Intelligent fusion terminal with accurate operation failure prediction function

Abstract

The application relates to the technical field of data processing, in particular to an intelligent fusion terminal with a running fault precise prediction function, which comprises a processor and a memory, and the processor executes the computer program of the memory to realize the following steps: the normal approximation degree of residual error distribution in a residual error sequence corresponding to an initial SARIMA model which has passed Ljung-Box test is acquired; if the normal approximation degree is less than a preset normal approximation degree threshold, the residual error sequence is divided into at least two subsequences, a secondary test judgment factor of the initial SARIMA model is acquired according to the abnormal value distribution and the trend intensity in each subsequence, the initial SARIMA model is subjected to secondary test to obtain an optimal SARIMA model; and the optimal SARIMA model is deployed into the intelligent fusion terminal, thereby improving the accuracy of running fault prediction of the distribution network station area.

Images