A Novel Hybrid PSO Identification Method Simultaneously Estimate Model Parameter and the Structure

Estimate model parameter and the structure simultaneously is a crucial and challenging problem in system identification. For solving the problem, a hybrid algorithm by integrating two-order oscillating particle swarm optimization with successive quadratic programming is proposed in this paper. The two-order oscillating particle swarm optimization is showed to converge rapidly to a near optimum solution, but the search process will become very slow around global optimum. On the contrary, successive quadratic programming is weak to escape local optimum but the ability of convergent speed around global optimum and the convergent accuracy is strong. In this case, the two-order oscillating particle swarm optimization is used to enhance global search ability and convergence speed of algorithm. When the change in fitness value is smaller than a threshold value, the searching process is switched to successive quadratic programming. In this way, this hybrid algorithm may find an optimum solution more accurately. To validate the performance of the proposed approach, it is evaluated on four optimal control problems. Results demonstrate the effectiveness and accuracy of the proposed algorithm.