Identification of nonlinear systems with outliers using modified quantum particle swarm optimization
Chia-Nan Ko, Ching-I Lee · 2016
This paper presents a modified quantum-behaved particle swarm optimization (MQPSO) based on hybrid evolution (HEMQPSO) approach is proposed to identify nonlinear systems with outliers. In HEMQPSO algorithm, combining the conceptions of genetic algorithm (GA) and adaptive annealing learning algorithm with the MQPSO algorithm is adopted to search optimal solutions. Simulation results are illustrated to verify the performance of identification for nonlinear systems. From the numerical simulations and comparisons with other extant evolutionary methods, the validity and superiority of the HEMQPSO approach are verified for identifying nonlinear systems with outliers.