A Study on the Detection of Sub-Metering Errors of Electrical Energy in Distribution Networks Based on Wavelet Neural Networks

The calculation of the error of sub-item metering of electric energy in the distribution network is conducive to improving the operational efficiency and energy management level of the distribution network, and the detection method of the error of sub-item metering of electric energy in the distribution network based on a wavelet neural network is studied. When improving operational efficiency and energy management level, calculating sub item measurement errors is extremely critical. This method utilizes the consolidation of sub item measurement data by merging units, and uses the dual laws of thermal and electrical energy to determine measurement errors. Using wavelet packet decomposition with db4 wavelet as the basis function, determine the characteristic frequency band with the highest energy value for the power measurement signal. Input the decomposed signal into a radial basis function neural network, which maps the input data using the radial basis function of the hidden layer and outputs the error detection result. The test results show that this method can accurately detect the error of the power sub-item metering in the distribution network, and the mean square error of the error detection of the power sub-item metering is less than 0.3.