【摘要】 基于动态神经网络构造了一种电力系统稳定器（ＮＮＰＳＳ）。通过单、多机数字仿真研究了神经网络控制器应用于电力系统实时控制所面临的两个问题：（１）未知扰动下控制输出的合理性；（２）单机－无穷大母线系统下训练的控制器在多机系统中的控制效果。文章详细介绍了ＮＮＰＳＳ的设计思想，训练样本选取单、多机仿真结果分析。更多还原

【Abstract】 A dynamic neural network based power system stabilizer(NNPSS)is developed.Bythe digital simulation tests of this NNPSS in single-machine infinite-bus power system and a multi-machine power system,this paper discusses the following two problems neural networkpower system controller facing①whether reasonable and effective control can be generated whenunknown disturbance occur :②whether NN controller trained in single machine power systemwork well in multi machine power systems. The controller designing,selection of trainingsamples and the single and multi machine power system simulation results analysis are allpresented in detailin this paper.更多还原