采用神经暂态元网络技术的电力系统暂态安全分析

NEURAL-NETWORK BASED TRANSIENT SECURITY ASSESSMENT FOR ELECTRICAL POWER SYSTEMS

提出了一种利用神经元网络技术的电力系统暂态安全分析的新方法。本方法基于多层感知器(perception)网络模型设计分类器,并用此分类器去直接判别电力系统暂态稳定性的类别。实际解算证明,神经元网络技术的应用对于提高模式识别的正确识别率起了较大的作用,并为电力系统实时暂态安全分析开辟了一条新的途径。

This paper proposes a neural-network based new method for transient security as sessment of electrical power systems. By using mutilayer perception network model, a classifier for directly assessing electrical power systems transient stability is designed Actual calculation shows that the application of neural-network technology plays a considerably important part in raising correct recognition rate and opens up a new way for real time transient security assessment or electrical power systems.