Tire Force Estimation Using CAN-Bus Sensor Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to accurately and robustly estimate tire normal, lateral, and longitudinal forces in real-time during vehicle operation, especially over the lifetime of a tire tread, due to reliance on indirect and load-dependent sensor measurements.
Innovation Solution
A method that accesses CAN-bus sensor data from multiple vehicle-mounted sensors to estimate tire forces using normal, lateral, and longitudinal force estimators, along with additional estimators for roll and pitch angles, center of gravity, and yaw inertia, excluding GPS and suspension displacement data, to provide robust and accurate force measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If indirect tire and vehicle sensor measurements are used to estimate tire forces, then the system complexity is reduced and standard vehicle sensors can be utilized, but the measurement precision and reliability of tire force estimation deteriorates over the lifetime of a tire tread
Solution Approach 1:
The system dynamically adapts inertial parameters (mass, center of gravity position, yaw inertia) based on operating conditions and tire wear state. This allows the estimation model to maintain accuracy despite changes in tire characteristics over time, resolving the contradiction between using simple sensor inputs and achieving precise force estimation throughout the tire lifecycle
Solution Approach 2:
The system implements continuous feedback through real-time estimation of tire forces and comparison with expected values. This feedback mechanism enables dynamic adjustment of inertial parameters and detection of tire wear states, maintaining measurement precision without increasing hardware complexity
2Adaptability or versatility
If load-dependent inertial parameters are used in the estimation model, then the model can adapt to different vehicle loading conditions, but the reliability of tire force estimation deteriorates due to susceptibility to erroneous sensor readings
Solution Approach 1:
The system transitions from static inertial parameters to dynamic, adaptive parameters that change with operating conditions. The mass, center of gravity, and yaw inertia are continuously updated based on sensor inputs and estimation algorithms, enabling the model to adapt to loading conditions while maintaining reliability through robust estimation techniques
Solution Approach 2:
The system prepares for potential sensor errors by implementing redundancy and cross-validation among multiple sensors. The adaptive inertial parameter estimation incorporates checks to detect and compensate for erroneous readings before they significantly impact the reliability of tire force estimation
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
A method for estimating a tire state including normal force, lateral force, and longitudinal force on a tire mounted to a wheel and supporting a vehicle is disclosed. The method comprises: accessing a vehicle CAN-bus for vehicle sensor-measured information; equipping the vehicle with a plurality of CAN-bus accessible, vehicle mounted sensors providing by the CAN-bus input sensor data, the input sensor data including acceleration and one or more angular velocities, steering wheel angle measurement, angular wheel speed of the wheel, roll rate, pitch rate, and yaw rate; deploying a normal force estimator operable to estimate a normal force on the tire from a summation of longitudinal load transfer, lateral load transfer and static normal force using as inputs lateral acceleration, longitudinal acceleration and roll angle derived from the input sensor data; deploying a lateral force estimator operable to estimate a lateral force on the tire from a planar vehicle model using as inputs measured lateral acceleration, longitudinal acceleration and yaw rate derived from the input sensor data; and deploying a longitudinal force estimator operable to estimate a longitudinal force on the tire from a wheel rotational dynamics model using as inputs wheel angular speed and drive/brake torque derived from the input sensor data.