Liquid Level Tracking Control Of A Three-Tank System Based On Differential Flatness

The liquid level control of multiple flow systems is widely used in modern petroleum and chemical industries.This paper takes the three-tank system as the research object and based on the model predictive control method to study the liquid level tracking control problem of multiple flow system.The main contents include:Firstly,based on the characteristics of the three-tank system,the nonlinear model of the system is established by using the mechanism method.By finding the flat output of the model,the system is proved to have differential flatness.Secondly,the feedforward controller is designed by using the differential flatness of the three-tank system to realize the small deviation adjustment of the feedback controller;the deviation system is linearized at the equilibrium point,then the linear model predictive controller is designed,and the control problem is transformed into Quadratic Program-ming problem to solve;in the Matlab environment,the simulation and verification are carried out.In order to improve the accuracy of liquid level tracking,a nonlinear model predictive control(NMPC)method based on differential flat feedforward is proposed.The Cuckoo Search Algorithm is adaptively stepped and integrated with the simplex algorithm.The adaptive penalty function is used to transform the nonlinear constrained optimization problem into an unconstrained optimization problem,and use the improved Cuckoo Search Algorithm to solve.The simulation and experimental results show that the nonlinear model predictive control strategy based on differential flat feedforward is better than the linear model predictive control(LMPC)strategy.In order to further improve the liquid level tracking performance of the three-tank system,Back Propagation(BP)neural network model is used as the prediction model,and the NMPC liquid level tracking control strategy based on BP neural network is proposed.The experimental results show that the two NMPC liquid level tracking control strategies proposed in this paper are superior to the LMPC and Linear Quadratic Regulator(LQR)control strategy.