随着电力系统的广泛发展,电力系统静态安全评估已变得越来越重要。文中比较了现在几种常用的人工智能方法,选择了支持向量机算法解决这一问题。由于解决大样本问题时,支持向量机所需训练时间显著增加,文中提出了约简样本的方法,并结合适合于电力系统的二叉树结构,提出了一种改进的简化二叉树支持向量机算法。将这种新的支持向量机算法应用于IEEE57节点电力系统,结果表明,文中提出的算法取得了比较好的结果,有效可行。
Power system static security assessment is becoming more and more important with the expansion of electrical power system. It compares several common artificial intelligence methods and then selects the support vector machine (SVM) algorithm. Due to the large number, of the training set, the training time of SVM increases significantly. To solve this problem, an improved simplifying SVM method is put forward in this article. It decreases the number of the samples and combines the suitable binary tree structure for power system. The propo~d simplified SVM algorithm has been applied to IEEE 57-bus power system. The simulation results demonstrate the effectiveness of the proposed algorithm.