Method for controlling a vehicle

Abstract

The invention relates to a method for controlling a vehicle (1) with at least one actuator along a trajectory, wherein the trajectory is planned within a search space taking into account a preview of at least one control variable of the actuator, the method comprising the following method steps: establishing an actuator model from the at least one control variable of the actuator; setting a time step of the preview; determining a change in the control variable of the actuator along the time step from the actuator model and a limit value of the control variable; limiting the search space from the limit value of the control variable of the actuator; determining an acceleration value and / or a deceleration value of the vehicle (1) by transforming the at least one control variable together with the mass of the vehicle and the wheel radius; and outputting the acceleration value and the deceleration value for limiting the search space for planning the trajectory.

Description

TECHNICAL FIELD

[0001] The present application relates to a method for trajectory planning of a vehicle, a control system arranged for implementing the trajectory planning according to the present application, a computer program comprising program code for implementing the method according to the present application, and a computer-readable storage medium which causes a computer to implement the method according to the present application when the computer-readable medium is run on the computer. BACKGROUND

[0002] Modern vehicles, such as passenger cars, heavy goods vehicles (trucks) or motorcycles, are increasingly equipped with driver assistance systems which can detect the surroundings and objects therein by means of sensors, identify traffic situations and support the driver, for example by a braking intervention or a steering intervention or by issuing a visual or acoustic warning. Radar sensors, lidar sensors, camera sensors or similar sensors are often used as sensors for detecting objects and / or the surroundings. From the sensor data determined by the sensors, conclusions about the surroundings can then be drawn. Here, the processed sensor information is used for surroundings recognition in order to issue, on this basis, a driver warning / driver information or an instruction for a controlled steering, braking and acceleration. By means of the assistance functions of the sensors and surroundings data processing, accidents with other traffic participants can be avoided, for example, or complex driving maneuvers are facilitated by supporting or even taking over the driving task or the vehicle guidance completely (in a semi-automated or fully automated manner). For example, the vehicle can carry out an autonomous emergency braking (AEB) by means of an emergency brake assist function (EBA), follow-up driving and speed control by means of an adaptive cruise control assist function (ACC), or keep the vehicle in the lane by means of an active lane-keeping function (LKA) with steering support. Furthermore, a plurality of these functions can also be combined in a system. In addition to emergency braking in dangerous situations, automatic braking interventions are also very important, in particular when driving a vehicle in an (fully) automated manner. In a critical traffic situation or traffic scenario, in particular when threatened by an impending collision, an automatic braking is initiated.

[0003] Control systems are usually used for the control of (semi-) autonomous driving, which use vehicle actuators such as steering, braking or power devices and comprise a trajectory planner (short: planner) for planning and selecting a trajectory and one or more controllers for implementing the trajectory. Here, the data flow control is basically implemented along the control chain "Sense-Plan-Act", i.e. from left to right. Here, the planner first processes information from the sensing part, i.e. surrounding data and target object data such as distances and velocities of vehicles driving ahead detected by means of surrounding sensors, but not information from the execution part such as current actuator states and actuator limits. Instead, the planner method is usually based on the simplification that no actuator limits occur and the dynamic behavior of the actuators is neglected. In fact, the only requirement on the actuators is a structure-determined dynamic behavior only within the actuator operating range after a specific actuator dead time, i.e. a time range between a change of the system input signal and a response of the system output signal of the control chain. If the planner does not have access to this information, it results in a planner planning an unfeasible trajectory. For example, the planned trajectory cannot be implemented with the dynamic behavior of the available actuators or is outside the operating range of the actuators. In this case, the controller behind the planning can either trigger a controller windup or a new planning. The windup is caused by a constant control deviation, e.g. a distance to a vehicle driving ahead that is lower than planned. In a transition from a trajectory that requires deceleration to a trajectory that requires acceleration (or vice versa), the windup results in an additional delay of the vehicle reaction. A reduction of the accuracy by a new planning that artificially reduces the control deviation can result in a high frequency of new planning by continuously violating the actuator limits, which leads to an increased phenomenon of a drift between the necessary trajectory and the actually driven trajectory.

[0004] Usually, generic power devices and brake systems do not provide feedback on actuator limits to the planner. If there are signals for actuator limits, these are often only valid for the current calculation cycle. For example, a power device control provides a maximum torque value and a minimum torque value of the power device (i.e. a maximum drag torque), but does not indicate when these maximum values can be achieved. Thus, the maximum torque gradient and thus the actual actuator potential cannot be predicted. The planner calculates an optimal trajectory with a time preview. If limits are to be introduced in the planner by means of formulation of boundary conditions, the limits must be provided with a time preview. The limits provided by the power device control at the current point in time are not sufficient. Therefore, actuator limits cannot be taken into account in the planner, which results in windup or new planning.

[0005] Furthermore, in addition to the missing prediction, the current power plant torque limits are not used in the planner, although the planner is able to take into account restrictions in the form of boundary conditions by means of optimization methods. The reason for this is that it is not possible to convert the actuator torque limits into vehicle acceleration limits. The planner calculates an optimized trajectory on the basis of the vehicle longitudinal acceleration. If the limits of the actuators only occur in the form of torque or axle torque, the planner cannot use them, since the planner does not include an actuator / vehicle model for converting the torque into acceleration. As a result, the limits of the actuators cannot be taken into account in the planner.

[0006] A vehicle control system with a plurality of actuators or actuators, such as steering devices, powertrains, service brakes and parking brakes, is known from DE 10 2016 221 723 A1. The control system comprises a module for controlling the movement of the vehicle, a module for controlling the actuators, a module for specifying a vehicle operating strategy to be implemented and a module for torque coordination, wherein a normalized requirement vector comprising a longitudinal component, a lateral component and a vertical component is formed from the movement requirements of the vehicle. Furthermore, the control system is configured to form the torques to be allocated to the actuators from the vehicle operating strategy and the requirement vector.

[0007] DE 10 2015 209 066 A1 describes a cost-reducing method for trajectory planning for a vehicle, wherein a search space for determining a trajectory is limited in accordance with an approximate end time point. In this case, the search space is limited to a specific range around the approximate end time point, in particular to a specific range of 10% around the approximate end time point, in order to reduce the computational effort required for determining the trajectory of the driving maneuver. SUMMARY

[0008] The task underlying the present application is therefore to provide a general method for controlling a vehicle, in which method the trajectory planning is improved in a simpler and more cost-effective manner.

[0009] In the method for controlling a vehicle along a trajectory according to the application, the vehicle comprises at least one actuator, preferably a plurality of actuators. Furthermore, the vehicle comprises a control device having a trajectory planner which plans a trajectory within a definable search space, taking into account a preview or prediction of at least one control variable of the actuator or controller. The method is particularly suitable for a method for planning a longitudinal trajectory (longitudinal trajectory planning). The method comprises the following method steps:

[0010] - establishing an actuator model of the actuator on the basis of at least one control variable of the actuator,

[0011] - setting a time step of the preview,

[0012] - determining a change of the actuator control variable along a time step depending on the actuator model and a limit value of the control variable,

[0013] - limiting the search space depending on a limit value of the actuator control variable,

[0014] - determining a vehicle acceleration value and / or a deceleration value by transforming at least one control variable with the vehicle mass and the wheel radius, and

[0015] - outputting the acceleration value and the deceleration value for limiting the planning trajectory or the search space of the trajectory planner for searching a possible trajectory.

[0016] The method according to the application has the advantage that by also taking into account the natural limits of the actuators, it is ensured that the trajectory planner only plans drivable trajectories. The application achieves a variable preview of the control variables in a particularly simple manner, whereby the method can also be used for a plurality of actuators and does not require a complex vehicle model. Furthermore, the method requires only a simple parameterization and only a low computing effort based on analytical calculations. The method is also compatible with different planning approaches and can therefore be implemented in a particularly simple and flexible manner.

[0017] A power device, such as an internal combustion engine or an electric machine, and / or a vehicle brake device is preferably predefined as an actuator. Alternatively or additionally, a steering device can also be predefined as one of the actuators.

[0018] According to a preferred embodiment of the application, a torque, in particular a power device torque of a power device or a brake torque of a brake device, can be predefined as a control variable.

[0019] The maximum possible change of the control variable in the positive and / or negative direction can advantageously be predefined as a limit value. A power device torque or a power device torque, for example, can be predefined as a control variable.

[0020] Depending on the intended use, the actuator model can be based on a function of the control variable, which is plotted over time, with a positive increase of the control variable for an increase of the control variable and a negative increase of the control variable for a decrease of the control variable. For example, a power device model and / or a brake device model can be predefined as an actuator model or a plurality of actuator models. As long as a power device torque or a brake torque is predefined as a control variable, the power device model, for example, comprises a power device torque increase as a positive increase and a power device torque decrease as a negative increase, while the brake device model preferably comprises a brake torque increase as a positive increase and a brake torque decrease as a negative increase. The respective increases can be determined in a practical manner by observing the time steps of the preview with respect to the time axis.

[0021] The possible maximum acceleration of the vehicle is preferably predefined as an acceleration value, and the possible maximum deceleration of the vehicle is preferably predefined as a deceleration value.

[0022] According to a preferred embodiment of the method, the driving resistance can be determined. In particular, the driving resistance can be determined by means of the previewed time step or the previewed duration.

[0023] The driving resistance can then also be taken into account in determining the acceleration value and / or the deceleration value, depending on the purpose of use. This further improves the safety reliability of the determination.

[0024] Furthermore, at least one sensor for detecting the surroundings, in particular a camera and / or a lidar sensor and / or a radar sensor and / or an ultrasonic sensor, can be predefined. The surroundings of the vehicle and the target objects and road users therein can be detected on the basis of the sensor data of the sensor or sensors. Here, the sensor data of a plurality of sensors can also be fused in order to further improve the detection of the surroundings and the target objects.

[0025] The detected surroundings of the vehicle, including the target objects and road users therein, can be used in a useful manner for setting the search space and / or for the trajectory planning. This can be implemented, for example, by additionally limiting the search space for searching for possible trajectories, since the target objects detected by the sensors are in the previously defined search space. Furthermore, the trajectory to be driven can be selected when planning the trajectory or after planning the trajectory, for example, taking into account the view of avoiding collisions, selecting a trajectory that follows the course of the road and does not collide with other target objects / road users.

[0026] Furthermore, the application also comprises a computer program with program code, which, when executed in a computer or another programmable computing device known from the prior art, implements the method according to the application. The method can thus also be provided as a purely computer-implemented method, wherein the term "computer-implemented method" in the sense of the application refers to a process planning or method step that is implemented or carried out by means of a computer. Here, a computing device such as a computer, a computer network or another programmable device known from the prior art, for example a computer device comprising a processor, a microcontroller or the like, can process data by means of programmable computing rules.

[0027] Furthermore, the application also comprises a computer-readable storage medium with instructions, which cause a computer executing the instructions to implement any one of the embodiments of the method described above.

[0028] The invention also comprises a control device for controlling a vehicle along a trajectory, in particular longitudinally, which is designed to implement vehicle control according to the method of the invention.

[0029] In the sense of the invention, the term "search space" refers to a spatial section in which the control unit searches for possible drivable trajectories, wherein a plurality of trajectories can be planned within the search space in order to subsequently select a respective trajectory that is suitable for the situation.

[0030] In the sense of the invention, the term "limit value" refers to a maximum or minimum value of a control variable, i.e. the maximum or minimum, which can be detected, for example, along the driving distance or the time t.

[0031] The invention also explicitly comprises combinations of features that are not explicitly mentioned, i.e. so-called dependent combinations. BRIEF DESCRIPTION OF DRAWINGS

[0032] The invention will be described in more detail below on the basis of suitable embodiments. Therein:

[0033] Figure 1 a schematic diagram of a vehicle is shown, in which a prediction of a maximum control variable is implemented according to the method of the invention, and

[0034] Figure 2 a simplified schematic diagram of a flow chart of the method according to the invention is shown. DETAILED DESCRIPTION

[0035] Figure 1The reference 1 in the drawing denotes a vehicle with different actuators (steering device 3, drive device 4, brake device 5), which vehicle has a control device 2 (ECU, Electronic Control Unit) by means of which a trajectory can be planned with respect to one or more dynamic properties of the actuators. Here, the trajectory is calculated by means of a trajectory planner, wherein a prediction of a maximum control variable, in particular along the longitudinal direction, for limiting the search space of the trajectory planner is carried out for each respective actuator and used for the trajectory planning. Here, the trajectory planner can be provided as a hardware module or a pure software module of the control device 2. Furthermore, the vehicle 1 has sensors (camera 6, laser radar sensor 7, radar sensor 8 and ultrasonic sensors 9a, 9b) for detecting the surroundings, the sensor data of which are used for the recognition of the surroundings and target objects in order to enable various assistance functions such as, for example, an emergency brake assist (EBA, Electronik Brake Assist), a distance tracking control (ACC, Automatic Cruise Control), a lane keeping control or a lane keeping assist (LKA, Lane Keep Assist) or the like. The assistance functions can also be implemented by means of the control device 2 or a separate control device in a utilitarian manner.

[0036] According to the method according to the application, the input signals are the vehicle acceleration a, the vehicle speed v, the brake torque T B , the drive device torque T M , the minimum drive device torque T M,最小 , the maximum drive device torque T M,最大 and the currently set gear, and two output vectors are provided, the maximum acceleration and the maximum deceleration, each of which has a variable preview. Here, the vehicle parameters transmission ratio i g , vehicle mass m, wheel radius r, frontal area A, air resistance coefficient c W (decreasing precision in this order) and air density p L (unit: kg / m3) are known. Figure 2 A design of a flow chart or data flow of the method is shown in simplified structure.

[0037] Figure 2As an example, the brake device and the power device are pre- configured as actuators, which are illustrated according to respective actuator models (power device model 10 and brake device model 11) having a positive increase of the torque T for a torque increase (power device torque increase 10a and brake device torque increase 11a) and a negative increase of the torque T for a torque decrease (power device torque decrease 10b and brake device torque decrease 11b) with respect to the time t, respectively. Herein, Figure 2 The torque or moment T is given in Newton meters (Nm) and the time is given in seconds (s). Herein, the brake device is represented according to a linear function having a dead time and a saturation, wherein the dead time to be included is included as an option for the first slope portion of the function and the saturation is included for the second slope portion of the function. In the power device model 10, the power device is represented as a piecewise defined, linear function with an optional dead time and saturation, because e.g. an internal combustion engine and an electric machine provide a lower torque gradient below the power device torque limit value, which depends on the gear, due to the construction form. In contrast, for an electric machine with a single gear transmission, the modeling can be chosen like the brake device model 11.

[0038] Within the preview 12, the maximum rising gradient and the maximum falling gradient of the actuator are multiplied with the vector of the preview, respectively, to determine the maximum positive change and the maximum negative change of the (torque) moment T within the time step (time t in seconds s) given by the vector, wherein the maximum positive change and the maximum negative change are determined according to Figure 2 The following time steps are chosen as shown: 0 seconds, 0.5 seconds, 1 second and 2 seconds.

[0039] Furthermore, the maximum torque change added to the current torque T cannot be larger than the absolute limit of the respective actuator (i.e. e.g. the maximum power device torque value T M,最大 , so that the maximum torque change is limited within the possible operating range of the actuator as shown by the saturation 13.

[0040] Subsequently, a conversion 14 is suitably performed, wherein the (four) torque changes are converted into a vehicle acceleration according to the following equation:

[0041] a = T • (m • r)

[0042] The sum of the converted power device torque increase and the brake torque decrease results in the maximum vehicle acceleration. Correspondingly, the sum of the converted power device torque decrease and the brake torque increase results in the maximum deceleration of the vehicle.

[0043] Furthermore, the determined vehicle deceleration is also considered with respect to the driving resistance. Here, the calculated vehicle acceleration is used in order to derive the change in driving resistance over the preview duration. This also leads to a change in acceleration. Here, the following equation can be used:

[0044]

[0045] The change in driving resistance 15a, 15b is taken into account.

[0046] Then, at the output (right arrow), the acceleration limit and deceleration limit for correcting the influence of the driving resistance change are output, for example, using a preview map or output of 0 seconds, 0.5 seconds, 1 second, 2 seconds or similar (variable) and are transmitted to the planner.

[0047] In summary, the present application provides a prediction method for estimating current and future, in particular longitudinal, actuator limits. These actuator limits are to be provided to the planner in order to ensure the calculation of drivable trajectories. Here, in particular, the available dynamic properties and the absolute limits of the actuators (power device and brake device) are to be estimated. Furthermore, the driving resistance is also to be taken into account. Here, the estimated actuator limits can be calculated for the current point in time and for a time prediction directed at the planner. Furthermore, in order to achieve compatibility with optimization-based and non-optimization-based planning methods, the estimated actuator limits can be transmitted to the planner in the form of vehicle longitudinal acceleration. The present application can also be adapted to systems with additional actuators, for example vehicles with front axle steering and rear axle steering or all-wheel drive vehicles, i.e. other longitudinal actuators are extended. Furthermore, it can also be applied to redundant actuators, for example a "brake-by-wire" device with a conventional brake device as a backup path. The method is not limited to longitudinal control, but can also be applied to "Steer-by-Braking" methods. Furthermore, the method can also be applied to planner solutions and controller methods and combined methods such as model predictive control (MPC).

[0048] List of reference signs:

[0049] 1 vehicle

[0050] 2 control device

[0051] 3 steering device

[0052] 4 power device

[0053] 5 brake device

[0054] 6 camera device

[0055] 7 lidar sensor

[0056] 8 radar sensor

[0057] 9a ultrasonic sensor

[0058] 9b ultrasonic sensor

[0059] 10 power plant model

[0060] 10a power plant torque increase

[0061] 10b power plant torque decrease

[0062] 11 brake model

[0063] 11a brake torque increase

[0064] 11b brake torque decrease

[0065] 12 preview

[0066] 13 saturation

[0067] 14 transition

[0068] 15a travel resistance

[0069] 15b travel resistance

[0070] A front area

[0071] a vehicle acceleration

[0072] c W air resistance coefficient

[0073] i G drive ratio

[0074] m vehicle mass

[0075] r wheel radius

[0076] T torque

[0077] t time

[0078] v vehicle speed

[0079] p L air density

Claims

1. A method of controlling a vehicle (1) with at least one actuator along a trajectory, wherein, In the case of a preview of at least one control variable of the actuator, a trajectory is planned within a search space, the method comprising the following method steps: - establishing an actuator model of the actuator as a function of at least one control variable of the actuator, - setting a time step of the preview, - determining a change of the actuator control variable along the time step from the actuator model and from limit values of the control variable, wherein the maximum possible change of the control variable in the positive and / or negative direction is predefined as a limit value, - limiting the search space from the limit values of the control variable of the actuator, - determining acceleration values and / or deceleration values of the vehicle (1) by converting at least one control variable together with the vehicle mass and the wheel radius, wherein the driving resistance is determined by the time step of the preview and is to be taken into account together when determining the acceleration values and / or deceleration values, - outputting the acceleration values and deceleration values for limiting the search space in which the trajectory planning is carried out.

2. The method of claim 1, wherein, The power device (4) and / or the brake device (5) of the vehicle (1) are predefined as actuators.

3. The method according to claim 1 or 2, characterized in that, The torque is predefined as a control variable.

4. The method of claim 3, wherein, The torque is a power device torque of the power device (4) or a brake torque of the brake device (5).

5. The method according to claim 1 or 2, characterized in that, The actuator model comprises a positive increase of the control variable for an increase of the control variable and a negative increase of the control variable for a decrease of the control variable as a function of the control variable with respect to time t.

6. The method of claim 1 or 2, wherein, The maximum possible acceleration of the vehicle (1) is predefined as an acceleration value and the maximum possible deceleration of the vehicle (1) is predefined as a deceleration value.

7. The method of claim 1 or 2, wherein, At least one sensor for the detection of the surroundings is predefined.

8. The method of claim 7, wherein, The at least one sensor for the detection of the surroundings is a camera (6) and / or a lidar sensor (7) and / or a radar sensor (8) and / or an ultrasonic sensor.

9. The method according to claim 7 or 8, characterized in that, The detected surroundings are used for setting the search space (9) and / or for the trajectory planning.

10. Computer program product with program code which, if the program code is executed on a computer, carries out the method according to one of claims 1 to 9.

11. Computer-readable storage medium comprising instructions which cause a computer executing the instructions to carry out the method according to one of claims 1 to 9.

12. A control device (2) for controlling a vehicle (1) to travel along a trajectory, characterized in that The method according to one of claims 1 to 9 carries out a control of the vehicle (1). The method according to one of claims 1 to 9 carries out a control of the vehicle (1).

Images