Application of an improved PSO-based PID neural network controller for VSC-HVDC

作者：Li, Shuang^*; Wang, Zhixin; Wang, Guoqiang

来源：Proceedings of the CSEE, 2013, 33(3): 14-21.

• Chaotic mutations
• Fitness sharing
• High-voltages
• Off-shore wind power
• PID neural network
• Restricted competition selection (RCS) niche
• Tent map

An improved niche chaotic particle swarm optimization (INCPSO) method for voltage source converter based high-voltage direct-current (VSC-HVDC) system was presented. It emerges the good features of niche and chaotic mutation evolutionary algorithms for searching the global best parameters of proportional-integral-derivative neural-network (PIDNN). With the introduction of restricted competition selection (RCS) method into the niche mechanism, the global exploration ability of PSO is enhanced. Simultaneously, the precise local exploitation performance can be well achieved by using the chaotic mutation algorithm with tent-mapping functions. The INCPSO can solve the premature phenomenon and local convergence problem of traditional PSO, and give a balance point between the exploration and exploitation capability. The detail steps of parameters optimizing for PIDNN controller in VSC-HVDC system was presented, and a numerical example was carried out to validate the feasibility and effectiveness of INCPSO. Simulation results show that the INCPSO search capability is more accurate, effective and robust. The INCPSO based PIDNN controller provides a feasible control scheme in the offshore wind farm VSC-HVDC system.

• 出版日期2013
• 单位上海交通大学