DE102022003544A1 - Assistance system and method for supporting a driver during a drift maneuver with a vehicle - Google Patents

Abstract

The invention relates to an assistance system (1) for supporting a driver during a drift maneuver with a vehicle. According to the invention, a data processing unit (3) comprises an artificial neural network (2), which determines the driver's driving skills with respect to the third party based on the driver's driving actions, taking into account input data (E1 to En) relating to the vehicle, and, depending on the driving skills, a steering assistance level (as output data) L) and an accelerator pedal assistance level (F). An output device (4) coupled to the data processing unit (3) outputs the output data optically, acoustically and/or haptically.
The invention further relates to a method for supporting a driver during a drift maneuver with a vehicle.

Description

The invention relates to an assistance system for supporting a driver during a drift maneuver with a vehicle.

The invention further relates to a method for supporting a driver during a drift maneuver with a vehicle.

From the DE 10 2013 200 027 A1 is a drift training assistance system for a motor vehicle for learning a third party necessary in highly dynamic driving training. Both a steering activity and an accelerator pedal movement or their combined action are taken over or supported by at least one drift assistance system.

The invention is based on the object of specifying a novel assistance system and method for supporting a driver during a drift maneuver with a vehicle.

The object is achieved according to the invention by an assistance system which has the features specified in claim 1, and by a method which has the features specified in claim 7.

Advantageous embodiments of the invention are the subject of the subclaims.

According to the invention, the assistance system for supporting a driver during a drift maneuver with a vehicle comprises a data processing unit with an artificial neural network, which determines the driver's driving skills with respect to the third party based on the driver's driving actions, taking into account input data relating to the vehicle, and as output data depending on the driving skills Steering assistance level and an accelerator pedal assistance level are determined. Furthermore, the assistance system includes an output device coupled to the data processing unit, which outputs the output data optically, acoustically and/or haptically.

Drivers of modern vehicles, especially high-powered sports vehicles, generally do not have the driving skills to drive the vehicle to the limit of its physical capabilities. The present assistance system is intended to help the driver improve his driving skills in terms of vehicle control, especially when performing drift maneuvers. The assistance system supports the driver in his driving maneuvers in order to improve his driving and reaction behavior in different driving dynamics situations and, for example, to offer a fallback level in critical driving situations. The assistance system thus enables the driving skills of a driver of a vehicle to be improved in a particularly efficient manner while at the same time minimizing practice time for the driver.

In a possible embodiment of the assistance system, it is designed such that the driver can define properties of the vehicle with regard to its dynamic behavior, particularly in the transverse direction when cornering. The driver can, for example, specify a maximum yaw rate, yaw acceleration and/or a sideslip angle of the vehicle.

In a further possible embodiment of the assistance system, it comprises a steering assistance unit coupled to the data processing unit, which generates steering assistance during a drift maneuver depending on the determined degree of steering assistance until predetermined drift parameters are reached. This means that the driver always receives exactly the right amount of steering support to carry out the drift maneuver safely and to keep his personal learning progress as high as possible.

In a further possible embodiment of the assistance system, the steering support includes a dynamic adjustment of a steering torque. This provides support that is intuitively perceivable to the driver and thus leads to rapid learning success while at the same time ensuring a high level of acceptance of the assistance function for the driver.

In a further possible embodiment of the assistance system, it comprises an accelerator pedal support unit coupled to the data processing unit, which generates accelerator pedal support during a drift maneuver depending on the determined level of accelerator pedal support until predetermined drift parameters are reached. This means that the driver always receives exactly the right amount of accelerator pedal support to carry out the drift maneuver safely and to keep his personal learning progress as high as possible.

In a further possible embodiment of the assistance system, the accelerator pedal support includes a dynamic adjustment of an accelerator pedal characteristic curve, which maps a dependence of a generated drive torque on an accelerator pedal position. This provides support that is intuitively perceivable to the driver and thus leads to rapid learning success while at the same time ensuring a high level of acceptance of the assistance function for the driver.

• - eine Fahrgeschwindigkeit des Fahrzeugs und/oder
• - eine Gierrate des Fahrzeugs und/oder
• - eine Gierbeschleunigung des Fahrzeugs und/oder
• - einen Schwimmwinkel des Fahrzeugs und/oder
• - einen Schwimmwinkelgradient des Fahrzeugs und/oder
• - einen Lenkradwinkel des Fahrzeugs und/oder
• - einen Lenkradwinkelgradient des Fahrzeugs und/oder
• - eine Fahrpedalstellung des Fahrzeugs und/oder
• - einen Fahrpedalgradient des Fahrzeugs und/oder
• - einen Radschlupf des Fahrzeugs und/oder
• - Raddrehzahlen des Fahrzeugs und/oder
• - Raddrehzahlgradienten des Fahrzeugs und/oder
• - eine Solltrajektorie des Fahrzeugs.

Diese Eingangsdaten bilden einen Zustand des Fahrzeugs besonders umfassend ab und ermöglichen eine zuverlässige und einfache Bestimmung der Fahrfähigkeiten des Fahrers.In a further possible embodiment of the assistance system, the input data optionally includes

• - a driving speed of the vehicle and/or
• - a yaw rate of the vehicle and/or
• - a yaw acceleration of the vehicle and/or
• - a slip angle of the vehicle and/or
• - a slip angle gradient of the vehicle and/or
• - a steering wheel angle of the vehicle and/or
• - a steering wheel angle gradient of the vehicle and/or
• - an accelerator pedal position of the vehicle and/or
• - an accelerator pedal gradient of the vehicle and/or
• - vehicle wheel slip and/or
• - Vehicle wheel speeds and/or
• - Wheel speed gradients of the vehicle and/or
• - a target trajectory of the vehicle.

This input data represents the state of the vehicle in a particularly comprehensive manner and enables the driver's driving skills to be reliably and easily determined.

In the method for supporting a driver during a drift maneuver with a vehicle, according to the invention, using an artificial neural network, a steering assistance level and an accelerator pedal assistance level are used as output data based on the driver's driving actions, taking into account input data relating to the vehicle, the driver's driving skills with regard to drifting and, depending on the driving skills, as output data determined. The output data is transmitted to an output device, the output data being output optically, acoustically and/or haptically by means of the output device.

The present method helps the driver to improve his driving skills in terms of vehicle control, especially when performing drift maneuvers. Using the method, the driver can be supported in his driving maneuvers in such a way that his driving and reaction behavior can be improved in different driving dynamics situations and, for example, a fallback level can be offered in critical driving situations. The use of the method thus enables the driving skills of a driver of a vehicle to be improved in a particularly efficient manner while at the same time minimizing practice time for the driver.

In a possible embodiment of the method, the output data is transmitted to a steering assistance unit, which generates steering assistance during a drift maneuver depending on the determined degree of steering assistance until predetermined drift parameters are reached. This means that the driver always receives exactly the right amount of steering support to carry out the drift maneuver safely and to keep his personal learning progress as high as possible.

In a further possible embodiment of the method, the output data is transmitted to an accelerator pedal support unit, which generates accelerator pedal support during a drift maneuver depending on the determined level of accelerator pedal support until predetermined drift parameters are reached. This means that the driver always receives exactly the right amount of accelerator pedal support to carry out the drift maneuver safely and to keep his personal learning progress as high as possible.

In a further possible embodiment of the method, the quality of drift maneuvers carried out is determined. The artificial neural network is trained based on the quality.

This enables simple and reliable training of the network and, as a result, optimized performance of the process.

In a further possible embodiment of the method, instructions regarding an optimal steering wheel position and/or an optimal accelerator pedal position are issued to the driver visually, acoustically and/or haptically while a drift maneuver is being carried out. On the one hand, this makes it possible for the driver to perform a drift maneuver optimally in a particularly simple manner. Furthermore, a learning effect can be maximized because the driver can easily see what he needs to change in his driving actions in order to carry out the drift maneuver as optimally as possible.

Exemplary embodiments of the invention are explained in more detail below with reference to drawings.

• 1 schematisch ein Blockschaltbild eines Assistenzsystems zur Unterstützung eines Fahrers bei einem Driftmanöver mit einem Fahrzeug und
• 2 schematisch einen Aufbau eines künstlichen neuronalen Netzwerks eines Assistenzsystems gemäß 1.

Show:

• 1 schematically a block diagram of an assistance system to support a driver during a drift maneuver with a vehicle and
• 2 schematically shows a structure of an artificial neural network of an assistance system 1 .

In 1 is a block diagram of an assistance system 1 for supporting a driver during a drift maneuver with a vehicle. 2 shows a structure of an artificial neural network 2 of the assistance system 1 according to 1 .

The assistance system 1 includes a data processing unit 3, an output device 4, a steering support unit 5 and an accelerator pedal support unit 6.

In order to provide the driver of the vehicle with the best possible support in learning driving skills to carry out drift maneuvers with the vehicle, a drift trainer based on artificial intelligence is implemented using the assistance system 1, which includes the artificial neural network 2.

The artificial neural network 2 determines the driver's driving skills with regard to drifting, that is, the driver's ability to drift, based on the driver's driving actions, taking into account input data E1 to En relating to the vehicle. For this purpose, the input data E1 to En are in a so-called input layer 2.1 or input layer of the artificial neural network 2.

The input data E1 to En are measured variables or estimated variables which, in particular, continuously describe the current driving dynamics of the vehicle. In addition, all relevant driver inputs, such as a steering wheel angle, a steering angular speed and/or an accelerator pedal position, can be delivered to the artificial neural network 2. For example, a number of input data E1 to En is flexible.

• - eine Fahrgeschwindigkeit des Fahrzeugs und/oder
• - eine Gierrate des Fahrzeugs und/oder
• - eine Gierbeschleunigung des Fahrzeugs und/oder
• - einen Schwimmwinkel des Fahrzeugs und/oder
• - einen Schwimmwinkelgradient des Fahrzeugs und/oder
• - einen Lenkradwinkel des Fahrzeugs und/oder
• - einen Lenkradwinkelgradient des Fahrzeugs und/oder
• - eine Fahrpedalstellung des Fahrzeugs und/oder
• - einen Fahrpedalgradient des Fahrzeugs und/oder
• - einen Radschlupf des Fahrzeugs und/oder
• - Raddrehzahlen des Fahrzeugs und/oder
• - Raddrehzahlgradienten des Fahrzeugs und/oder
• - eine Solltrajektorie des Fahrzeugs.

The input data E1 to En include, for example

• - a driving speed of the vehicle and/or
• - a yaw rate of the vehicle and/or
• - a yaw acceleration of the vehicle and/or
• - a slip angle of the vehicle and/or
• - a slip angle gradient of the vehicle and/or
• - a steering wheel angle of the vehicle and/or
• - a steering wheel angle gradient of the vehicle and/or
• - an accelerator pedal position of the vehicle and/or
• - an accelerator pedal gradient of the vehicle and/or
• - vehicle wheel slip and/or
• - Vehicle wheel speeds and/or
• - Wheel speed gradients of the vehicle and/or
• - a target trajectory of the vehicle.

From the input data E1 to En, the artificial neural network 2 determines two results using a number of hidden layers 2.3 to 2.m or hidden layers, which produce an output layer 2.2 or output layer of the artificial neural network 2 in the form of two signals outputs.

A first signal is a steering assistance level L. This is an indicator of how much support the driver needs from the assistance system 1 when moving the steering wheel in order to carry out the drift maneuver. A size of the steering assistance level L is selected based on an individual driver level, that is, the driver's determined driving skills with regard to drifting.

A second signal is an accelerator pedal assistance level F. This is an indicator of how high the required intensity of the driver's accelerator pedal assistance by the assistance system 1 must be so that he can carry out the drift maneuver. A size of the accelerator pedal assistance level F is selected based on an individual driver level, that is, the driver's determined driving skills with regard to drifting.

The signals are fed to the output device 4, which outputs them optically, acoustically and/or haptically as a recommendation and for information to the driver. For example, the steering assistance level L and the accelerator pedal assistance level F provide a signal value between 0 and 1, with signal value 1, for example, indicating the greatest possible respective assistance. It is also possible to divide individual value ranges into classes and to connect these classes to form a community via an application, or app for short, running on a device. From this, a ranking list can be created with different drivers in order to establish competition between these drivers regarding the execution of drift maneuvers.

Additionally, during the drift maneuver, instructions regarding an optimal steering wheel position and/or an optimal accelerator pedal position can be issued to the driver visually, acoustically and/or haptically using the output device 4.

In addition, the signals are directly linked to assistance functions of vehicle steering or drive torque distribution.

For this purpose, the degree of steering assistance L is transmitted to the steering assistance unit 5 coupled to the data processing unit 3, which generates steering assistance, in particular a dynamic adjustment of a steering torque, during a drift maneuver depending on the determined degree of steering assistance L until predetermined drift parameters are reached.

Furthermore, the accelerator pedal support level F is transmitted to the accelerator pedal support unit 6 coupled to the data processing unit 3, which generates accelerator pedal support during a drift maneuver depending on the determined accelerator pedal support level F until predetermined drift parameters are reached. The accelerator pedal support includes a dynamic adjustment of an accelerator pedal characteristic curve, which maps a dependence of a generated drive torque on an accelerator pedal position.

This means that the determined steering assistance level L and accelerator pedal assistance level F are used to support the driver when steering by changing the steering or steering wheel torque and, when the accelerator pedal position is adjusted, by modifying the drive torque. This means the driver always receives exactly the right level of support to keep their personal learning progress as high as possible.

A structure of the artificial neural network 2 can be flexibly selected. This means that a number of hidden layers 2.3 to 2.m and a selection of specific networks are not specified. Combinations are also possible.

The artificial neural network 2 is trained using recorded “good and bad measurements” of various drift maneuvers. This means that a quality of drift maneuvers carried out is determined and the artificial neural network 2 is trained based on the quality determined. It is particularly important that the widest possible range of drift techniques is taken into account. On the one hand, this refers to the drift figures themselves, but also to different types of execution. A training process of the network 2 can be carried out both monitored and unsupervised. During monitored training, the input data E1 to En and output data, i.e. the steering assistance level L and the accelerator pedal assistance level F, are also transmitted with corresponding training data. The training varies the weights of the artificial neural network 2 until the output data of the same has approached predetermined target values.

In one possible embodiment, an expansion of the number of output data is also possible. For example, steering assistance that depends on the steering direction can be taken into account if the driver has more deficits in one steering direction than in the other when drifting.

A visualization of the artificial neural network 2 can also be displayed by means of the output device 4, for example a so-called head unit of the vehicle, in which signal flows through the network 2 are displayed.

Reference symbol list

1

Assistance system

2

network

2.1

Input layer

2.2

Output layer

2.3 to 2.m

Hidden layer

3

Data processing unit

4

Dispensing device

5

Steering support unit

6

Accelerator pedal support unit

E1 to En

Input data

F

Accelerator pedal assistance level

L

Steering assistance level

QUOTES INCLUDED IN THE DESCRIPTION

This list of documents listed by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 102013200027 A1 [0003]

Claims (10)

Assistance system (1) for supporting a driver during a drift maneuver with a vehicle, characterized by - a data processing unit (3) with an artificial neural network (2), which is based on the driver's driving actions, taking into account input data relating to the vehicle (E1 to En) The driver's driving skills are determined with respect to the third party and, depending on the driving skills, a steering assistance level (L) and an accelerator pedal assistance level (F) are determined as output data, and - an output device (4) coupled to the data processing unit (3), which outputs the output data optically, acoustically and/or or haptic. Assistance system (1). Claim 1 , characterized by a steering support unit (5) coupled to the data processing unit (3), which generates steering support during a drift maneuver depending on the determined level of steering support (L) until predetermined drift parameters are reached. Assistance system (1). Claim 2 , characterized in that the steering assistance includes a dynamic adjustment of a steering torque. Assistance system (1) according to one of the preceding claims, characterized by an accelerator pedal support unit (6) coupled to the data processing unit (3), which generates accelerator pedal support during a drift maneuver depending on the determined accelerator pedal support level (F) until predetermined drift parameters are reached. Assistance system (1). Claim 4 , characterized in that the accelerator pedal support includes a dynamic adjustment of an accelerator pedal characteristic curve, which maps a dependence of a generated drive torque on an accelerator pedal position. Assistance system (1) according to one of the preceding claims, characterized in that the input data (E1 to En) - a driving speed of the vehicle and / or - a yaw rate of the vehicle and / or - a yaw acceleration of the vehicle and / or - a slip angle of the vehicle and/or - a slip angle gradient of the vehicle and/or - a steering wheel angle of the vehicle and/or - a steering wheel angle gradient of the vehicle and/or - an accelerator pedal position of the vehicle and/or - an accelerator pedal gradient of the vehicle and/or - a wheel slip of the vehicle and/or or - wheel speeds of the vehicle and/or - wheel speed gradient of the vehicle and/or - a target trajectory of the vehicle. Method for supporting a driver during a drift maneuver with a vehicle, characterized in that the driving skills of the driver with respect to the third party are determined by means of an artificial neural network (2) based on the driver's driving actions, taking into account input data (E1 to En) relating to the vehicle , - depending on the driving skills, a steering assistance level (L) and an accelerator pedal assistance level (F) are determined as output data and - the output data are transmitted to an output device (4), and by means of the output device (4) the output data optically, acoustically and / or tactilely be issued. Procedure according to Claim 7 , characterized in that - the output data are transmitted to a steering support unit (5), which generates steering support during a drift maneuver depending on the determined level of steering support (L) until predetermined drift parameters are reached, and / or - the output data to an accelerator pedal support unit (6) are transmitted, which generates accelerator pedal support during a drift maneuver depending on the determined accelerator pedal support level (F) until predetermined drift parameters are reached. Procedure according to Claim 7 or 8th , characterized in that a quality of drift maneuvers carried out is determined and the artificial neural network (2) is trained based on the quality. Procedure according to one of the Claims 7 until 9 , characterized in that while a drift maneuver is being carried out, instructions regarding an optimal steering wheel position and/or an optimal accelerator pedal position are issued to the driver visually, acoustically and/or haptically.

Images