Tire Wear Estimation Using Driving-State Statistical Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing tire wear monitoring methods, whether direct or indirect, face challenges such as sensor durability issues, high cost, inaccuracy, difficulty in adaptation to different vehicle types, and manpower consumption, necessitating a more reliable and adaptable approach for tire wear estimation.
Innovation Solution
A method that predicts tire wear based on driving states using statistical representations of vehicle accelerations, yaw rates, and other driving parameters, employing a machine learning model to minimize errors and provide accurate tire wear indicators without additional sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If direct monitoring approaches using sensors located in the tire are used, then tire wear can be monitored directly, but sensor durability decreases due to large temperature variations and forces
Solution Approach 1:
The patent uses an intermediary estimation model that indirectly determines tire wear from vehicle driving data rather than placing sensors directly in the tire. This mediator (the estimation algorithm) translates easily measurable vehicle parameters into tire wear indicators without exposing any sensor to the harsh tire environment, thus resolving the contradiction between direct measurement accuracy and sensor durability.
Solution Approach 2:
The patent replaces the mechanical sensor system inside the tire with a computational estimation system that uses vehicle-level sensors and algorithms. Instead of mechanically placing durable sensors in the tire, the system substitutes a software-based estimation approach that achieves comparable accuracy without the durability problems of physical sensors in the tire.
2Measurement precision
If direct monitoring approaches using sensors located in the tire are used, then tire wear can be monitored directly, but additional cost increases significantly
Solution Approach 1:
The patent makes the tire wear monitoring system universal by using vehicle data that already exists for multiple other functions (driving state monitoring, performance analysis, etc.). The same vehicle sensors used for general vehicle operation are repurposed for tire wear estimation, eliminating the need for additional dedicated sensors and reducing implementation cost while maintaining accuracy.
Solution Approach 2:
The vehicle's existing data collection infrastructure serves the additional function of tire wear monitoring. The vehicle's own sensors and processing systems provide the data and computational power needed for tire wear estimation without requiring external or additional dedicated resources, making the system cost-effective and easy to manufacture.
3Ease of manufacture
If indirect estimation approaches are used, then implementation cost is reduced, but measurement precision decreases
Solution Approach 1:
The patent improves indirect estimation accuracy by transforming the input parameters through statistical processing (histograms, percentiles, time-weighted averages) before feeding them to the estimation model. This parameter transformation extracts more meaningful wear-indicative features from the raw vehicle data, significantly improving measurement precision while keeping the system cost-effective and easy to manufacture.
4Ease of manufacture
If indirect estimation approaches are used, then implementation cost is reduced, but adaptation difficulty to new vehicle types increases
Solution Approach 1:
The patent makes the estimation system dynamic and adaptable to different vehicle types through configurable parameters and vehicle-specific calibration options. The system can adjust its input data selection, statistical processing parameters, and estimation model coefficients based on the specific vehicle type, enabling easy adaptation across different vehicle platforms while maintaining cost-effectiveness.
Data Source
Figure 1~3
Figure 4~5
Figure 6A~6B
AI summary
A computer-implemented method for determining a wear indicator (W*) of a tire of a vehicle (VEH) so-called test vehicle based on driving states (DS) of said test vehicle during a time period, said method comprising steps of: • determining (S10) at least one statistical representation (HIST) of said driving states over said time period; • determining (S20) at least one parametrized function (FPp, FPn) representative of said at least one statistical representation; and • determining (S30) said wear indicator with an estimation model (PM) taking as input at least one parameter (Ap, Bp, An, Bn) of the at least one parametrized function.