Sparse Identification and M.P.C. Based Closed Loop Control of Nonlinear Quadruple Tank Process

The technique known as Sparse Identification of nonlinear dynamics (SINDy) has demonstrated effectiveness in determining system's underlying equations from empirical data. SINDy is a sparsity-promoting technique that identifies the most important terms in the dynamical system equations. This paper discusses the nonlinear model identification of nonlinear dynamics of a Quadruple Tank process using SINDy. After identifying the appropriate dynamics, the dynamical model of Quadruple Tank process can be determined by the evolution of states which can be compared with the original dynamical system. Then, linear and nonlinear model predictive control based techniques are applied to the identified model for achieving the control objectives. The passivity based approach is utilized in nonlinear model predictive control to ensure the closed-loop stability of the discovered nonlinear model. Our findings pave the way for accurately modelling complex physical systems and using the findings for applying different control strategies.