An Improved Grasshopper Optimization Algorithm for Optimizing Hybrid Active Power Filters' Parameters

The selection of parameters for the hybrid active power filter (HAPF) is essential to harmonic compensation. To optimize HAPF parameters, this paper presents an improved grasshopper optimization algorithm (IGOA). In the IGOA, the whole population is divided into two sub-populations which focus on exploration and exploitation respectively. An improved social interaction mechanism is proposed to balance global and local searches. Furthermore, a learning strategy is introduced and an exemplar pool is built to replace the target in the original GOA, which can enhance the global search ability and escape local optima. The proposed IGOA is employed to optimize the parameters of two prevalent HAPF topologies for several cases. The experimental results show that the IGOA can get a promising performance compared with previous studies and other meta-heuristic algorithms.