电压稳定极限附近神经网络模拟及静态稳定裕度判定

NEURALNETWORK BASED SIMULATION NEIGHBORING THEAREAOF VOLTAGESTABILITY LIMITAND STABILITY MARGIN ASSESSMENT

基于系统电压静态稳定的观点，对给定系统给定运行状态，在求取某节点一对相关邻近潮流解的基础上，用误差反向传播模型的人工神经网络来模拟该节点电压静态稳定临界状态附近的过渡过程，较为准确地确定了节点电压稳定的临界状态，并可以确定过渡过程中任意状态下节点电压静态稳定裕度。

According to acertain system operating in a given m ode, an errorback-propagation neuralnetw ork with a variable learning factor is used to sim ulate pow ersystem transientofa load busaround the static voltage stability lim it. Sam ples used to train the neuralnetwork are obtained from the load flow m ulti-solution and the relative nearby solutions. As a result, the static voltage stability lim it can be estim ated accurately, and the static voltage stability m argin atthe load buses for a given operation condition can be assessed correctly.