Search

DE-102024132599-A1 - Computer-implemented method for operating a vehicle using at least one autonomous driving function

DE102024132599A1DE 102024132599 A1DE102024132599 A1DE 102024132599A1DE-102024132599-A1

Abstract

The invention relates to a computer-implemented method (100) for operating a vehicle by means of at least one autonomous driving function, comprising at least the following steps: identifying (102) a plurality of control data records of a driver of the vehicle, wherein each control data record indicates a journey along a route with control actions of the at least one autonomous driving function and/or the driver; identifying (104) a subset of the control data records that includes control data records indicating at least one non-traffic-related control action of the driver in at least one section; modifying (106) at least one parameter of a desired behavior of the at least one autonomous driving function for the at least one section based on the subset; and providing (108) at least one control signal to the vehicle based on the desired behavior with the at least one modified parameter of the at least one autonomous driving function when the vehicle is used by the driver and is moving in the at least one section. The method (100) can reduce the number of interventions by a driver to override the driving function.

Inventors

  • Robin Schwager
  • Matej Svaral

Assignees

  • DR. ING. H.C. F. PORSCHE AKTIENGESELLSCHAFT

Dates

Publication Date
20260513
Application Date
20241108

Claims (10)

  1. A computer-implemented method (100) for operating a vehicle by means of at least one autonomous driving function, comprising at least the following steps: a. Retrieving (102) a plurality of control data records of a driver of the vehicle, wherein each control data record indicates a journey along a route with control actions of the at least one autonomous driving function and/or the driver; b. Retrieving (104) a subset of the control data records that includes control data records indicating at least one non-traffic-related control action by the driver in at least one segment; c. Modifying (106) at least one parameter of a desired behavior of the at least one autonomous driving function for the at least one segment based on the subset; and d. Providing (108) at least one control signal to the vehicle based on the desired behavior with the at least one modified parameter of the at least one autonomous driving function when the vehicle is being used by the driver and is moving in the at least one segment.
  2. Computer-implemented method (100) according to Claim 1 , characterized in that the step: Determining (104) a subset from the control data records, further comprises at least the following sub-step: a. Determining (110) a subset of the plurality of control data records that contain control actions of the driver; and b. Determining (112) control data records of the subset whose control actions are not necessary for traffic, to determine the subset.
  3. Computer-implemented method (100) according to Claim 2 , characterized in that the sub-step: Determining (112) control data records of the subset is carried out using a classifier, preferably using a neural network.
  4. Computer-implemented method (100) according to one of the preceding claims, characterized in that the step: Changing (106) at least one parameter, further comprises at least the following sub-step: a. Determining (114) position data of at least one section in at least one control data record of the subset in which at least one non-traffic-related control action was performed.
  5. Computer-implemented method (100) according to Claim 4 , characterized in that the at least one section is determined using position data of the vehicle, preferably based on a global navigation satellite system.
  6. Computer-implemented method (100) according to Claim 4 or 5 , characterized in that the step: Change (106) at least one parameter, further comprises at least the following sub-steps: a. Determine (116) a first group of tax data records in which at least one non-essential tax action was carried out in the at least one section and a second group of tax data records in which no non-essential tax action was carried out in the at least one section; b. Determine (118) a third group of tax data records from the second group in which a non-essential tax action was possible due to a traffic situation in the at least one section; c. Determine (120) an intervention rate based on the first group and the third group; and d. Change (122) the at least one parameter if the intervention rate lies within a significance interval.
  7. Computer-implemented method (100) according to Claim 6 , characterized in that the significance interval extends in the range between 90% and 100%, preferably between 95% and 100%, more preferably between 99% and 100%.
  8. Computer-implemented method (100) according to one of the preceding claims, characterized in that the method (100) after the step: changing (106) at least one parameter, further comprises at least the following, in particular repeatable, step: a. Changing (124) the at least one parameter when, in a further journey along the route, a non-traffic-related control action of the driver is determined in the at least section.
  9. Computer program product, comprising instructions which, when executed by a computer, cause the computer to perform the steps of procedure (100) according to one of the Claims 1 until 8 to execute.
  10. Vehicle (50) comprising at least one control unit (52) for operating at least one autonomous driving function, wherein the control unit (52) is further configured to execute the steps of the procedure (100) according to one of the Claims 1 until 8 to execute.

Description

The invention relates to a computer-implemented method for operating a vehicle by means of at least one autonomous driving function. Vehicles can be equipped with driver assistance and/or automated driving functions. The levels of automation for these functions are divided into five different stages, with stage 5 representing fully autonomous vehicle control. From stage 3 onwards, continuous monitoring of the driving function by the driver is no longer required, although driver intervention can be expected after a warning period. When using stage 1 and 2 driver assistance and automated driving functions, the driver relinquishes some control over the vehicle to the driving function. Because the system is a stage 1 or 2 system, the driver is then responsible for continuously monitoring the function's behavior and intervening if necessary. These interventions occur, for example, if the driving function deviates from its system design. Out of US 9815478 B2 A driver assistance system is known that compares a driver's current driving behavior with recordings of the driver's previous driving behavior in order to determine an unsafe driving condition. Location data can be associated with the driving behavior. The object of the invention is to provide a computer-implemented method with which automated driving errors can be improved in such a way that the number of interventions by a driver to override the driving function can be reduced. The problem is solved by the features of the independent claims. Advantageous further developments are the subject of the dependent claims and the following description. According to a first aspect, a computer-implemented method for operating a vehicle by means of at least one autonomous driving function is described, comprising at least the following steps: Identifying a plurality of control data records of a driver of the vehicle, each control data record indicating a journey along a route with control actions of the at least one autonomous driving function and/or the driver; identifying a subset of the control data records that includes control data records indicating at least one non-traffic-related control action by the driver in at least one section; modifying at least one parameter of a desired behavior of the at least one autonomous driving function for the at least one section based on the subset; and providing at least one control signal to the vehicle based on the desired behavior with the at least one modified parameter of the at least one autonomous driving function when the vehicle is used by the driver and is moving in the at least one section. The computer-implemented method analyzes a driver's historical driving behavior for non-traffic-related control actions in order to adapt the target behavior of at least one autonomous driving function. The autonomous driving function can have at least an automation level of Level 1 or Level 2. Therefore, the driving function can also have an automation level of Level 3 or 4. A large number of control data sets can be used for the analysis, with each control data set being assigned to a trip along a specific route by a specific driver. The control actions of the autonomous driving function and the driver can be assigned to positions along the specific route. From the control data sets assigned to the same route and the same driver, those control data sets in which non-traffic-related control actions by the driver were performed can be identified as a subset. Traffic-related control actions are those actions necessitated by traffic, such as stopping at a red light and/or braking due to a vehicle ahead or a right-of-way situation. Non-traffic-related control actions are not dictated by traffic conditions and can therefore be based on the driver's preferences regarding the vehicle's behavior. Once this subset has been identified, the autonomous driving function can be adapted to a new target behavior by modifying at least one parameter. This modified target behavior then approximates, and can even correspond to, the driver's preferences, thus requiring a reduced number of interventions in the vehicle's control system on a given route when the autonomous driving function uses control signals to steer the vehicle based on the modified target behavior. This allows the driving behavior of the autonomous driving function to be further individualized for a driver. A vehicle in this context can be, for example, a land vehicle, a watercraft, an aircraft and/or a spacecraft. A multitude of tax records can be defined by at least two tax records, preferably by at least three tax records. According to some embodiments, it is conceivable that the step: Determining a subset from the control data records, may further include at least the following sub-step: Determining a subset of the multitude of control data records that contain control actions of the driver, and determining control data records of the subset whose control actions are not nece