METHOD AND SYSTEM FOR CONTROLLING A VEHICLE

Description

[0001] The invention relates to a method and a system for controlling a vehicle. In particular, a method for controlling the lateral guidance of a vehicle is disclosed which avoids the overloading of integrating components of a control device by limiting or saturating vehicle dynamic parameters.

[0002] Regulations governing the lateral control of a vehicle during at least partially autonomous driving maneuvers are generally well-established. Due to legal requirements or for reasons of comfort or safety, limitations or saturation of vehicle dynamic parameters may occur. For example, gradient limitations may be imposed for steering torque, steering angle, vehicle curvature, lateral acceleration, or tire lateral force. Limitations can also arise due to physical constraints, such as the steering angle limit during parking maneuvers or the maximum tire lateral force in highly dynamic scenarios. It is known that such limitations or saturation of vehicle dynamic parameters negatively impact lateral control behavior, as the unusable components of the vehicle dynamic parameter lead to a overload of the integrating components of the control system.

[0003] To avoid these negative effects, a so-called back-calculation anti-windup method is used in the prior art. This method determines the difference between the limited and unlimited manipulated variable. This difference is then fed back to the integral component of the controller with an applied time constant, thus "discharging" it.

[0004] A disadvantage of this method is that it can only be applied to simple PID controllers or controllers with clearly separated integral (I) components. State-based or model-based controllers cannot readily utilize this method. Another problem is the time constant for the feedback of the difference between the limited and unlimited manipulated variable. This must be determined experimentally or through simulation, which is time-consuming and often yields insufficient results. Finally, the backcalculation anti-windup method often fails to adequately compensate for the negative effects resulting from the limitation, as cascaded controllers frequently have multiple integral components, and charging can also occur due to implicit integrators in the vehicle trajectory kinematics, which the backcalculation anti-windup method cannot prevent.

[0005] Furthermore, CN 1 13 126 623 A discloses a path tracking method for an autonomous vehicle for tracking vehicle paths taking into account a problem with the saturation of the control input.

[0006] German patent DE 10 2019 114 585 A1 describes a method for lane keeping or steering intervention in a vehicle. Based on a kinematic model that considers vehicle yaw rate, lateral acceleration, speed, lane marking position, lane marking path, and lane curvature, the vehicle's movement relative to the lane is predicted. A trajectory approximation algorithm used in this process includes a vehicle model and defines vehicle movement restrictions or physical constraints. From this prediction, a lateral deviation between the vehicle position and the lane marking position is determined. Based on this determined lateral deviation, a time until the vehicle leaves the lane (so-called lane departure time) is calculated.The time-to-lane crossing (TTLC) is determined, and this TTLC is then projected over several time steps using a predictive model and assigned a confidence value that describes the predictive accuracy. Based on the projected TTLC and the assigned confidence value, a steering intervention is then activated.

[0007] German patent DE 41 24 654 A1 describes a method for the continuous and automatic guidance of a vehicle based on an estimated relative position of the vehicle to the road surface. Image data from a camera system of the vehicle is processed, with the evaluation being model-based using a recursive estimation method that incorporates various sub-models, such as a vehicle lateral dynamics model and a geometric road surface model. Parameters of the road surface, in particular horizontal and vertical curvature, are determined and used for state estimation to ascertain the vehicle's position relative to the road surface. The determined state variables, such as lateral drift, yaw angle, and road surface curvature, serve as the basis for deriving control data for vehicle guidance. The relative vehicle position to the road surface is used to determine a future trajectory along the modeled road surface.The difference between the desired lane position and the current vehicle position is interpreted as a control deviation and processed by a control system that calculates the necessary steering angle commands based on a vehicle model. This control deviation can manifest as a lateral offset or a heading error and is converted into a steering input by suitable control algorithms, such as state controllers or model-based controllers. Vehicle dynamics and road geometry are taken into account to enable stable and predictive guidance along the lane. Control is achieved through continuous adjustment of the steering angle commands, with the control system reacting to the estimated state variables and updating them cyclically.

[0008] German patent DE 10 2009 050 777 A1 describes a method for determining a vehicle-related state variable based on a control variable influencing the direction of travel of the motor vehicle. A lateral force at the vehicle's wheels is calculated from the steering torque and the steering transmission characteristics. The calculated lateral force is fed into a vehicle dynamics model based on a single-track model, which serves to determine vehicle dynamics state variables such as yaw rate and lateral velocity. The lateral force is also used, in addition to other sensor signals, as a measured variable in an observer model to estimate quantities or states such as vehicle mass and inertia. To increase accuracy, alternative measurement equations are used in the observer model that do not require differentiation and instead use directly measurable quantities such as lateral force and steering angle.This should reduce the dependence on vehicle parameters that are difficult to determine and improve the robustness of the model.

[0009] The object of the invention is to provide a method for controlling a vehicle that enables a technically simple and effective suppression of charging effects, even with state controllers or model-based controllers.

[0010] The problem is solved by a method having the features of independent claim 1. Preferred embodiments are the subject of the dependent claims. A system for controlling a vehicle is the subject of dependent claim 11.

[0011] According to a first aspect, a method for controlling a vehicle is disclosed. The vehicle includes a driver assistance system by means of which at least partially autonomous driving functions are performed. A control loop is provided for regulating the vehicle's driving functions, which receives information as a setpoint, for example, from a trajectory planner. The control loop has at least one controller that limits at least one vehicle dynamics parameter by means of a limiter or by saturation. Using a vehicle dynamics model, the change in the vehicle dynamics parameter resulting from the limitation is converted into a residual curvature and / or yaw rate parameter of the vehicle. In other words, the difference between the limited and unlimited vehicle dynamics parameter, which corresponds to the unconverted vehicle dynamics parameter, is determined. This difference is converted into a residual curvature and / or yaw rate parameter, which indicates the residual vehicle curvature or yaw rate parameter.Yaw rate could not be implemented due to limitations or saturation.

[0012] Subsequently, the residual curvature and / or yaw rate of the vehicle is converted into a lateral and / or heading angle deviation resulting from the limitation using a kinematic model of the vehicle. In other words, this determines the lateral trajectory deviation or heading angle deviation caused by the limitation or saturation of at least one vehicle dynamic parameter.

[0013] Subsequently, a measured lateral and / or heading angle deviation of the vehicle is modified based on the lateral and / or heading angle deviation caused by the limitation, thereby generating a modified lateral and / or heading angle deviation. Specifically, the lateral and / or heading angle deviation caused by the limitation is subtracted from the measured lateral and / or heading angle deviation, thus compensating for the deviation resulting from the limitation or saturation of at least one vehicle dynamic parameter.

[0014] Finally, a control error is calculated based on the setpoint and the modified lateral and / or heading angle deviation. In other words, the measured lateral and / or heading angle deviation is not fed back as the controlled variable and used to determine the control error (difference between setpoint and controlled variable), but rather the modified lateral and / or heading angle deviation.

[0015] The method according to the invention has the technical advantage of being independent of the controller structure used. For example, it can be used with both a simple PID controller and a state-based controller. Furthermore, no application parameters need to be determined in the method according to the invention, since a model for the vehicle-road kinematics is used as the feedback transfer function. Finally, it is possible to effectively prevent the controller from becoming overloaded due to the limitation or saturation of one or more different vehicle dynamic variables.

[0016] According to one embodiment, the vehicle dynamics parameter is the steering angle, the steering torque, the vehicle curvature, the vehicle's lateral acceleration, and / or the vehicle's tire lateral force. The disclosed method can compensate for the limitation or saturation of one or more of these vehicle dynamics parameters.

[0017] According to one embodiment, the at least one controller comprises a control cascade with a first and a second control unit. The first control unit is, for example, an outer, superimposed control unit, and the second control unit is an inner, subordinate control unit. In particular, the feedback in the control loop is modified such that a limitation or saturation in the area of ​​the second control unit does not cause charging effects in the area of ​​the first control unit.

[0018] According to one embodiment, the first control device comprises a kinematic control system and the second control device comprises a vehicle dynamics control system.

[0019] According to one embodiment, the first control device provides a target curvature value for the vehicle. Thus, the first control device converts a lateral or heading angle deviation into a target curvature value for the vehicle.

[0020] According to one embodiment, the second control unit converts a target curvature value of the vehicle into a target steering angle. Thus, through the interaction of the first and second control units, a lateral or heading angle deviation can be converted into steering angle information. The second control unit can also include steering angle control.

[0021] According to one embodiment, the limitation of the vehicle dynamic parameter takes place in the second control unit. Alternatively, the limitation of the vehicle dynamic parameter can also take place between the second control unit and the controlled system, where the controlled system includes, for example, the vehicle steering.

[0022] According to one embodiment, the vehicle dynamics model is configured to determine a residual curvature and / or yaw rate from a difference value of the vehicle dynamics parameter that arises due to a limitation. This difference value is the difference between the unlimited and limited values ​​of the vehicle dynamics parameter and thus indicates the proportion of the vehicle dynamics parameter that cannot be realized due to the limitation or saturation. This unrealized proportion is then converted into the residual curvature and / or yaw rate parameter by the vehicle dynamics model.

[0023] According to one embodiment, a single-track model, in particular a linear single-track model of the vehicle, is implemented in the vehicle dynamics model. The vehicle dynamics model allows the difference value of the vehicle dynamics parameter arising from the constraints to be converted into a vehicle-specific residual curvature and / or yaw rate parameter.

[0024] According to one embodiment, the change in the vehicle dynamics parameter resulting from the limitation is calculated by differentiating between the unlimited and limited vehicle dynamics parameters. This difference indicates the proportion of the vehicle dynamics parameter that cannot be implemented and is thus a measure of the control error that causes the integrating components to increase.

[0025] According to one embodiment, the measured lateral and / or heading angle deviation is provided by the vehicle's odometry system. If the vehicle dynamics parameter is limited, the measured lateral and / or heading angle deviation deviates from the predefined lateral and / or heading angle deviation. The influence of the limitation is compensated for by subtracting the lateral and / or heading angle deviation resulting from the limitation from the measured lateral and / or heading angle deviation.

[0026] According to a further aspect, the invention relates to a system for controlling a vehicle, comprising a driver assistance system by means of which at least partially autonomous driving functions can be performed. A control loop is provided for regulating the vehicle's driving functions, which receives setpoint information from a trajectory tracking controller. At least one controller is provided that limits at least one vehicle dynamics parameter by means of a limiter or by saturation. The control loop is configured as follows: to convert the change in the vehicle dynamics parameter resulting from the limitation into a residual curvature and / or yaw rate parameter of the vehicle using a vehicle dynamics model; to convert the residual curvature and / or yaw rate parameter of the vehicle into a lateral and / or heading angle deviation resulting from the limitation using a kinematic model of the vehicle; to modify a measured lateral and / or heading angle deviation of the vehicle based on the lateral and / or heading angle deviation resulting from the limitation; to generate a modified lateral and / or heading angle deviation; and to calculate a control error based on the target parameter and based on the modified lateral and / or heading angle deviation.

[0027] The terms "approximately", "essentially" or "about" mean, within the meaning of the invention, deviations from the respective exact value by + / - 10%, preferably by + / - 5% and / or deviations in the form of changes that are insignificant for the function.

[0028] Further developments, advantages and application possibilities of the invention also result from the following description of exemplary embodiments and from the figures.

[0029] The invention will be explained in more detail below with reference to exemplary embodiments shown in the figures. The figures show: Fig. 1 shows an exemplary schematic representation of a control loop for vehicle lateral guidance; Fig. 2 shows an exemplary schematic representation of the control loop according to Fig. 1 , which has a further feedback path to compensate for limiting effects, by means of which the controlled variable is modified; Fig. 3 shows an exemplary schematic representation of the control loop similar to the Fig. 2, where the limitation of the steering angle and the calculation of the change in the vehicle dynamics parameter resulting from the limitation are indicated; and Fig. 4 shows an exemplary schematic representation of an exemplary embodiment of the vehicle dynamics model of the Figures 2 and 3 .

[0030] Figure 1 Figure 1 shows an exemplary and schematic block diagram of a control loop 1, which can be used to control at least partially autonomous driving functions of a vehicle's driver assistance system. Control loop 1 is designed for lateral vehicle guidance.

[0031] The control loop comprises at least one controller, in the illustrated embodiment a first and a second control unit 2, 3. The first control unit 2 is, for example, a kinematic motion controller (KMC), also known as a trajectory tracking controller. This first control unit 2 provides, for example, a target curvature K as its output. This output is provided to the second control unit 3 as its input. The second control unit 3 is, for example, a dynamic motion controller (DMC) that provides a target steering angle δv as its output. The second control unit 3 can, in particular, also include a steering angle controller as a subordinate control loop, which provides a target torque for the steering actuator as its output. The second control unit 3 can, for example, be a model-based dynamic motion controller.

[0032] The target steering angle δv is then transferred to the system to be controlled, i.e., the controlled system. The controlled system includes, in particular, the vehicle's steering system. The controlled system can be divided, for example, into two transfer functions: the first transfer function GFzg and the second transfer function Gtraj. GFzg describes the transfer behavior from the target steering angle δv (output of the DMC) to the resulting vehicle curvature, while Gtraj describes the dynamic structure of the heading angle Δθ and the lateral deviation Δy (lateral direction) of the point to be guided along the trajectory in a locally fixed coordinate system.

[0033] For feedback, a measured lateral and / or heading angle deviation Δy is used. , Δθ ,The data provided, for example, by the vehicle's odometry system is fed back to the input of control loop 1, so that the setpoints Δy target , Δθ target (for example, provided by a trajectory planning unit) are determined using the measured lateral and / or heading angle deviation Δy , Δθ is converted into a control error Δy Err , Δθ Err, which is provided to the first control device 2 as input information.

[0034] Due to legal requirements, but also for other reasons, the second control unit 3 may limit vehicle dynamic parameters. For example, the steering angle, steering torque, vehicle curvature, lateral acceleration, or tire lateral force may be limited. These limitations can lead to a build-up of integrating components (so-called windup) in the first control unit 2, which can result in overcompensation and vibrations of the vehicle, potentially leading to instability.

[0035] Fig. 2 Figure 1 shows a modified control loop 1, by means of which the disturbing effects resulting from the limitation can be advantageously prevented or mitigated.

[0036] First, the change in the vehicle dynamics parameter δRest resulting from the limitation, hereinafter also referred to as residual parameter δRest, is determined. This residual parameter, in this case the steering angle, is limited. This residual parameter δRest is then transferred to a vehicle dynamics model 4. Additionally, in the case of a model-based vehicle dynamics control system, it is possible, as a second control system 3, to determine a residual curvature from the remaining value of the vehicle dynamics parameter that cannot be implemented due to the limitation.

[0037] The vehicle dynamics model 4 can be an integral part of the second control unit 3. Alternatively, the vehicle dynamics model 4 can be implemented as a separate control module.

[0038] The vehicle dynamics model 4 is designed to map the residual quantity δresidual and, if applicable, the resulting residual curvature into a resulting residual curvature and / or yaw rate quantity Kvehicle,residual of the vehicle. The vehicle dynamics model 4 can, for example, be formed by a linear single-track model.

[0039] The vehicle dynamics model 4 converts the change in the vehicle dynamics quantity δrest resulting from the limitation into a vehicle dynamics residual quantity, for example, the curvature and / or yaw rate residual quantity KFzg,Rest. This is then fed back to a vehicle trajectory model 5. The vehicle trajectory model 5 describes, for example, the kinematic behavior of a point mass on a path. This vehicle trajectory model 5 can be an integral part of the first control unit 2. Alternatively, the vehicle trajectory model 5 can be implemented as a separate control module.

[0040] The vehicle trajectory model 5 is designed to determine a lateral and / or heading angle deviation Δy Rest , Δθ Rest resulting from the curvature and / or yaw rate residual quantity K Fzg,Rest.

[0041] The transfer function of the vehicle trajectory model 5 can be represented as follows: G ΔΘ , κ ^ = ΔΘ κ Fzg = v s ; G Δy , κ ^ = Δy κ Fzg = v 2 s 2 ; where v is the longitudinal speed and Kfzg is the curvature traveled.

[0042] The lateral and / or heading angle deviation Δy Rest , Δθ Rest caused by the limitations is then subtracted from the measured lateral and / or heading angle deviation Δy ,Δθ is subtracted at subtraction point 7, resulting in a modified lateral and / or heading angle deviation Δy', Δθ'. This means that the manipulated variable of the first control unit 2 is reduced by precisely the amount that exceeds the respective saturation limit of the vehicle dynamics variable and cannot be implemented anyway. The first control unit 2 therefore always operates at the limit as soon as a limitation or saturation of a vehicle dynamics variable is reached.

[0043] Fig. 3Figure 1 shows a specific embodiment of the invention with a limitation of the target steering angle δv by means of a limiter 6, a method frequently encountered in practice. The limiter 6 is configured, for example, to limit the target steering angle δv vertically to a steering angle value δv,lim. ​​Limiting the target steering angle δv is done, for example, for reasons of comfort and safety. It is often limited to a predetermined rate of change.

[0044] By calculating the difference between the target steering angle δv and the limited steering angle δv,lim, the residual quantity δRest, by which the steering angle was limited, can be determined. Due to this limitation, the control system cannot implement this residual quantity δRest. The vehicle dynamics model 4, which models the vehicle (here, the transfer function from target steering angle to vehicle curvature), determines the residual curvature and / or yaw rate KFzg,Rest, which arises due to the limitation and also cannot be implemented. This is passed on to the vehicle trajectory model 5. As described previously, the vehicle trajectory model 5 converts the residual curvature and / or yaw rate KFzg,Rest into a lateral and / or heading angle deviation ΔyRest, ΔθRest resulting from the limitation.

[0045] This lateral and / or heading angle deviation Δy Rest , Δθ Rest is derived from a measured lateral and / or heading angle deviation Δy , Δθ and the resulting modified lateral and / or heading angle deviation Δy', Δθ' is used to modify the setpoint at the input of control loop 1.

[0046] Fig. 4 shows an example schematic diagram illustrating the vehicle dynamics model 4 for the conversion of the steering angle δ v into the vehicle curvature K Fzg.

[0047] The input variable for vehicle dynamics model 4 is the steering angle at the front axle δv. This steering angle δv is used for the transfer functions. G Ψ̇, δ and G β,δ The input variable is provided, where β denotes the swim angle and Ψ the yaw rate. These transfer functions are taken, for example, from the linear single-track model.

[0048] About the connection κ Fzg = ψ ˙ + β ˙ v The resulting vehicle curvature K Fzg is then obtained, which serves as a feedback to the vehicle trajectory model 5.

[0049] As previously explained, δv denotes the steering angle at the front axle of the vehicle. It follows that the steering wheel angle can also be limited in the same way. In this case, the steering ratio would have to be taken into account. Additionally, a model for torsional stiffness and steering play could optionally be included.

[0050] The preceding section described the compensation of effects caused by limiting or saturating a single parameter, such as the steering angle. It is understood that the proposed method also allows for the compensation of the negative effects of limiting or saturating multiple parameters. In each case, the difference between the limited and unlimited parameter is calculated. The remaining parameter not realized due to the limitation is then converted into a residual curvature and / or yaw rate parameter. This residual curvature and / or yaw rate parameter is then converted into a lateral and / or heading angle deviation resulting from the limitation using a kinematic model of the vehicle. Finally, the lateral and / or heading angle deviations determined for all limited parameters are subtracted from the measured lateral and / or heading angle deviation. This compensates for the negative effects caused by the limitation or saturation of multiple parameters.Saturation of multiple sizes is possible.

[0051] The invention has been described above using exemplary embodiments. It is understood that numerous modifications and adaptations are possible without thereby departing from the scope of protection defined by the patent claims.

Claims

1. A method for controlling a vehicle comprising a driver assistance system, by means of which at least semiautonomous driving functions are carried out, wherein a control loop (1) is provided to regulate the driving functions of the vehicle, which loop receives information (Δysoll, Δθsoll) as a setpoint variable, that at least one controller is provided which limits at least one driving dynamics variable by means of a limiter (6) or by saturation, that by means of a vehicle dynamics model (4), the change in the driving dynamics variable (δRest) that arises as a result of the limitation is converted into a curvature and / or yaw rate residual variable (KFzg,Rest) of the vehicle, that the curvature and / or yaw rate residual variable (K,Fzg,Rest) of the vehicle is converted, by means of a kinematic model (5) of the vehicle, into a lateral and / or course angle deviation (ΔyRest, ΔθRest) that arises due to the limiting, and that a measured lateral and / or course angle deviation (Δy, Δθ) of the vehicle is modified based on the lateral and / or course angle deviation (Δysoll, Δθsoll) that arises due to the limiting, by which a modified lateral and / or course angle deviation (Δy', Δθ') is formed, and wherein a control error (ΔyErr, ΔθErr) is calculated based on the setpoint variable (Δysoll, Δθsoll) and based on the modified lateral and / or course angle deviation (Δy', Δθ').

2. The method as claimed in claim 1, characterized in that the driving dynamics variable is the steering angle, the steering torque, the vehicle curvature, the lateral acceleration of the vehicle, and / or the tire lateral force of the vehicle.

3. The method as claimed in claim 1 or 2, characterized in that the at least one regulator comprises a regulating cascade having a first and a second regulating device (2, 3).

4. The method as claimed in claim 3, characterized in that the first regulating device (2) comprises a kinematics controller and the second regulating device (3) comprises a driving dynamics controller.

5. The method as claimed in claim 3 or 4, characterized in that a setpoint curvature value (Ksoll) is provided by the first regulating device (2).

6. The method as claimed in any one of claims 3 to 5, characterized in that a setpoint curvature value (Ksoll) is converted into a setpoint steering angle (δv) by the second regulating device (3).

7. The method as claimed in any one of claims 3 to 6, characterized in that the limitation of the driving dynamics variable takes place in the second regulating device (3).

8. The method as claimed in any one of the preceding claims, characterized in that the vehicle dynamics model (4) is designed to determine a curvature and / or yaw rate residual variable (KFzg,Rest) from a differential value of the driving dynamics variable (δRest) that arises due to limiting.

9. The method as claimed in claim 8, characterized in that a single-track model of the vehicle is implemented in the vehicle dynamics model (4).

10. The method as claimed in any one of the preceding claims, characterized in that the change in the driving dynamics variable (δRest) that arises due to the limiting is formed by calculating the difference between the unlimited driving dynamics variable (δv) and the limited driving dynamics variable (δv, lim).

11. The method as claimed in any one of the preceding claims, characterized in that the measured lateral and / or course angle deviation (Δy, Δθ) is provided by an odometry system of the vehicle.

12. A system for controlling a vehicle comprising a driver assistance system, by means of which at least partially autonomous driving functions can be carried out, wherein, to regulate the driving functions of the vehicle, a control loop (1) is provided, which receives information (Δysoll, Δθsoll) as a setpoint variable, that at least one controller is provided which limits at least one driving dynamics variable by means of a limiter (6) or by saturation, that the control loop is designed: - to convert, by means of a vehicle dynamics model (4), the change in the driving dynamics variable (δRest) arising due to the limiting into a curvature and / or yaw rate residual variable (KFzg,Rest) of the vehicle; - to convert the remaining curvature and / or yaw rate variable (KFzg,Rest) of the vehicle into a lateral and / or course angle deviation (ΔyRest, ΔθRest)that arises due to the limiting by means of a kinematic model (5) of the vehicle; - to modify a measured lateral and / or course angle deviation (Δy, Δθ) of the vehicle based on the lateral and / or course angle deviation (Δyrest, Δθrest) that arises due to the limiting; - to form a modified lateral and / or course angle deviation (Δy', Δθ'); and - to calculate a control error (ΔyErr, ΔθErr) based on the setpoint variable (Δysoll, Δθsoll) and based on the modified lateral and / or course angle deviation (Δy', ΔΘ').