一种自适应多策略行为粒子群优化算法

针对粒子群优化算法收敛速度慢、局部搜索能力差等缺点,提出一种自适应多策略行为粒子群优化算法.算法中每个粒子拥有4种行为进化策略,在迭代过程中通过计算每种进化策略的立即价值、未来价值和综合奖励来决定粒子的进化行为,并通过策略行为概率变异算法提升个体寻优速度或避免陷入局部最优解.在经典的基准测试函数上,对新算法与其他7个群智能进化算法的测试结果进行比较分析,结果表明所提出算法具有很好的求解精度和收敛速度,尤其适合应用于一些高维优化问题.

An adaptive multi-strategy behavior particle swarm optimization algorithm

Aiming at the shortcomings of slow convergence rate and poor local search ability of particle swarm optimization algorithms, an adaptive multi-strategy particle swarm optimization algorithm is proposed. Each particle has four behavioral evolution strategies in the algorithm. In the iteration process, the evolutionary behavior of the particles is determined by calculating the immediate value, the future value and the comprehensive reward of each evolutionary strategy, and the strategy behavioral mutation algorithm is proposed to improve the individual search speed or to avoid falling into the local optimal solution. Comparison of the results of the proposed algorithm with the other 7 swarm intelligence evolutionary algorithms for the classical benchmark function show that the algorithm has better accuracy and convergence speed, especially suitable for some high-dimensional optimization problems.