Tread Wear Model Calibration with Neural Network Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for estimating tire tread wear and Remaining Tread Material (RTM) do not accurately account for irregular wear patterns, which can lead to inaccurate tire replacement timing, potentially compromising safety.
Innovation Solution
An improved tread wear monitoring method and system that incorporates a preliminary step for calibrating a Tread Wear Model with correction factors for irregular wear due to tire features and usage, utilizing an Artificial Neural Network (ANN) to refine RTM estimation based on driving-related data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a standard Tread Wear Model is used to estimate RTM, then the estimation process is simple and quick, but the accuracy is insufficient due to irregular wear patterns
Solution Approach 1:
The patent segments the tread wear estimation process into two distinct phases: a preliminary calibration phase that establishes baseline wear characteristics, and an operational monitoring phase that uses simplified models. This segmentation allows complex calibration to be performed once offline, while maintaining simple real-time estimation
Solution Approach 2:
The patent applies preliminary action by performing tread wear tests and calibrating the TWM during the preliminary calibration step before actual use. The calibration values and correction factors are determined in advance, allowing the operational monitoring to use pre-configured models without real-time complexity
2Measurement precision
If correction factors for irregular wear are incorporated into the TWM, then RTM estimation accuracy improves, but the system complexity increases
Solution Approach 1:
The patent applies local quality by introducing correction factors that specifically address irregular wear patterns at critical locations. Instead of complicating the entire model, correction factors are applied locally to adjust for specific wear deviations caused by tire features and usage conditions
Solution Approach 2:
The patent uses correction factors as intermediary elements that bridge the gap between the simple standard TWM and the complex reality of irregular wear. These correction factors act as mediators that adjust the basic model's output without requiring a complete redesign of the estimation system
3Measurement precision
If comprehensive tread wear testing and calibration are performed, then model accuracy improves, but time and resources are consumed
Solution Approach 1:
The patent performs comprehensive tread wear testing and calibration as preliminary actions during the calibration step, before the tire is put into service. This allows accurate model configuration to be completed offline, so that during operational monitoring, the system can quickly estimate RTM without repeated time-consuming tests
Solution Approach 2:
The patent applies partial action by performing wear tests on representative samples of tires during calibration, rather than testing every individual tire. The calibration values derived from these partial tests are then applied to similar tires, reducing the time and resources required while maintaining adequate accuracy
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
The invention concerns a tread wear monitoring method comprising a preliminary step (6) and a tread wear monitoring step (7). The preliminary step (6) includes: performing tread wear tests on one or more tires; measuring tread-wear-related quantities and first frictional-energy-related quantities, wherein the tread wear-related quantities are indicative of tread wear resulting from the performed tread wear tests, and the first frictional-energy-related quantities are related to frictional energy which the tested tire(s) is/are subject to during the performed tread wear tests; and determining a calibrated tread wear model based on the measured tread-wear-related and first frictiona1-energy-related quantities. The tread wear monitoring step (7) includes: acquiring, from a vehicle bus (40) of a motor vehicle (4) equipped with two or more wheels fitted, each, with a tire, driving-related quantities related to driving of the motor vehicle (4); computing, based on the acquired driving-related quantities and a predefined vehicle dynamics model related to the motor vehicle (4), second frictional-energy-related quantities related to frictional energy experienced, during driving, by at least one tire of the motor vehicle (4); estimating, based on the second frictional-energy-related quantities and the calibrated tread wear model, tread wear experienced by said at least one tire of the motor vehicle (4) during driving; and estimating a current average remaining tread material amount of said at least one tire of the motor vehicle (4) based on the estimated tread wear. Additionally, the preliminary step (6) further includes: determining, based on one or more of the measured tread-wear-related quantities, a first correction factor related to irregular tread wear due to tire features; and training an artificial neural network to provide second correction factors related to irregular tread wear due to tire usage. Finally, the tread wear monitoring step (7) further includes: providing a second correction factor by means of the trained artificial neural network based on one or more of the acquired driving-related quantities; and computing a corrected remaining tread material amount based on the current average remaining tread material amount, the first correction factor and the second correction factor provided by the trained artificial neural network based on the one or more acquired driving-related quantities.