Vehicle Tire Stiffness Estimation via Particle Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the stiffness of vehicle tires during operation are inefficient, particularly in real-time applications, as they rely on indirect measurements and are prone to errors due to varying road conditions and tire characteristics, and often converge to local optima or fail to account for stochastic disturbances.
Innovation Solution
A system and method that jointly estimate the state of a vehicle and its tire stiffness using a combination of deterministic and probabilistic motion models, representing tire stiffness with particles that include mean and variance to account for uncertainty, allowing for real-time estimation without a pre-defined model of tire stiffness evolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If indirect friction determination methods are used with sensors, then friction can be determined during driving, but measurement precision deteriorates due to difficulty in direct measurement
Solution Approach 1:
The patent uses tire stiffness as an intermediary parameter to indirectly determine friction characteristics. Instead of attempting to directly measure friction between tire and road, the system measures tire deformation (stiffness) which correlates with friction conditions, thereby solving the measurement difficulty while maintaining acceptable precision
Solution Approach 2:
The patent replaces direct mechanical friction measurement with a mechanical model-based approach. By using a tire model that relates tire stiffness to friction coefficients, the system substitutes complex direct measurement with model-based calculation, improving measurement precision
2Reliability
If precomputing tire parameters is done, then a starting point or nominal value is obtained, but reliability deteriorates due to varying road conditions, tire pressure, loading, temperature, and wear
Solution Approach 1:
The patent transitions from static precomputed tire parameters to dynamic real-time estimation. The system continuously updates tire stiffness and friction parameters during vehicle operation, adapting to changing conditions such as road surface, temperature, and tire wear, thereby maintaining reliability across diverse operating scenarios
Solution Approach 2:
The patent implements feedback-based parameter adjustment by using measured vehicle dynamics data to continuously refine tire parameter estimates. The system compares model predictions with actual measurements and adjusts parameters accordingly, ensuring reliability under varying conditions through closed-loop adaptation
3Measurement precision
If nonlinear optimization is used to estimate tire stiffness, then friction and stiffness parameters can be approximated, but productivity deteriorates due to lack of convergence or convergence to local optimum
Solution Approach 1:
The patent applies preliminary linearization to the tire stiffness estimation problem before optimization. By linearizing the relationship between tire parameters and measured quantities around the operating point, the system creates a simplified estimation problem that converges reliably and quickly, maintaining both accuracy and real-time performance
Solution Approach 2:
The patent changes the estimation approach from direct nonlinear optimization of friction and stiffness to a two-stage process: first estimating tire stiffness through linearized methods, then using that to determine friction. This parameter transformation simplifies the mathematical problem, ensuring convergence while maintaining accuracy and enabling real-time operation
4Reliability
If tire stiffness is determined using multiple parameters, then comprehensive tire-to-road interaction is captured, but device complexity increases
Solution Approach 1:
The patent extracts and focuses on the most critical parameter - tire stiffness - as the primary indicator of tire-to-road interaction. By identifying stiffness as the dominant characteristic that encapsulates the effects of multiple underlying factors, the system reduces complexity while maintaining reliable characterization of road conditions and tire behavior
Data Source
Figure 1A
Figure 1B
Figure 1C
AI summary
A method jointly estimates a state of a vehicle including a velocity and a heading rate of the vehicle and a state of stiffness of tires of the vehicle including at least one parameter defining an interaction of at least one tire of the vehicle with a road on which the vehicle is traveling. The method uses the motion and measurement models that include a combination of deterministic component independent from the state of stiffness and probabilistic components dependent on the state of stiffness. The method represents the state of stiffness with a set of particles. Each particle includes a mean and a variance of the state of stiffness defining a feasible space of the parameters of the state of stiffness. The method updates iteratively the mean and the variance of at least some particles using a difference between an estimated state of stiffness estimated using the motion model of the vehicle including the state of stiffness with parameters sampled on the feasible space of the particle and the measured state of stiffness determined according to the measurement model using measurements of the state of the vehicle. The method outputs a mean and a variance of the state of stiffness determined as a function of the updated mean and the updated variance in at least one particle.