Classical PID control in presence of missing data using compressed sensing techniques

摘要

Most of the controllers used in process industries are proportional, integral and derivative (PID) controllers. Missing data is common in these industries due to many reasons such as sensor malfunctioning, data transmission loss etc. The performance of PID control scheme will degrade if operated in presence of missing data. Hence in this paper a method based on compressed sensing techniques is proposed for online reconstruction of missing data and performance is shown with an application to process control. The proposed algorithm is demonstrated on three simulated studies and results show that the proposed algorithm gives comparable results with classical PID control scheme with no missing data.