基于自适应混沌粒子群优化算法的多目标无功优化

Multi-objective reactive power optimization based on adaptive chaos particle swarm optimization algorithm

针对粒子群无功优化中由于随机生成代表控制变量值的粒子,使得在优化迭代过程中易陷入局部最优解,而且后期收敛速度慢等问题,将混沌优化算法融合到粒子群算法中,提出了混沌粒子群算法求解多目标无功优化问题。该算法在初始化粒子即无功优化控制变量值时,采用混沌思想,增加控制变量取值的多样性;通过粒子群无功优化算法计算各个粒子对应的适应值即无功优化目标函数值,并按照其大小择优选取控制变量值进行混沌优化以帮助无功优化控制变量跳出局部极值区域;并根据无功优化目标函数值自适应地调整其惯性权重系数以提高全局与局部搜索能力。通过算例分析表明,采用自适应混沌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。

Particle swarm algorithm used in reactive power optimization always falls into local optimal solution and final slow convergence due to generating particles as controlling variable values randomly.Consequently,by integrating the chaotic opitimization algorithm into the particle swarm algorithm, a new adaptive chaotic particle swarm optimization based on chaos theory is adopted to solve the problem.Through the using of chaos ergodicity firstly,the control variables in the system are initialized to enhance the diversity of particle populations.For each iteration update of the group,the individual particle＇s fitness value, namely the reactive power optimization objective function value is calculated,and according to their sizes some particles are selected to be treated with chaos optimization to help the reactive power optimization controlling variables to jump out of the local extreme regions; according to each particle＇s fitness value, its inertia weigh coefficient is adjusted adaptively to enhance the entire group of global and local search capabilities.Through calculation and analysis of cases,the results show that adaptive chaotic particle swarm algorithm used in reactive power optimization can jump out of local optimum in time to find the global optimal solution and complete fast convergence.