具有量子行为的协同粒子群优化算法

以分布估计算法(EDA)的角度,从理论上指出,具有量子行为的粒子群优化算法(QPSO)本质上是EDA算法与原始粒子群算法(SPSO)的综合.针对进化类算法普遍遇到的过早熟问题,将协同搜索策略引入传统的QPSO算法,提出了具有量子行为的协同粒子群优化算法(MQPSO).通过实验确定了最适合MQPSO算法的通信频率以及子种群大小.实验结果表明,该算法较QPSO及SPSO算法具有更快的收敛速度和更强的搜索精度,其优势在高维优化问题中更为明显.

Quantum-behaved particle swarm optimization algorithm with cooperative approach

Quantum-behaved particle swarm optimization algorithm(QPSO) is investigated from the perspective of
estimation of distribution algorithms(EDAs) for the first time, which proves that QPSO is a combination of EDAs and original
particle swarm optimization. A quantum-behaved particle swarm optimization algorithm based on cooperative search strategy
is presented, which helps prevent the evolutionary algorithms’ universal tendency to be easily trapped into local optima as
a result of the rapid decline in diversity. Communication frequency and the size of each sub-swarm are ensured through
experiments to obtain the most effective setting for this algorithm. Experiment results show that this algorithm is able to find
better solutions than the original QPSO and particle swarm optimization algorithm with higher efficiency.