花粉授粉机制在改进粒子群算法研究

Research on Pollen Pollination Mechanism Employed in Improved Particle Swarm Optimization

针对柔性作业中多目标优化问题,首先构建多目标任务满意度数学模型,该模型以最小加工时间、最低制造成本和最短运输时间为目标,去量纲操作后利用几何平均法求解综合满意度评价值。然后,提出一种改进的粒子群算法(LFPSO),该算法为平衡算法全局和局部搜索能力,惯性权重采用幂函数自适应调节,为改变粒子群前期的搜索性能,在惯性权重中加入了Logistic混沌映射丰富粒子多样性,为平衡全局搜索能力与局部搜索能力,引入花粉授粉机制作为全局搜索阈值。最后,将LFPSO算法与其他算法进行仿真对比,结果验证了LFPSO算法具有良好的性能及解决柔性作业多目标优化问题的有效性。

Aiming at the multi-objective optimization problem in flexible operations,a multi objective task satisfaction mathematical model is first constructed.The model takes the minimum machining time,the lowest manufacturing cost and the shortest transportation time as its objectives,and uses the geometric average method to solve the comprehensive satisfaction evaluation value after dedimensional operation.Secondly,an improved Logistic chaotic map and flower pollination mechanism employed in particle swarm optimization(LFPSO)is proposed.The inertia weight of the algorithm adopts power function adaptive adjustment to balance global search ability and local search ability.In order to increase the search performance of particle swarm in the early stage,Logistic chaos mapping is added to the inertia weight to enrich particle diversity.In order to balance global search ability and local search ability,pollen pollination mechanism is introduced as the global search threshold.Finally,the LFPSO algorithm is simulated and compared with the other three algorithms,and the results verify that the LFPSO algorithm has good performance and the effectiveness of solving the multi-objective optimization problem of flexible operation.