Vehicle multi-dimensional force sensor decoupling method based on PSO optimized LSSVM

Abstract

The invention discloses a vehicle multi-dimensional force sensor decoupling method based on a PSO optimized LSSVM, which comprises the steps of 1, acquiring calibration data of a vehicle multi-dimensional force sensor; 2, building a least square support vector machine model for decoupling of the vehicle multi-dimensional force sensor; and 3, optimizing and training a PSO algorithm to obtain an optimal vehicle multi-dimensional force sensor decoupling model. According to the invention, the least square support vector machine decoupling model is established, and the PSO (particle swarm optimization) is used to globally search the optimal characteristic to optimize the parameters of the least square support vector machine, so that the least square support vector machine decoupling model has good convergence and adaptability, and the multi-dimensional force sensor has good measurement precision to meet the actual application requirements.

Description

technical field [0001] The invention relates to the field of methods related to multidimensional force sensors, in particular to a decoupling method for vehicle multidimensional force sensors based on PSO optimized LSSVM. Background technique [0002] During the actual driving process of the car, the wheels are affected by six-dimensional forces (lateral force, vertical force, longitudinal force, roll moment, yaw moment, torque), these forces are important for the monitoring of the driving state of the car and the control of the vehicle. important. The traditional method uses single-dimensional force measurement sensors in all directions, the axial size is large, and the installation requires more space, which cannot meet the requirements of wheel multi-dimensional force measurement. application is becoming more and more widespread. The wheel multi-dimensional force sensor adopts the spoke structure and the elastic body directly connected with the wheel drum and the wheel ...

AI Technical Summary

Problems solved by technology

However, due to the integrated elastic body structure of the vehicle multi-dimensional force sensor and the error in the manufacturing process, there is an inevitable coupling on the conversion channels of the sensor in different dimensions, that is, inter-dimensional coupling, which seriously affects the sensor. Therefore, it is very important to decouple the multi-dimensional force sensor for vehicles

The decoupling method based on the least squares method can solve the linear decoupling problem very well, but in the process of calculating the nonlinear decoupling, the method often has low numerical accuracy, poor robustness, and easy local optimum. solution

Neural network-based methods require a large number of calibration samples for training, otherwise it is difficult to achieve satisfactory decoupling accuracy

Images