A Non-Intrusive Household Load Monitoring Algorithm for High Frequency Based on Multi-Source Data

This paper proposes a high-frequency, non-intrusive household load monitoring algorithm using multi-source data. By constructing a multi-source time series data set with time as the coordinate axis, combining the adaptive splitting module, the multi-source analysis architecture model and the variable convolution module, the algorithm realizes the accurate decomposition of the household load. The algorithm considers environmental impact on electricity habits and performs dynamic high-frequency data analysis, enhancing load monitoring accuracy and efficiency.