Analysis and prediction of wideband electromagnetic radiation in residential environment

Aiming at the problem that the electromagnetic radiation in the residential environment is mixed with many high-frequency components in the periodic law, which leads to the limited prediction performance of traditional time series modeling methods and neural network methods, a hybrid prediction method combining wavelet decomposition and seasonal difference autoregressive moving average (SARIMA) model is proposed. According to the time-frequency characteristics of radiation data, the hybrid method uses SARIMA model to predict the major periodic components and detail components obtained by wavelet decomposition, so as to adapt to the complex electromagnetic radiation conditions caused by multiple emission sources in the residential environment. The experimental results show that this method not only has higher prediction accuracy than single time series modeling method and neural network method, but also has stronger adaptability and stability of outliers.