Vehicle Tire Stiffness Calibration via Probabilistic Motion Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for calibrating tire parameters in vehicles, such as friction coefficient and stiffness, are challenging due to the difficulty in direct measurement during driving and reliance on high-precision sensors or test rigs, which are costly and limited in real-world applicability.
Innovation Solution
A system and method using low-cost sensors available in standard vehicles to estimate tire parameters probabilistically, employing a combination of deterministic and probabilistic motion models to iteratively update the probability distribution of tire stiffness, allowing for real-time calibration and adjustment based on road surface changes and tire pressure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-precision sensors or test rigs are used to calibrate tire parameters, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent replaces expensive high-precision sensors with low-cost sensors that are already available in standard vehicles. Instead of using specialized test equipment, the system utilizes inexpensive existing sensors (accelerometers, gyroscopes, wheel speed sensors) to achieve tire parameter calibration, effectively substituting costly components with cheaper alternatives that suffice for the application
Solution Approach 2:
The system enables the vehicle to calibrate its own tire parameters using its existing sensor suite without requiring external test rigs or specialized equipment. The calibration process is performed autonomously by the vehicle's control system using data from its own sensors, eliminating the need for external calibration facilities
2Measurement precision
If test rigs are used for tire calibration, then measurement precision is improved, but adaptability to real-world conditions deteriorates
Solution Approach 1:
Instead of attempting to replicate real-world conditions in a controlled test rig environment, the patent inverts the approach by bringing the measurement system into the real world. The calibration is performed in actual driving conditions using the vehicle's existing sensors, accepting real-world variability rather than trying to eliminate it through controlled testing
Solution Approach 2:
The system transitions from static test rig measurements to dynamic in-vehicle measurements. The calibration process adapts to changing driving conditions, road surfaces, and vehicle states, allowing tire parameters to be calibrated dynamically during normal operation rather than in fixed test environments
3Adaptability or versatility
If nonlinear optimization is used to estimate tire parameters, then adaptability is improved, but reliability deteriorates due to convergence issues
Solution Approach 1:
The patent replaces complex nonlinear optimization algorithms with a simpler linear regression approach. Instead of using iterative optimization methods that are prone to convergence problems, the system uses direct linear estimation techniques that are computationally simpler and more reliable, substituting a problematic computational mechanism with a more stable alternative
Data Source
Figure 1A
Figure 1B
Figure 1C
AI summary
A tire of a vehicle is calibrated based on a motion model of the vehicle relating control inputs to the vehicle with a state of the vehicle and a measurement model of the vehicle relating measurements of the motion of the vehicle with the state of the vehicle. The motion model of the vehicle includes a combination of a deterministic component of the motion and a probabilistic component of the motion, wherein the deterministic component of the motion is independent from the state of stiffness and defines the motion of the vehicle as a function of time. The probabilistic component of the motion includes the state of stiffness having an uncertainty and defining disturbance on the motion of the vehicle. The tire is calibrated based on motion data that include a sequence of control inputs to the vehicle that moves the vehicle according to the trajectory and a sequence of measurements of the motion of the vehicle moved along the trajectory by updating iteratively a probability distribution of the state of stiffness until a termination condition is met. An iteration determines a first state trajectory of the vehicle according to the motion model using the sequence of control inputs and one or multiple samples of the probability distribution of the state of stiffness, determines a second state trajectory of the vehicle according to the measurement model using the sequence of measurements, and updates the probability distribution of the state of stiffness to reduce an error between the first state trajectory of the vehicle and the second state trajectory of the vehicle.