To address the robust output feedback tracking problem for total outflow rate of two-stage anaerobic digestion (AD) process with input saturation, a novel performance guaranteed ultra-local model (ULM)-based predictive control (PG-ULMPC) is proposed. First, model of the two-stage AD process is considered which presents the characteristics of the production rate of hydrogen and methane. Then, a new control strategy is developed, which relies only on input/output data and does not require any model knowledge of two-stage AD process. Based on ULM, the designed PG-ULMPC strategy consists of time-delay estimation (TDE), prescribed performance function (PPF) as well as predictive control using feedback correction (FC). The ULM is utilized to approximate the complex model of two-stage AD process in a short sliding time window and simplify controller design. The TDE is employed to estimate the lumped disturbance. The PPF guarantees that the tracking error is limited within the performance bounds whose convergence time and precision can be artificially preset. The predictive control using FC stabilizes the closed-loop system and provides saturated control input. To demonstrate the PG-ULMPC performances, numerical simulations of two-stage AD process in presence of unknown parameter disturbances with nonlinear Proportional–Integral (PI) controller, with ULMPC and with PG-ULMPC are realized by using Matlab/Simulink. The obtained results illustrate the better performance of the proposed PG-ULMPC strategy. Moreover, under the measurement noise, by using filtered derivative technique, the proposed method can still achieve satisfactory performance and has little control chattering.
Read full abstract