AN IMPROVED DIAGNOSTIC TOOL FOR CONTROL VALVE STICTION BASED ON NONLINEAR PRINCIPLE COMPONENT ANALYSIS AND AUTOCOVARIANCE

Control valves suffer from wear and aging when opening and closing the valves which causes stiction nonlinearity. The existing stiction detection techniques are only applicable to loops from a certain type of processes or a certain amount of dataset. The nonlinear principle component analysis (NLPCA) method requires many ensemble runs per data, a large quantity of raw data and shows uncertain results when applied to loops from integrating processes.