针对遗传算法在求解多目标无功优化方面存在的缺陷,本文提出了基于蜜蜂双种群进化型云自适应遗传算法(double bee population evolutionary cloud adaptive genetic algorithm,BEPE-CAGA)。该算法根据蜜蜂双种群进化思想,引入了雄峰通过竞争参与交叉及雄峰与决定双峰群优秀遗传基因的蜂后交叉的策略,并结合正态云模型云滴的随机性和稳定倾向性特点对其进行改进,改善了算法陷入早熟的问题,提高了算法的收敛速度。建立了以有功网损最小、电压偏差最小及电压稳定裕度最大为目标的无功优化数学模型,并以BEPE-CAGA算法求解该模型。最后通过对IEEE14和IEEE30节点系统进行算例仿真,仿真结果验证了本文所提算法的有效性,同时也证明了该算法在收敛速度和优化效果上具有比基本GA算法和CAGA算法更佳的性能。
英文摘要:
In this paper, a double bee population evolutionary cloud adaptive genetic algorithm is proposed to cope with the limitation of genetic algorithms in solving the multi-objective reactive power optimization. Based on double bee population evolutionary thought, after the introduction of competition to participate in a cross by drones and drones and decided doublets excellent cross bee genetic strategies, and then used the normal cloud model cloud droplet characteristics of randomness and stable tendency, this algorithm solved the problem of genetic algorithm"s premature convergenceSandSspeed up the convergence speed; established a minimum active power loss, voltage deviation minimum and maximum voltage stability margin goals reactive power optimization model and BEPE-CAGA algorithm to solve the model. Finally, through the IEEE14 and IEEE30 node system examples simulation results verify the effectiveness of the proposed algorithm, but also proved that the algorithm is better than the basic GA algorithm and CAGA algorithm performance on the convergence speed and optimization results.