中文摘要：

在遮蔽情况下,光伏阵列输出的P-U特性曲线呈现出多峰状态,传统最大功率点跟踪（MPPT）算法可能会陷入局部极值。提出了一种基于改进萤火虫算法（IFA）的MPPT算法,首先在原始算法步长更新公式中引入具有变权重系数的惯性项,使得追踪过程能很好地实现全局搜索能力与局部寻优的平衡;其次在步长更新公式中引入加速系数,在功率值较低的区域加大步长,减少追踪所用时间和能量损失。最后,通过仿真和实验,与经典粒子群优化（PSO）算法进行了比较,证明了所提算法跟踪能力更优。

英文摘要：

Photovoltaic array has multimodal state on the P-U curve under the shaded condition.Traditional maximum power point tracking（MPPT） may get into local maximum.A MPPT is proposed based on the improved firefly algorithm（IFA）.Firstly,inertia item with the variable weight coefficient is introduced the into the original algorithm step update formula,so that the balance of the global search ability and local optimization can be well in the process of tracking.Secondly,acceleration coefficient is introduced into step update formula and increase the step length in the area of low power value in order to reduce the time of tracking and the energy loss.Finally,simulation and experiment confirm the ability of the tracking algorithm by compared with classic particle swarm optimization（PSO） algorithm.