Polymer Properties Robust Estimation And Control For Industrial Polyethylene Process And Application

Melt index and resin density are two important properties in polymer properties in industrial polyethylene process. Accurate controls of melt index and resin density according to request of production grade are significant to the profit for the enterprise. Due to the lack of reliable on-line analyzer, the control of the polymer properties is operated according to the analysis the Product sampled from the reactor. So it is necessary to design the online estimator and advanced process controller on the polyethylene plant.Main contents are as followed:(1) According to the polymerization characteristics, a polyethylene quality indexes nonlinear Hammerstein model is established. The nonlinear steady-state part model is established based on the Polymerization reaction kinetics equation. The linear dynamic model is established based on the feature modeling method. The model parameters are got by the PSO algorithm. Then the nonlinear Hammerstein model is established. (2) According to the established nonlinear Hammerstein model, considering the uncertainty of the measurement noise and introducing the maximum deviation between the laboratory analysis value and model predictive value, a robust estimator is proposed. According to the approximation error upper bound of the predictive model, the gain of the filter is gained by LMI method.(3) According to the established nonlinear Hammerstein model, a nonlinear inverse is induced. The nonlinear control problem is transformed into linear controller design problem. The linear model part adopts model predictive controller, and the PSO-Quasi-newton iteration method is adopted to gain the controlled variable. Then the nonlinear Hammerstein model predictive control of the polyethylene quality indexes is realized.Simulation results and ease study prove the validity and applicability of above methods.