Adaptive inferential control of wood chip refiner

Abstract

In the chip refining process, the measurement of primary output, the freeness, is often hindered by the limitation of the availability of reliable on line sensor and the long sampling time to obtain the test results. Adaptive inferential control method overcomes the above limitation by measuring the rapidly sampled secondary output and the result is inferred to primary output for feedback control or rejection of disturbance. The task is solved by online identification of two models with different sampling periods. One of them predicts the controlled output between sampling instances using actual values of secondary output which is currently at disposal, while the control synthesis is based on the other model which uses controlled output predicted by the first model. Using the proposed method, simulation results indicate that the control scheme is capable of rejecting the process disturbance in order to keep the freeness set point in the refining process.<>