A PSO Method of GM to Fault Diagnosis in NC Machine Tool

Abstract

The reliability tests of NC machine tools are with great importance for supply corporations to learn and further improve their products quality. But the overall and long-term tests are often with many costs. The grey model could forecast the long-term fault information from few short-term samples. During modeling, the grey model adopts a mean approximation to disperse the first order differential equation. The particle swarm optimization, as a multi-dimension search method is adopted here to find the optimal proportion point. The experiments show that the forecasting result of the optimized grey model is improved.