Improved QPSO with Selected Random Mean for Electromagnetic Problems
Anton Duca · 2021
A new QPSO (Quantum-behaved Particle Swarm Optimization) with selected random mean is proposed in this paper for solving electromagnetic optimization problems. The selection procedure is a probabilistic mechanism indirectly based on the fitness measure of the solutions, solutions which are candidates for the QPSO mean term. The selection is controlled through a locality search parameter which is adapted during the optimization process to improve the overall performance of the algorithm. To test the performances two benchmark electromagnetic functions are considered, namely TEAM22 and Loney's solenoid.