion prediction state is calculated by differentiating the dynamics of the vehicle by one or more orders of magnitude ;
In the step (b), an actuator command predicted value that minimizes a difference between the motion target value and the future vehicle motion is calculated based on the first motion predicted state calculated in the step (a) . The vehicle control method according to claim 10,
In the step (a), a second motion predicted state is calculated by integrating the first motion predicted state by one or more orders;
a step (b) of calculating an actuator command predicted value that minimizes a difference between the motion target value and the future vehicle motion based on the first motion predicted state and the second motion predicted state ; The vehicle control method according to claim 11,
In the step (a) , the first motion prediction state and the second motion prediction state are combined;
a step (b) of calculating an actuator command predicted value that minimizes a difference between the motion target value and the future vehicle motion based on a motion predicted value that combines the first motion predicted state and the second motion predicted state . The vehicle control method according to claim 12,
In the step (a) , an integral value of the change amount of the actuator command prediction value is further combined;
the step (b) calculating an actuator command predicted value that minimizes a difference between the motion target value and the future vehicle motion, based on a motion predicted value that combines the first motion predicted state, the second motion predicted state, and an integral value of an amount of change in the actuator command predicted value . The vehicle control method according to claim 12,
In the step (b), an optimization result is calculated based on a change amount of an actuator command prediction value obtained by differentiating the vehicle dynamics by one or more orders of magnitude ;
A vehicle control method , comprising generating and outputting actuator command values based on the calculated optimization results . The vehicle control method according to claim 12,
A vehicle control method in which the first motion prediction state and the second motion prediction state predict one or more of the vehicle's forward/backward, up/down, and left/right translational motions, and roll, pitch, and yaw, which are rotational motions around each axis, for a total of six degrees of freedom.

Images