Corelation between Conventional and Data-Driven Control Performance Assessment Indices for Heating Process

In this paper, the comparison between conventional and data-driven control performance assessment (CPA) indices in application to the heating process is presented. Based on the realistic simulation results for differently tuned single-loop PI/PID based control systems, the chosen CPA measures are computed and compared in terms of ability to load disturbance rejection. Then, the correlation between conventional CPA indices (maximum overshoot, settling time and integral indices) and data-driven methods (minimum variance (MV) and general minimum variance (GMV) indices) is presented. The results show that in some cases, data-driven CPA methods can be replaced by less complex conventional CPA indices.