Analysis and Prediction Method for the Ground Wave Propagation Delay Characteristics of Long‐Wave Time Signals

ABSTRACT At present, researchers often focus on various aspects such as the meteorological conditions along the long‐wave propagation path and the conductive characteristics of the Earth's surface, and analyse the long‐wave ground wave propagation delay one by one. If a unified analysis and modelling can be carried out for the long‐wave propagation delay affected by multiple factors, it will significantly improve its prediction efficiency and engineering implementation efficiency. Firstly, this paper conducts analysis from the time domain and the frequency domain, and discovers its obvious periodic time‐varying characteristics. Based on this, a quadratic polynomial model with periodic terms is constructed and used as the first prediction model. Next, in order to improve the accuracy and make full use of historical data, the new predicted values are inserted into the sequence of measured values to construct a quadratic polynomial long‐wave prediction model based on the sliding window and with periodic terms (Model 2). Finally, these two models are used to make predictions for different durations on the 7‐day delay data. The actual measurements show that in the short term, the accuracies of the two models are equivalent, and when the prediction duration increases, the advantages of Model Two become more apparent.