改进的量子行为粒子群优化算法及其应用

为提高量子行为粒子群算法的优化能力，提出了一种改进的算法．该算法也采用量子势阱作为寻优机制，但提出了新的势阱中心建立方法．在每步迭代中，首先计算粒子适应度，然后取前K个适应度最好的粒子作为候选集．采用轮盘赌策略在候选集中选择一个粒子作为势阱中心，调整其它粒子向势阱中心移动．在优化过程中，通过使K值单调下降，获得探索与开发的平衡．将提出的算法应用于标准函数极值优化和量子衍生神经网络权值优化，实验结果表明提出算法的优化能力比原算法确有明显提高．

Improved Quantum-behaved Particle Swarm Optimization Algorithm and Its Application

To enhance the performance of quantum-behaved particle swarm optimization, we propose an improved algorithm that uses quantum potential as well as an optimization mechanism and establishes the center of the potential well. In each iteration, we first calculate the fitness of each particle and then take the first K particles with the greatest fitness as the candidate set. We use a roulette strategy to select a particle from the candidate set as the center of the potential well and move other particles toward the center of the well. During optimization, the K value is monotonically decreased to achieve a balance between exploration and exploitation. We apply the proposed approach in the extremum optimization of benchmark functions and weight optimization of a quantum-inspired neural network. Experimental results demonstrate that the optimization ability of the proposed algorithm is quite competitive with that of the original algorithm.