The Method for Extraction and Identification of Non-intrusive Household Appliances Load Features Based on WPT Algorithm

To address the problems of non-intrusive household appliances load identification, such as the difficulty of accurately extracting feature parameters and low accuracy of the identification algorithm, a method for extracting and identifying load features of household appliances based on wavelet packet transform is proposed. First, the wavelet packet decomposition algorithm is used to decompose the voltage and current waveforms into eight energy features, and the effective energy features of household appliances are extracted through the difference analysis method. Then, the accurate identification of different household appliances load is achieved by the error reverse propagation neural network method. Experiments show that the proposed method can effectively identify four types of household appliances load, and it proves good reliability and robustness of this method compared to common methods of signal feature extraction for household appliances load.