Fault Detection of Artillery Automatic Loading System Based on PCA

摘要

The motion process of the automatic loading system is a high overloading and intermittent motion environment will bring about motor windings loosening, transmission system wear and tear, fracture, sensor failure and other security risks or system failures. In the paper no-stationary signal analysis by wavelet transform through wavelet decomposition and non-linear threshold de-noising. And use PCA established system model for on-line monitor. By calculate SPE and T~2 analysis four kind of result to find fault source. Finally through the experimental prove the reliability of the method.