DE102018217300A1 - Procedure for adapting a control system - Google Patents

Abstract

The present invention relates to a method for adapting a control system for an autonomous control, comprising the steps of electronically recording (S101) operating data via a human operator during operation of the device to be controlled; and adapting (S102) configuration parameters of the control system on the basis of the operating data obtained.

Description

The present invention relates to a method for adapting a control system for an autonomous control, a control system, computer program and a computer-readable storage medium.

State of the art

The manual adjustment of the setting parameters of a control system for an autonomous control, such as a driver assistance system for an autonomous control of a motor vehicle, is tedious and time-consuming for humans. In addition, manual adjustment can lead to dangerous situations in traffic if this is carried out during operation. An abstract numerical value is often poorly suited for setting the behavior of the driver assistance system that is perceived as pleasant. For example, every motorist has his own feeling for a comfortable distance from the vehicle in front, but cannot easily specify the distance in meters or in relation to the speed.

In order to take this observation into account, control units often have setting parameters that are set manually by humans, i.e. can be changed via a display and keyboard or manual controls in order to adapt to a specific environment, a specific task and preferences of the user. An example of such a setting parameter is the distance to vehicles in front with adaptive cruise control (ACC).

The publication DE 10 2010 063 006 A1 relates to a method in a driver assistance system with a front camera in a vehicle, and corresponding driver assistance systems. Here it can be checked whether the driver is actively pressing the accelerator pedal. If this is the case, this is taken as an indication that automatic stopping is not desired. The driver assistance system can generally be configured in such a way that an already initiated automatic stopping process is interrupted by an active actuation of the accelerator pedal by the driver (ie an actuation that lies above a predefined force value). The driver can therefore take control of the vehicle again at any time and can, for example, carry out the stopping process himself.

The publication DE 10 2013 209 064 A1 relates to a driver assistance device for a motor vehicle. The driver assistance device is configured to adapt the limit value that is applicable in the speed limit mode as a function of the prescribed maximum speed determined by the traffic sign monitoring device, taking into account an offset value that can be specified by the driver.

The publication DE 103 58 498 A1 Driver assistance device and method for predictive control of a vehicle. Driving strategy data are offered by a driver assistance device and by the real, human driver. The human driver operates the vehicle's brakes, accelerator and steering wheel, for example. In principle, the human driver could now completely commission the driver assistance device according to the invention to control the vehicle, so that the driving strategy data provided by the driver assistance device according to the invention are used to control the vehicle.

The publication DE 103 58 494 A1 relates to a driver assistance device and a method for predictive control of a vehicle. A predictive driver assistance system can control the vehicle fully autonomously. Usually, however, the driver wants, as far as possible, to have control over the vehicle itself and to carry out the control tasks himself, for example by actuating the accelerator pedal and brake and by steering. However, incorrect driving strategies can be specified by the human driver, which leads to critical driving situations. Both the driver assistance system and the human driver then offer driving strategies.

The patent CN105667574 considers the automatic adjustment of the parameters of an electric power steering depending on specific sizes on the vehicle that indicate the driver's driving style.

Disclosure of the invention

Advantages of the invention

A first aspect relates to a method for adapting a control system for an autonomous control, comprising the steps of electronically recording operating data via a human operator during operation of the device to be controlled; and adapting configuration parameters of the control system on the basis of the operating data obtained. Operating data are to be understood as all data relating to human operation during human operation. Configuration parameters are understood to mean all parameters that serve to configure the control system so that it can autonomously mimic the behavior of the operator. The method achieves the technical advantage that the control system is simple Way automatically adapted to the behavior of a user.

In a technically advantageous embodiment of the method, the recorded operating data are converted into configuration parameters using a calculation model in which the configuration parameters to be determined are incorporated linearly into the model equations. This achieves the technical advantage, for example, that the calculation can be simplified and carried out quickly.

In a further technically advantageous embodiment of the method, the calculation model comprises a gradient method or a recursive least squares method. This achieves the technical advantage, for example, that the configuration data can be determined with little computation effort.

In a further technically advantageous embodiment of the method, the control system is a driver assistance system in a motor vehicle. This achieves the technical advantage, for example, that the driver assistance system can be quickly adapted to specific users.

In a further technically advantageous embodiment of the method, the average distance is observed during an operator-controlled journey to a motor vehicle in front and an automatic distance control of the driver assistance system is configured on the basis of the average distance. This has the technical advantage, for example, that manual controls for setting the desired distance can be dispensed with.

In a further technically advantageous embodiment of the method, steering wheel movements during lane keeping or a lane change are recorded and an automatic lane keeping function or lane changing function of the driver assistance system is configured on the basis of the average steering wheel movement.

In a further technically advantageous embodiment of the method, operating data are recorded as to how and at what times the steering wheel is moved.

In a further technically advantageous embodiment of the method, operating data are recorded as to how and at what times the steering wheel is moved and / or the accelerator pedal and a brake of the motor vehicle are actuated.

In a further technically advantageous embodiment of the method, operating data are recorded, in what manner and at what times, an accelerator pedal and a brake of the motor vehicle are actuated. This has the technical advantage, for example, that the driving behavior of a user can be imitated in a simple manner.

In a further technically advantageous embodiment of the method, a configuration parameter for a shift point in time for a gear change is calculated on the basis of the operating data obtained. This also achieves the technical advantage, for example, that the driving behavior of a user can be imitated in a simple manner.

A second aspect relates to a control system which is set up to carry out the method according to the first aspect. The same technical advantages are achieved by the control system as by the method for adapting the control system according to the first aspect.

A third aspect relates to a computer program, comprising instructions which, when the program is executed by a computer, cause the computer to carry out the method according to the first aspect. The same technical advantages are achieved by the computer program as by the method for adapting the control system according to the first aspect.

A fourth aspect relates to a computer-readable storage medium, comprising instructions which, when executed by a computer, cause the computer to carry out the method according to the first aspect. The same technical advantages are achieved by the storage medium as by the method for adapting the control system according to the first aspect.

• 1 ein Konzept im Blockschaltbild;
• 2 eine schematische Ansicht einer Ausführung einer Steuerungseinheit;
• 3 eine Übersicht und ein Blockschaltbild der Modellgleichung am Beispiel eines Abstandstem pomats;
• 4 einen PI-Regelalgorithmus; und
• 5 ein Blockdiagramm eines Verfahrens zum Anpassen eines Fahrassistenzsystems.

Embodiments of the invention are shown in the drawings and explained in more detail in the following description. Show it

• 1 a concept in the block diagram;
• 2nd a schematic view of an embodiment of a control unit;
• 3rd an overview and a block diagram of the model equation using the example of a distance tomat;
• 4th a PI control algorithm; and
• 5 a block diagram of a method for adapting a driver assistance system.

1 shows a concept in the block diagram of a driver assistance system as a control system 100 for motor vehicles 115 , such as cars and trucks. The driver assistance system is a technical control system 100 that man 109 operated in the sense of a controller and alternating with an electronic control unit 101 of the control system 100 is active.

For example, when operating a vehicle, it is obvious that people 109 each have individual preferences with regard to a desired operation of the motor vehicle, such as for a sporty driver or an anxious driver. Therefore, an adaptation of the electronic control unit 101 of the control system 100 individual preferences of people 109 be made as a human being 109 otherwise quickly feel uncomfortable during operation. For example, an anxious passenger does not feel comfortable with sporty driving behavior.

Therefore, human behavior within the driver assistance system 109 compared with a model and the parameters of the model using an adaptation device 103 configured to minimize the difference between human behavior and the behavior of the model. The parameters determined in this way are in the control system 100 used to customize and configure the control unit 101 to people 109 to make.

The vehicle control or a part of it can either be done by humans 109 or the control unit 101 be taken over. There is usually a switchover to automatically controlled operation by humans 109 instead of. The electronic control device 101 closes the control loop by accessing measured values from sensors 107 of the driver assistance system while man 109 closes the control loop via its perception, for example through visual perception, sense of acceleration or sense of touch.

To control the parameters of the controller 101 the adjustment unit is suitable 103 intended. The adjustment unit 103 accesses the measurements available from the driver assistance system and those from humans 109 input variable to obtain the required operating data. For this purpose is a sensor 107 installed for the respective measured variable or a calculation from other measured variables is possible.

The adjustment unit 103 Calculates variables from the available measurement variables to describe the human perceptions relevant for the specific control task, such as the optical perception of the distance to the person in front in a distance control system in subsequent operation. In some driver assistance systems, the relevant human perception is also directly known, since it is read off a display, such as on a speedometer. The perception and the resulting adjustment on the driver assistance system can next be compared with a mathematical model.

The parameters of the mathematical model are determined while humans 109 is active, ie during operation of the device to be controlled by humans. The parameters are determined so that the difference between the model and human behavior over time is minimal.

On the one hand, the mathematical model is designed to best represent human behavior as a controller of the present technical system, or on the other hand it corresponds to the control algorithm implemented in the electronic control unit. In the first case, the model parameters are converted to the electronic control unit 101 to adapt to human behavior. In the second case, the parameters of the control algorithm are determined directly in order to best adapt it to human behavior. The determined parameters for adapting the driver assistance system are within certain safety restrictions in the control unit 101 accepted.

A technically advantageous structure of the mathematical model exists when the parameters a ₀ , a _1, ..., b ₀ , b ₁ , ... are linearly included in the model equations, for example in the case of a model with the inputs e and the exit u: $$a_0+a_{1}e+a_{\mathrm{2nd}}\dot{e}+a_{\mathrm{3rd}}\ddot{e}+\cdots =b_{´0}+b_{1}u+b_{\mathrm{2nd}}\dot{u}+b_{\mathrm{3rd}}\ddot{u}+\cdots$$

In this case, known estimation methods, such as a gradient method or a recursive least squares method (least squares method) can be used to determine the parameters that minimize the difference between the model and human behavior.

The technical system 100 , the control unit 101 , and the adjustment unit 103 are each formed, for example, by a software module or computer program that is executed on a computer device with a processor and an electronic memory for storing the software module and the data records.

2nd shows a schematic view of an embodiment of the control algorithm in the control unit 101 . The control unit 101 includes for example a controller 111 . The regulator 111 uses the difference from a setpoint 113 and the measurement of a controlled variable in order to determine an actuating signal for influencing the controlled variable. The setpoint 113 is either constant, for example a target distance to the right lane boundary, or dependent on measured or calculated auxiliary variables, such as a speed-dependent target distance to the vehicle in front in an ACC. Depending on the application, the setpoint can also be calculated dynamically depending on the time, for example with a filter or trajectory planning. The parameters of the filter or the planning algorithm can then also be part of the mathematical model and can be determined by humans on the basis of the operating data.

3rd shows an overview and a block diagram of the model equation using the example of a cruise control in the driver assistance system. The cruise control is in a follow-up mode, ie in a mode as a distance controller to a motor vehicle in front.

The adjustment unit 103 is active while the motor vehicle is following a motor vehicle in front. The distance d The relevant human perception for this control task is determined by a distance measurement. The manipulated variable of the person is the actuation of the accelerator pedal and brake, which can be summarized as input e.

The adjustment unit 103 evaluates the sizes d and u off, ie the distance to the vehicle in front and the actuation of the accelerator pedal and brake in order to determine the configuration parameters of the mathematical model. The configuration parameters are parameters that are used to configure the control system 100 serve so that it can autonomously mimic the behavior of the operator. The configuration parameter is, for example, an average distance that a self-driving vehicle should maintain from a vehicle in front.

Umgeformt ergibt sich die zuvor erläuterte, besonders günstige Modellstruktur $$su=k_l{d}_{soll}-k_{l}d+s{k}_p{d}_{soll}-s{k}_{p}d,$$

woraus mit sk_pd_soll = 0 (zeitliche Ableitung konstanter Terme) die Gleichung $$su=\left[\begin{array}{ccc}{k}_1{d}_{soll}& k_l& k_p\end{array}\right]\ast \underset{sd}{\overset{1}{\left[d\right]}}$$

zur Anwendung bekannter Parameterschätzverfahren resultiert.A PI control algorithm of the electronic control unit can be used as the mathematical calculation model 101 with the setpoint d _should and the parameters k _P and k _I be used. The model equation in the Laplace area is: $$u=\frac{1}{s}{k}_l\left(d_{sOll}-d\right)+k_p\left(d_{sOll}-d\right)$$

The previously described, particularly favorable model structure is formed $$su=k_l{d}_{sOll}-k_{l}d+s{k}_p{d}_{sOll}-s{k}_{p}d,$$

from which with sk _p d _soll = 0 (time derivative of constant terms) the equation $$su=\left[\begin{array}{ccc}{k}_1{d}_{sOll}& k_l& k_p\end{array}\right]\ast \underset{s\mathrm{<figure-callout\; id="1"\; label="d"\; filenames="DE102018217300A1\_0006.png"\; state="\{\{state\}\}">d</figure-callout>}}{\overset{1}{\left[d\right]}}$$

results in the use of known parameter estimation methods.

This estimation equation automatically determines the distance from the person in front that is comfortable for humans d _should and the aggressiveness of how rigidly to keep this distance, via the control parameters k _P and k _I certainly. Sufficient excitation must be taken into account for which parameter determination is active. If, for example, the target distance is reached and the person in front does not brake or accelerate, it makes sense to carry out the parameter determination only partially or not at all.

The automatically determined configuration parameters are before the electronic control device is operated 101 Check for safety restrictions, such as legal requirements regarding a minimum distance d _should and requirements regarding the stability of the control loop, and if necessary to limit it to maximum or minimum values.

The correspondingly limited and checked configuration parameters can then be set as setting parameters in the electronic control device 101 (Adaptive cruise control implemented as PI controller). With automatic transmissions, the shift times can be modified depending on the recognized driving style. In another example, an automatic lane change behavior of the driver assistance system is adapted to a driver's previously recorded preferences.

4th shows a PI control algorithm in which a manipulated variable u via an integral part with reinforcement k _I and a proportional portion with gain k _P from the difference of a target quantity d _should and an actual size d is calculated.

5 shows a block diagram of a method for adapting the control system 100 for autonomous control, such as a driver assistance system. In a first step S101 operating data about human operation during human operation of the motor vehicle is electronically recorded as the device to be controlled. Under operating data in the case of a driver assistance system, data about a Distance to a motor vehicle in front or data about an actuation of an accelerator pedal and the brake and / or data about the steering wheel movement can be understood. In general, operating data is understood to mean all data relating to human operation during human operation. These can be recorded using appropriate sensors. For example, a distance to a motor vehicle in front can be detected via a distance sensor or an actuation of an accelerator pedal and the brake and / or steering wheel movement by a corresponding sensor.

In step S102 configuration parameters of the driver assistance system as a control system 100 adjusted or configured on the basis of the operating data obtained and the control device 101 of the driver assistance system.

This makes it easy to adjust the driver assistance system to a driver's preferences. The method enables the driver assistance system to be more precisely adapted to people than would be possible manually with just a few setting parameters.

The configuration parameters of the electronic control unit 101 are automatically derived from the behavior of humans while they operate the system themselves. In the event of large deviations in the configuration parameters from expected or previously determined configuration parameters, a warning function can be implemented which indicates unusual behavior of the person, for example through fatigue or stress.

The procedure is used when used in control systems 100 with changing human activity 109 and electronic control unit 101 an adaptation of the control unit 101 to the previously observed human behavior. This can be checked through specifically different behavior as a human operator. The driver assistance system is configured on the basis of the previously recorded operating data in such a way that it imitates the driving style of the human driver as precisely as possible in automatic operation.

The scope of the present invention is given by the claims and is not limited by the features explained in the description or shown in the figures.

All of the features explained and shown in connection with individual embodiments of the invention can be provided in different combinations in the subject matter of the invention in order to simultaneously realize their advantageous effects.

QUOTES INCLUDE IN THE DESCRIPTION

This list of documents listed by the applicant has been generated automatically and is only included for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Patent literature cited

• DE 102010063006 A1 [0004]
• DE 102013209064 A1 [0005]
• DE 10358498 A1 [0006]
• DE 10358494 A1 [0007]
• CN 105667574 [0008]

Claims (10)

Method for adapting a control system (100) for an autonomous control, comprising the steps: - Electronic acquisition (S101) of operating data about a human operator while the device to be controlled (100) is in operation; and - Adapting (S102) configuration parameters of the control system (100) on the basis of the operating data obtained. Procedure according to Claim 1 , the recorded operating data being converted into configuration parameters using a calculation model, in which the configuration parameters to be determined are incorporated linearly into the model equations. Procedure according to Claim 2 , wherein the calculation model comprises a gradient method or a recursive least squares method. Method according to one of the preceding claims, wherein the control system (100) is a driver assistance system in a motor vehicle. Procedure according to Claim 4 , wherein the average distance is observed during an operator-controlled journey to a preceding motor vehicle (200) and an automatic distance control of the driver assistance system is configured on the basis of the average distance. Procedure according to one of the Claims 4 or 5 , wherein operating data are recorded, in what manner and at what times, an accelerator pedal and a brake of the motor vehicle are actuated. Procedure according to one of the Claims 4 to 6 , wherein a configuration parameter for a shift time for a gear change is calculated on the basis of the operating data obtained. Control system (100), which is set up, the method according to one of the Claims 1 to 7 to execute. Computer program, comprising instructions which cause the computer to execute the program, the method according to one of the Claims 1 to 7 to execute. Computer-readable storage medium, comprising instructions which, when executed by a computer, cause the computer to execute the method according to one of the Claims 1 to 7 to execute.

Images