A Novel Approach for Sensor Noise Elimination in Real Time Process

A review on the current literature on sensor noise elimination algorithms, reports the existence of several adaptive noise reduction strategies. The well known, Wiener filter, which is an optimum filter, belongs to this category and removes the noise, in the mean square error sense. A neural network based independent component analysis (ICA) algorithm is incorporated into the existing system to remove the sensor noise. The proposed intelligent controller, along with this ICA algorithm for sensor noise elimination, can be used to handle plant exceptions, such as poor performance of controller due to faulty sensor readings reduced life of actuators due to frequent false actuating signals to the control valve, to regulate the manipulated variable even when the process is at steady state.