Application of PSO and PIDNN controller for VSC-HVDC

作者：Wang, Guoqiang^*; Wang, Zhixin

来源：Proceedings of the CSEE, 2011, 31(3): 8-13.

• Conventional-PI controller
• Feasible control strategy
• Grid-side converter
• PID neural network
• Simulation research
• Steady state performance
• Structure characteristic
• VSC-HVDC

The double closed-loop PI regulator is widely used for the control of wind farm side and grid side converters in the VSC-HVDC system for offshore wind farms. But there are too many PI regulators, so it is difficult to adjust all the parameters. In this paper we designed PID neural network controller combined with particle swarm optimization (PSO). And it was used to control the converters of VSC-HVDC. First simple improvement was carried out based on PIDNN structure characteristic. We set the linking weights of the first two layers as constant values. Only the interlayer and output layer linking weights of PID neural network (PIDNN) were used as the optical parameters of PSO. It reduced dimension of particles and improved training speed obviously. Then the conventional PID controller was replaced with PIDNN controller. Simulation research was carried out based on transfer function of converters control system. Simulation results show that the transient and steady state performance is improved compared with conventional PI controllers. Training times can be reduced significantly compared with conventional PIDNN and PSO algorithm. It lays a foundation for on line training and improves a feasible control strategy for VSC-HVDC.

• 出版日期2011
• 单位上海交通大学