考虑地理因素的改进量子粒子群算法在多目标电网规划中的应用

针对电网规划的多目标权衡优化问题,建立以可靠性和经济性为目标的电网规划模型,提出改进的量子粒子群算法,采用Pareto支配关系来更新粒子的个体和局部最优值,定义粒子紊流极大极小间距,并采用紊流间距方法裁剪非支配解,引入收敛因子K加快粒子跳出局部最优后的收敛速度。同时考虑电网规划存在的地理环境不确定因素的影响,在规划目标函数中引入地理障碍罚因子。通过18节点电网规划算例仿真结果表明,提出的改进算法与基于非支配遗传算法和基于多目标进化算法相比,所得的Pareto解数目,解的优劣情况以及分布效果都有明显提升。

Application of Improved Quantum Particle Swarm Optimization in Power Network Planning Considering Geography Factor

In allusion to the balance optimization of multi-object power network planning and to built power network planning model taking reliability and economy as objects,an improved quantum particle swarm optimization(QPSO) is proposed;the Pareto domination is adopted to update the individual and local optimal of particle as well as to define the max-min distance of particle turbulence,and the non-dominated solution is clipped by turbulent distance;a convergence factor K is led in to speed up the convergence speed of particle that jumps out of local optimal.Meanwhile,the impacts of uncertain factors of geographic environment where the power network being planned is located are taken into account,thus the penalty factor of geographic barrier is led into the objective function of the planning.Simulation results of an 18-bus power network planning show that the number of obtained Pareto optimal solutions,the quality of solutions and their distribution by the proposed improved(QPSO) algorithm are much better than the solutions solved by non-dominated sorting generic algorithm(NSGA) and multi-objective evolution algorithm(MOEA).