Method and apparatus for operating an automated vehicle

Abstract

A method for operating an automated vehicle comprises the following steps: sensing the environment of the vehicle; providing environmental data of the sensed environment; the environmental data is fed to a situation sensing device having a defined number of situation sensing elements greater than one; computationally modeling the environment of the vehicle with the aid of the situation sensing elements; using output data of the model of the situation sensing elements to control a driver assistance system; deciding with the aid of a decision device: which output data of the model of the situation sensing elements are used to control an actuation device of the vehicle; and using the decided output data to control the actuation device of the vehicle.

Description

[0001] Technical Area

[0002] The invention relates to a method for operating an automated vehicle. The invention also relates to a device for operating an automated vehicle. The invention also relates to a computer program product. Background Art

[0003] Nowadays, data from environmental sensors are used to perform specific driving tasks for the vehicle. For example, environmental sensor data are used to control emergency brake assistance (AEB, English: Autonomous Emergency Braking) or adaptive cruise control (ACC, English: Adaptive Cruise Control) for example. Typically, the environmental sensor data for implementing functions such as AEB or ACC are determined by radar sensors and / or cameras. What all these functions have in common is that their range of action is severely limited.

[0004] DE102015222605A1 discloses a method for assisted driving, partially automated driving, highly automated driving, fully automated driving or unmanned driving of a motor vehicle. It can be provided that all subsystems have a monitoring function and transmit their operating capabilities to a controller, wherein it can also be provided that individual subsystems transmit operating capabilities in quantity while other subsystems transmit operating capabilities only in quality.

[0005] EP1369288A2 discloses a vehicle control device, which includes a vehicle control subsystem. Summary of the invention

[0006] The object of the present invention is to provide an improved method for operating an automated vehicle.

[0007] According to a first aspect, the object is achieved by a method for operating an automated vehicle, the method comprising the following steps:

[0008] -Sensing the vehicle’s environment;

[0009] -Provide environmental data of the sensed environment;

[0010] - feeding the environmental data to a situation sensing device having a defined number of situation sensing elements greater than one;

[0011] -Computational modeling of the vehicle's environment using situation sensing elements;

[0012] - using the output data of the model of the situation sensing element to control the driver assistance system;

[0013] - deciding by means of a decision device which output data of the model of the situation sensing element are to be used for controlling an actuating device of the vehicle; and

[0014] - Using the determined output data to control the vehicle's actuators.

[0015] Advantageously, in this way, automated vehicles up to level 5 (autonomous vehicles) can be realized. Advantageously, this can be achieved with the invention by economical use of hardware resources, whereby the method can be implemented with a relatively small controller. As a result, the computing effort is reduced and faster computing cycles can be implemented. As a result, a real-time-capable system for operating an automated vehicle is provided.

[0016] According to a second aspect, the object is achieved with a device for operating an automated vehicle, the device having:

[0017] - an environmental sensor device for sensing the environment of the vehicle and for providing environmental data of the sensed environment;

[0018] a situation sensing device having a defined number of situation sensing elements greater than one, to which environmental data are fed, wherein computational modeling of the vehicle's environment can be carried out with the aid of the situation sensing elements;

[0019] a driver assistance system which can be controlled using output data of the model of the situation sensing element;

[0020] a decision device, by means of which it is possible to decide which output data of the model of the situation sensing element are to be used for controlling the actuating device of the vehicle; and

[0021] Actuation devices of the vehicle, which can be controlled using the determined output data.

[0022] Advantageous developments of the method are the subject matter of the dependent claims.

[0023] An advantageous embodiment of the method provides that the models of the situation sensing elements are determined independently and synchronously, wherein the models of the situation sensing elements at least partially overlap. This increases the robustness of the assistance system, since an active driver assistance system can continue to operate in the event of a failure of a situation sensing element.

[0024] Another advantageous embodiment of the method provides that the driver assistance systems are executed independently of one another and synchronously, thereby also assisting the independence and robustness of the method.

[0025] Another advantageous embodiment of the method provides that the sensors of the surrounding sensor device provide different data for different situation sensing elements. For example, it can be provided that a video camera is used for lane and object recognition, wherein these different information are used by different situation sensing elements.

[0026] Another advantageous embodiment of the method is characterized in that the self-estimation of each situation sensing element is used to determine how plausible the model of the corresponding situation sensing element is, thereby providing the decision support with a decision aid on which situation sensing element is plausible and which is to be selected.

[0027] Another advantageous embodiment of the method is characterized in that data is provided to all driver assistance systems by means of a model of the situation sensing element. As a result, all driver assistance systems are activated using the same situation sensing element data, which also assists the robustness of the method.

[0028] Another advantageous embodiment of the method provides that the driver assistance system decides which data of which situation sensing element the driver assistance system processes. This also advantageously increases the robustness of the method.

[0029] Another advantageous embodiment of the method is characterized in that the decision device of the functional device decides which driver assistance system processes which data, thereby ultimately deciding which data are to be processed and which actuators are to be fed with the data.

[0030] Another advantageous development of the method provides that each situation sensing element evaluates the behavior data of the vehicle using the same metric, wherein the metric of the situation sensing element comprises at least one of the following: probability of collision, resource conservation, environmental sustainability, legality. The most important metric is the smallest possible probability of collision, by means of which a collision of the vehicle can be avoided as far as possible. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The invention will be described in detail below with further features and advantages with the aid of a number of drawings. The technical features and advantages of the device are derived in a similar manner from the technical features and the device of the method.

[0032] The accompanying drawings show:

[0033] Figure 1 A schematic diagram showing conventional equipment for operating an automated vehicle;

[0034] Figure 2 Another diagram showing another conventional apparatus for operating an automated vehicle;

[0035] Figure 3A schematic diagram showing a first specific embodiment of the proposed device for operating an automated vehicle; and

[0036] Figure 4 Schematic diagram showing a flow chart of the proposed method for operating an automated vehicle. DETAILED DESCRIPTION

[0037] The core idea of ​​the invention is, in particular, to implement driving tasks within the largest possible range of automated vehicle use. Advantageously, in this way, methods for autonomous driving up to level 5 (level 5 is defined as autonomous driving) can be implemented (level 5 is defined as "all aspects of the dynamic driving task are performed without exception by the automated driving system under all driving and environmental conditions that a human driver can handle").

[0038] To this end, the invention proposes a so-called "multi-agent approach" that enables autonomous driving up to level 5. In this case, individual so-called "agents" in the form of situation sensing elements 51a...51n use a mathematical model that is to be determined approximately accurately to simplify the so-called "open world problem" (level 5 autonomous driving possible worldwide). In this case, the different agents or situation sensing elements 51a...51n are designed so that their models use approximately independent states. For example, one situation sensing element 51a...51n is dedicated to lane sensing, another situation sensing element 51a...51n is dedicated to object sensing, etc. The sum of all situation sensing elements 51a...51n thus covers the entire range of action required by the automated vehicle for autonomous driving.

[0039] It is thus advantageously possible to achieve that, by reducing to a model to be determined approximately mathematically, the solution of an open-world problem can be realized in real time with limited hardware resources.

[0040] Nowadays, vehicles are usually equipped with different environmental sensors (e.g. radar, camera, ultrasound, lidar, infrared camera, etc.). Environmental data are sensed with the help of these environmental sensors. Then, functions for completing specific driving tasks can be realized based on these environmental data. Usually, certain environmental sensors are used specifically for specific driving tasks, such as ultrasonic sensors for parking tasks, radar sensors for distance control speed controllers, etc.

[0041] In the context of autonomous driving functions, it is proposed to provide a combination of sensors from a plurality of different surroundings in order to achieve hardware redundancy. A method is proposed that enables autonomous driving with hardware and software redundancy.

[0042] A driver assistance system is a system that is intended to solve driving tasks such as lane keeping. Figure 1 Works as shown in .

[0043] The environmental sensor device 10 is used to determine environmental data 20. The environmental data 20 are analyzed in the driver assistance system 30a, wherein a solution to the driving task is proposed based on the result of the analysis. The driving task solution is then used to control an actuating device 40, for example in the form of a steering device and / or a braking device of the vehicle (not shown). If the vehicle has a plurality of driver assistance systems 30a ... 30n, there can also be a plurality of parallel levels of such logic, for example, such as Figure 2 As shown in .

[0044] Figure 3 An embodiment of a device 100 for operating an automated vehicle according to the invention is shown. It can be seen that in the method proposed here, a situation sensing device 50 is additionally embedded in the closed loop information path. As in the case of the above-described driver assistance systems 30a ... 30n, the information flow starts at the environment sensor device 10.

[0045] The environment sensor device determines environment data 20, wherein the environment data from different environment sensors (e.g. ultrasound, radar, lidar, camera, etc.) of the environment sensor device 10 are plausibly checked and fused. The environment data 20 are preferably provided by different types of sensors, whereby the sensor data of the different sensors are compared. If an object is detected, for example, by a radar sensor and by a video sensor, then the probability that "the object is actually present" is high. The information thus fused and plausibly checked is forwarded to the situation sensing device 50. In the situation sensing device 50, these pre-processed environment data 20 are further processed by a plurality of agents in the form of situation sensing elements 51a . . . 51n.

[0046] The situation sensing elements 51a ... 51n of the situation sensing device 50 abstract the environment of the vehicle into a simplified mathematical model. Here, each situation sensing element 51a ... 51n of the situation sensing device 50 is constructed or programmed so that the situation sensing element uses approximately independent information. For example, the degree of freedom of an object can be limited by means of the situation sensing elements 51a ... 51n, for example, the movement of the vehicle in the Z direction when it is driving straight on a flat highway can be excluded. Therefore, for example, one situation sensing element 51a ... 51n can be set to identify lane markings or lanes based on the environmental data preprocessed as described above, while another situation sensing element 51a ... 51n can be used to determine the relationship between moving objects based on the environmental data, etc. Therefore, each situation sensing element 51a ... 51n has the following task: evaluate different vehicle behaviors with the help of specific metrics (such as collision probability, resource conservation, environmental sustainability, vehicle behavior legality, etc.). Here, the metric "collision probability between objects" is usually given the highest priority. Preferably, all situation sensing elements 51a . . . 51n use the same metric, so that different situation sensing elements 51a . . . 51n and the behaviors generated thereby can thereby be compared with one another.

[0047] Exemplary vehicle behaviors are “driving in a lane”, “implementing a lane change”, “braking”, “accelerating”, “driving to a parking space”, etc. For example, the following situation sensing elements 51a . . . 51n may have overlapping action areas:

[0048] Situation sensing element A: observes all objects that can be assigned to a lane;

[0049] Situation sensing element B: observes all objects.

[0050] In this example, all objects sensed by condition sensing element A are within the active area of ​​condition sensing element B. The opposite is not true.

[0051] The evaluation of different behaviors is achieved by a common set of metrics, wherein each behavior is evaluated by each situation sensing element 51a . . . 51n using its own specific metric.

[0052] In addition to the evaluation of the behavior, a further task of the individual situation sensing elements 51a ... 51n is to output a self-estimation with the aid of a self-estimation device 52a ... 52n. This self-estimation indicates how reliable the evaluation of the behavior by the corresponding situation sensing element 51a ... 51n is. Preferably, the self-estimation of each situation sensing element 51a ... 51n is calculated based on a measure of the quality of the input signal used by the corresponding situation sensing element 51a ... 51n for its calculation. In addition, it can be checked whether the basic assumptions of the situation sensing elements 51a ... 51n can be complied with in the current situation. The evaluated behavior and the self-estimation of the situation sensing elements 51a ... 51n are forwarded by the situation sensing device 50 to a functional device 60 having a decision device 61.

[0053] The evaluated behavior is used in the functional device 60 to implement various driver assistance systems 30a ... 30n (e.g. AEB, ACC, LKA (English: lane keeping assist, lane keeping assist), etc.), which are to perform the driving task. In this case, only a single driver assistance system 30a ... 30n can always be active within a limited time period, because the various driver assistance systems 30a ... 30n are in competition with each other. In this case, the different evaluated vehicle behaviors are compared with each other in the form of environmental data that have been checked for plausibility and fused. In this case, the function is selected that satisfies the desired driving task as well as possible with the subordinate behavior and represents the lowest collision probability or collision risk for the vehicle.

[0054] If a situation sensing element 51a ... 51n reports that its self-estimation is very low, the functional device 60 can, for example, ignore the evaluation of this situation sensing element 51a ... 51n and use the evaluation of another situation sensing element 51a ... 51n. Finally, based on the decision for a behavior and the subordinate functions, an actuator of the actuation device 40, for example in the form of a steering device and / or a braking device of the vehicle, is controlled. Typically, in the case of ACC as the selected driver assistance system 30a ... 30n, the motor of the vehicle is controlled. Typically, in the case of LKA as the selected driver assistance system, the steering device of the vehicle is controlled. Here, for example, it can be provided that the steering device is controlled using trajectory data. As a result, the vehicle thus performs its driving task according to the evaluated behavior.

[0055] Optionally, it may also be provided that the driver assistance system 30a . . . 30n is controlled via a plurality of different situation sensing elements 51a . . . 51n. However, it is not absolutely necessary for every situation sensing element 51a . . . 51n to be able to perform every function.

[0056] The proposed method has the following advantages over conventional classic driver assistance systems:

[0057] - abstraction from the environment to a simple mathematical model is implemented through the situation sensing elements 51a ... 51n, thereby saving hardware resources;

[0058] - The situation sensing elements 51a ... 51n can reduce the computation time and thus provide a system with real-time capabilities;

[0059] - In addition to the hardware redundancy caused by the multiple environmental sensor types of the environmental sensor device 10, software redundancy is provided by the multiple situation sensing elements 51a ... 51n with overlapping areas of action. As a result, the system continues to be functional even in the event of a failure of one situation sensing element 51a ... 51n;

[0060] - Specific tasks can be solved by the development of dedicated agents. For example, a situation sensing element can be provided that is based only on the own vehicle data when, for example, all environmental sensors fail. This situation sensing element can be used to brake the vehicle or to completely hand over the driving task to the human driver.

[0061] The device 100 is preferably designed as an electronic control unit of a vehicle, on which a computer program product (for example in the form of a program code in C++) is executed.

[0062] Figure 4 A basic flow chart of the proposed method for operating an automated vehicle is shown.

[0063] In step 200 , sensing of the vehicle's environment is performed.

[0064] In step 210, environmental data 20 of the sensed environment is provided.

[0065] In step 220, environmental data 20 is fed to a situation sensing device 50 having a defined number of situation sensing elements 51a...51n greater than one.

[0066] In step 230 , computational modeling of the vehicle's environment is performed with the aid of situation sensing elements 51 a . . . 51 n .

[0067] In step 240 , the output data of the model of the situation sensing elements 51 a . . . 51 n are used to implement the control of the driver assistance system 30 a . . . 30 n .

[0068] In step 250 , a decision is made with the aid of decision device 61 as to which output data of the model of situation sensing elements 51 a . . . 51 n are to be used for the activation of actuator device 40 of the vehicle.

[0069] In step 260 , the determined output data is used to control the actuation devices 40 of the vehicle.

[0070] Advantageously, the method according to the invention can be implemented as software, which runs, for example, on an electronic control unit in the vehicle or on a server device based on the cloud. This facilitates the method's simple adaptability.

[0071] Those skilled in the art may modify and / or combine the features of the present invention in an appropriate manner without departing from the core of the present invention.

Claims

1. A method for operating an automated vehicle, the method comprising the following steps: -Sensing the vehicle’s environment; - providing environmental data (20) of the sensed environment; - feeding said environmental data (20) to a situation sensing device (50), said situation sensing device having a defined number of situation sensing elements (51a . . . 51n) greater than one; - computational modeling of the vehicle's environment by means of the situation sensing elements (51a...51n), wherein: The models of the situation sensing elements (51a ... 51n) are determined independently and synchronously, wherein the models of the situation sensing elements (51a ... 51n) overlap at least partially, wherein each situation sensing element (51a ... 51n) is designed in such a way that it uses independent information, wherein the sum of all situation sensing elements (51a ... 51n) covers the entire range of action required by the automated vehicle for autonomous driving; - using the output data of the model of the situation sensing element (51a...51n) to control the driver assistance system (30a...30n); - using a decision device (61) to decide which output data of the model of the situation sensing element (51a . . . 51n) is used to control the actuation device (40) of the vehicle; and - using the determined output data to control the actuation device (40) of the vehicle.

2. The method according to claim 1, wherein: The driver assistance systems (30a . . . 30n) are implemented independently and synchronously with one another.

3. The method according to any one of the preceding claims, characterized in that Different data are provided by sensors of the environment sensor device (10) to different situation sensing elements (51a...51n).

4. A method according to any one of the preceding claims, wherein: With the aid of a self-estimation device (52a...52n) of each situation sensing element (51a...51n), it is determined how reliable the model of the corresponding situation sensing element (51a...51n) is.

5. A method according to any one of the preceding claims, wherein: Data are provided to all driver assistance systems (30a . . . 30n) by means of the model of the situation sensing elements (51a . . . 51n).

6. The method according to claim 5, wherein: The driver assistance system (30a...30n) decides which data from which situation sensing element (51a...51n) the driver assistance system processes.

7. The method according to claim 6, wherein: A decision device (61) of the functional device (60) decides which driver assistance system (30a . . . 30n) processes which data.

8. A method according to any one of the preceding claims, wherein: Each condition sensing element (51a...51n) evaluates the behavior data of the vehicle using the same metric, wherein the metric of the condition sensing element (51a...51n) includes at least one of the following: collision probability, resource conservation, environmental sustainability, and legality.

9. A device (100) for operating an automated vehicle, the device being configured to perform the method according to any one of claims 1 to 8, the device comprising: - an environmental sensor device (10) for sensing the environment of the vehicle and for providing environmental data (20) of the sensed environment; a situation sensing device (50) having a defined number of situation sensing elements (51a . . . 51n) greater than one, to which the environmental data (20) can be fed, wherein By means of the situation sensing elements (51a...51n), computational modeling of the environment of the vehicle can be performed; a driver assistance system (30a . . . 30n) which can be controlled using output data of the model of the situation sensing element (51a . . . 51n); - a decision device (61), by means of which it is possible to decide which output data of the model of the situation sensing element (51a . . . 51n) are used to control the actuation device (40) of the vehicle; and - an actuation device (40) of the vehicle, which can be controlled using the determined output data.

10. The device (100) according to claim 9, characterized in that The device is designed as an electronic control unit of the vehicle.

11. A computer program product having program code means for carrying out the method according to claim 1, when the program code means is run on an electronic device (100) for operating an automated vehicle or is stored on a computer-readable data carrier.

Images