卷积神经网络的电力系统设备状态智能识别研究

为提升电力系统设备状态识别效果,文中提出了卷积神经网络的电力系统设备状态智能识别方法.首先采集电力系统设备状态图像,采用卷积神经网络获取图像特征,然后根据图像内状态信息修正卷积神经网络参数,更新权值公式提取修正误差后的设备状态特征,将特征输入神经网络进行学习,建立电力系统设备状态智能识别模型,最后仿真测试结果表明,该方...

Intelligent recognition of power system equipment state based on CNN

In order to improve the results of power system equipment state recognition,an intelligent recognition method of power system equipment status based on convolution neural network is proposed.Firstly,the equipment state image of power system is collected,and the image features are obtained by using convolution neural network.Then,according to the state information in the image,the parameters of convolution neural network are modified,the weight formula is updated,and the equipment state characteristics after error correction are extracted.The features are input into the neural network for learning,and the intelligent recognition model of power system equipment status is established.The method can accurately identify the status of power system equipment,and the accuracy rate of identifying 10 kinds of equipment status is higher than 99%,and the real-time performance is high.