基于改进离散粒子群算法的多目标无功优化

针对离散粒子群算法直接应用于无功优化后存在优化迭代过程易陷入局部最优解且后期收敛速度慢等问题,结合混沌算法,提出更加有效的改进离散粒子群算法求解多目标无功优化问题。同时,对每次迭代后产生的控制变量进行混沌优化,从而避免无功优化控制变量陷入局部极值区域。通过算例分析表明,采用改进离散粒子群算法进行无功优化,能够及时跳出局部最优得到全局最优解,且收敛速度快。

Multi-objective Reactive Power Optimization Based on Improved Discrete Particle Swarm Algorithm

Discrete particle swarm algorithm used in reactive power optimization always falls into local optimal solution and final slow convergence.Consequently,a more effective improved discrete particle swarm optimization based on the chaos theory is adopted to solve the problem.Setting the initial particle as reactive power compensated disposition,present gears of adjustable transformations and terminal voltage of generations to ensure the first result is not greater than the adaptive result of initial system status.Meanwhile,control variables are optimized by the chaos theory to avoid falling into the local extreme regions.Through calculation and analysis of cases,the results show that improved discrete particle swarm optimization algorithm used in reactive power optimization can jump out of local optimum in time to find the global optimal solution and complete fast convergence.