SELF‐TUNING CONTROL OF AN INDUSTRIAL PHOSPHATE DRY PROCESS

Abstract

SUMMARYThis paper presents experimental results of an application of self‐tuning control to an industrial process. The process involved is an industrial phosphate dryer. The main control objective is to keep the moisture content of the dried phosphate close to a constant desired value (1‐5 per cent) and at the same time to minimize the energy consumption, despite feed flow rate variations (100~240t/h) and variable moisture content of the damp phosphate (7 ~ 20 per cent).A simplified model of the complex dynamics of the physical plant was assumed. The controller uses a recursive factorization parameters estimator (either square root or upper diagonal factorization form) based on the method of least squares, to estimate an implicit plant‐controller model for the control design. The algorithm employed is based on a generalized minimum variance strategy. Both set point following and feedforward control actions are included in the tuning strategy.The results of the self‐tuning control experiment reported in this paper are compared with the controlled performances achieved when using conventional multi‐loop PID regulators, operating on steady state. The results obtained over several weeks of continuous experimentation illustrate the key features of self‐adaptive control, especially its inherent capability to adapt to the change of the operating point, where conventional control fails.A 4‐3 per cent saving was gained by the use of the presented algorithm compared to the best controlled performances that were achieved when using conventional PID regulators. Thus this provides a heavy motive to introduce advanced computer control to the Moroccan phosphate industry in an economic way.