Quantum Particle Swarm Optimization Algorithms Based On Truncated Mean Stability And Diversity Migration Strategy

With the development of artificial intelligence,engineering design,etc,the related optimization problems are becoming increasingly complex.Therefore,many advanced swarm intelligence algorithms are utilized to solve the optimization problems,and the particle swarm optimization algorithm with high flexibility and strong adaptability is the basis for improving most optimization algorithms.However,the search performance of the standard particle swarm optimization algorithm is poor in the complex high-dimensional optimization problems,and the particle swarm optimization algorithm cannot update the population samples during the iterative process,which makes the population more likely to be in a local optimal state.By combining with quantum information theory,quantum particle swarm optimization algorithms have been investigated.Based on the existing quantum particle swarm optimization algorithm,two improved quantum particle swarm optimization algorithms were proposed by introducing the truncated mean stabilization strategy and the diversity migration strategy,respectively.The specific work of this dissertation is as follows.To improve the diversity of population and the convergence accuracy of the quantum particle swarm optimization algorithm while dealing with the complex high-dimensional optimization problems,a quantum particle swarm optimization algorithm based on a truncated mean stabilization strategy was designed.In the proposed algorithm,the processed particles were intercepted with equal proportions and the average of the indices of the remaining particle positions was calculated.Then the worst particle was replaced by the new particle during the iterative process.The proposed strategy can improve the population diversification and the convergence efficiency.The convergence performance of the quantum particle swarm optimization algorithm with truncated mean stabilization strategy was researched on the benchmark function optimization problems of different dimensions.It is shown that the search accuracy and the convergence of the quantum particle swarm optimization algorithm with the truncated mean stabilization strategy are better than those of the typical particle swarm optimization algorithms.By introducing a new migration mechanism,a quantum particle swarm optimization algorithm based on diversity migration strategy was presented.This strategy can capture individuals of different ranges in the population,and the selection of migrating individuals not only depended on the size of the fitness value,but also was affected by the position of the particle in the population.The individual with the smallest average Hamming distance can indicate the direction of population iterative optimization.The fitness values and the average Hamming distance between particles were compared and the particles that deviate from the center range of the population were replaced.In the simulation experiments under different conditions,the diversity migration strategy was compared with the traditional migration strategy.It is demonstrated that the performance of the quantum particle swarm optimization algorithm with the diversity migration strategy is better than that of the algorithm based on the standard migration strategy.