基于人工神经网络的发电机故障暂态稳定性预测方法研究

刘  辉1; 郑  剑2; 刘行行1

基于人工神经网络的发电机故障暂态稳定性预测方法研究

引用本文：刘辉1，郑剑2，刘行行1.基于人工神经网络的发电机故障暂态稳定性预测方法研究[J].计算技术与自动化,2022,(4):7-11

摘要点击次数: 157

全文下载次数: 0

作者	单位
刘辉1，郑剑2，刘行行1	（1.河北冀研能源科学技术研究院有限公司，河北石家庄 050000；2.国网河北石家庄供电公司，河北石家庄 050000）

中文摘要:提出了一种利用径向基函数神经网络（RBFNN)预测大扰动后发电机转子转角值的方法，来实时判断系统暂态稳定状态，并对相干发电机组的辨识进行了研究。在故障后将相量测量单元（PMU）同步采样的前六个周期的发电机的转子角度和电压等数据作为神经网络的输入，以预测系统未来的状态。该方法还可以实时判断发电机的同步状态。在测试系统上对该方法以不同运行条件进行了评估，实验结果证明所提出的径向基函数神经网络对扰动后的转子转角值具有良好的预测性能，适合于实时应用。

中文关键词:故障预测暂态分析人工神经网络

Research on Fault Transient Stability Prediction Method of Generator Based on Artificial Neural Network

Abstract:A method using radial basis function neural nework (RBFNN) to predict the rotor angle of the generator after large disturbances is proposed to judge the transient and stable state of the system in real time, and to identify the coherent generator set. Research. After a fault, the phasor measurement unit (PMU) synchronously sampled the rotor angle and voltage of the generator in the first six cycles as the input of the neural network to predict the future state of the system. This method can also judge the synchronization status of the generator in real time. The method is evaluated under different operating conditions on the test system. The experimental results prove that the proposed radial basis function neural network has good predictive performance on the rotor angle value after disturbance, and is suitable for real-time applications.

keywords:fault prediction transient analysis artificial neural network

查看全文 查看/发表评论 下载pdf阅读器