Method and device for determining the readiness of a driver assistance function

Abstract

Method for determining the readiness of a driver assistance function, wherein a first release signal (20) is generated by means of a rule-based decision logic (12) based on detected and / or queried first status data (10), wherein a second release signal (21) is generated by means of a machine learning-based decision logic (13) based on detected and / or queried second status data (11), wherein an indication signal (30) for a release is generated and provided, wherein the display signal (30) is activated when the first enable signal (20) is present, wherein the display signal (30) is deactivated after activation by the first enable signal (20) only when the first enable signal (20) and the second enable signal (21) are no longer present.

Description

The invention relates to a method and a device for determining a readiness of a driver assistance function.In modern vehicles, drivers are supported by driver assistance systems that provide driver assistance functions. Examples of this are an automated or partially automated longitudinal and / or transverse guidance of the vehicle. At present, it is still customary for the driver to arrange his hands on the steering wheel at regular intervals, so that automated or partially automated operation is possible. Otherwise, the driver assistance function is deactivated. Various sensors and / or methods can be used to detect at least one hand on the steering wheel (also referred to as hands-on detection).For example, a capacitive sensor on a steering wheel of the vehicle can be used to determine whether or not at least one hand of the driver is on the steering wheel. A multistage status can be determined from the capacitively captured sensor data, for example (e.g. no contact, slightly surrounded on the left / right to strongly surrounded on both sides, etc.), which is transmitted to support functions, such as a longitudinal and / or transverse guidance assistance system, for example. For example, such a status can be associated with an adaptive cruise control (or a closed-loop adaptive cruise control. Adaptive cruise control, ACC) is supplied as input, wherein, when a contact of at least one hand with the steering wheel is detected, for example, a starting process for following a front vehicle is enabled. Driver observation cameras are also known which can detect driver activity.Furthermore, there are methods that recognize driver activity, in particular with respect to hands-on detection, by means of artificial intelligence (machine learning) methods or rule-based approaches. For example, a method of artificial intelligence, in particular an artificial neural network, can be trained to perform hands-on detection on the basis of a torque detected at a steering system, a detected steering angle and / or changes in these variables. Such a method is known, for example, from U.S. Pat. No. 2022 / 0 161 846 A1. The rule-based approach, on the other hand, passively evaluates steering signals (e.g. a steering torque, a steering speed and / or a steering angle) with respect to threshold values and / or signal curves and / or actively operates with a torque circuit, in which a test signal is impressed on the steering wheel by means of an actuator, which test signal is intended to cause a counter-torque of hands arranged on the steering wheel, which counter-torque can be detected by means of a sensor. As soon as such a counter torque has been detected, it is assumed that at least one hand is arranged on the steering wheel. Such a method is known, for example, from DE 10 2013 209 459 A1.DE 10 2009 028 647 A1 discloses a method for detecting the operating state of a steering wheel in a vehicle. For this purpose, a current, high-resolution steering wheel angle and a current steering torque are determined. Depending on the steering wheel angle and the steering torque, a driver hand torque currently applied by the driver is determined. Depending on a curve of the driver hand torque, the operating state of the steering wheel is then determined or it is detected whether the driver has his hands on the steering wheel.DE 10 2011 109 711 A1 discloses a method for detecting a hands-on or hands-off situation of a driver of a motor vehicle, wherein the motor vehicle is in the operating state.If, for example, a capacitive sensor on the steering wheel is to be dispensed with for cost reasons, a rule-based approach and a machine-learning-based approach can be used. The rule-based approach is less sensitive, so that a state in which a hand is arranged on the steering wheel but in particular no steering movements take place cannot always be reliably detected. Although a machine learning-based approach is more sensitive, it is currently not possible to enable the driver assistance function exclusively on the basis of the machine learning-based approach on regulatory requirements.The invention is based on the object of improving a method and a device for determining a readiness of a driver assistance function, in particular with respect to feedback to a driver.The object is achieved according to the invention by a method with the features of claim 1 and a device with the features of claim 6. Advantageous embodiments of the invention are evident from the dependent claims.In particular, a method for determining a readiness of a driver assistance function is provided, wherein a first enable signal is generated by means of a rule-based decision logic on the basis of detected and / or queried first state data, wherein a second enable signal is generated by means of a machine-learning-based decision logic on the basis of detected and / or queried second state data, wherein an indication signal for an enable is generated and provided, wherein the indication signal is activated if the first enable signal is present, wherein the indication signal, after activation by the first enable signal, is only deactivated if the first enable signal and the second enable signal are no longer present.Furthermore, in particular, a device for determining a readiness of a driver assistance function is provided, comprising a data processing device, wherein the data processing device is configured to obtain captured and / or queried first state data and to provide a rule-based decision logic which generates a first enable signal on the basis of the captured and / or queried first state data, to further obtain captured and / or queried second state data and to provide a machine-learning-based decision logic which generates a second enable signal on the basis of the captured and / or queried second state data, to further generate and provide an indication signal for an enable, and to activate the indication signal if the first enable signal is present, and to deactivate the indication signal after activation by the first enable signal only if the first enable signal and the second enable signal are no longer present.The method and the device make it possible to provide a display signal which makes feedback about an availability of the driver assistance function, in particular in connection with hands-on detection, less confusion for a driver. This is based on the idea that the display signal for indicating a readiness of the driver assistance function is provided or activated exclusively as a function of the first enable signal, that is to say the display signal is activated only when the first enable signal is present. In order to nevertheless signal a fundamental availability to the driver, the display signal is deactivated after activation by the first enable signal provided by the rule-based decision logic after this first enable signal is omitted (because the rule-based decision logic can no longer recognize a hand on the steering wheel) only if the second enable signal provided by the more sensitive machine-learning-based decision logic is no longer present. As a result, a display can be actuated by means of the display signal, which display can signal the availability of a driver assistance function of the driver assistance system in an improved manner. In particular in the case of a multiple state change, that is to say when the rule-based decision logic switches back and forth multiple times in a short time between detection and non-detection, the display can thereby remain active in order to indicate to the driver the fundamental availability of the driver assistance function. A repeated change to and fro of the display can thereby be avoided. However, the real availability and provision of the driver assistance function is still determined in particular by the presence of the first enable signal. The second enable signal acts exclusively on the display signal and the display. A state representation can thereby be configured to be less confusion and more reliable for the driver. This increases ride comfort and confidence in the technique.The rule-based decision logic in particular passively evaluates the first state data and, starting from this, derives the first enable signal in a rule-based manner, for example by comparing the first state data with predefined threshold values. For example, steering signals (e.g., a steering torque, a steering speed and / or a steering angle) can be evaluated with respect to threshold values and / or signal curves. As described in the introduction, a test signal can also be actively impressed on a steering wheel in order to induce and detect a counter-torque. For example, a hands-on detection can be carried out on the basis of a detected torque, a detected steering angle and / or changes in these variables, the first enable signal encoding a hands-on state.The machine learning-based decision logic can be provided on the basis of a method of artificial intelligence, in particular an artificial neural network, wherein the artificial intelligence or the neural network is trained to estimate the second enable signal on the basis of the first state data. The training and application of the method of artificial intelligence, in particular of the neural network, are carried out in a manner known per se. For example, a hands-on detection can be carried out on the basis of a detected torque, a detected steering angle and / or changes in these variables, wherein the second enable signal encodes a hands-on state.Parts of the apparatus, in particular the data processing device, can be configured individually or collectively as a combination of hardware and software, for example as program code which is executed on a microcontroller or microprocessor. However, it can also be provided that parts are configured individually or combined as an application-specific integrated circuit (ASIC) and / or field-programmable gate array (FPGA).In one specific embodiment, it is provided that the driver assistance function is enabled only when the first enable signal is present. This makes it possible to ensure that the driver assistance function-independently of the display signal and an output of the display signal on a display-is provided, that is to say is activated and deactivated, exclusively on the basis of the presence of the first enable signal.In one specific embodiment, it is provided that the first enable signal and the second enable signal represent a contact of at least one hand with a steering wheel. In particular, the first enable signal and the second enable signal include and / or encode a result of a hands-on detection.In one specific embodiment, it is provided that the driver assistance function includes a cruise control (CLOSED) system. Adaptive Cruise Control, ACC). The display signal can then indicate, for example, a readiness for starting the adaptive cruise control from standstill to follow a vehicle in front.In a further embodiment, it is provided that an active test signal is impressed on the steering wheel if the first enable signal is not present and a following of a front vehicle is possible. As a result, a torque can be imposed on the steering wheel in order to carry out hands-on detection. In particular, this allows a hand on the steering wheel to be recognized in a short time in order to activate the driver assistance function almost instantaneously (after recognition of at least one hand on the steering wheel) even when the driver assistance function is deactivated and to carry out the starting for following a front vehicle.Further features for the configuration of the device will become apparent from the description of configurations of the method. The advantages of the device are in each case the same as in the embodiments of the method.The invention is explained in more detail below with reference to preferred exemplary embodiments with reference to the figures. The following are shown here: FIG. 1 shows a schematic illustration for illustrating embodiments of the device; FIG. 2 shows a schematic flow diagram of an embodiment of the method.FIG. 1 shows a schematic illustration for illustrating embodiments of the device 1. the device 1 is arranged in a vehicle 50, in particular a motor vehicle. The device 1 can itself provide a driver assistance function or control the provision of such a function.The apparatus 1 comprises a data processing device 2. the data processing device 2 comprises, for example, a computing device, for example a microprocessor, and a memory (both not shown) for carrying out method steps of the method described in this disclosure.The data processing device 2 is configured to receive acquired and / or interrogated first state data 10 and to provide a rule-based decision logic 12 which generates a first enable signal 20 on the basis of the acquired and / or interrogated first state data 10. For this purpose, the data processing device 2 has, for example, a module 2- 1, which can also be designed as a software module. The acquired and / or queried first state data 10 originate, for example, from sensors of the vehicle 50, for example, a steering angle sensor and / or a torque sensor of a steering system of the vehicle 50. The first enable signal 20 can contain, in particular, the states active (or "on" or "1") or inactive (or "off" or "0") and can be provided as a corresponding signal level.Furthermore, the data processing device 2 is configured to obtain acquired and / or queried second state data 11 and to provide a machine learning-based decision logic 13 which generates a second enable signal 21 on the basis of the acquired and / or queried second state data 11. For this purpose, the data processing device 2 has, for example, a module 2- 2, which can also be designed as a software module. The detected and / or queried second state data 11 also originate, for example, from sensors of the vehicle 50, for example a steering angle sensor and / or a torque sensor of a steering system of the vehicle 50. The second enable signal 21 can contain, in particular, the states active (or "on" or "1") or inactive (or "off" or "0") and can be provided as a corresponding signal level.The first state data 10 and the second state data 11 can basically be the same or different. For example, the status data 10, 11 can originate from the same sensors or different sensor types and / or sensors can be used.Furthermore, the data processing device 2 is configured to generate and provide an indicator signal 30 for an enable, and to activate the indicator signal 30 when the first enable signal 20 is present, and to deactivate the indicator signal 30 after activation by the first enable signal 20 only when the first enable signal 20 and the second enable signal 21 are no longer present. For this purpose, the data processing device 2 has, for example, a module 2- 3, which can also be designed as a software module and which evaluates the first enable signal 20 and the second enable signal 21 and generates the display signal 30 according to the logic described.The display signal 30 can be used, for example, to actuate a display 51 in the vehicle 50 or can be displayed on a display device 52 of the vehicle 50.It can be provided that the data processing device 2 is configured to enable the driver assistance function only if the first enable signal 20 is present. For this purpose, the data processing device 2 can have a module 2- 4 which checks the presence of the first enable signal 20 and, starting from this, enables or disables the driver assistance function. For example, a control signal 31 can be generated and provided for this purpose. The module 2- 4 can also be designed as a software module.It can be provided that the first release signal 20 and the second release signal 21 represent a contact of at least one hand with a steering wheel of the vehicle 50. The first state data 10 and the second state data 11 can then comprise sensor data of a steering angle sensor and / or of a torque sensor, for example.It can be provided that the driver assistance function includes a cruise control (CLOSED) template. Adaptive Cruise Control, ACC). In this case, the display signal 20 can signal a readiness for following a front vehicle.It can be provided that the data processing device 2 is configured to cause an active test signal to be impressed on the steering wheel if the first enable signal 20 is not present and a following of a front vehicle is possible. For this purpose, the data processing device 2 has, for example, a module 2- 5, which can also be designed as a software module. In particular, a trigger signal 32 is generated in order to cause the application of the active test signal in a manner known per se. For this purpose, module 2- 5 may also receive a stand-by signal 40 of the distance control template with the content that a following is possible (or not) at the current point in time.FIG. 2 shows a schematic flow diagram of an embodiment of the method for determining a readiness of a driver assistance function.In a method step 100, method steps 100 aand 100 bare carried out. In method step 100 a, a first enable signal is generated by means of a rule-based decision logic on the basis of acquired and / or queried first state data. In method step 100 b, a second enable signal is generated by means of machine-learning-based decision logic on the basis of detected and / or queried second state data. The enable signals can be generated, for example, in the form of signal levels, wherein two signal levels are provided in each case in order to code two states (for example "yes" or "no" or "1" or "0").In a method step 101, it is checked whether or not the first enable signal generated in method step 100 ais present. If this is the case, a driver assistance function of the driver assistance system is activated in a method step 102. Furthermore, in a method step 103, a generated and provided display signal is activated for enabling, for example by setting a signal level of the display signal provided for this purpose. The method then jumps back to method step 100.If, on the other hand, the check in method step 101 reveals that the first enable signal is not present, then a check is made in a method step 104 as to whether or not the second enable signal generated in method step 100 bis present. If this is the case, the method jumps back to method step 100. If, on the other hand, this is not the case, then the display signal is deactivated in a method step 105, for example by setting a signal level provided for this purpose. The method then jumps back to method step 100.The activation of the display signal can therefore only take place when the first enable signal is present. The deactivation in the event of the first enable signal being omitted, on the other hand, is delayed if the second enable signal is still present after the first enable signal being omitted. If the second enable signal is also omitted, the display signal is deactivated. A renewed activation can then take place only by a renewed presence of the first enable signal.It is provided in particular that the first enable signal and the second enable signal represent a contact of at least one hand with a steering wheel.Furthermore, it is provided in particular that the driver assistance function is a cruise control. The display signal can indicate, for example, a readiness of the adaptive cruise control for following a vehicle in front.It can be provided that an active test signal is impressed on the steering wheel if the first enable signal is not present and a following of a front vehicle is possible.List of reference characters1 Device 2 Data processing device 2- 1 Module 2- 2 Module 2- 3 Module 2- 4 Module 10 First state data 11 Second state data 12 Rule-based decision logic 13 Machine learning-based decision logic 20 First enable signal 21 Second enable signal 30 Display signal 31 Control signal 32 Trigger signal 40 Ready signal 50 Vehicle 51 Display 52 Display device 100- 105 Method steps

Claims

Method for determining a readiness of a driver assistance function, wherein a first enable signal (20) is generated by means of a rule-based decision logic (12) on the basis of detected and / or interrogated first state data (10), wherein a second enable signal (21) is generated by means of a machine-learning-based decision logic (13) on the basis of detected and / or interrogated second state data (11), wherein an indication signal (30) for an enable is generated and provided, wherein the indication signal (30) is activated if the first enable signal (20) is present, wherein the indication signal (30), after activation by the first enable signal (20), is only deactivated if the first enable signal (20) and the second enable signal (21) are no longer present.Method according to Claim 1, characterized in that the driver assistance function is enabled only if the first enable signal (20) is present.Method according to Claim 1 or 2, characterized in that the first enable signal (20) and the second enable signal (20) represent a contact of at least one hand with a steering wheel.Method according to one of the preceding claims, characterized in that the driver assistance function is a cruise control.Method according to Claim 4, characterized in that an active test signal is impressed on the steering wheel if the first enable signal (20) is not present and a following of a front vehicle is possible.Device (1) for determining a readiness of a driver assistance function, comprising: a data processing device (2), wherein the data processing device (2) is configured to obtain acquired and / or interrogated first state data (10) and to provide a rule-based decision logic (12) which, on the basis of the acquired and / or interrogated first state data (10), generates a first enable signal (20), further acquires acquired and / or interrogated second state data (11) and to provide a machine-learning-based decision logic (13) which, on the basis of the acquired and / or interrogated second state data (11), generates a second enable signal (21), further generates and provides an indication signal (30) for an enable, and to activate the indication signal (30) if the first enable signal (20) is present and to deactivate the indication signal (30) only after activation by the first enable signal (20), when the first enable signal (20) and the second enable signal (21) are no longer present.Device (1) according to Claim 6, characterized in that the data processing device (2) is configured to enable the driver assistance function only if the first enable signal (20) is present.Device (1) according to Claim 6 or 7, characterized in that the first enable signal (20) and the second enable signal (21) represent a contact of at least one hand with a steering wheel.Device (1) according to one of Claims 6 to 8, characterized in that the driver assistance function is a cruise control system.Device (1) according to Claim 9, characterized in that the data processing device (2) is configured to cause an active test signal to be impressed on the steering wheel if the first enable signal (20) is not present and a following of a front vehicle is possible.

Images