余芸, 明哲, 崔焱, 甘莹. 云PaaS层下的数字化电网中台快速响应模式自动化识别方法[J]. 电网与清洁能源, 2023, 39(4): 105-110.
引用本文: 余芸, 明哲, 崔焱, 甘莹. 云PaaS层下的数字化电网中台快速响应模式自动化识别方法[J]. 电网与清洁能源, 2023, 39(4): 105-110.
YU Yun, MING Zhe, CUI Yan, GAN Ying. An Automatic Identification Method for Rapid Response Mode of Digital Power Grid under Cloud PaaS Layer[J]. Power system and Clean Energy, 2023, 39(4): 105-110.
Citation: YU Yun, MING Zhe, CUI Yan, GAN Ying. An Automatic Identification Method for Rapid Response Mode of Digital Power Grid under Cloud PaaS Layer[J]. Power system and Clean Energy, 2023, 39(4): 105-110.

云PaaS层下的数字化电网中台快速响应模式自动化识别方法

An Automatic Identification Method for Rapid Response Mode of Digital Power Grid under Cloud PaaS Layer

  • 摘要: 现有方法未能对数字化电网中台响应信号进行预处理,导致响应模式的识别结果不准确、识别时间较长。提出一种云PaaS层下的数字化电网中台快速响应模式自动化识别方法。在云PaaS层下,通过Bayes估计的小波阈值去噪方法对含有噪声的电网中台快速响应信号进行预处理;设置模型的可见层、隐藏层和输出层,计算各层间的权重矩阵,对预处理后的响应信号进行训练,并将训练后的数据输入到RBM模型中,获取PaaS层下的数字化电网中台快速响应模式的识别结果。实验可知,所提数字化电网中台响应模式的识别时间最多为53 s,识别精度最高可达99%,召回率最高为98%,F1值最高为99%,表明该方法能够以较快的速度和较高的精度完成模式识别。

     

    Abstract: Because the existing methods fail to preprocess the response signal of digital power grid,resulting in inaccurate response pattern recognition results and long recognition time,an automatic recognition method of fast response pattern of digital power grid under cloud PAAS layer is proposed. Under the cloud PAAS layer,the wavelet threshold denoising method based on Bayes estimation is used to preprocess the fast response signal of power grid with noise. Set the visible layer,hidden layer and output layer of the model,calculate the weight matrix between each layer,train the preprocessed response signal,and input the trained data into the RBM model to obtain the fast response pattern recognition results of digital power grid under PAAS layer. The experimental results show that the response pattern recognition time of the proposed method is 53s,the recognition accuracy is up to 99%,the recall rate is 98%,and the F1 value is 99%.

     

/

返回文章
返回