Method and control device for influencing the drive dynamics of a vehicle
Patent Information
- Authority / Receiving Office
- EP · EP
- Patent Type
- Applications
- Current Assignee / Owner
- ROBERT BOSCH GMBH
- Filing Date
- 2024-06-14
- Publication Date
- 2026-04-29
AI Technical Summary
Current vehicle driving dynamics controllers rely on time-consuming rule-based algorithms that require extensive parameter setting by application engineers, limiting adaptability and functional safety, while data-driven algorithms offer better interventions but lack direct security verification.
Implementing a dual-algorithm approach where a data-driven algorithm created through machine learning operates in parallel with a conventionally secure, rule-based algorithm, with the conventionally secure algorithm serving as a fallback to ensure functional safety by comparing and selecting interventions based on performance.
This hybrid approach enhances driving dynamics interventions by leveraging the adaptability of data-driven algorithms while ensuring functional safety through the oversight of a conventionally secure algorithm, reducing the risk of under- or over-braking and maintaining reliable vehicle control.
Smart Images
Figure EP2024066607_26122024_PF_FP_ABST
Abstract
Description
[0001] Description
[0002] title
[0003] Method and control unit for influencing the driving dynamics of a vehicle
[0004] Field of the invention
[0005] The invention relates to a method for influencing the driving dynamics of a vehicle, a corresponding control unit, and a corresponding computer program product.
[0006] State of the art
[0007] A vehicle's dynamics controller influences the vehicle's driving dynamics through interventions. The interventions are calculated using a rule-based algorithm and executed by the vehicle's actuators. For example, the interventions can be braking, steering, and / or drivetrain interventions. The rule-based algorithm is applied specifically for a model or model variant of the vehicle during test drives. The rule-based algorithm is comprehensible and can therefore be verified with regard to its functional safety.
[0008] The application is time-consuming because a large number of algorithm parameters must be adjusted by an application engineer. Only a small number of parameters can be changed at a time to understand the effects of the change.
[0009] Disclosure of the invention
[0010] Against this background, the approach presented here provides a method for influencing the driving dynamics of a vehicle, a corresponding control unit, and a corresponding computer program product in accordance with the
[0011] KU:GR presented in independent claims. Advantageous further developments and improvements of the approach presented here emerge from the description and are described in the dependent claims.
[0012] Advantages of the invention
[0013] A data-driven algorithm is not configured by a calibration engineer. The data-driven algorithm is created through machine learning or artificial intelligence using training data or training drives without human intervention. The data-driven algorithm may not be comprehensible and cannot be conventionally secured with regard to functional safety. However, during vehicle operation, the data-driven algorithm, just like the rule-based algorithm, issues interventions for the vehicle's actuators. The interventions of the data-driven algorithm can influence the vehicle's driving dynamics more effectively than those of the rule-based algorithm. However, a residual risk remains.
[0014] In the approach presented here, two algorithms are executed in parallel in a vehicle. One of the algorithms was created automatically for the vehicle using machine learning and is therefore not directly verifiable. This algorithm is referred to here as a data-driven algorithm. The other algorithm is a conventionally created algorithm and is therefore verifiable. This algorithm is referred to as a conventionally verifiable algorithm. Both algorithms are fed with input variables from the vehicle and calculate interventions for the vehicle's actuators in parallel. As long as the interventions of the data-driven algorithm influence the vehicle's driving dynamics more effectively than the interventions of the conventionally verifiable algorithm, the interventions of the data-driven algorithm are executed.If the interventions of the data-driven algorithm have a worse impact on the vehicle's driving dynamics than the interventions of the conventionally secure algorithm, the interventions of the conventionally secure algorithm are executed.
[0015] Using the approach presented here, a data-driven algorithm that is actually not securable can be indirectly secured to achieve functional safety by executing a conventionally securable algorithm in parallel and switching to the interventions of the conventionally securable algorithm if the interventions of the data-driven algorithm are worse than the interventions of the conventionally securable algorithm.
[0016] A method for influencing the driving dynamics of a vehicle is proposed, wherein data-driven interventions for actuators of the vehicle are determined using input variables of the vehicle and a data-driven algorithm, and in parallel, conventional interventions for the actuators are determined using the input variables and a conventionally verifiable algorithm, wherein a comparison is carried out between the data-driven interventions and the conventional interventions, and interventions to be carried out for the actuators are determined and sent to the actuators using a result of the comparison.
[0017] Ideas for embodiments of the present invention can be considered, among other things, to be based on the thoughts and findings described below.
[0018] A vehicle can be a two-track vehicle, such as a car or truck. The vehicle can, in particular, be a single-track vehicle, such as a motorcycle. Actuators of the vehicle can be installed in a braking system, a drive system, a steering system, and / or a chassis of the vehicle. In particular, the actuators can be the vehicle's brakes. In this case, a corresponding driving dynamics controller can be referred to as ABS / ESP.
[0019] Input variables can be measured values from the vehicle. The input variables can also be inputs from a driver. The inputs can also come from a vehicle assistance system.
[0020] A data-driven algorithm can be called an AI (artificial intelligence) algorithm. A conventionally securable algorithm can be called a deterministic algorithm. For simplicity, the conventionally securable algorithm can be called a conventional algorithm.
[0021] An intervention can be triggered by a control signal. The control signals are coded specifically for the respective controlled actuator. Both algorithms output identically coded control signals for the same actuator. For example, for a vehicle's braking system, both the data-driven algorithm and the conventionally verifiable algorithm for the braking system determine uniformly coded control signals. Comparable control signals can therefore be compared to compare the interventions. The selected control signals can be sent to the actuators to trigger the interventions.
[0022] The comparison can, for example, be part of the conventionally secured algorithm. The comparison can also be performed using a separate comparison algorithm.
[0023] As a rule, the data-driven interventions can be determined as the interventions to be carried out. The conventional interventions can be determined as the interventions to be carried out if the conventional interventions are better than the data-driven interventions. An intervention can be described as "better" if it ensures greater deceleration while simultaneously preventing locking. The intervention can also be described as "better" if it can be assessed as safer in terms of functional safety. The data-driven algorithm can determine better interventions in many areas than the conventional algorithm. However, the conventional algorithm can reliably determine the interventions up to the limits of driving dynamics. This can prevent the vehicle from being under-braked, i.e. braked too weakly, for example, by the data-driven interventions.This can also prevent the vehicle from being over-braked by the data-driven interventions, which would lead to the wheels locking.
[0024] Conventional interventions can be determined as the interventions to be executed if the conventional interventions deviate from the data-driven interventions by more than a predefined tolerance. Small deviations can be tolerated. If the deviations are greater than the tolerance, a switch to conventional interventions may be necessary. The tolerance can, for example, be 10 percent of the conventional interventions. A distinction can be made here as to whether this would lead to greater or lesser deceleration of the vehicle. Such a tolerance can, for example, be measured as braking torque / Newton meter [Nm], braking force / Newton [N], braking pressure [bar], or actuator signals, such as valve opening times. Conventional interventions can be determined as the interventions to be executed if the conventional interventions are better than the data-driven interventions for longer than a predefined period of time.Within the time period, the deviation can be greater than the tolerance. This temporal tolerance can, for example, buffer an overshoot from conventional interventions. The time period can range from 100 milliseconds to 500 milliseconds.
[0025] The data-driven algorithm can be trained specifically for a vehicle configuration. This allows the data-driven algorithm to be very well adapted to the vehicle.
[0026] The conventionally secure algorithm can be generic for different vehicles, eliminating the application effort required for a specific vehicle model. The conventionally secure algorithm can be essentially the same for different vehicles. This means that the conventionally secure algorithm may be less well-adapted to the vehicle, the vehicle model, or a vehicle configuration. In other words, the conventionally secure algorithm can be secure, but not specifically optimized for the vehicle.
[0027] At least one parameter of the conventionally verifiable, generic algorithm can be parameterized for the vehicle. For example, basic parameters such as the vehicle's curb weight, wheelbase, or center of gravity can be parameterized to at least roughly characterize the vehicle.
[0028] A deviation between the data-driven interventions and the conventional interventions can be integrated to form a deviation sum. The conventional interventions can be determined as the interventions to be carried out if the deviation sum is greater than a tolerance sum. By integrating the deviation, it can be detected if the conventional interventions deviate from the data-driven interventions in one direction within the tolerance over a longer period of time. If the deviation fluctuates between positive and negative, the integral does not rise above the tolerance sum. By integrating, a small systematic error in the data-driven algorithm can also be detected. Then, for safety reasons, the conventional interventions can be switched to. The tolerance sum can, for example, be 10 percent of the conventional interventions.A distinction can be made here as to whether this would lead to greater or lesser deceleration of the vehicle. This tolerance sum can, for example, be given a unit of [Nms] (Newton-meter-seconds) or [Ns] Newton-seconds.
[0029] The interventions to be performed can be a mix of data-driven interventions and conventional interventions. The ratio of the interventions can depend on the result of the comparison. For example, the ratio can depend on the integrated sum of the deviations. Then, as the deviation increases, the conventional interventions can be increasingly reflected in the interventions to be performed. However, the optimized component of the data-driven interventions cannot be lost.
[0030] If conventional interventions are determined to be performed, a message can be stored in the vehicle's diagnostic memory. A diagnostic memory or error memory can be read out during a workshop visit. Alternatively, the error message can also be transmitted directly via radio to the vehicle manufacturer or supplier. This allows, for example, the data-driven algorithm to be updated. Normally, data-driven interventions should be used because the data-driven algorithm is optimized for the vehicle. The use of conventional interventions should be the exception. The messages in the diagnostic memory can be used to verify the quality of the data-driven algorithm. If no messages are stored in the diagnostic memory for many vehicles over a longer period of time, the functional safety of the data-driven algorithm can also be demonstrated.
[0031] The method is preferably computer-implemented and can be implemented, for example, in software or hardware or in a mixed form of software and hardware, for example in a driver assistance system.
[0032] In particular, individual or preferably all of the process steps to be performed within the framework of the method can be automated, i.e., performed by one or more machines. For example, to automatically influence the driving dynamics, data-driven interventions and / or conventional interventions can be determined automatically. The data-driven interventions and the conventional interventions can also be automatically compared with each other, and the interventions to be performed can be automatically determined.
[0033] The approach presented here further creates a control unit, wherein the control unit is designed to carry out, control or implement the steps of a variant of the method presented here in corresponding devices.
[0034] The control unit can be an electrical device with at least one computing unit for processing signals or data, at least one memory unit for storing signals or data, and at least one interface and / or a communication interface for reading in or outputting data embedded in a communication protocol. The computing unit can be, for example, a signal processor, a so-called system ASIC, or a microcontroller for processing sensor signals and outputting data signals depending on the sensor signals. The memory unit can be, for example, a flash memory, an EPROM, or a magnetic storage unit. The interface can be designed as a sensor interface for reading in the sensor signals from a sensor and / or as an actuator interface for outputting the data signals and / or control signals to an actuator.The communication interface can be configured to read or output data wirelessly and / or via a wired connection. The interfaces can also be software modules, which are present, for example, on a microcontroller alongside other software modules.
[0035] Also advantageous is a computer program product or computer program with program code that can be stored on a machine-readable carrier or storage medium such as a semiconductor memory, a hard disk memory or an optical memory and is used to carry out, implement and / or control the steps of the method according to one of the embodiments described above, in particular when the program product or program is executed on a computer, in a control unit or a device. It should be noted that some of the possible features and advantages of the invention are described herein with reference to different embodiments. A person skilled in the art will recognize that the features of the control unit and the method can be combined, adapted or exchanged as appropriate to arrive at further embodiments of the invention.
[0036] Short description of the drawing
[0037] Embodiments of the invention are described below with reference to the accompanying drawings, wherein neither the drawings nor the description are to be interpreted as limiting the invention.
[0038] Fig. 1 shows a block diagram of a vehicle dynamics control system according to an embodiment.
[0039] The figure is merely schematic and not to scale. Like reference numerals denote like or equivalent features.
[0040] Embodiments of the invention
[0041] Fig. 1 shows a block diagram of a vehicle dynamics control system 100 of a vehicle, in particular a motorcycle. The vehicle dynamics control system 100 uses input variables 102 and, based on these, outputs interventions 106 to be performed on actuators 104 of the vehicle. The input variables 102 can be measured variables 108 of the vehicle and / or requests 110 for interventions via vehicle control elements and / or vehicle assistance functions.
[0042] For the vehicle dynamics control system 100 presented here, two algorithms 112, 114 are executed in parallel. The first algorithm 112 is a data-driven algorithm 112. The second algorithm 114 is a conventionally secure algorithm 114. The data-driven algorithm 112 determines data-driven interventions 116. The conventionally secure algorithm 114 determines conventional interventions 118. The conventionally secure algorithm 114 is also referred to as the conventional algorithm 114.
[0043] Both algorithms 112, 114 process the at least substantially identical input variables 102. The input variables 102 of the conventional algorithm 114 can be filtered in a filter 120 before processing to avoid unrealistic input variables 102. The algorithms 112, 114 can be executed on separate computing units of the vehicle or on a common computing unit of the vehicle.
[0044] The data-driven algorithm 112 is based on artificial intelligence specifically trained for the vehicle. Due to the training or machine learning, the data-driven interventions 116 often intervene more effectively in the driving dynamics than the conventional interventions 118. However, with the data-driven algorithm 112, a relationship between the input variables 102 and the data-driven interventions 116 is not directly traceable. Therefore, proving the functional safety of the data-driven algorithm 112 is difficult.
[0045] The conventionally secure algorithm 114 is rule-based and deterministic. The relationship between the input variables 102 and the conventional interventions 118 is directly traceable. Therefore, the functional safety of the conventionally secure algorithm 114 is demonstrated.
[0046] In the approach presented here, the data-driven interventions 116 and the conventional interventions 118 are compared with each other. Depending on the result of the comparison, the data-driven interventions 116 or the conventional interventions 118 are sent to the vehicle's actuators 104 as the interventions 106 to be executed and executed.
[0047] In one embodiment, the data-driven interventions 116 are typically used as the interventions 106 to be executed. Only if the conventional interventions 118 are better than the data-driven interventions 116 are the conventional interventions 118 used as the interventions 106 to be executed. Thus, the vehicle can typically benefit from the training of the data-driven algorithm 112. In critical situations, the conventional algorithm 114 serves as a fallback for the data-driven algorithm 112.
[0048] In one embodiment, the system only switches to the conventional interventions 118 when the conventional interventions 118 are better than the data-driven interventions 116 by one tolerance. Since both algorithms 112, 114 essentially process the same input variables 102 and have the same goal, both interventions 116, 118 can be very similar. A minor deviation can be tolerated.
[0049] In one embodiment, the system switches to conventional interventions 118 only after a predefined period of time. When the algorithms 112, 114 react to changes in the input variables 102, the changes in the interventions 116, 118 may differ. For example, the conventional interventions 118 may exhibit an overshoot if the control parameters of the conventional algorithm 114 are less well adapted to the vehicle than the data-driven algorithm 112 specifically trained for the vehicle.
[0050] In one embodiment, a deviation between the data-driven interventions 116 and the conventional interventions 118 is integrated. If the integral becomes greater than a deviation sum, the conventional interventions 118 are selected at least partially for the interventions 106 to be performed. This allows a minor but persistent deviation in one direction to be detected and corrected.
[0051] In one embodiment, the conventional algorithm 114 performs the comparison of the interventions 116, 118. The conventional algorithm 114 thus controls the output of the interventions 106 to be executed. For the comparison, the conventional algorithm 114 reads in the data-driven interventions 116 and compares them with the self-calculated conventional interventions 118.
[0052] In an alternative embodiment, an independent comparator 122 performs the comparison and controls the output of the interventions 106 to be executed. The comparator 122 reads in the data-driven interventions 116 and the conventional interventions 118, compares them, and outputs the interventions 106 to be executed.
[0053] In one embodiment, the interventions 106 to be executed are mixed in a mixer 124 using the data-driven interventions 116 and the conventional interventions 118. The interventions 106 to be executed are mixed proportionally from the data-driven interventions 116 and proportionally from the conventional interventions 118. A mixing ratio is set based on the result of the comparison and is variable, for example, between zero and one. The mixing ratio can also depend, for example, on the integral of the deviation.
[0054] In the following, possible embodiments of the invention are summarized again or presented with slightly different wording.
[0055] A safety mechanism is presented, particularly for an anti-lock braking system based on artificial intelligence.
[0056] Anti-lock braking systems (ABS) on motorcycles are traditionally applied specifically for each vehicle type. The ABS is parameterized to adapt to the specific characteristics of the respective motorcycle. This process is complex and costly, as it can only be performed by experienced application engineers.
[0057] By using AI-based approaches, this calibration effort can be significantly reduced. The algorithm learns the motorcycle's specific characteristics directly and does not need to be manually adapted to the respective model. Furthermore, the resulting decelerations can exceed those of conventional ABS.
[0058] One problem with the AI approach is proving that the "learned" algorithm is "safe." This is a fundamental problem with AI algorithms. The approach presented here allows such an AI approach to be implemented in the vehicle while simultaneously ensuring its safety.
[0059] In motorcycles, one of the worst error cases is when the learned algorithm "underbrakes" the motorcycle. In this case, the algorithm prevents or reduces deceleration because it expects the wheel to lock or the rear wheel to lift off, even though greater deceleration would be easily possible and a conventionally applied controller could also accomplish this.
[0060] The approach presented here implements two algorithms. Normally, the Kl algorithm is used to control the ABS. However, this algorithm is monitored by a conventional, generic algorithm. If the motorcycle's deceleration falls below the level guaranteed by the conventional algorithm, the control regime changes, and the conventional controller takes over control of the ABS. In this case, it is no longer necessary to configure the conventional, generic algorithm for optimal performance, but only for safety. This significantly reduces the application effort.
[0061] A motorcycle's ABS uses specific sensor signals as inputs to regulate the slip between the tire and the road. The goal is to set the amount of slip that results in maximum deceleration while also preventing the wheel from locking.
[0062] In the approach presented here, two controllers control the slip, with only one controller in command at any given time. The first controller is a KL algorithm that has learned the slip that allows maximum deceleration without wheel locking. This controller normally has command. The second controller is a conventional algorithm that controls the slip. These algorithms are generally very safe (from a functional safety perspective) because they are well-known through years of experience.
[0063] The second controller monitors the first controller at all times. As long as the first controller sets a better slip than the second controller would set itself, the second controller remains in the observation state. However, if the first controller sets a lower slip or causes the wheels to lock, the second controller actively switches from observation to command and actively controls the wheel slip itself.
[0064] By applying this combined control strategy, all 'safety loads' are transferred to the conventional controller and not to the Kl controller, making the entire control strategy easier to approve for road traffic.
[0065] Several embodiments are conceivable for the second controller. The second controller can be implemented identically for every motorcycle. The second controller can also be a generic algorithm that can be adapted to different motorcycles through parameterization. Parameters can include, for example, weight, center of gravity, and / or wheelbase. Alternatively, the second controller can be applied motorcycle-specifically. The command change between the two controllers can, for example, only occur after a confirmation time interval has elapsed. The command change can also only occur after a certain error magnitude has been exceeded. Furthermore, continuous blending can occur based on the difference between the approaches.
[0066] The two controllers can be hosted on the same control unit. The two controllers can also be hosted on different control units.
[0067] The switch can be operated either by the second control or an independent monitoring control.
[0068] The driver can be informed of the regime change via the HMI (Human-Machine Interface). Alternatively, or in addition, the event can be written to the diagnostic memory or transmitted to the manufacturer via radio.
[0069] In the approach presented here, two control algorithms are used on the motorcycle: a conventional one and a Kl-based one, for ABS control.
[0070] The approach presented here can be used particularly for two-wheeled motorized vehicles with ABS. Likewise, the approach presented here can be used in similar applications where Cl approaches offer advantages in control, but functional safety cannot be demonstrated, for example, ESP or ABS in two-track vehicles.
[0071] Finally, it should be noted that terms such as "comprising," "having," etc., do not exclude other elements or steps, and terms such as "a" or "an" do not exclude a plurality. Reference signs in the claims are not to be considered limiting.
Claims
Claims 1 . Method for influencing driving dynamics of a vehicle, wherein data-driven interventions (116) for actuators (104) of the vehicle are determined using a data-driven algorithm (112) and, in parallel, conventional interventions (118) for the actuators (104) are determined using a conventionally secure algorithm (114), wherein a comparison is carried out between the data-driven interventions (116) and the conventional interventions (118) and interventions (106) to be carried out for the actuators (104) are determined using a result of the comparison.
2. The method according to claim 1, wherein the conventional interventions (118) are determined as the interventions (106) to be performed if the conventional interventions (118) are better than the data-driven interventions (116).
3. The method according to claim 2, wherein the conventional interventions (118) are determined as the interventions (106) to be performed if the conventional interventions (118) deviate from the data-driven interventions (116) by more than a predefined tolerance.
4. The method according to one of claims 2 to 3, wherein the conventional interventions (118) are determined as the interventions (106) to be carried out if the conventional interventions (118) are better than the data-driven interventions (116) for longer than a predefined period of time.
5. Method according to one of the preceding claims, in which a deviation between the data-driven interventions (116) and the conventional interventions (118) is integrated to form a deviation sum, wherein the conventional interventions (118) are considered to be the interventions to be carried out (106) if the deviation sum is greater than a tolerance sum.
6. Method according to one of the preceding claims, in which the interventions (106) to be carried out are mixed from the data-driven interventions (116) and the conventional interventions (118), wherein a mixing ratio is dependent on the result of the comparison.
7. Method according to one of the preceding claims, wherein the conventionally securable algorithm (114) is generic for different vehicles.
8. The method according to claim 7, wherein at least one parameter of the generic algorithm (114) is parameterized for the vehicle.
9. The method according to any one of the preceding claims, wherein the data-driven algorithm (112) is trained for a configuration of the vehicle.
10. Method according to one of the preceding claims, in which a message is stored in a diagnostic memory of the vehicle when the conventional interventions (118) are determined as the interventions (106) to be carried out.
11. Control device, wherein the control device is designed to execute, implement and / or control the method according to one of the preceding claims in corresponding devices.
12. A computer program product configured to instruct a processor, upon execution of the computer program product, to execute, implement and / or control the method according to one of claims 1 to 10.
13. A machine-readable storage medium on which the computer program product according to claim 12 is stored.