Fault Forecast of Electronic Equipment Based on ε -SVR

摘要

In order to ensure security and reliability of the equipment, so as to decrease the maintenance cost, combining with the characteristics of fault data, this paper adopts ε - support vector regression to establish a fault forecast model and evaluation system to prediction model effect which are proper to the electronic equipment. Selecting multi-electronic equipment and training on the ε - SVR with different kernel functions. It is demonstrated that the prediction effect is better and it is still of vital realistic significance for realizing condition-based maintenance of modern electronic equipment.