Method for tracking control of a vehicle, control device, vehicle and computer program

Abstract

The invention relates to a method for the tracking control of a vehicle (1), in which a control variable is determined by means of a lateral control unit (7) on the basis of a lateral position error which describes a lateral deviation (4) of the vehicle (1) from a predetermined target trajectory (5), the lateral position of the vehicle (1) is adjusted to the target trajectory (5) on the basis of the control variable, at least one calibration value is additionally applied to the control variable, the calibration value being determined by at least one calibration function (12) at least partially depending on a time integral of the lateral position error.

Description

[0001] The invention relates to a method for tracking control / trajectory control of a vehicle, wherein a control variable / manipulation variable is determined by a lateral controller as a function of a lateral position error, the lateral position error describing a lateral deviation of the vehicle relative to a predetermined target trajectory, wherein the lateral position of the vehicle is adjusted to the target trajectory as a function of the control variable. The invention also relates to a control device, a vehicle and a computer program.

[0002] The vehicle may be equipped with a system or function for automatic tracking control. This makes it possible to automatically set the lateral position of the vehicle according to a predetermined target trajectory during driving. In this case, the lateral position of the vehicle, i.e. the sideways position of the vehicle on the lane, can be changed or adjusted to the target position, for example by an automatic steering movement.

[0003] In this case, such lane keeping functions usually guide the vehicle along a planned trajectory with the aid of sensors for detecting the vehicle's surroundings, such as cameras, radar or lidar systems, and / or with the aid of environmental data obtained by navigation systems, such as the global positioning system (GPS). In this case, the trajectory is designed, for example, for high driving comfort, i.e., the lateral position of the vehicle changes as little as possible, either abruptly or rapidly. The vehicle's driving along such a trajectory is usually guided by a control device, in which at least the lateral deviation of the vehicle from the planned trajectory is continuously determined, wherein, depending on the deviation, control instructions of appropriate size and direction are issued, for example, to the vehicle's electric steering assistance system. Criteria for evaluating the control quality of such systems are, for example, the amplitude and dynamics of the lateral control deviation and the frequency of the jerk movements of the vehicle's steering wheel. Ideally, the lateral control error is zero in all driving situations, so that the steering intervention caused by the tracking control device or the resulting steering wheel movement is determined only by the direction of the planned trajectory.

[0004] The performance of such tracking control systems depends on the quality and performance of the vehicle sensors and actuators used. The performance of the overall system can usually be improved to a certain extent by using more sensors or more actuators and / or by using suitable software measures. This applies in particular to driver assistance systems such as lane keeping assist functions.

[0005] However, the sensor error of the sensor, i.e. any deviation of the sensor signal from its physical equivalent, affects the control characteristics and thus, in particular, the control quality of the vehicle tracking control due to the occurrence of error propagation. For this reason, different influence chains, sometimes even superimposed influence chains, are generated depending on the sensor information.

[0006] Imaging errors caused by the camera device when detecting lane markings can, for example, lead to deviations of the determined lane curvature from the actual curvature, which is the so-called ground truth error. The tracking control device usually includes at least one control device and / or at least one pilot control device, which adjusts the lateral position of the vehicle as a function of the determined lane curvature. The curvature information describing the lane curvature is usually determined at the forward point of view of the vehicle, where it can be used to calculate the steering angle required for each respective curvature using an inverted vehicle model and introduced as a target steering angle component into the steering angle control device, for example as part of the pilot control.

[0007] If the curvature information is biased relative to the actual curvature due to an error, the target steering angle of the pilot control will also contain a bias component, which will more or less interfere with the lateral control. As a result, when the lane keeping assist system is activated, the vehicle does not follow the expected trajectory, but, depending on the design of the lateral control device or cruise control device, that is, the type of control device used has or does not have fixed accuracy, the planned trajectory will produce a continuous lateral deviation, or at least cause a short-term steering before the planned trajectory is driven after the vehicle balancing process has weakened. In this case, the continuous or short-term lateral deviation will affect the driving comfort, thereby adversely affecting the driver's acceptance of the driving assistance system.

[0008] The degree of camera imaging error usually changes over time as the driver assistance system is operating, since the camera calibration procedure is typically performed continuously. This means that the level of the bias of the curve pilot control and the degree to which the vehicle's lateral control is affected changes over time, depending on the camera calibration state, even though the camera's transmission characteristics and curvature bias do not change suddenly, but rather change within a certain range or gradually.

[0009] In most cases, the steering angle is an important auxiliary variable or control variable in the traction control assistance system, so a routine is usually used to calibrate the steering angle offset. Due to the limited accuracy of the calibration routine, the steering angle offset determined here always has an uncompensated part. The residual offset error in the steering angle signal can have various reasons. Typical calibration routines are based on the signals of yaw rate sensors or wheel speed sensors, for example. Errors in these sensors, such as offset and / or linearity errors, or in the case of wheel speed sensors, speed differences due to different tire pressures on the axles, are propagated into the determined steering angle offset. The influence of the steering angle offset is similar to the above-mentioned curvature offset: here, when the lane keeping assistance function is activated, the vehicle does not follow the planned trajectory as expected, but, depending on the design of the lateral controller or cruise control, shows a permanent lateral deviation or a transient lateral deviation from the planned trajectory, which has an adverse effect on the comfort and acceptability of the driver assistance system.

[0010] The same applies to uncompensated offsets of the front and rear axle steering angles.

[0011] In this case, the sensor signals of the yaw rate, the front-axle steering angle, the rear-axle steering angle and the lateral acceleration can also be used in particular for other functions within the vehicle, which also have an influence on the lateral position or lateral control of the vehicle. For example, these sensor signals can also be used to estimate the disturbance forces and load torques acting on the vehicle and to estimate the vehicle's side slip angle, for example due to a lane inclination and / or a crosswind. In this case, under the influence of the disturbance variables, the determined variables can also act on the lateral control of the vehicle as a component of the target steering angle, wherein in particular the corresponding offset of the sensor signals can also be propagated as an error to the target steering angle and thus to the lateral position of the vehicle. This can also result in unexpected transient or permanent lateral deviations of the vehicle under tracking control.

[0012] DE 10 2008 026 233 B4 discloses a method for offset compensation of the steering angle of a motor vehicle with the aid of a vehicle model. In this case, the yaw rate is calculated as a function of the detected steering angle, the vehicle speed and an assumed steering wheel angle offset. Furthermore, the calculated yaw rate is compared with the measured yaw rate, wherein the difference between the two forms an error equation, from which the steering wheel angle offset is then determined and fed back into the vehicle model in order to determine and compensate for the steering wheel angle offset in an iterative manner.

[0013] Such bias calibrations of yaw rate or lateral acceleration have limited accuracy and always lag behind the actual bias, so that even in this case there is an additional undesired steering intervention.

[0014] The technical problem to be solved by the present invention is to improve the tracking control accuracy of the vehicle, thereby improving the driving comfort and the driver's acceptance of the auxiliary system even under the influence of the bias error.

[0015] To solve this problem, the invention provides for the method described in the introduction that at least one calibration value is additionally applied to the control variable, wherein the calibration value is determined at least partially by at least one calibration function as a function of the time integral of the lateral position error.

[0016] The vehicle transverse position is thus adjusted by means of the sum of the control variable and the at least one calibration value. Depending on the design of the transverse controller and / or the tracking control, in addition to the at least one calibration value, further components may be added to the control variable.

[0017] The control variable to which the calibration value is applied influences the lateral position of the vehicle. The control variable may be, for example, a target steering angle. The target steering angle may be pre-set, for example, at a steering angle interface of an electric power steering system (EPS). Alternatively, the control variable may also be pre-set at a torque interface of an electric power steering system (EPS) that generates a steering torque or a steering angle.

[0018] The calibration function determines the calibration value, in particular as a function of the sign and magnitude of the integrated lateral position error, in such a way that the calibration value, when applied to the control variable, always aims at reducing the current lateral position error. By applying the calibration value to the control variable, it is achieved that the target steering angle is corrected or compensated as a function of the time integral of the lateral position error.

[0019] The lateral position error describes the lateral or sideways deviation of the vehicle relative to a predefined target trajectory, i.e. the deviation in the transverse direction of the vehicle. The lateral position error can therefore also be referred to as lateral deviation or lateral deviation. In other words, the lateral position error represents the instantaneous deviation between the actual position of the vehicle and the desired target position of the vehicle described by the target trajectory. In this context, the determination of the lateral deviation of the vehicle relative to the target trajectory can be related to an agreed point along the longitudinal axis of the vehicle. Thus, the lateral deviation can be advantageously associated with the geometric center of the rear axle or the front axle, but also with any other predicted point along the longitudinal axis of the vehicle.

[0020] In this case, the lateral position error can be determined continuously, for example, based on sensor data of at least one vehicle surroundings sensor. The surroundings sensor can be, for example, a camera, a lidar sensor, a radar sensor and / or an ultrasonic sensor. As an additional measure or alternative, the lateral position error can also be determined based on position data of a navigation system. In this case, the value of the lateral position error is time-dependent and is integrated by means of a calibration function.

[0021] The lateral controller for the lateral control can be, for example, a cruise control or include such a control device. There is also a calibration function in addition to the cruise control, which can also be called centering calibration or centering function, and thus forms a peer control loop or superordinate control loop, which also influences the control variable and thus also the lateral position of the vehicle. In this case, the dynamics of the calibration function are particularly significantly lower than the dynamics of the lateral control, so that the lateral position error integrated over time has a very slow influence on the control variable.

[0022] As a calibration function or as a basic calculation rule for determining the calibration value, for example, an integrator with a relatively small integration factor can be used, so that a low-dynamic integration of the lateral position error and thus a low-dynamic influencing of the calibration value is achieved.

[0023] According to the invention, at least one calibration function can be set as an integration unit / integration link or a PT1 unit / PT1 link. Compared with a pure integration unit, a PT1 unit also has an integration characteristic, but has the advantage that the measures required for limiting the integrator state in the case of an integration unit are not required. In the corresponding application, this advantage should be weighed against the disadvantage of unstable precise compensation of the bias parameter in order to fully utilize the parameterization freedom of the PT1 unit.

[0024] The dynamics of the adaptation of the calibration value is determined in such a way that it is in the order of magnitude of the dynamics of the change of the bias error. For this purpose, the maximum expected temperature-dependent bias drift of the sensor for the lateral control is converted, for example, to the level of the corresponding steering angle. This has the advantage that no overshooting of the lateral control occurs, which would, for example, lead to intervention in other compensation functions that act on the steering angle, such as lane inclination compensation and / or crosswind compensation, if the dynamics of the calibration function were higher.

[0025] By using at least one calibration function or applying at least one calibration value to the control variable, a continuous influence on the traction control can be achieved in a beneficial way, so that the bias errors are partially or completely reduced over time, or these bias errors are partially or completely compensated over time. Compensation functions that act on the steering angle are usually not activated continuously. For example, the lane keeping function usually only takes effect after being activated by the driver. If the compensation function generates an assumed control error due to the bias error at the time of activation, the compensation function will output a control variable in an attempt to reduce the control error. Depending on the dynamics of the compensation function, this can lead to jerky steering movements. By compensating for the bias errors in advance, jerky steering movements to correct accumulated lateral deviations can be avoided in a beneficial way, thereby improving the comfort of the occupants when the traction control is activated. This also improves the acceptability of the traction control and the driver assistance system that provides the traction control.

[0026] In summary, the basic idea of ​​the superordinate calibration function can also be interpreted as a low-dynamic fixed value control with a target value of zero for the lateral position error. In terms of its effect, it reduces the sum of all direct and indirect deviation effects of the sensors used in the network on the lateral deviation of the vehicle during automatic lateral control or tracking control.

[0027] According to the invention, it can be provided that the time constant of the calibration function is greater than the main time constant of the lateral controller, in particular at least twice the main time constant. The time constant of the calibration function can preferably also be at least 5 times, at least 10 times, at least 100 times or at least 1000 times the main time constant of the lateral controller. In principle, greater differences between the time constants are also conceivable. For example, the integration factor of the calibration function implemented as an integration unit can be, for example, 0.01 degrees / (m*s) to 0.00005 degrees / (m*s).

[0028] In a preferred design scheme of the present invention, it can be provided that the integral value of the calibration function is determined by the time integral of the lateral position error, wherein the integral value is retained when the tracking control is not working and is used as the starting value of the time integral of the lateral position error when the tracking control is reactivated.

[0029] In this case, the integral value describes a total value which is formed by integrating the lateral position error over time up to the point in time to which the integral value corresponds. The integral value thus represents, for a specific point in time, the result of integrating the lateral position error over time up to this point in time. If a plurality of calibration functions are used, in particular for each calibration function, an integral value assigned to the respective calibration function is formed separately. In this case, the at least one integral value can be stored, for example, in a retrievable manner in a storage device of a computing device, for example a control device, which executes the method.

[0030] The state of the calibration function, i.e. the state of the integrator or similar functional unit forming the calibration function, is therefore not reset, for example, when the vehicle is currently in operation with the ignition on, but remains unchanged during repeated deactivation and activation of the driver assistance system or driver intervention. However, the integrator is also reset if an abnormal activity is detected, such as a jump in the output of one or more offset calibrations of the vehicle sensors, or a request is made via the vehicle diagnostic interface. The integrated value can also be retained after the vehicle is in operation with the ignition on, i.e. during multiple driving operations.

[0031] According to the invention, it can be provided that the transverse controller comprises a pilot control device, wherein the pilot control device determines a pilot control variable as a function of measured and / or predicted curvature values ​​of a lane in which the vehicle is traveling, wherein the pilot control variable is included as a component in the control variable.

[0032] In this case, the measured curvature value and / or the predicted curvature value can be determined in particular from sensor data of one or more vehicle surroundings sensors. The pilot control device can in particular use the measured curvature value and / or the predicted curvature value to calculate the steering angle required for the respective curvature by means of an inverted vehicle model and use it as a pilot control variable, for example, as a control variable of a cruise control device switched to a lateral controller. The vehicle model changes its transformation / transfer characteristics depending on the vehicle speed. This means that errors in the measured curvature value have an influence on the vehicle speed-dependent lateral control of the vehicle.

[0033] According to the invention, it can be provided that the pilot control is performed based on a vehicle model describing the vehicle's self-steering gradient / natural steering trapezoid, wherein the self-steering gradient is adjusted during vehicle operation as a function of at least one vehicle parameter, in particular vehicle mass and / or tire rigidity.

[0034] In the case of a vehicle turning, errors in the pilot control lead to lateral position errors that are not only related to the bias error of the curvature, but also to inaccuracies in the stored vehicle parameters, in particular the uncertainty of the self-steering gradient used in the vehicle model. Errors or deviations in the self-steering gradient cause disturbances in the vehicle lateral control and thus also proportionally cause additional lateral position errors, leading to erroneous deviations in the calibration values ​​generated by the calibration function. In this case, the calibration values ​​do not only react to the sensor bias variable. As long as the errors or deviations in the self-steering gradient are limited, their influence is also small due to the low dynamics of the calibration function.

[0035] In order to avoid the influence of large errors and / or deviations in the self-steering gradient, it can be adjusted, in particular continuously, as a function of at least one vehicle parameter. Here, for example, the adjustment can be made as a function of the current vehicle mass, the current tire stiffness of one or more vehicle tires and / or other parameters of the vehicle.

[0036] In a preferred embodiment of the invention, it can be provided that the time integration of the lateral position error is performed with a weighting factor, in particular an integration factor, which is dependent on the speed of the vehicle and / or increases with increasing magnitude of the lateral position error.

[0037] While the use of a constant integral factor, for example, can represent an easily implementable compromise for the entire speed range of the vehicle, the use of a speed-dependent integral factor or a comparable integral weighting factor has the advantage that speed-dependent offsets can be better compensated. For example, based on the commonly used setting of the control parameters of the lateral control, an influence of the offset variable on the lateral position error can occur in dependence on the vehicle's driving speed. This also changes the disturbance compensation characteristics of the lateral control and thus the influence of the sensor offset variable on the lateral position error. Taking this into account, the time integral is also made dependent on the driving speed by using a speed-dependent weighting factor, for example a speed-dependent integral factor. This makes it possible to achieve an ideal, individually adjusted calibration dynamics for each driving speed range in an advantageous manner.

[0038] As an additional measure or alternative, the weighting coefficient may also be related to the size of the lateral position error, i.e., to the lateral position error value to be integrated and / or the lateral position error value previously integrated, wherein the larger the lateral position error value, the larger the weight. To this end, for example, the weighting coefficient or integration coefficient assigned to each current lateral position error may be determined based on a nonlinear allocation rule and / or one or more thresholds of the lateral position error, for the time integration of the lateral position error. By gradually increasing the weight of the lateral position error size, i.e., increasing the weighting coefficient or integration coefficient as the lateral position error size increases, the calibration speed may be effectively increased.

[0039] For this purpose, for example, the integration factor and thus the adaptation rate of the calibration function can be switched between two or more values, wherein a higher integration factor is selected for a larger absolute value of the lateral position error. Advantageously, when the magnitude of the lateral position error exceeds 0.2 meters, doubling the integration factor or the integration constant can effectively and quickly initially adapt to the existing bias situation when the driver assistance function is activated.

[0040] Furthermore, in the case of a vehicle having an additional sensor offset calibration function, which enables the offset drift of at least one of the vehicle sensors to be eliminated after a validation time or a debounce time by means of a step-shaped correction of the corresponding offset, an increase in the adaptation rate of the centering calibration as a function of the magnitude of the lateral position error enables a faster adaptation to new offset situations. This is all the more important the longer the debounce time of the sensor offset calibration function is, because during this debounce time the calibration function increasingly compensates for the integration that must be repeated again after the signal of the sensor offset calibration function is updated, due to the time integration of the lateral position error. The speed of the compensation process can be increased in a beneficial manner by using a greater weight or a greater integration value.

[0041] Since the step-like correction of the vehicle sensor offset calibration function directly causes a lateral jerk of the vehicle under tracking control that is related to the step height, it can be stipulated for a vehicle that includes a control device configured to implement the method according to the invention that the changes to the corresponding offsets of all sensor offset calibration functions in the vehicle are not performed in a step-like manner, but only within a limited range.

[0042] According to the present invention, it can be provided that

[0043] - the integral value of the calibration function determined by the time integration of the lateral position error is equivalent to / equal to a predetermined limit value,

[0044] - a measured and / or predicted curvature value of the lane on which the vehicle is travelling exceeds a predetermined threshold value,

[0045] - the curvature value of the target trajectory exceeds a predetermined threshold,

[0046] - the product of the square of the current vehicle speed and the lane curvature value exceeds a predetermined threshold, and / or

[0047] - the product of the square of the current vehicle speed and the curvature value of the target trajectory exceeds a predetermined threshold,

[0048] In this case, the time integration of the lateral position error by the calibration function is stopped.

[0049] The magnitude of the calibration value of the control variable can be limited to a limit value by at least briefly stopping the time integration of the lateral position error in the calibration function or by limiting the total value of the integration to a limit value. The limit value here represents a maximum amplitude and depends, for example, on the sum of the maximum influences of all offset variables on the steering angle target value and the steering angle actual value. By this method step, measures for avoiding integrator saturation effects can be advantageously omitted in the calibration function.

[0050] As an alternative to limiting the integrator value to a limit value, the calibration function can also include a low-pass filter function, for example a PT-1 filter. Saturation effects can also be avoided in this way. Although the sensor offset cannot be completely compensated for in this case, sufficient centering accuracy can still be achieved by the calibration function.

[0051] As an additional measure or alternative, the calibration function may stop the time integration of the lateral position error if measured and / or predicted curvature values ​​of the lanes in which the vehicles are traveling or of the lanes in which the vehicles are traveling exceed a predetermined threshold and / or the curvature value of the target trajectory exceeds a predetermined threshold.

[0052] In the case of a vehicle driving on a curve, a possible error in the pilot control causes a lateral position error that depends not only on the offset error of the curvature, but also on the inaccuracy of the stored vehicle parameters, in particular the uncertainty of the vehicle's self-steering gradient. An erroneous vehicle pilot control on a curve can interfere with the lateral control of the vehicle and thus proportionally cause an additional lateral position error. This additional lateral position error can lead to an erroneous intervention of the calibration function, which can be avoided or reduced in a beneficial way by stopping the calibration function or at least stopping the time integration of the lateral position error by the calibration function.

[0053] The influence of the error in the curve pilot control on the driving behavior or the lateral position of the vehicle is in particular dependent on the vehicle speed. Therefore, in order to take this speed dependency into account, it can be provided that the time integration of the lateral position error of the calibration function is stopped if the product of the square of the current vehicle speed and the lane curvature value exceeds a predetermined threshold value and / or the product of the square of the current vehicle speed and the target track curvature value exceeds a predetermined threshold value.

[0054] In a preferred embodiment of the invention, if the steering torque generated by the steering intervention of the vehicle driver is higher than a predetermined threshold value and / or the generated steering torque increases the current lateral position error, the time integration of the lateral position error is stopped and / or the application of the calibration value to the control variable is abandoned. This has the advantage that lateral position errors caused by intentional driving interventions can be prevented from being included in the calibration value determined by the calibration function, because these lateral position errors are not caused by offsets to be compensated for by sensors or similar devices. In this way, the influence of lateral deviations of the target trajectory intentionally caused by the driver on the calibration value can be avoided in a beneficial way, which deviations are not related to the tracking control function.

[0055] According to the invention, provision can be made for the lateral position error to be time-integrated with increased weighting, in particular increased integration factor, at least over a predetermined period of time, if a steering torque generated by a steering intervention by the vehicle driver reduces the current lateral position error.

[0056] If the steering torque resulting from the driver's active steering intervention is in the same direction as the calibration value generated by the calibration function, this is considered a confirmation of the adaptation direction of the calibration function. In this case, the calibration dynamics can be improved by the calibration function using an increased weighting or an increased integration value. From the driver's perspective, this means that the driver can move the vehicle's driving path more laterally towards the lane center, and the driving path will then remain in the lane center even if the correction steering torque is reduced.

[0057] In a preferred design scheme of the present invention, it can be provided that multiple integration functions are used, wherein each integration function is respectively assigned to a different speed range, and the integration functions respectively perform time integration of the lateral position error when the vehicle speed is within the speed range correspondingly assigned to it, wherein a calibration value of the integration function whose speed range includes the current vehicle speed, or a total calibration value determined based on the calibration value, is applied to the relevant control parameter.

[0058] As already mentioned above, the effect of the bias variable on the lateral position error depends in part on the vehicle speed. This applies, for example, to bias errors contained in the output variable of the sensor disturbance compensation or as a component of the curve pilot control. In order to keep the lateral position error as low as possible, the calibration value associated with the control variable can also be determined as a function of the speed. In particular, due to the high integration time constant and the associated low dynamics, the adaptation of the calibration value of a single calibration function to the changing driving speed can require a relatively long time, during which the accuracy of the tracking control can be reduced.

[0059] Taking this into account, it is very beneficial to use multiple calibration functions instead of a single calibration function over the entire vehicle speed range. In this case, each calibration function can determine its own calibration value, which can then be stored separately from the other calibration values. The ability to store the calibration values ​​separately, for example in a non-volatile memory, means that during further operation, these calibration values ​​can be continuously extended by time integration, or all states of multiple calibration functions can be initialized separately when the ignition is restarted.

[0060] In this case, the entire possible vehicle speed range can be divided into several sub-intervals, each of which covers only a limited vehicle speed range, and a dedicated calibration function is assigned to each sub-interval. In this case, the calibration value of the calibration function assigned to the speed interval in which the current vehicle speed lies is applied to the control variable. As a result, only a single calibration function is always activated, while the other calibration functions, in particular the time integration of the lateral position error of the other calibration functions, are stopped or switched to a zero input signal at their input. On the other hand, the correspondingly activated calibration function can provide the current lateral position error. In this case, the lateral position error can be weighted with a constant weighting factor or integration factor or with a weighting factor or integration factor that depends on the driving speed. Then, the calibration value of the integration function for the speed interval including the current vehicle speed, or a total calibration value determined from this calibration value, is applied to the control variable. Compared with the use of a single calibration function, the advantages of using a plurality of calibration functions or calibration function groups are particularly evident when accelerating or braking, and therefore when the vehicle speed range changes frequently.

[0061] According to the invention, an overall calibration value can be determined from two or more calibration values ​​generated by a plurality of integration functions in such a way that the overall calibration value is stable over time for a period of time even when changing between two or more speed intervals. In this way, abrupt changes in the calibration value associated with the control variable and thus abrupt interventions when changing the calibration function due to changes in vehicle speed are avoided.

[0062] In a preferred design scheme, the calibration function integral value associated with the calibration function is determined by the calibration function through the time integration of the lateral position error, wherein the integral values ​​of one or more integral functions whose current speed is outside the corresponding speed ranges are continuously adapted, in particular at a constant ratio or a ratio related to the vehicle speed, to the integral values ​​of the calibration functions whose current speed of the vehicle is within its speed range.

[0063] In order to avoid that, when the automatic tracking control function is mainly used in a narrow speed range, only a few or only a single calibration function are loaded with the lateral position error accordingly, so that only these calibration functions can ultimately correctly reflect the current bias situation, the non-active calibration functions that are not used at the current speed can also continue to integrate the current lateral position error over time. In this case, the integral value of the calibration function that is respectively assigned to the adjacent area of ​​the current speed range, i.e. the integrator state, is limited to the maximum value of the integral value of the calibration function assigned to the current speed range. In addition, the weighting coefficient or integration coefficient assigned to the calibration function of the non-current speed range can be reduced in an advantageous manner compared to the active case, because if the integration coefficient remains unchanged, the advantage of dividing into a plurality of integrator states disappears. The degree of reduction of the adjacent integration coefficient depends on the expected maximum sensor drift variable, for example in the order of 10% to 20%. In this way, it is prevented that the typical drift of the bias variables of curvature, yaw rate and lateral acceleration leads to the state of the non-active calibration function becoming increasingly less realistic, thereby avoiding a temporary increase in the lateral position error and a reduction in driving comfort when entering the adjacent speed range, and achieving adaptation to the current bias situation.

[0064] Adapting the integral value of one or more integral functions, whose current speed is outside their respective speed interval, in particular at a constant rate or a rate dependent on the vehicle speed, to the integral value of the calibration function, whose current vehicle speed is within its speed interval, can be seen as implementing a forgetting factor. In this case, the state of the inactive calibration function is close to the state of the currently active calibration function, for example with a uniform or individually predefined gradient for each respective speed range.

[0065] A control device according to the invention for generating control variables for at least one lateral control actuator of a vehicle is proposed, wherein the control device is configured to execute a method according to any of the above claims. The control device can also be configured to implement lateral control, to determine a target trajectory and / or to execute a curve pilot control. As an alternative, the control device can only implement some of the functions. The control device can be connected to at least one sensor of the vehicle, in particular to at least one environmental sensor for at least partially detecting the vehicle environment and / or at least one speed sensor for detecting the vehicle speed and / or a vehicle navigation system.

[0066] Provision is made for a vehicle according to the invention to include at least one lateral control actuator and a control device according to the invention, wherein the control device is configured to control the at least one lateral control actuator using a control variable. The lateral control actuator may be, for example, a steering actuator, such as an electric servo drive for a front axle steering system and / or a rear axle steering system.

[0067] It is provided that the computer program according to the invention comprises instructions which enable a control unit to carry out the method according to the invention. The control unit may in particular be a control unit which can be connected to at least one lateral control actuator of the vehicle.

[0068] All advantages and configurations described above in connection with the method according to the invention also apply to the control device according to the invention, the vehicle according to the invention and the computer program according to the invention, and vice versa.

[0069] Further advantages and details of the present invention can be obtained from the embodiments described below and from the accompanying drawings. The accompanying drawings are schematic:

[0070] Figure 1 An embodiment of a motor vehicle according to the invention is shown,

[0071] Figure 2 A block diagram showing an embodiment of the method of the present invention for vehicle tracking control,

[0072] Figure 3 A block diagram illustrating various calibration functions of an embodiment of the method of the present invention.

[0073] Figure 1 An embodiment of a vehicle 1 is shown. The vehicle 1 can be a motor vehicle, such as a passenger car or a commercial vehicle. Furthermore, the vehicle 1 can also be a vehicle combination, which has, for example, a tractor and one or more trailers. Alternatively, the method can also be used for other types of vehicles 1.

[0074] The vehicle 1 comprises a control device 2 which is configured to execute a method for tracking control of the vehicle 1. The vehicle 1 also comprises at least one lateral control actuator 3, by means of which the lateral position of the vehicle 1 can be adjusted.

[0075] The lateral control actuator 3 is designed, for example, as a steering actuator, by means of which the front axle steering angle of the vehicle 1 can be automatically adjusted. The steering actuator can be designed, for example, as an electric servomotor. As an additional measure or alternative to this, the lateral control actuator 3 can also be a steering actuator for automatically adjusting the rear axle steering angle.

[0076] The control device 2 is configured to determine a control variable by means of a transverse controller as a function of a transverse position error, and to adjust the transverse position of the vehicle 1 to a target trajectory 5 as a function of the control variable, the transverse position error describing a transverse deviation 4 of the vehicle 1 relative to a predetermined target trajectory 5 (indicated by a dashed line). To this end, the control device 2 controls a transverse control actuator 3 with the control variable. In this case, the transverse deviation 4 describes a deviation between the target trajectory 5 based on the vehicle center of the vehicle 1 and the actual direction of movement 6 of the vehicle center in the transverse direction of the vehicle, for example.

[0077] Here, if Figure 1 As shown, the deviation from the target trajectory can be determined at a position at a predetermined distance from the front end of the vehicle 1 (forward length) or at other reference points. Other possible reference points are, for example, at the height of the rear axle or the front axle of the vehicle 1. These reference points are preferably located on the longitudinal axis of the vehicle, but other reference points of the vehicle 1 that are fixed during the observation are also conceivable.

[0078] Figure 2 Shown is a block diagram of an embodiment of an automatic tracking control method for a vehicle 1 executed by a control device 2. The automatic tracking control includes a lateral controller 7, which adjusts the lateral position of the vehicle 1 to maintain a target trajectory 4. The lateral controller 7 generates a control variable delta_target for a lateral control actuator 3, which can be controlled directly or via an interposed steering angle controller (not shown), for example.

[0079] The lateral controller 7 comprises a cruise control 8 for generating a control variable delta_soll, a curve pilot control 9, one or more disturbance compensators 10, for example for compensating for side winds, road inclination and / or other influences, and a steering angle offset calibrator 11. The control variable component delta_regler is determined by means of the cruise control 8.

[0080] The curve pilot control device 9 determines a pilot control variable delta_vorst based on the measured and / or predicted curvature value of the lane on which the vehicle 1 is traveling, wherein the control variable delta_soll includes the pilot control variable delta_vorst as a control variable component. Accordingly, one or more disturbance compensators 10 generate one or more control variable components delta_komp and a control variable component delta_offset is generated by a steering angle offset calibrator 11.

[0081] Furthermore, at least one calibration function 12 is provided for determining a calibration value delta_kalib. In this case, the calibration value delta_kalib is applied to the control variable delta_soll generated by the transverse controller 7. The yaw rate g and the lateral acceleration a_lateral of the vehicle 1 are transmitted as input variables to at least one disturbance compensator 10 via a signal processor 13. For this purpose, the signal processor 13 can evaluate measured values ​​acquired by means of various sensors of the vehicle 1, which describe the yaw rate and / or the lateral acceleration.

[0082] The lateral position error delta_y is provided by the position recognition and trajectory planning unit 14 to the calibration function 12 and the cruise device 8. Here, the lateral position error delta_y describes Figure 1 Shown is a lateral deviation 4 between a target trajectory 5 and an actual direction of movement 6 or actual position of vehicle 1 .

[0083] In addition, the position recognition and trajectory planning unit 14 provides the curve pilot control device 9 with at least one measured and / or predicted curvature information kappa, which describes the curvature of the target trajectory 5 of the section of the target trajectory 5 located in front of the vehicle 1 in the driving direction. The position recognition and trajectory planning unit 14 determines the target trajectory 5 and the corresponding curvature information kappa based on environmental data, which are provided by the environment detection unit 15 of the vehicle 1. The environment detection unit 15 generates environmental data based on the environmental detection data, and the environmental detection data is generated in particular by at least one environmental sensor (not shown) that detects at least a local environmental area of ​​the vehicle 1.

[0084] The functions of modules 8 to 15 can be implemented in full or in part in the controller 7. The output variables of modules 8 to 11 represent all components of the control variable delta_soll. In addition, at least one calibration value delta_kalib of a calibration function 12 is applied to the control variable delta_soll.

[0085] The independent components of the control parameter delta_soll and at least one calibration value delta_kalib have different effects on the tracking control of the vehicle 1, which are described in detail as follows.

[0086] The target steering angle component delta_vorst of the curve pilot control device 9 is calculated, for example, as:

[0087] (1)delta_vorst=kappa*(l+EG*v^2),

[0088] The parameter l corresponds to the wheelbase, and EG corresponds to the self-steering gradient of vehicle 1. The self-steering gradient depends on

[0089] (2)EG=m*(ch*lh-cv*lv) / (ch*cv*(lh+lv))

[0090] The distance lh of the center of gravity from the front axle or rear axle of the vehicle 1, the vehicle mass m of the vehicle 1, and the cornering stiffness cv of the front tires and the cornering stiffness ch of the rear tires.

[0091] If the curvature information kappa now has an offset kappa_offs compared to the actual curvature due to an error, the target steering angle of the pilot control also contains an offset component delta_vorst_kappaoff, which is obtained from the following formula:

[0092] (3)delta_vorst_kappaoff=kappa_offs*(l+EG*v^2)

[0093] The result of the control value delta_vorst including the error delta_vorst_kappaoff is that it more or less interferes with the lateral controller 7, so that when the lane keeping assist is activated, the vehicle 1 does not travel along the target trajectory 5 as expected, but, depending on the implementation of the cruise control 8 - that is, the type of regulator used has or does not have steady-state accuracy - there is a permanent lateral deviation 4 from the planned target trajectory 5, or at least causes a momentary steering, and the vehicle 1 then travels along the planned target trajectory 5 only after the compensation process has subsided.

[0094] The camera device used as an environmental sensor provides environmental measurement data to the position recognition and trajectory planning unit 14. The degree of imaging error of the camera device usually changes during the operation time of the driver assistance system, because the calibration routine of the camera device is generally carried out continuously. This means that, depending on the calibration state, the magnitude of the offset of the curve pilot control device 9 and thus the degree of influence on the lateral control of the vehicle changes over time, although the transmission characteristics of the camera device and thus the curvature offset generally do not change suddenly in a step, but change within a limited range.

[0095] It is possible to use the steering angle as an auxiliary variable or control variable in a traction control driver assistance system. The steering angle can be influenced or corrected by a steering angle calibration routine 11. Due to the limited accuracy of the steering angle calibration routine 11, the steering angle offset delta_offset determined here usually has an uncompensated component.

[0096] Here, the residual offset error delta_offset_error in the steering angle signal can have various causes. Typically, the steering angle calibration routine 11 is largely based on the signal of the yaw rate sensor of the vehicle 1 and / or the signal of the wheel speed sensor of the vehicle 1. Therefore, errors in these sensors, such as offset errors and linearity errors, are propagated into the determined steering angle offset. The influence of the steering angle offset is similar to the above-mentioned curvature offset case, where the vehicle 1 does not follow the target trajectory 5 as expected when the tracking control driver assistance system is activated, but, depending on the design of the cruise control device 8, has a permanent or instantaneous lateral deviation 4 from the planned target trajectory 5, which can have an adverse effect on the comfort and acceptability of the driver assistance system.

[0097] For example, if a proportional derivative (PD) control device is used as the cruise device 8, and the transfer function

[0098] (4)Gr(s)=Kp(v)+s*Kd(v) / (1+s*T),

[0099] Kp(v) and Kd(v) are speed-dependent control coefficients, whereby the lateral position error delta_y describing the lateral deviation 4 is an input variable of the cruise control 8 and delta_regler is an output variable of the cruise control.

[0100] (5)delta_regler(s)=Gr(s)*delta_y(s).

[0101] Therefore, in a steady-state speed state during straight-line driving, that is, when the physical steering angle is zero degrees, the cruise control 8 outputs an uncompensated steering angle offset delta_offset_error, ignoring the lateral disturbance forces acting on the vehicle 1. The lateral deviation 4 required for this is obtained in the time range by the following formula:

[0102] (6)Delta_y(0)*Gr(0)=Delta_offset_error(0)

[0103] arrive

[0104] (7)delta_y=delta_offset_error / Kp(v)

[0105] The control factor Kp cannot be chosen to be arbitrarily large during design, so that when the uncompensated steering angle offset delta_offset_error is not equal to zero, a lateral deviation 4 of the vehicle 1 relative to the target trajectory 5 always occurs. In this case, in the case of rear-axle steering of the vehicle 1, the influence of the uncompensated rear-axle steering angle offset on the lateral deviation 4 of the vehicle 1 is similar to that of the uncompensated front-axle steering angle offset.

[0106] The sensor signals of the yaw rate, the front axle steering angle, the rear axle steering angle and the lateral acceleration are also used in particular to estimate disturbance forces and load moments on the vehicle 1, for example due to a lane inclination and / or a crosswind acting on the vehicle 1. The determined forces and moments then proportionally / partially influence the steering angle target value delta_target during the disturbance variable application delta_target that is carried out.

[0107] Furthermore, these sensors can also be used to estimate the slip angle of the vehicle 1, which can also be used as an auxiliary variable for load estimation. Bias errors in the yaw rate g, the steering angle or the lateral acceleration a_lateral usually have an effect on the estimated variables and thus also on the application of the disturbance variable in a steady-state state in a proportion to the corresponding bias. Bias errors in the yaw rate g, the steering angle and the lateral acceleration a_lateral are therefore propagated to the disturbance compensation component of the steering angle target value delta_soll. As a result, the tracking controlled vehicle 1 has a transient lateral deviation or a permanent lateral deviation in an undesirable manner relative to the planned target trajectory 5. The disturbance compensator 10 for the yaw rate g and the lateral acceleration a_lateral usually has a limited accuracy similar to the steering angle compensator 11 and can therefore lag behind the actual bias, so that even if a disturbance compensator is present, the above-mentioned adverse effects on the driving comfort of the vehicle 1 when the tracking controlled driver assistance system is used.

[0108] This problem can be solved or at least partially compensated by at least one calibration function 12, thereby improving the driving comfort of the vehicle 1. The calibration function 12 is superimposed on the lateral controller 7 and generates at least one calibration value delta_kalib, which is applied to the control variable delta_soll. The calibration value delta_kalib is determined by at least one calibration function 12 at least partially as a function of the time integral of the lateral position error delta_y.

[0109] Depending on the sign and amplitude of the lateral position error delta_kalib, a calibration value delta_kalib for the target steering angle delta_soll is determined by the calibration function 12, which calibration value is always used to reduce the lateral position error delta_y, especially in terms of the direction of action. Then, this calibration value delta_kalib is finally superimposed on the control variable delta_soll from the cruise control 8, from the curve pilot control device 9, from one or more disturbance compensators 10 and from the output of the steering angle offset calibration device 11, and is transmitted as a calibrated control variable delta_soll_kalib, for example, to the target steering angle interface of the lateral control actuator 3. As an alternative, the calibrated control variable can also be transmitted to the torque interface of the lateral control actuator 3.

[0110] The calibration function 12 can be designed as an integrator or an integration unit, for example, with a relatively small integration coefficient ki, so that the calibration function 12 integrates the lateral position error with low dynamics. As an alternative, the calibration function can also be implemented as a PT1 unit or other types of functions with at least partial integration characteristics.

[0111] Thus, in the case where the calibration function 12 is implemented as an integrator with an integration coefficient ki, the calibration value delta_kalib can be determined by the calibration function 12, for example, as follows:

[0112] (8) Delta_kalib = Delta_y*ki / s

[0113] In this case, the time constant of the calibration function 12 is in particular at least 2 times, at least 5 times, at least 10 times, at least 100 times or at least 1000 times the main time constant of the lateral controller 7 or the cruise control device 8. For this purpose, the integral coefficient ki is of the order of magnitude of 0.001 degrees / (meter*second), so that, for the sake of illustration, in the example of a hypothetical lateral position error delta_y of 0.2 meters, the gradient of the calibration value 12 is only 0.0002 degrees / second.

[0114] The dynamics of the adaptation of the calibration value 12 is preferably set to the same order of magnitude as the above-mentioned dynamics of the change of the bias error. The value used for this is, for example, the maximum, expected, temperature-dependent bias drift of the sensor used for the lateral controller 7, which is converted, for example, to the corresponding steering angle level. If the dynamics selected are significantly greater, typical overshoots will occur in the signal of delta_kalib, because the calibration would then partially conflict with the transient behavior of the compensation function 10 for road inclination or side wind. However, this would also have a limiting effect on the comfort improvement achieved by the calibration function 12.

[0115] The calibration function 12 determines the integral value of the calibration function 12 by integrating the lateral position error delta_y over time. The integral value represents the result of the time integration of the lateral position error delta_y up to a specific time point. When the tracking control is not working, the integral value is retained, and when the tracking control is reactivated, the integral value is used as the starting value for re-integrating the lateral position error delta_y over time. The state of the calibration function 12, which is designed as an integrator, for example, is not reset in the current ignition operation, but remains unchanged at the end of the trip and / or when the driver intervention is covered by the automatic tracking control even when the driving assistance function is repeatedly stopped and activated.

[0116] Since the influence of the bias variable on the lateral position error delta_y is dependent on the driving speed, the use of a constant integration factor ki in the calibration function 12 makes concessions over the entire achievable speed range of the vehicle 1. This can be due, for example, to the usual control parameter control of the cruise control 8 as a function of the driving speed. This also changes the disturbance compensation characteristics of the cruise control 8 and thus also the influence of the sensor bias variable on the lateral position error delta_y.

[0117] In order to take this into account and expand the potential of the calibration function, it can be provided that the time integration of the lateral position error delta_y is carried out with a weighting factor, in particular an integration factor ki that is dependent on the vehicle speed. The integration factor ki can thus be dependent on the current speed of the vehicle 1 as ki(v). This advantageously enables the calibration function 12 to have an ideal, individually adaptable calibration dynamics for each speed range.

[0118] As an additional measure or alternative, it can also be provided that the weighting factor or integration factor ki increases with the magnitude of the lateral position error delta_y. Here, the calibration speed of the calibration function 12 can be effectively increased by additional nonlinear or progressive weighting of the magnitude of the lateral position error delta_y.

[0119] Here, the integral coefficient ki and thus the adaptation rate are switched between two or more values, for example, wherein the integral coefficient ki is selected to be higher when the absolute value of the lateral position error delta_y is higher. For example, when the driver assistance function is activated, the integral constant ki is doubled for the case where the magnitude of the lateral position error delta_y is greater than 0.2 meters, which can achieve faster initial adaptation to the existing bias situation.

[0120] The limitation of the adaptation rate of the calibration function 12 as a function of the magnitude of the lateral position error delta_y ensures a faster adaptation to new bias conditions even if the sensor compensator 10 and / or other sensor bias calibration functions in the vehicle 1 eventually eliminate known sensor bias drifts by means of a step-shaped correction of the corresponding bias after a debounce time. This is particularly advantageous for relatively long debounce times of the sensor compensator 10 or other sensor bias calibration functions, since during such debounce processes the compensations incrementally performed by the calibration function 12 must be re-integrated after a signal update of the sensor compensator or sensor bias calibration function.

[0121] Since a step-shaped correction of the sensor bias directly leads to a step-height-dependent lateral jerk of the vehicle 1, it can be provided as an advantageous development that all sensor bias calibration functions in the vehicle 1 do not change the corresponding sensor bias in a step-shaped manner, but rather within a limited range. If, for example, the offset compensation of the vehicle sensors occurs at most at a defined maximum gradient, and the output delta_kalib of the center calibration can always follow this gradient with a small tracking error of the order of 0.03 degrees, then a nonlinear or progressive weighting of the magnitude of the lateral position error delta_y in the calibration function 12 can also be omitted because of the increased dynamics.

[0122] The value of delta_kalib for the total controlled variable delta_soll_kalib can also be further limited to a maximum amplitude in an advantageous manner by limiting the integrator state in the calibration function 12. To this end, it can be provided that the time integration of the lateral position error delta_y by the calibration function 12 is stopped if the integral value of the calibration function 12 determined by the time integration of the lateral position error delta_y corresponds to a predetermined limit value.

[0123] In this case, the limit value or maximum amplitude of the integrator or calibration function 12 is based, for example, on the sum of the maximum influences of all offset variables on the steering angle target value delta_set and the steering angle actual value. If delta_komp, delta_vorst and delta_offset each have a maximum offset error of, for example, 0.05 degrees, then delta_kalib must be limited to an interval of + / -0.15 degrees. The angle specification here refers to the vehicle steering angle, for example the steering angle of the vehicle's front wheels. With this method step, further measures that must be taken in the calibration function 12 to avoid integrator saturation effects can also be advantageously omitted.

[0124] As an alternative to limiting the integrator state, a low-pass filter function, such as a PT-1 filter, can also be used. Saturation effects can also be avoided in this way. Although offset conditions cannot be completely eliminated, sufficient centering accuracy can still be achieved.

[0125] When driving on a curve, the lateral position error caused by the error of the pilot control is not only related to the bias error of the curvature, but also to the inaccuracy of the stored vehicle parameters, especially the uncertainty of the self-steering gradient EG. By fully differentiating equation (1) and evaluating EG or its error EG_err, the error component of the pilot control caused by the error of the self-steering gradient EG can be obtained, that is,

[0126] (8) delta_vorst_EG_err= (d delta_vorst / d EG) * EG_err

[0127] =(d(kappa*(l+EG*v^2)) / d EG)*EG_err

[0128] =kappa*v^2*EG_err

[0129] The magnitude of delta_vorst_EG_err has a disturbing effect on the lateral control of the vehicle 1 and proportionally causes a lateral position error delta_y, which results in an error-containing deviation of delta_kalib. Therefore, the delta_kalib signal does not only react to the sensor bias parameter. However, as long as EG_err is limited, the influence of EG_err on the vehicle level is small due to the large integration time constant of the center calibration. Here, the integration time constant is defined as the reciprocal of the integration factor ki.

[0130] This also applies, since the erroneous deflection of delta_kalib in the curve is corrected on the straight section of the lane. In the case of large changes in EG_err, as compensation for the value of EG in the vehicle, the adaptation algorithm is constantly adapted to the current situation, such as the vehicle mass m or the tire stiffness, which reduces EG_err and thus the value of delta_vorst_EG_err.

[0131] Furthermore, in the case of large changes in EG_err, the calibration function 12 can be suspended at high values ​​of the product kappa*v^2. It can therefore be provided that the time integration of the lateral position error delta_y by the calibration function 12 is stopped if the product of the square of the current vehicle speed v and the curvature value kappa of the target trajectory 5 and / or the lane curvature value exceeds a predetermined threshold value.

[0132] For this purpose, it is advantageous to first low-pass filter the curvature kappa, for example by means of the PT-1 algorithm, and then compare the absolute value with a threshold value that depends on the vehicle speed v. If the filtered curvature is above this threshold value, the calibration function 12 is temporarily stopped.

[0133] Furthermore, it may be provided that the time integration of the lateral position error delta_y by the calibration function 12 is stopped if a measured and / or predicted curvature value kappa of the lane in which the vehicle 1 is travelling exceeds a predetermined threshold value and / or if a curvature value of the target trajectory 5 exceeds a predetermined threshold value.

[0134] If the curvature itself temporarily has a greater offset error, this can temporarily reduce the calibration work of the calibration function 12 at higher driving speeds until the calibration of the camera device compensates for this error. However, the adaptation range only shifts the range of the curvature offset, so that the calibration work is not completely stopped. In addition, the calibration work is carried out again more intensively when the driving speed v decreases and thus the product kappa*v^2 decreases.

[0135] Disturbances in the calibration can also be caused in particular by an intervening driver, who can, for example, directly influence the lateral position error delta_y by turning the steering wheel of the vehicle 1. For compensation, the calibration function 12 can be stopped, i.e. the integration of the lateral position error delta_y can be stopped, when the driver intervenes in the steering wheel. In particular, it is possible to stop the time integration of the lateral position error delta_y and / or abandon the application of the calibration value delta_kalib to the control variable delta_soll if the steering torque M caused by the steering intervention by the driver of the vehicle 1 is above a predetermined threshold value and / or causes an increase in the current lateral position error delta_y.

[0136] However, by stopping the calibration function 12 only when the driver steers in the direction of increasing the lateral position error delta_y, a greater availability of the calibration function 12 and thus a faster convergence can be achieved. This can be achieved by comparing the signs of the driver's steering torque M and the lateral position error delta_y. For example, if the direction of the driver's steering torque M is defined in such a way that steering to the left in the direction of travel produces a positive torque, and a positive lateral positioning error means that the vehicle is positioned to the right of the planned target trajectory 5 in the direction of travel, then as a stop condition for the calibration function, the signs of the driver's steering torque M and the lateral position error must be different.

[0137] Initially, for the implementation of lane assistance, for driver steering interventions below a defined intensity, the integral coefficient ki is evaluated, for example, as a basis for design according to the level of the measured driver steering torque M. As an additional measure or alternative, it can also be provided that when the steering torque M generated by the steering intervention of the driver of the vehicle 1 reduces the current lateral position error delta_y, the lateral position error delta_y is time-integrated with an increased weight, in particular an increased integral coefficient, at least over a predetermined time period. Thus, compared to the basic design, the integral coefficient is specifically increased when the driver intervention is identified and the driver simultaneously turns in the direction of reducing the lateral position error delta_y. Here, the driver's steering can be regarded as a confirmation of the calibration process of the calibration function 12, which means that it is reasonable to increase the calibration dynamics by increasing the integral coefficient ki.

[0138] As already mentioned above, the effect of the bias variables on the lateral position error delta_y is partly dependent on the driving speed v. This applies, for example, to the bias error contained in the output of the disturbance compensator delta_komp or as part of the curve pilot control delta_vorst. Therefore, when the driving speed changes, delta_kalib must also be constantly changed in order to keep the lateral position error low, although the bias variables causing the lateral position error from the yaw rate g, the lateral acceleration a_lateral and the pilot curvature kappa are assumed to be constant in the analysis. Due to the high integration time constant of the calibration function 12, the dynamic performance of the calibration function 12 is low and the adaptation to the changing driving speed v requires a long time, during which the accuracy of the tracking control is reduced.

[0139] To take this into account, for example Figure 3 As shown, it is beneficial to use a combination of multiple integrators or memories over the entire vehicle speed range rather than a single integrator as the memory for delta_kalib.

[0140] Figure 3 The use of a plurality of integration functions 12_1 to 12_N is shown. Furthermore, an interpolation unit 16 , an initialization unit 17 , an integration coefficient determination unit 18 and a sequence control unit 19 are provided as part of the control device 2 .

[0141] The integration functions 12_1 to 12_N are each assigned to a different speed interval, wherein the integration functions 12_1 to 12_N each perform a time integration of the lateral position error delta_y at the speed of the vehicle 1 within the speed interval assigned to them. The entire vehicle speed range is thus divided into N>1 sub-intervals, each of which covers only a limited speed range, and a calibration function 12 or a dedicated integrator is assigned to each sub-interval. As a result, only a single calibration function 12 is always working at a point in time, while the remaining calibration functions 12 are stopped or switched to a zero input signal at their input. On the other hand, the correspondingly activated calibration function, i.e. the calibration function 12_1 to 12_N, provides a weighted lateral position error delta_y, and each integrator generates its own delta_kalib_v(k), k=[1...N]. The calibration values ​​delta_kalib_v(1) to delta_kalib_v(N) of the corresponding calibration functions 12_1 to 12_N for the speed interval including the current speed of the vehicle 1 can be directly applied to the control variable delta_soll as the resulting delta_kalib. The weighting of the lateral position error delta_y can be performed using a constant integration factor ki or a factor ki(v) that is dependent on the driving speed.

[0142] Integral coefficients ki or ki(v) can be determined for each calibration function 12_1 to 12_N by an integral coefficient determination unit 18. To this end, the integral coefficient determination unit 18 can receive, for example, the vehicle speed v, the steering torque M of the driver of the vehicle 1, the current lateral position error delta_y and the measured or predicted curvature kappa as input variables.

[0143] The centers of the N driving speed intervals represent, to a certain extent, the support points of the output value delta_kalib_v(k). In order to obtain the result delta_kalib from the individual values ​​delta_kalib_v(k), in addition or alternatively, a total calibration value delta_kalib_ges determined from the calibration values ​​delta_kalib_v(1) to delta_kalib_v(N) by the interpolation unit 16 can be applied to the control variable delta_soll. Here, the total calibration value delta_kalib_ges is determined by two or more calibration values ​​delta_kalib_v(1) to delta_kalib_v(N) generated by the integration function 12, in particular so that the total calibration value delta_kalib_ges is also continuous in time even when switching between two or more speed intervals. This can be achieved by the interpolation unit 16, for example by linear interpolation, wherein the interpolation unit receives the vehicle speed v as input variable.

[0144] For this purpose, for example, delta_kalib_ges is obtained from the linear combination of delta_kalib_v(k) and delta_kalib_v(k+1) based on the two distances between the current driving speed and the two centers of the adjacent speed ranges.

[0145] (9)delta_kalib_ges=delta_kalib_v(k)*a+delta_kalib_v(k+1)*(1-a)

[0146] in

[0147] a=(vv(k)) / (v(k+1)-v(k)),v>v(k),v <v(k+1)

[0148] Interpolation between the integrator states delta_kalib_v(k) is also advantageous because no delta_kalib_ges jumps occur when changing from one speed range to another. Jumps in the steering angle target value delta_soll_kalib would lead to lateral jerking of the vehicle 1, which must be avoided for reasons of driving comfort. Using a set of calibration functions or a set of integrators from 12_1 to 12_N is particularly advantageous when accelerating or braking, whereby the speed range changes frequently, compared to using a single integrator.

[0149] If the driver assistance function is mainly operated within a narrow speed range, it may happen that only one calibration function 12 acts accordingly on the lateral position error delta_y, so that only this calibration function 12 correctly reflects the current bias situation. Due to the typical drift of the bias variables of the curvature kappa, the yaw rate g and the lateral acceleration a_lateral, the states of the other integrators become less and less realistic.

[0150] When entering an adjacent speed range, a larger lateral position error delta_y may temporarily result, and driving comfort may be reduced at the same time, until the adaptation to the current bias situation is completed. As compensation, a forgetting factor can be implemented in a beneficial manner. To this end, it can be provided that the integral values ​​assigned to the calibration functions 12_1 to 12_N are determined by the calibration functions 12_1 to 12_N respectively by time integration of the lateral position error, wherein the integral values ​​of one or more integral functions whose current speed is outside the corresponding speed interval are continuously adapted, in particular using a constant rate or a rate with the vehicle speed, so that they match the integral values ​​of the calibration functions 12_1 to 12_N corresponding to the speed interval in which the current vehicle 1 speed is located.

[0151] In this case, the goal is to bring all integrator states or integrated values ​​of, for example, inactive calibration functions 12_1 to 12_N to the states or integrated values ​​of currently active calibration functions 12_1 to 12_N with a uniform or individually specified gradient for each speed interval. This gradient can be based, for example, on the dynamics of the change of the bias variable. In application, it can be advantageously quantified to correspond to 25% of the gradient to be adjusted for delta_kalib_v(k) for a lateral position error of 0.2 meters. If the gradient to be compensated is chosen too high, the advantage of using a group of calibration functions 12_1 to 12_N over using a single calibration function 12 is reduced.

[0152] As an alternative to the gradient method described above, the calibration function of the currently irrelevant speed range can be continued with an integration factor reduced to, for example, 10% to 20%. In this method, the integrator state outside the currently operating speed range can be avoided by corresponding limiting measures.

[0153] Depending on the dynamics of the bias influence, it may be advantageous to write the state of a single calibration function 12 or all states of a plurality of calibration functions 12_1 to 12_N to a non-volatile memory, for example a memory of the control device 3, so that when the ignition of the vehicle 1 is restarted, the calibrated state of one or more calibration functions 12 or 12_1 to 12_N can be set. This is particularly advantageous if most of the bias errors have only a small time drift or no time drift.

[0154] The storage of the states or integration values ​​of one or more calibration functions 12_1 to 12_N can be performed by the initialization unit 17. Furthermore, the initialization unit 17 can also initialize the calibration functions 12_1 to 12_N and perform state balancing between the calibration functions 12_1 to 12_N or their current integration values.

[0155] The process control unit 19 can, for example, take into account the activation and / or deactivation of automatic tracking control, the conditions for stopping time integration and / or the conditions for applying at least one calibration value delta_kalib to the control variable delta_soll, the change between speed intervals and / or the initiation of other events and accordingly run at least one calibration function 12.

[0156] The calibration function 12 can also be regarded as an integral component in the cruise control device 8 in terms of its basic function, but preferably has essential characteristics or essential differences, namely, the integral coefficient or other weighting factors for the time integral of the lateral position error delta_y are very low, which is different from the typical design criteria of the lateral controller 7; and / or the calibration function 12 or the time integral of the lateral position error delta_y is not initialized with zero when the assistance system is restarted, but is set to the last state reached or takes over this state. This has the advantage that when the driver assistance function is activated, the vehicle 1 has almost no lateral oscillation process, but can directly drive along the planned target trajectory 5. The expansion of the at least one calibration function 12 mentioned above also reflects other differences from the classical integral part of the lateral control device.

[0157] The at least one calibration function 12 can be implemented as a separate function or, as an alternative, can also be implemented in addition to an existing integral part in the cruise control device 8. In this case, in particular, the at least one calibration function 12 is adapted only during the operating phase of the driver assistance system until the integral part of the cruise control device 8 reduces the lateral position error delta_y to zero, or the at least one calibration function is adapted always when the integral part of the cruise control device 8 is reset and must be rebuilt. This occurs, for example, in the case of a Level 2 autonomous driving system when the driver is steering.

[0158] In summary, the calibration dynamics of the calibration function 12, i.e. the time required for the lateral deviation 4 of the vehicle 1 to adapt to the current bias situation, may be greater than in the case of a driver assistance system without an integral part in the cruise device 8. When one or more calibration functions 12 are used together with a cruise device 8 without an integral part, this situation can be at least partially addressed by increasing one or more integral coefficients or integral values. Furthermore, when using the integral part in the cruise device 8, one or more disturbance compensators 10 for side wind and / or lane inclination can be omitted. As a result of this, an important source of bias errors is also eliminated, whereby the bias effects that need to be taken into account by the calibration function 12 are also reduced, which is advantageous in reducing the requirements on the calibration dynamics of the at least one calibration function 12.

Claims

1. A method for tracking a vehicle (1), wherein: A control parameter is determined by a lateral controller (7) based on a lateral position error, the lateral position error describing a lateral deviation (4) of the vehicle (1) relative to a predetermined target trajectory (5), and the lateral position of the vehicle (1) is adjusted to the target trajectory (5) based on the control parameter. It is characterized in that At least one calibration value is additionally applied to the control variable, wherein the calibration value is determined by at least one calibration function (12) at least partially as a function of the time integral of the lateral position error.

2. The method according to claim 1, It is characterized in that The time constant of the calibration function (12) is greater than a main time constant of the lateral controller (7), in particular at least twice the main time constant.

3. The method according to claim 1 or 2, It is characterized in that The integral value of the calibration function (12) is determined by the time integration of the lateral position error, wherein the integral value is retained when the tracking control is not in operation and is used as a starting value for the time integration of the lateral position error when the tracking control is reactivated.

4. A method according to any one of the preceding claims, It is characterized in that The lateral controller (7) comprises a pilot control device (9), wherein the pilot control device (9) determines a pilot control variable based on measured and / or predicted curvature values ​​of a lane in which the vehicle (1) is traveling, wherein the pilot control variable is included as a component in the control variable.

5. The method according to claim 4, It is characterized in that The pilot control is performed based on a vehicle model that describes a self-steering gradient of the vehicle, wherein the self-steering gradient is adapted during operation of the vehicle as a function of at least one vehicle parameter, in particular the vehicle mass and / or the tire rigidity.

6. A method according to any one of the preceding claims, It is characterized in that The time integration of the lateral position error is performed with a weighting factor, in particular an integration factor, which is dependent on the speed of the vehicle (1) and / or increases with increasing magnitude of the lateral position error.

7. A method according to any one of the preceding claims, It is characterized in that exist - the integral value of the calibration function (12), determined by the time integration of the lateral position error, corresponds to a predetermined limit value, - a measured and / or predicted curvature value of the lane on which the vehicle (1) is travelling exceeds a predetermined threshold value, - the curvature value of the target trajectory (5) exceeds a predetermined threshold, - the product of the square of the current vehicle speed and the lane curvature value exceeds a predetermined threshold, and / or - If the product of the square of the current vehicle speed and the curvature value of the target trajectory (5) exceeds a predetermined threshold, the time integration of the lateral position error by the calibration function (12) is stopped.

8. A method according to any one of the preceding claims, It is characterized in that If the steering torque generated by the steering intervention of the driver of the vehicle (1) exceeds a predetermined threshold value and / or the generated steering torque increases the current lateral position error, the time integration of the lateral position error is stopped and / or the application of a calibration value to the control variable is abandoned.

9. The method according to any one of the preceding claims, It is characterized in that If a steering torque resulting from a steering intervention by the driver of the vehicle (1) reduces the current lateral position error, the lateral position error is time-integrated with increased weighting, in particular an increased integration factor, at least over a predetermined period of time.

10. The method according to any one of the preceding claims, It is characterized in that A plurality of integral functions (12) are used, wherein the integral functions (12) are each assigned to a different speed interval, wherein the integral functions (12) respectively perform a time integration of the lateral position error when the speed of the vehicle (1) is within the speed interval assigned thereto, wherein a calibration value of the integral function (12) whose speed interval includes the current vehicle (1) speed, or a total calibration value determined based on the calibration value, is applied to the relevant control variable.

11. The method according to claim 10, It is characterized in that The total calibration value is determined from two or more calibration values ​​generated by the integration function (12), so that the total calibration value is continuous in time even when switching between two or more speed intervals.

12. The method according to claim 10 or 11, It is characterized in that The calibration function (12) is used to determine the integral value of the calibration function (12) associated with the calibration function (12) by respectively integrating the lateral position error over time, wherein the integral value of one or more integral functions whose current speed is outside the respective speed intervals is continuously adapted, in particular at a constant ratio or a ratio related to the vehicle speed, to the integral value of the calibration function (12) whose current speed of the vehicle (1) is within its speed interval.

13. A method according to any one of the preceding claims, It is characterized in that At least one calibration function (12) is implemented as an integration unit or a PT1 unit, respectively.

14. A control device for generating a control variable for at least one lateral control actuator (3) of a vehicle (1), wherein: The control device (2) is configured to carry out the method according to one of the preceding claims.

15. A vehicle comprising at least one lateral control actuator (3) and a control device (2) according to claim 14, wherein: The control device (2) is provided for actuating the at least one lateral control actuator (3) using a control variable.

16. A computer program comprising instructions for causing a control device to execute the method according to any one of claims 1 to 14.

Images