Fault identification and isolation in pneumatic valve using image processing

This paper presents a technique for analysis of fault in pneumatic control valve. The objective of the paper is to develop a non-contact process for detection of faults in control valve if any, using a camera. The proposed design is designed to identify the faults in valve. Once the fault is identified, the next step is to isolate the fault by indicating the type of fault so that necessary action can be taken. The proposed technique is programmed to identify stiction, and stem alignment faults. Video of the process captured from camera is acquired on to the system for processing, identifying and isolating the faults by analyzing the fault signature. The process once designed is tested under real life situation. Results obtained show that the proposed technique was able to identify the faulty control valve, along with indicating the type of fault.