Prediction of Optical Power Data Based on Optimized ARIMA Model

Abstract

In order to make the operation of optical fiber protection system more stable and improve the accuracy of time series prediction for a small amount of optical power data samples, this paper presents an ARIMA model prediction method based on improved wavelet decomposition. This method uses the improved wavelet decomposition is multistage discrete wavelet decomposition SWT and improved it. Different from the conventional method of decomposition and reconstruction of signals, this paper directly uses wavelet decomposition coefficient for modeling, which simplifies the process of input data construction and reduces the data loss caused by data reconstruction. ARIMA is autoregressive integrated moving Average model. Building a combination model and using the data to conduct simulation experiments. Experimental results verify that the prediction accuracy of this optimization model is higher than that of the ARIMA model alone and prove that this model is superior to a small amount of sample optical power data.