Backstepping Control With Online Estimation Towards Complex Chemical Process

As one of important manufacturing,chemical industry plays a key role in China’s national economic and social development.Chemical processes feature multivariable,nonlinear,strong coupling,time-varying,time-delay,etc.,so it is difficult to establish a relatively accurate system model for them only by using conventional physico-chemical mechanism methods.Therefore,traditional model-based control methods are hardly effective to the control of complex chemical processes.Data-based control methods have become a new development direction of the complex industrial system control,because they do not require designers to know too much details about the targets,but only need inputs and outputs of the targets to complete the online estimation of the system model,so as to obtain the control scheme.In this paper,control schemes based on online estimation are used to study the control problem of complex chemical processes.and on this basis the time-delay problem in chemical process is deeply studied.Furthermore,the distributed online estimation backstepping control scheme of complex chemical system with time-delay is studied on Tennessee-Eastman process model.The research topics of this paper are three folds:First,for nonlinear multivariable chemical system,we employ data-driven backstepping control method,exploring control mechanisms for chemical system when procise models are not available or unknown perturbation exists.The method achieves high stability and efficiency of chemical processing.We take the Williams-Otto reactor as the target chemical process,and verify the robustness and effectiveness of online estimation based backstepping control algorithm.Second,in view of the time-delay features of complex chemical system,the continuous-time approximation method based on low-pass filtering is used to equate the time-delay system to a system without time-delay.On this basis,the control design is carried out by using the online estimation and backstepping control method to solve the instability of controller and performance degradation due to time-delay in chemical system.Third,taking the TE process model as the target chemical process,based on data-driven backstepping control algorithm,a distributed cooperative control algorithm derived from multi-agent consensus is studied to realize the cooperative control between the subsystems of the TE model,and the stability of the cooperative control between subsystems is improved.The advantages of the online estimation and backstepping algorithm based on the distributed control are demonstrated,and it brings convenience to the practical application of chemical system.