基于模糊神经网络的电力系统连锁故障风险评估

为了减少大规模停电事故，通过分析由保护装置隐性故障造成的电力系统连锁故障的基本过程和基本原理，建立了风险模型，并提出了应用模糊神经网络进行电力系统连锁故障风险评估的指标和方法.利用训练后的模糊神经网络，实现了通过升级保护装置性能来降低电力系统连锁故障风险的预防策略.针对电力系统连锁故障的随机性和严重性的特点，该方法较好地描述了电力系统的安全状态，对系统风险的分类效果好，评估速度快，具有较好的实用性.IEEE 118 bus系统的应用结果表明了该方法的可行性.

Risk assessment of cascading failure in power system based on fuzzy neural network

The basic process and theory of power system cascading failure considering protection system hidden failures were analyzed to decrease bulk blackout failure. The risk model was built, and the risk assessment index and assessment approach of power system cascading failure were proposed based on fuzzy neural network. Some preventive steps were realized by using the trained fuzzy neural network, which could reduce the system cascading failure risk by updating the characters of protection system. This ap- proach grasps the probability and severity of power system cascading failure, so it accurately describes the security state of power system, effectively classifies the system risk, and has faster assessment velocity and good practicality. The feasibility of this approach was illustrated by application in IEEE 118-bus system.