Methods for detecting the wear of a vehicle tire and a vehicle

Abstract

The invention relates to a method for detecting the wear of a vehicle tire, in which data from sensors (7, 9, 11, 13, 15) installed in the vehicle (1) for wear detection are evaluated to create a model (21) that is continuously updated. In a method that does not require sensors installed directly on or in the vehicle tire, the measurement data from the sensors (7, 9, 11, 13, 15) detecting the current driving state of the vehicle (1) are continuously processed by a pre-trained artificial intelligence model (21) for wear detection. The artificial intelligence model (21) correlates the sensor data with different wear states and determines the current wear state of the vehicle tire using an anomaly detection method.

Description

[0001] The invention relates to a method for detecting the wear of a vehicle tire, according to the preamble of claim 1, and to a vehicle.

[0002] From WO 2022 / 214175 A1, a method for controlling the movement of a heavy-duty vehicle is known, in which input data regarding one or more tire parameters are transmitted to the control unit, or the tire parameters are estimated based on the input data, and a tire model is configured from the tire parameters. The tire model defines a relationship between a tire wear rate and the vehicle's state of motion, whereby the vehicle's movement is controlled depending on the tire parameters determined by the tire model. The input data is transmitted to the control unit by sensors that detect tire pressure, tire temperature, and / or tire tension.

[0003] WO 2020 / 005 346 A1 discloses a method for estimating the wear of a vehicle tire using a hybrid machine learning system. This learning system comprises a communication unit with a control module and a machine learning model. The machine learning model receives speed and wheel speed signals from several vehicle sensors. Additionally, the machine learning model receives signals from vehicle sensors regarding steering wheel angle, brake pressure, longitudinal acceleration, and total distance, and uses these signals to determine a tire tread wear value.

[0004] US Patent 2023 / 0011981 A1 discloses a method in which information on the tire wear condition of a vehicle tire is determined using a tire wear model created by machine learning. The tire wear model includes an artificial intelligence component that is applied to past tire wear information representing the wear of older tires. The artificial intelligence component comprises an artificial intelligence module and / or a machine learning model. The artificial intelligence component can learn to calibrate various sensors.

[0005] The object of the invention is to provide a method for detecting the wear of a vehicle tire, in which sensors installed directly on or in the vehicle tire can be dispensed with.

[0006] The invention is defined by the features of the independent claims. Advantageous further developments and embodiments are the subject of the dependent claims. Further features, applications, and advantages of the invention will become apparent from the following description and the explanation of exemplary embodiments of the invention illustrated in the figures.

[0007] The problem is solved by the subject matter of claims 1 and 6.

[0008] In the previously described method for detecting tire wear, data from vehicle-installed sensors are evaluated to create a continuously updated model. The measurement data from the sensors detecting the vehicle's current driving state is continuously processed by a pre-trained artificial intelligence (AI) model. This AI model correlates the sensor data with various wear states and uses an anomaly detection method to determine the current wear state of the tire. After the pre-trained AI model is installed in the vehicle or after a tire change, it undergoes an initial calibration to update its performance with the data from the vehicle's sensors.By solely analyzing sensor signals that detect vehicle behavior, such as steering, tilt, vehicle speed, and, in the case of electric vehicles, energy consumption, the system can determine the wear of a vehicle's tires. This eliminates the need for individual sensors attached to each tire, for example, to check tire pressure, thus reducing hardware costs. The method requires only one system to monitor the wear status of all vehicle tires simultaneously. Continuous monitoring of tire wear can potentially prevent accidents caused by worn tires. This allows for the simple implementation of the artificial intelligence model in the vehicle, as the data from the vehicle sensors only needs to be transmitted to the AI ​​model via the vehicle bus.The pre-trained artificial intelligence model can be implemented in different vehicle models with thus different sensor types and software architectures.

[0009] In one implementation, the artificial intelligence model is trained using various historical data sets, including different tire wear conditions. These training datasets contain sensor data correlated with known tire wear conditions, enabling the AI ​​model to reliably recognize the sensor data patterns associated with tire wear.

[0010] In a further refinement, the pre-trained artificial intelligence model compares the currently determined wear state of the vehicle tire with a predefined baseline of a wear-free tire using an anomaly detection method. Based on the identified deviations between the current wear state and the predefined baseline, it infers the presence of tire wear. The anomaly detection method allows for the precise determination of tire anomalies and, through the continuous learning process of the artificial intelligence model within the vehicle, also improves the detection of minor tire wear.

[0011] In a further configuration, if anomalies indicating tire wear are detected, information is sent to the vehicle's driver or maintenance personnel. The driver can then immediately adjust their driving behavior or stop the vehicle to check the tire condition. The information sent to maintenance personnel can be stored in the vehicle's fault memory, where it can be read out and the tire condition checked during the next service visit.

[0012] In a further configuration, the sensors installed in the vehicle communicate with an in-vehicle computing unit containing an artificial intelligence model via an internal bus network. Access to the vehicle's existing bus network, such as CAN, LIN, or FlexRay, allows for real-time data transmission, enabling very rapid monitoring of the vehicle's tire wear.

[0013] Another aspect of the invention relates to a vehicle with a tire wear detection system, comprising a control unit that communicates with a plurality of sensors installed in the vehicle via an internal vehicle bus network, and a tire wear detection model. In a vehicle where sensors installed directly on or in the tire are not required, the tire wear detection control unit comprises an artificial intelligence model configured to perform at least one of the method features described in this patent application. By integrating the artificial intelligence model into the control unit's software architecture, existing computer resources and interfaces in the vehicle can be used for communication and warnings.

[0014] In a further embodiment, the sensors installed in the vehicle monitor the vehicle's current driving status independently of the tires. Since the system uses existing vehicle bus signals, such as steering angle, steering angle speed, vehicle speed, level signals, and vehicle power consumption, to detect tire wear, reliable real-time monitoring of tire wear is possible.

[0015] In a further embodiment, the control unit is connected to a vehicle display system to issue a warning signal to the driver when tire wear is detected. This allows the driver to react proactively to potential hazardous situations and directly check the wear condition of the vehicle's tires.

[0016] Further advantages, features, and details will become apparent from the following description, in which at least one embodiment is described in detail. The described features can, individually or in any meaningful combination, constitute the subject matter of the invention, optionally also independently of the claims, and can, in particular, also be the subject matter of one or more separate applications.

[0017] This shows: Fig. 1 An embodiment of a vehicle for carrying out the method according to the invention.

[0018] In Fig.Figure 1 shows an embodiment of a vehicle for carrying out the method according to the invention. The vehicle 1 comprises a control unit 3, which communicates via a vehicle bus 5, for example a CAN bus, with a steering angle sensor 7, a steering angle velocity sensor 9, a vehicle speed sensor 11, a tilt sensor 13 for the vehicle suspension, and a current consumption measurement unit 15 of a high-voltage battery of the vehicle 1. In addition, the control unit 3 is connected to a display unit 19 visible to the driver 17. The control unit 3 includes a model of artificial intelligence 21, which is pre-trained to detect a correlation between the wear of the vehicle tires and the driving behavior of the vehicle.

[0019] The artificial intelligence model 21 is trained before installation in the vehicle 1 using historical data indicating the vehicle's driving condition in relation to tire wear states, by applying anomaly detection methods. The training datasets contain sensor data correlated with known tire wear states, whereby the artificial intelligence model 21 learns the tire wear patterns associated with different driving conditions of the vehicle 1.

[0020] The pre-trained AI 21 model is implemented as a software function in the control unit 3 and must be calibrated before its first use in the vehicle 1. During installation, the control unit 3, with the integrated, pre-trained AI 21 model, undergoes an initial calibration phase. In this phase, the AI ​​21 model is updated with the current sensor data for new or recently replaced vehicle tires, provided via the vehicle bus 5. This initial calibration state of the AI ​​21 model establishes a baseline state for the vehicle's tires.

[0021] After initial calibration, the control unit 3 continuously monitors the sensor data during vehicle 1 journeys. The artificial intelligence model 21 processes the incoming sensor data in real time to detect deviations from the baseline condition of the tires that could indicate tire wear. The artificial intelligence model 21 uses anomaly detection methods to determine whether the currently supplied sensor data deviates significantly from the baseline. Any deviations are analyzed to determine whether tire wear is present. If tire wear is suspected, the control unit 3 sends a warning message, which is displayed to the driver 17 on the display unit 19 or in a corresponding connected vehicle app. Simultaneously, the warning message can be stored in the vehicle 1's fault memory for readout at a workshop.

Claims

[1] Method for detecting the wear of a vehicle tire, in which sensor data from sensors (7, 9, 11, 13, 15) installed in the vehicle (1) for wear detection are evaluated to create an artificial intelligence model (21) which is continuously updated, wherein the sensor data from the sensors (7, 9, 11, 13, 15) detecting the current driving state of the vehicle (1) are continuously processed by the pre-trained artificial intelligence model (21) for wear detection, wherein the artificial intelligence model (21) correlates the sensor data with different wear states and determines the current wear state of the vehicle tire using an anomaly detection method, characterized by , that After installation of the pre-trained artificial intelligence model (21) in the vehicle (1) or after a tire change, an initial calibration of the pre-trained artificial intelligence model (21) is performed to update it with the sensor data from the sensors (7, 9, 11, 13, 15) installed in the vehicle (1). [2] Method according to claim 1, characterized by , that the artificial intelligence model (21) is trained using various historical sensor data, which include different tire wear conditions. [3] Method according to claim 1 or 2, characterized by, that the pre-trained artificial intelligence model (21) compares the currently determined wear state of the vehicle tire with a predefined base state of a wear-free vehicle tire using the anomaly detection method and concludes the presence of tire wear based on the determined deviations between the current wear state and the predefined base state. [4] Method according to claim 3, characterized by , that if anomalies indicating tire wear are detected, information is issued to the driver (17) of the vehicle (1) or to maintenance personnel. [5] Method according to at least one of the preceding claims, characterized by , that the sensors (7, 9, 11, 13, 15) installed in the vehicle (1) communicate via a vehicle-internal bus network (5) with a computing unit (3) of the vehicle (1) comprising the model of artificial intelligence (21). [6] Vehicle (1) with a system for detecting the wear of a vehicle tire, comprising a Control unit (3) which communicates with a variety of sensors (7, 9, 11, 13, 15) installed in the vehicle (1) via a vehicle-specific bus network (5) and includes an artificial intelligence model (21) for detecting wear of vehicle tires, characterized by , that the control unit (3) for wear detection comprises a model of artificial intelligence (21) which is configured to carry out the method according to at least one of the preceding claims. [7] Vehicle (1) according to claim 6, characterized by , that the sensors (7, 9, 11, 13, 15) installed in the vehicle (1) monitor the current driving condition of the vehicle (1) independently of the tires. [8] Vehicle (1) according to claim 6 or 7, characterized by, that the control unit (3) is connected to a display device (19) of the vehicle (1) to issue a warning signal to the driver (17) when tire wear is detected.

Images