Analytical fault detection and diagnosis (FDD) for pneumatic systems in robotics and manufacturing automation

Abstract

Pneumatic systems are often found in manufacturing floors for automation and robotic systems. Early and intelligent faults detection and diagnosis (FDD) of such systems can prevent failure of devices that causes shutdown and loss of precious production time and profits. In this paper, we introduce analytical FDD for pneumatic systems. The diagnosis system presented in this paper focuses on the signal-based approach which employs multi-resolution wavelet decomposition of various sensor signals such as pressure, flow rate, etc., to determine leak configuration. Pattern recognition technique and analytical vectorized maps are developed to diagnose an unknown leakage based on the established FDD information using affine mapping. Experimental studies and analysis are presented to illustrate the FDD system.