论文部分内容阅读
针对无功优化中连续变量和离散变量共存的特点,对简单遗传算法编码方式、交叉算子和变异算子进行了重新确定。为了使算法能较快收敛并且能以较大的概率跳出局部最优,在总结前人成果的基础上,提出了一种改进自适应变异概率遗传算法。改进的自适应变异概率既考虑了种群中个体的适应度值情况,同时也计及了算法所处的阶段。论文提出的方法既能保持典型自适应概率较快收敛的特性,又通过在算法后期产生的更多新基因维持种群多样性,从而拥有更强的寻优能力,所求得的有功网损更小。最后通过对某地区电网的仿真计算证明了所提算法在求解无功优化问题时的正确性和有效性。
Aiming at the coexistence of continuous variables and discrete variables in reactive power optimization, the coding methods of simple genetic algorithm, crossover operator and mutation operator were re-determined. In order to make the algorithm converge faster and jump out of the local optimum with larger probability, a genetic algorithm based on improved adaptive mutation probability is proposed based on the conclusion of previous achievements. The improved probability of adaptive mutation not only considers the fitness value of individuals in the population, but also considers the stage of the algorithm. The proposed method not only maintains the characteristic of quick convergence of typical adaptive probability, but also maintains the diversity of population through more new genes generated in the latter part of the algorithm, so as to have better searching ability. The obtained active network loss small. At last, the correctness and effectiveness of the proposed algorithm in solving reactive power optimization problems are proved through the simulation of a local power network.