中文摘要：

光伏阵列的P-U特性曲线在局部阴影的情况下会呈现多个极值点,传统算法容易陷于局部极值,智能算法追踪耗时过多.在研究两类算法的基础上,提出一种基于扰动观察法和粒子群算法的3步最大功率点跟踪（MPPT）算法.该算法采用大步长扰动观察法缩小搜索范围并确定粒子数目;采用改进粒子群算法实现全局搜索寻找最优局部;采用逐步逼近的扰动观察法在最优局部内搜索最大功率点.仿真结果表明：该算法在均匀光照和局部遮阴情况下均能准确迅速地跟踪到最大功率点,相比于粒子群算法,追踪时间缩短35%,以上.

英文摘要：

Multiple local maximums would be exhibited on the power-voltage characteristic curve of photovoltaic array,under partially shaded conditions.Conventional maximum power point tracking（MPPT）algorithms tend to get into local maximum,while the intelligent MPPT methods would spend much time in tracking.After studying the conventional and intelligent MPPT algorithms,a three-step MPPT algorithm based on perturbation and observation（PO）algorithm and particle swarm optimization（PSO） algorithm was proposed in this paper.The three-step algorithm used PO method with large step to narrow the search range and determined the number of particles,used PSO algorithm to achieve global search and to find the global optimal local,and used approximation PO method to find the global maximum power point in the optimal local.Simulation results proved the three-step algorithm can track the global maximum power point quickly and accurately under uniform condition as well as partially shaded conditions.The tracking time of the three-step MPPT algorithm is 35%, or more shorter than that of PSO algorithm.