Based on the braking test data of different driving condition in many cities across China

this paper studies a recognition method for braking intention based on PSO-SVM. Firstly

after preprocessing the data of the automotive braking test

a set of effective braking data for 8 parameters is obtained. Then

the principal component analysis (PCA) algorithm is used to select the maximum deceleration

maximum pedal stroke

and maximum pedal force as the identification parameters for braking intention

and the braking intention is divided into mild braking

moderate braking and severe braking. A support vector machine (SVM) model for brake intention recognition was constructed

and particle swarm optimization algorithm was used to optimize the penalty factor c and kernel function parameter g. Finally

the PSO-SVM algorithm is offline verified according to the real vehicle braking test data. The results show that the PSO-SVM algorithm can effectively identify the braking intention

and the training and recognition speed is faster than the BP neural network algorithm and the short-term memory network algorithm optimized by the genetic algorithm when the recognition success rate is high.