Research on the control of the yellow river diversion project based on the decentralized model predictive control

Multivariable and long distance open channel water transmission project in a large-scale system is hard to control in a centralized way. A centralized model predictive control system is incapable of solving the issue effectively of regulating channels' flow and water-level, this paper establishes a typical water transmission model based on the state space equation of channel transition process. The decentralized model predictive control algorithm is presented and the channel operating characteristics are analyzed. The channel operating in the initial conditions is simulated by using MATLAB. The comparison of two results between the channel centralized control and decentralized control show that: the decentralized model predictive controller under the complicated operating conditions can effectively solve the coupling of the channels and the large hysteresis problems. The dynamic and robust performances of the system are verified better with the decentralized model predictive control.