基于FIS和RBFN的预想事故自动选择  

Automatic Contingency Selection Based on Fuzzy Inference System and Radial Basis Function Network

作　　者：陈刚[1] 田志平[1] 

机构地区：[1]重庆大学输配电装备及系统安全与新技术国家重点实验室,重庆400044

出　　处：《电力系统及其自动化学报》2011年第5期80-85,共6页Proceedings of the CSU-EPSA

摘　　要：针对电力系统预想事故自动选择问题,提出了一种基于模糊推理系统FIS(fuzzy inference system)和径向基函数网络RBFN(radial basis function network)算法。定义了一种有功行为指标PIpf,该指标添加了一个模糊补偿系数用以改善遮蔽现象;同时构造了一个三层的RBFN,该网络以发电机功率、负荷功率和网络拓扑结构作为输入,以PIpf作为输出,并通过离线潮流计算获得训练样本;对算例进行计算并与其他算法比较,结果显示该算法能使事故排序更为合理,且计算精度和速度都令人满意。In view of the problems of automatic contingency selection of power system, an advanced algorithm is proposed, which is based on fuzzy inference system（FIS） and radial basis function network（RBFN）. Firstly an active performance index is defined, which adds a fuzzy compensation factor coefficient to improve shelter phenomenon. Meanwhile a three-layer RBFN is constructed, which treats generator power, load power and network topology as inputs, while treats the active performance index as output. The results of off-line load flow calculation are used to train the RBFN. Finally, the proposed method is demonstrated by an example, compared with several other algorithms. And the results show that the ranking of contingency is much more reasonable, and the calculation accuracy and speed are satisfied.

关 键 词：静态安全分析 预想事故 模糊推理系统 径向基神经网络 反向传播神经网络 

分 类 号：TM715[电气工程—电力系统及自动化]