A Study Of Improved Cooperative Quantum-behaved Particle Swarm Optimization Algorithm And Its Application In Image Segmentation

Particle swarm optimization algorithm have some flaws, such as it easily falls into local optimal, namely the pre-mature convergence phenomenon. Therefore, the researchers have combined the particle swarm optimization algorithm with quantum mechanics. Sun Jun proposed the quantum-behaved particle swarm optimization algorithm, namely QPSO algorithm. The QPSO algorithm is based on the indefinite principle of quantum mechanics and does not need the velocity of the particle. It only need the location information and only has a control factor. The QPSO algorithm also has certain disadvantages, for example,when QPSO algorithm searches in high-dimensional complex problems, it has poor search ability. Based on the basic principle, model of the particle swarm optimization algorithm and the quantum-particle swarm optimization algorithm, several improved QPSO algorithms are proposed. Primarily this thesis and contribution are as follows:(1) To improve the global search capability of quantum-behaved particle swarm optimization algorithm and guide the search, we propose a modified Quantum-Particle Swarm Optimization algorithm. This method combines dynamic mutation and background collaborative strategies. Among them, we also select dynamic mutation based on the Cauchy mutation and the shrinkage factor, with the purpose to improve the search capabilities of particle. And then, the background of cooperative mainly combining the multiple measurement update with a background variable iteration, its purpose is to improve the efficiency of the algorithm, and change the updating pattern with particle in the search space. In order to verify the algorithm's performance, the algorithm are applied to the optimization of functions and medical image segmentation.(2) When the MCQPSO algorithm are applied to Benchmark function optimization,several functions get poor results, in view of these shortcomings, once again, we put forward the Partitioned-cooperative quantum-behaved particle swarm optimization algorithm. During the process of initialization, the population is divided into several partitions, each partition's population increases exponentially, and each partition makes an iterative optimization, at last, each partition finds the global optimum, and then the globally optimal set up as the primary species. In order to verify the algorithm's performance, the algorithm are applied to the optimization of functions and medical image segmentation.(3) When SCQPSO algorithm are applied into the complex functions optimization, as the function dimension increases, the optimization abilities of SCQPSO algorithm have declined, therefore, some complex function do not converge to the global optimum because of the diversity of population declined sharply at the end of iteration. To solve this problem,we propose an algorithm based on the reverse learning and the cooperative strategies,namely RSCQPSO algorithm. In order to verify the algorithm's convergence, the algorithm is applied to the optimization of complex functions and medical image segmentation.