Identification of Dynamical Systems Using a Broad Neural Network and Particle Swarm Optimization

System identification plays an important role in improving the structure and parameters of a system, but there are many problems encountered in actual operation. The identification of dynamic systems is not as simple as it is for static systems; thus, choosing effective model structures and parameters is the key to solving this problem. This paper proposes a novel algorithm based on a combination of a broad learning system (BLS) and particle swarm optimization (PSO) to identify nonlinear dynamical systems. The proposed method first uses the dimension expansion of the data set as the input of the BLS and then optimizes the model weight by the PSO algorithm. To verify the effectiveness of our proposal, we use four second-order systems for simulation experiments. The simulation results clearly show the efficiency and anti-interference ability of the proposed method.