The Approximation of the Train Resistance Based on Improved PSO-RBF

The calculation and research on the traction braking performance of trains is the basic work that must be performed in the train design, and it is also an important foundation for the effective control of the train traction systems. This paper studies the traction and braking characteristic of high-speed trains, and simulates the changes in the basic resistance coefficients of the following vehicles under different passenger loads and gradients. The basic resistance of the train is approximated by the improved PSO-RBF (Particle swarm optimization radial basis function) neural network. Compared with the improved PSO-RBF neural network and the RBF neural network, the proposed method has better accuracy in approaching the basic resistance of trains.