摘要
提出了神经网络用于多端直流输电系统控制器参数寻优的一种方法。这种直流控制器与传统的PI调节器相比有更好的鲁棒性和自适应学习能力,具有更好的动态性能。控制器采用该策略后,能实现定有功功率控制模式与定直流电压模式之间的自动切换,确保定直流电压控制的换流站故障退出后,仍能实现对直流电压的控制和有功的平衡。通过Matlab仿真实验证明了该控制器对交直流系统稳定有良好的效果。
An optimization method for NN parameters of MTDC controller is proposed. Compared with the traditional PI controller, this controller has better performance on robust, self-adaptive study ability and the dynamic response of the controller. This approach can realize automatic switching active power and DC voltage control modes,so that control of DC voltage and active power balance can still be realized after voltage control mode. Simulation results acquired by using Matlab are show that bility of HVDC system. lost of the converter which is in DC the controller can improve the sta-
出处
《科学技术与工程》
2008年第9期2344-2347,共4页
Science Technology and Engineering
