Vehicle dynamics control and vehicle dynamics control system for vehicles

Abstract

The present invention relates to a vehicle dynamics control for a vehicle (102), wherein an open-loop control algorithm (108) of the vehicle dynamics control is parameterized with a model behavior (110) of a virtual surrogate model (112) of the vehicle (102), the model behavior (110) of the surrogate model (112) is compared with a real behavior (116) of the vehicle (102), and the surrogate model (112) is adapted when the model behavior (110) deviates from the real behavior (116).

Description

[Technical Field]

[0001] The present invention relates to a vehicle dynamics control for a vehicle, a corresponding vehicle dynamics control system and a corresponding computer program product. [Background technology]

[0002] A vehicle may have a vehicle dynamics control system that may control actuators of the vehicle to affect the driving dynamics of the vehicle. To this end, the vehicle dynamics control system may, for example, read sensor signals from inertial sensors of the vehicle and, in response to vehicle vibrations mapped in the sensor signals, control the actuators to act to counteract these vibrations.

[0003] The inertia of the vehicle then introduces a certain time lag into the events that occur, since the vibrations can only be mapped into the sensor signal once a corresponding movement of the vehicle has actually occurred.

[0004] To counter this, vehicle dynamics control systems may be equipped with model-based open-loop controllers. In this case, the vehicle driver's inputs are processed by a virtual surrogate model of the vehicle, and the actuators are controlled based on the calculated response of the model. In this case, the model may predict that the vehicle is expected to reach a critical region at a predetermined input. The vehicle dynamics control system may therefore open-loop control the actuators even before any vibrations are visible in the sensor signals.

[0005] US Pat. No. 5,629,999 describes a method and device for operating a vehicle. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] European Patent Application Publication No. 2832599 Summary of the Invention

[0007] Against this background, the approach presented here provides a vehicle dynamics control for a vehicle according to the independent claims, a corresponding vehicle dynamics control system and a corresponding computer program product. Advantageous developments and improvements of the approach presented here can be seen from the description and are set forth in the dependent claims. [Effects of the Invention]

[0008] Conventional virtual surrogate models cannot completely map the behavior of real vehicles because they are preconfigured for all vehicles of a certain vehicle type or vehicle configuration, and individual differences between vehicles that are the same per se, such as different tires and / or different loads, cannot therefore be taken into account.

[0009] In the approach presented here, the actual behavior of an individual vehicle is evaluated during driving and compared to the model behavior of a surrogate model. The vehicle's individual idiosyncrasies are mapped into the individual's actual behavior. If there are deviations between the actual behavior and the model behavior, the surrogate model is modified to minimize these deviations.

[0010] The approach presented here allows a surrogate model to map the individual idiosyncrasies of the vehicle and thus to accurately predict the vehicle's behavior. This predicted behavior will map very well to the actual behavior, even in the limiting range of driving dynamics. This allows the predicted behavior to be advantageously used to parameterize the open-loop control algorithms of the vehicle dynamics control.

[0011] A vehicle dynamics control for a vehicle is proposed, in which an open-loop control algorithm of the vehicle dynamics control is parameterized using a model behavior of a virtual surrogate model of the vehicle, the model behavior of the surrogate model is compared with the real behavior of the vehicle, and when the model behavior deviates from the real behavior, the surrogate model is adapted.

[0012] The ideas that led to the embodiments of the present invention may be considered to be based, inter alia, on the thoughts and realizations described below.

[0013] Vehicle dynamics control for a vehicle can be performed by a vehicle dynamics control system of the vehicle. Vehicle dynamics control can affect, among other things, the lateral dynamics of the vehicle. The lateral dynamics determine, for example, the vehicle's turning or yawing behavior. The vehicle dynamics control system can comprise an open-loop control algorithm and a closed-loop controller. The vehicle dynamics control system can act on at least one actuator that affects the vehicle's lateral dynamics. The actuator can be, for example, the vehicle's braking system. The actuator can also be rear-axle steering. The actuator can also be a controllable differential. These actuators can affect the vehicle's yaw rate.

[0014] The open-loop control algorithm and the closed-loop controller can each send control commands to at least one actuator. The open-loop control algorithm then acts proactively, while the closed-loop controller acts reactively. The open-loop control algorithm intervenes before something happens, so that it doesn't happen. The closed-loop controller intervenes when the thing does happen.

[0015] For proactive intervention, the open-loop control algorithm requires prior knowledge of the vehicle's predictable driving behavior when subjected to inputs, such as steering inputs, acceleration inputs, and / or deceleration inputs. This prior knowledge is mapped into a virtual surrogate model of the vehicle. The surrogate model processes the same inputs as the vehicle and maps the expected driving behavior based on these inputs into the model behavior as an output. The parameters of the model behavior can be input quantities for the open-loop control algorithm.

[0016] Here, the model behavior is compared to the actual behavior of the vehicle in response to these inputs. The actual behavior is then referred to as the real behavior. When the real behavior deviates from the model behavior, the surrogate model is modified until the real behavior and the model behavior substantially match.

[0017] The real behavior can be observed using at least one current measured quantity detected in the vehicle. The real behavior can be observed by an observer of a vehicle dynamics control system. The measured quantity can be, for example, mapped into a sensor signal of a sensor of the vehicle. The measured quantity can be, for example, a yaw rate, a steering angle, or a lateral acceleration. The real behavior can be recognized, in particular, using a plurality of measured quantities. At least one measured quantity can be evaluated by the observer. The observer can recognize the real behavior based on the at least one measured quantity. The real behavior is influenced by an actual vehicle state. The observer can observe at least one measured quantity and an input to recognize the actual vehicle state. The vehicle state can be, for example, the position of the center of gravity of the vehicle, additional cargo on the vehicle, the state of the vehicle's tires, or the state of the road surface under the vehicle. The actual vehicle state can be compared with a model state of a surrogate model, and a deviation between the vehicle state and the model state can be output as a state correction.

[0018] The vehicle state can be recognized by comparing at least one measured quantity with an ideal value that maps to an ideal behavior, in which the ideal vehicle responds to inputs and drives as if it were on rails, which is impossible to achieve in reality.

[0019] The adaptation of at least one model state of the surrogate model may be performed using a state correction for the model state, where the adaptable model state may be predefined. The adaptable model state may be a choice of model state of the surrogate model. The direction of change of the state correction may be predetermined based on observed deviations of the real behavior from the model behavior.

[0020] Adaptation of up to two model states can occur simultaneously. By limiting simultaneous changes to two, a unique relationship between cause and effect can be observed. If more than two model states change, it is no longer possible to uniquely identify which change led to which effect. If there are more than two adaptable model states, the model state to be adapted for each can be iteratively switched.

[0021] The state correction may be limited to a predefined value range centered on the current model state. The state correction may be limited to the limited state correction. The limiting may be implemented within a limiter in the vehicle dynamics control system. The limiting may prevent excessive changes in the model state. Relatively large changes may be implemented incrementally. After each incremental change in the model state, the new model behavior may be compared to the real behavior to check the success of the change.

[0022] The value range can be set depending on the vehicle's driving situation. In this case, the value range can be limited, especially when the vehicle approaches a limit range. When the vehicle dynamics control system is performing stabilization intervention, no state correction can be performed. As a result, relatively large state corrections can only be performed when the vehicle is far from the limit range. In this way, it can be ensured that the state correction does not change the model behavior into a critical or unstable range.

[0023] State corrections can be limited to physically meaningful values. The meaningful values ​​can be based on prior knowledge. The meaningful values ​​can be limited by meaningful limits. The limits can ensure that model behavior stays outside critical or unstable regions.

[0024] The method is preferably computer-implemented and may be implemented, for example, in software or hardware or in a mixed form consisting of software and hardware, for example in a driver assistance system.

[0025] The approach presented herein further provides a vehicle dynamics control system for a vehicle, the vehicle dynamics control system configured to perform, control, or translate steps of a variant of the method presented herein within a corresponding mechanism.

[0026] The vehicle dynamics control system may be an electrical device having at least one computing unit for processing signals or data, at least one storage unit for storing signals or data, and at least one interface and / or communication interface for reading or outputting data embedded in a communication protocol. The computing unit may be, for example, a signal processor, a so-called system ASIC, or a microcontroller that processes sensor signals and outputs data signals in response to the sensor signals. The storage unit may be, for example, a flash memory, an EPROM, or a magnetic storage unit. The interface may be configured as a sensor interface for reading sensor signals from sensors and / or as an actuator interface for outputting data and / or control signals to actuators. The communication interface may be configured for reading or outputting data wirelessly and / or via a wired connection. The interface may be, for example, a software module residing alongside other software modules on a microcontroller.

[0027] Advantageously, there is also provided a computer program product or a computer program comprising a program code, which may be stored on a machine-readable carrier or storage medium, such as a semiconductor memory, a hard disk storage device or an optical memory, and which is used for executing, transforming and / or controlling the operation of the steps of the method according to one of the aforementioned embodiments, in particular when said program product or program is run on a computer or any device.

[0028] Additionally, some of the possible features and advantages of the present invention will now be described with reference to different embodiments, and it will be obvious to those skilled in the art that the features of the control device and method may be suitably combined, adapted or interchanged to arrive at further embodiments of the present invention.

[0029] DETAILED DESCRIPTION OF THE INVENTION

[0013] The present invention will now be described in detail with reference to the accompanying drawings, in which:

[0014] Neither the drawings nor the description should be construed as limiting the present invention. [Brief explanation of the drawings]

[0030] [Figure 1] FIG. 1 is a block diagram of a vehicle dynamics control system according to an embodiment. [Figure 2] FIG. 2 is a detailed diagram of a vehicle dynamics control system according to one embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0031] The drawings are only schematic and are not to scale. Like numbers refer to features of the same or identical function.

[0032] 1 shows a block diagram of a vehicle dynamics control system 100 of a vehicle 102. The vehicle dynamics control system 100 acts on at least one actuator of the vehicle 102 to affect the yaw rate of the vehicle 102. Examples of actuators may be the braking system of the vehicle 102, a steerable rear axle of the vehicle, or a motion-controllable torque-vectoring axle of the vehicle.

[0033] The vehicle dynamics control system 100 includes an open-loop control unit 104 and a closed-loop controller 106. Both the open-loop control unit 104 and the closed-loop controller 106 generate control commands for actuators to affect the yaw rate. An open-loop control algorithm 108 of the open-loop control unit 104 proactively outputs control commands to prevent undesirable driving behavior of the vehicle 102. If the vehicle still exhibits undesirable driving behavior, the closed-loop controller 106 outputs control commands to counteract the undesirable driving behavior. To adjust the control commands as closely as possible to the vehicle 102, the open-loop control algorithm 108 and the closed-loop controller 106 are parameterized by an online adaptive algorithm 114 using a model behavior 110 of a virtual surrogate model 112 of the vehicle 102.

[0034] The approach presented here compares the modeled behavior 110 of the surrogate model 112 with the real behavior 116 of the vehicle 102. Based on the results of the comparison, the surrogate model 112 is adapted to bring the modeled behavior 110 as close as possible to the real behavior 116.

[0035] In one embodiment, the online adaptation algorithm 114 reads at least one model state 118 of the surrogate model 112 and, based on the results of the comparison, outputs a state correction 120 for the model state 118 to adapt the surrogate model 112.

[0036] In one embodiment, at least one measured quantity 124 detected by a measurement device 122 on the vehicle 102 is evaluated by an online adaptive algorithm 114 to recognize the real behavior 116. The model behavior 110 can be derived from the output quantities and / or model states 118 of the surrogate model 112.

[0037] Figure 2 shows a detailed view of the vehicle dynamics control system 100 according to one embodiment. This detailed view shows the online adaptive algorithm 114 from Figure 1. Here, the online adaptive algorithm 114 includes a model-based observer 200 that observes measurements 124 and model states 118. The observer 200 determines state corrections 120 from the observed measurements 124 and the observed model states 118.

[0038] In one embodiment, the online adaptive algorithm 114 includes a limiting unit 202 that reads the unconstrained state corrections 120 and limits the state corrections 120 to prevent overly strong changes in the surrogate model.

[0039] In one embodiment, the unconstrained state correction 120 is constrained to a realistic value to obtain the constrained state correction 120 .

[0040] In one embodiment, the limiting unit 202 limits the state correction 120 depending on the current stability state of the vehicle. Thus, the state correction 120 is limited at least when the vehicle is in an unstable driving state. That is, the limiting unit 202 limits the state correction 120 when the vehicle dynamics control system 100 is operating.

[0041] In the following, possible configurations of the invention are summarized once again, or expressed in slightly different wording.

[0042] This paper presents an online adaptation of model-based vehicle dynamics control to improve performance and robustness.

[0043] Vehicle dynamics control systems that ensure vehicle stability today typically have a model-based approach with fixed parameters that control the lateral dynamics. Typically, the parameters are determined once and then stored in a fixed manner in the controller code. This means that the model behavior is fixedly prescribed or applied. Typical vehicle dynamics models, in this case, are linear and nonlinear single-track models that include multiple states of yaw rate and body side-slip angle.

[0044] The vehicle dynamics model has a two-degree-of-freedom structure with an open-loop control (feedforward, FF) and a closed-loop controller (feedback, FB), as described in, for example, the above-mentioned Patent Document 1. This control circuit, in this case, is typically closed via the yaw rate. Deviations always occur between reality and the model. Typical reasons for this are, for example, tire variations (wear, model, summer tires / winter tires, ...), load variations (low load, high load, high center of gravity, ...), road obstacles, model inaccuracies (unaccounted dynamics, unknown friction values), and lateral road gradients. These uncertainties lead to deviations between the target yaw rate and the actual yaw rate.

[0045] This can lead to incorrect control initiation via the closed-loop controller path. This is compensated for by increasing the control initiation threshold. This may result in delayed control initiation in the application, thus degrading performance. Furthermore, the open-loop control path can no longer fully guarantee the guidance behavior, since the model, and therefore the states, no longer match reality.

[0046] The approach presented here extends the vehicle dynamics control described in the above-mentioned patent application WO 2007 / 024999 by the possibility of real-time adaptation to the current situation, taking into account the inaccuracies of the system to be controlled (e.g. tires, aging, loading). This results in an improved model validity compared to the above-mentioned patent application WO 2007 / 024999 and therefore an improved guidance behavior via the open-loop control path.

[0047] The approach presented here allows for precise and early intervention of the open-loop control to achieve the desired driving behavior. The effectiveness of the open-loop control is thereby significantly increased. Hard and uncomfortable closed-loop controller interventions are prevented. Furthermore, robustness against false control initiation during normal driving is improved.

[0048] Due to the intervention of the open-loop control, feedback intervention is not necessary because the actual yaw rate and the target yaw rate match, which reduces the hardware load. Furthermore, the control initiation threshold can be reduced, which leads to higher performance in certain applications.

[0049] In the approach presented here, a vehicle model is adapted to reality in real time via a model-based observer approach. The observer, in this case, permanently compares the measurable states of the real vehicle with the states of the vehicle model, and based on this, and on the physical model, estimates the disturbance levels to the states (beta, dPsi) and corrects them to better match the true vehicle behavior. This allows the model behavior to be adapted to the actual behavior in real time. The effects of the sum of all uncertainties in the model are corrected.

[0050] The observer must then operate only within a predefined uncertainty range. The uncertainty is based on physical assumptions about the influence of various disturbances on the dynamics, such as tire variations. These assumptions apply in the linear range; they are automatically very small at large tire sideslip angles. This allows each project to individually determine which disturbances occur and at what magnitude. During normal driving, the observer thus always remains within the defined uncertainty band and can fully correct the state. In this way, closed-loop control deviations cannot occur, false control starts are avoided, and the modeled state remains valid for a longer period, meaning the open-loop control can achieve the desired guidance behavior for a longer period.

[0051] In application cases, disturbances may have a magnitude significantly larger than the uncertainty band due to dynamic effects. In this case, the disturbances can be limited. This limit may reach a level where disturbances are no longer tolerated in the application case. In this way, the adaptation of the target behavior (target yaw rate) remains possible, since it is not influenced by the observer approach. Adaptation via feedback remains possible in this case.

[0052] In the approach presented here, the adaptation is performed significantly faster compared to known parameter adaptations (e.g., vch). The effect of road imperfections, such as potholes, bumps, and friction value changes, on the vehicle can thus be directly recognized and modeled.

[0053] The approach presented here describes an online adaptive algorithm that adapts a vehicle model to a real vehicle in real time via a model-based observer approach. The algorithm uses measured quantities, such as yaw rate, steering angle, and lateral acceleration, on the one hand, and the state of an open-loop control vehicle model on the other. Based on this information, the algorithm calculates appropriate state corrections for the vehicle model, so that the vehicle model better adapts to the current environment.

[0054] The algorithm consists of two blocks: the first block is a model-based observer, and the second block is a constraint calculator.

[0055] Within the model-based observer, calculation rules are derived based on the vehicle's underlying physical equations, which match the model state with the measurements and, from the match, calculate the necessary corrections to adapt the model state to the actual state.

[0056] In the limit calculation unit, the maximum range in which the disturbance variables, and therefore the state corrections, can move is calculated in real time using predetermined uncertainties and physical assumptions. This range also depends on the extent to which the vehicle is in the application case. This ensures that the algorithm only corrects in the non-application case and ensures robustness, while still allowing the target behavior and feedback to be applied in the application case.

[0057] Blocks 104 to 106 are standard control circuit structures as described in the aforementioned patent application. Block 114 relates to the extension described herein for online adaptation of the vehicle model. In the detailed diagram, block 114 comprises a model-based observer 200 and a constraint calculator 202.

[0058] With the approach presented here, the target yaw rate is updated directly (immediately after uncertainties appear) when physical vehicle parameters change.

[0059] Finally, it should be noted that the word "comprises" does not exclude other elements or steps, and the word "indefinite article" does not exclude a plurality. Reference signs in the claims are not to be regarded as limiting. [Explanation of symbols]

[0060] 100 Vehicle Dynamics Control System 102 vehicles 104 Open Loop Control Unit 106 Closed Loop Controller 108 Open-Loop Control Algorithm 110 Model Behavior 112 Virtual Surrogate Models 114 Online Adaptive Algorithms 116 Realistic Behavior 118 Model State 120 Status Correction 122 Measuring Equipment 124 Measured Quantity 200 Observer 202 Restricted Section

Claims

1. 1. A vehicle dynamics control for a vehicle (102), comprising: parameterizing an open-loop control algorithm (108) of the vehicle dynamics control with a model behavior (110) of a virtual surrogate model (112) of the vehicle (102); comparing the model behavior (110) of the surrogate model (112) with a real behavior (116) of the vehicle (102); and adapting the surrogate model (112) when the model behavior (110) deviates from the real behavior (116).

2. 2. The vehicle dynamics control of claim 1, wherein the real behavior (116) is observed using at least one current measured quantity (124) sensed on the vehicle (102).

3. 3. The vehicle dynamics control of claim 1, wherein the adaptation of at least one model state of the surrogate model is performed using a state correction for the model state.

4. 4. The vehicle dynamics control of claim 3, wherein the vehicle dynamics control adapts up to two model states (118) simultaneously.

5. 5. A vehicle dynamics control according to claim 3 or 4, wherein the state corrections (120) are limited to a predefined range of values ​​centred on the model state (118).

6. 6. A vehicle dynamics control according to claim 3, wherein the value range is set depending on a driving situation of the vehicle (102).

7. 7. A vehicle dynamics control according to any one of claims 3 to 6, wherein the state corrections (120) are limited to physically meaningful values.

8. A vehicle dynamics control system (100), configured to implement, translate and / or control the operation of a vehicle dynamics control according to any one of claims 1 to 7 in a corresponding mechanism.

9. 8. A computer program product configured to provide a processor with instructions to implement, transform and / or control the operation of a vehicle dynamics control according to any one of claims 1 to 7 when the computer program product is executed.

10. 10. A machine-readable storage medium having stored thereon the computer program product of claim 9.

Images