A new particle swarm optimization algorithm with dynamically changing the inertial weight and threshold value based on the improved particle swarm optimization is proposed

in which the inertial weight of the particle is adaptively adjusted based on the premature convergence degree of the swarm and the fitness of the particle.The diversity of inertial weight makes a compromise between the global convergence and local convergence

so it can effectively alleviate the problem of premature convergence.The algorithm is applied to train neural network and a model of fault diagnosis for steer-by-wire is established.The simulation results illustrate that compared with particle swarm optimization algorithm and genetic algorithm

the proposed algorithm can effectively improve the training efficiency of neural network

speed up the convergence rate and obtain good diagnosis results.