Method for the automatic autonomous control of a motor vehicle

Abstract

Method for the automatic autonomous control of a motor vehicle (3), comprising the following steps: - Calculation of a vehicle path using an external controller (1) using a state (6) of the motor vehicle (3), wherein the vehicle path defines a route along which the motor vehicle (3) is to be moved; - Output of values ​​(4) by the outer controller (1) to an inner controller (2), where the values ​​(4) correspond to geometric points along the vehicle path; - Calculation of a vehicle dynamics (5) using the output values ​​(4), wherein the vehicle dynamics (5) is designed to influence the state (6) of the motor vehicle (3); - Application of vehicle dynamics (5) to the motor vehicle (3); - Recording the state influenced by the applied vehicle dynamics (5); - Output of the influenced state (6) to the external controller (1); wherein the process steps are performed repeatedly, wherein the output influenced state (6) is used as the state (6) of the motor vehicle (3) when the calculation of the vehicle path is performed again, characterized in that the vehicle path is optimized by the repeated execution of the process steps, wherein the optimization can be aborted at any time, wherein, upon abortion, the vehicle path calculated at that time and the vehicle dynamics (5) calculated at that time are used to control the motor vehicle (3), wherein there is a first range and a second range for each of the values ​​(4) output by the external controller (1) and / or for the influenced state (6),wherein first optimization costs are calculated for the values ​​(4) output by the external controller (1) and / or second optimization costs for the influenced state (6), and wherein the first optimization costs and / or the second optimization costs are higher when the influenced state (6) and / or the values ​​(4) output by the external controller (1) are in the first range than when the influenced state (6) and / or the values ​​(4) output by the external controller (1) are in the second range.

Description

[0001] The invention relates to a method for the automatic autonomous control of a motor vehicle according to the preamble of claim 1.

[0002] Methods for automatic autonomous control are known from the state of the art. US 2014 / 0 277 895 A1 discloses such a method for a motorcycle. The motorcycle is controlled autonomously using GPS recognition on a test track and reports driving dynamics data back to a control center.

[0003] US 8,170,739 B2 discloses a system for providing path generation for automatic lane centering and / or lane changing. The system includes a desired path generation processor that receives signals indicative of the roadway the vehicle is traveling on, a lane change request, vehicle condition information, and a vehicle steering angle. The system also includes a path prediction processor that predicts the vehicle path based on vehicle condition information, including vehicle longitudinal velocity, vehicle lateral velocity, vehicle yaw rate, and vehicle steering angle. The desired path information and the predicted path information are compared to generate an error signal, which is sent to a lane change controller, which provides a steering angle signal to steer the vehicle and reduce the error signal.

[0004] DE 10 2018 009 927 A1 discloses a control system for use in a motor vehicle. The control system is configured and intended to update the information regarding the current driving situation of the motor vehicle and / or the provided environmental data and to adapt the trajectory for the motor vehicle using a target function and based on the updated provided environmental data and / or based on the updated information regarding the current driving situation of the motor vehicle.

[0005] In contrast, the present invention is based on the object of improving the real-time capability of the method. Furthermore, a motor vehicle with improved real-time capability is to be created.

[0006] This object is achieved by a method according to claim 1 and a motor vehicle according to claim 6. Embodiments of the invention are specified in the dependent claims.

[0007] For the purposes of this description, automatic autonomous control is understood to mean, in particular, that the motor vehicle is controlled while driving without human interaction. It is therefore controlled neither by a person inside the vehicle nor by a person outside the vehicle.

[0008] The method according to claim 1 can, in particular, be a method that enables the automatic, autonomous control of the motor vehicle during a continuous run on a test track. This refers to the operation of the motor vehicle on the test track over a relatively long period of time in order to test the function of various components of the motor vehicle during operation. Such a continuous run on the test track differs, for example, from the automatic, autonomous operation of the motor vehicle in regular road traffic. During a continuous run, for example, it is not important to observe traffic regulations or other road users. During a continuous run, however, it can be important that the motor vehicle is driven on the test track within the physical limits with regard to cornering speeds, acceleration, and braking. For testing purposes, it is often important during a continuous run that a specific driving profile (e.g.extremely sporty or designed for comfort) is used over a comparatively long period of time.

[0009] First, a vehicle path is calculated using an external controller based on the state of the vehicle. The state of the vehicle can include the current position, current orientation, and / or current speed of the vehicle. The vehicle path defines a route along which the vehicle is to be moved.

[0010] The outer controller outputs values ​​to an inner controller. The values ​​correspond to geometric points along the vehicle's path. These can, in particular, be the geometric points that define the part of the vehicle's path that the vehicle is to reach in the immediate future. The values ​​can, for example, include a curvature of the vehicle's path, starting from the vehicle's current position. In this description, geometric points are understood to be virtual points that are used solely to define the vehicle's path. They do not have to be real points.

[0011] The inner controller calculates vehicle dynamics using the output values. The vehicle dynamics are designed to influence the state of the vehicle when applied. The calculated vehicle dynamics are then applied to the vehicle. In this way, the state of the vehicle is influenced. The state influenced by the applied vehicle dynamics is recorded and output to the outer controller.

[0012] In addition to the values ​​output by the outer controller, the inner controller may also include the coefficient of friction between the tires of the motor vehicle and the road surface, a sideslip angle of the motor vehicle, a pitch angle of the motor vehicle, a roll angle of the motor vehicle and / or a yaw angle of the motor vehicle when calculating the vehicle dynamics.

[0013] By using two feedback controllers, continuous optimization is achieved when the process is executed multiple times, even though the computational effort for each individual run is relatively low. Furthermore, the optimization can be aborted at any time without interrupting vehicle operation. In this case, the vehicle can be controlled with incompletely optimized vehicle dynamics. This means that the vehicle continues to operate automatically and autonomously, and, for example, the endurance test does not need to be aborted.

[0014] The process steps are performed repeatedly. For the purposes of this description, the process steps include, in particular, calculating the vehicle path, outputting the values, calculating the vehicle dynamics, applying the vehicle dynamics, recording the influenced state, and outputting the influenced state. The output influenced state is then used as the state of the motor vehicle when the vehicle path calculation is performed again.

[0015] By repeatedly executing the process steps, the vehicle path is optimized. In this context, optimization is understood in particular to mean that the vehicle path is optimized according to the driving profile selected by a user. This can mean, for example, that the vehicle path is optimized for a particularly fast drive over the test track.

[0016] The repetitive execution of the process steps can be aborted at any time. If aborted, the vehicle path and vehicle dynamics calculated at that time are used to control the motor vehicle. Due to the use of the inner and outer controllers, this vehicle path and these vehicle dynamics are not the optimal solution. However, they can be used without concern and are better suited for endurance testing than any vehicle paths and vehicle dynamics previously calculated as part of the process. In particular, tests have shown that this type of vehicle path and vehicle dynamics can also have a stabilizing effect on the motor vehicle if the motor vehicle is in danger of entering an uncontrollable driving situation.

[0017] For the values ​​output by the external controller and / or for the influenced state, there is a first range and a second range. First optimization costs are calculated for the values ​​output by the external controller. Alternatively or additionally, second optimization costs can be calculated for the influenced state. The first optimization costs are higher when the values ​​output by the external controller are in the first range than when the values ​​output by the external controller are in the second range. The second optimization costs are higher when the influenced state is in the first range than when the influenced state is in the second range. In the context of this description, optimization costs are understood to mean parameters that should be kept as low as possible in a method according to one embodiment of the invention.These may in particular be optimization costs which are known to the person skilled in the art from various optimization methods.

[0018] In particular, it is possible that the optimization costs for output values ​​and influenced states in the first area are higher than third optimization costs associated with other objectives, so that the output values ​​and the influenced state are returned to the second area relatively quickly.

[0019] The outer controller therefore tries to keep the output values ​​in the second range. The inner controller tries to keep the influenced state in the second range. However, if it is not possible to keep the output values ​​or the influenced state in the second range, values ​​in the first range or an influenced state in the first range can also be permitted. Therefore, both the outer controller and the inner controller can be referred to as “Relaxed Barrier Functions MPC controllers”. There are restrictions on values ​​and states in the second range. These restrictions are the boundaries between the first and second ranges. However, the process is also carried out with values ​​and states from the first ranges. The values ​​and states within the second range can be referred to as permissible values.The values ​​and states within the second range can be called illegal values, although they can also be used in exceptional cases.

[0020] According to one embodiment of the invention, the inner controller can use all values ​​output by the outer controller to calculate the vehicle dynamics. This is particularly advantageous for a particularly accurate calculation of the vehicle dynamics, which is suitable for influencing the state of the motor vehicle in such a way that the motor vehicle follows the vehicle path as precisely as possible.

[0021] According to one embodiment of the invention, the state of the motor vehicle may include a position, an orientation, a speed, and / or an acceleration of the motor vehicle. These parameters largely determine how accurately the motor vehicle follows the vehicle path. The position may, for example, be used as a value from a differential satellite navigation system. Furthermore, the state may include a coefficient of friction between the tires of the motor vehicle and a road surface on which the motor vehicle is located. Furthermore, the state may include data generated by an inertial measurement unit (IMU). The inertial measurement unit may include multiple inertial sensors such as acceleration sensors and yaw rate sensors.

[0022] According to one embodiment of the invention, the outer controller and / or the inner controller can be designed as a model predictive controller (MPC). This enables the outer controller to calculate an optimized input signal for the inner controller, which in turn leads to an optimized output signal from the inner controller. Limitations of the values ​​output by the outer controller and the applied vehicle dynamics can be taken into account. While the model behavior is predicted up to a certain time horizon, it is possible to use only the values ​​and the vehicle dynamics for the next time step, after which the optimization is repeated. The optimization is then performed in the next time step with the then influenced state, which can be viewed as feedback and transforms the model predictive controller into a controller.This allows, in particular, the consideration of disturbances to the vehicle dynamics caused by external factors.

[0023] According to one embodiment of the invention, the vehicle dynamics can include an acceleration value, a steering angle, and / or a gear ratio. The acceleration value can be positive or negative. For example, it can be an equivalent value to a position of a brake or accelerator pedal. The steering angle can be a value for how steerable wheels of the motor vehicle are steered. The gear ratio can be a value for which gear ratio is used to transfer the energy generated by a motor vehicle engine to the drive wheels of the motor vehicle.

[0024] The motor vehicle according to claim 6 comprises a control unit for automatic, autonomous control of the motor vehicle. The control unit is designed to carry out a method according to an embodiment of the invention. The control unit can, for example, comprise the inner and outer controllers.

[0025] Further features and advantages of the present invention will become clear from the following description of a preferred embodiment with reference to the accompanying drawing.

[0026] This shows Fig. 1 is a schematic block diagram of a method according to an embodiment of the invention using an inner and an outer controller.

[0027] The outer controller 1 calculates a vehicle path using a state 6 of the motor vehicle 3. Values ​​4 defining geometric points along this vehicle path are output from the outer controller 1 to an inner controller 2. The inner controller 2 calculates vehicle dynamics 5 using the values ​​4. This vehicle dynamics 5 is applied to the motor vehicle 3, thereby influencing the state 6 of the motor vehicle 3.

[0028] Both the inner controller 2 and the outer controller 1 can be designed as model-predictive controllers. In particular, they can be controllers that can also output suboptimal solutions. For example, the inner controller 2 can abort the optimization of the vehicle dynamics 5 and output suboptimal vehicle dynamics 5. This suboptimal vehicle dynamics 5 has a stabilizing effect on the motor vehicle 3, even though the optimization was aborted.

[0029] The vehicle dynamics 5 can include, for example, a positive or negative acceleration value, a steering angle, and / or a gear ratio. The state 6 of the motor vehicle 3 can include a current position, a current orientation, and / or a current speed of the motor vehicle 3.

[0030] Out of Fig. 1 clearly shows that the process is implemented by a control loop with an inner controller 2 and an outer controller 1. The values ​​4 calculated by the outer controller 1, which define the vehicle path, are fed to the inner controller 2 as input signals. The inner controller 2 uses these values ​​4 to calculate the vehicle dynamics 5, which then influences the state 6 of the motor vehicle 3. The influenced state 6 is then fed back to the outer controller 1 as an input signal.

[0031] In addition, the state 6 of the motor vehicle can also be fed to the internal controller 2. The same applies to the influenced state 6 if the method has been performed at least once. This is advantageous because the internal controller 2 can then also use the state 6 or the influenced state 6 to calculate the vehicle dynamics. For this purpose, the state 6 can, in particular, comprise a coefficient of friction between the tires of the motor vehicle 3 and a road surface on which the motor vehicle 3 is located, a sideslip angle of the motor vehicle 3, a pitch angle of the motor vehicle 3, a roll angle of the motor vehicle 3, and / or a yaw angle of the motor vehicle 3.

Claims

[1] Method for the automatic autonomous control of a motor vehicle (3), comprising the following steps: - Calculating a vehicle path by means of an external controller (1) using a state (6) of the motor vehicle (3), wherein the vehicle path defines a route along which the motor vehicle (3) is to be moved; - output of values ​​(4) by the outer controller (1) to an inner controller (2), wherein the values ​​(4) correspond to geometric points along the vehicle path; - Calculating vehicle dynamics (5) using the output values ​​(4), wherein the vehicle dynamics (5) is designed to influence the state (6) of the motor vehicle (3); - application of vehicle dynamics (5) to the motor vehicle (3); - recording the state influenced by the applied vehicle dynamics (5); - outputting the influenced state (6) to the external controller (1); wherein the method steps are carried out repeatedly, wherein the output influenced state (6) is used as the state (6) of the motor vehicle (3) when the calculation of the vehicle path is carried out again, characterized bythat the vehicle path is optimized by repeatedly executing the method steps, wherein the optimization can be aborted at any time, wherein, upon aborting, the vehicle path calculated at that time and the vehicle dynamics (5) calculated at that time are used to control the motor vehicle (3), wherein there is a first range and a second range for the values ​​(4) output by the external controller (1) and / or for the influenced state (6), wherein first optimization costs are calculated for the values ​​(4) output by the external controller (1) and / or second optimization costs are calculated for the influenced state (6), and wherein the first optimization costs and / or the second optimization costs are higher if the influenced state (6) and / or the values ​​(4) output by the external controller (1) are in the first range,than if the influenced state (6) and / or the values ​​(4) output by the external controller (1) are in the second range., [2] Method according to claim 1, characterized by that the inner controller (2) uses all values ​​(4) output by the outer controller (1) to calculate the vehicle dynamics (5). [3] Method according to one of the preceding claims, characterized by that the state (6) of the motor vehicle (3) comprises a position, an orientation, a speed and / or an acceleration of the motor vehicle (3). [4] Method according to one of the preceding claims, characterized by that the outer controller (1) and / or the inner controller (2) are / is designed as a model predictive controller. [5] Method according to one of the preceding claims, characterized by that the vehicle dynamics (5) include an acceleration value, a steering angle and / or a gear ratio. [6] Motor vehicle (3), comprising a control unit (1; 2) for the automatic autonomous control of the motor vehicle (3), wherein the control unit (1; 2) is designed to carry out a method according to one of the preceding claims.

Images