Prioritized Multi-objective Model Predictive Control Strategies Of Constrained Nonlinear Systems

Multi-objective model predictive control(MO-MPC)gains considerable attention in industrial control field due to its ability to explicitly deal with constraints,nonlinearity and multi-objective optimization.In recent years,the stability issues of MO-MPC closed-loop systems are studied and a rich of stability synthesis schemes have been proposed by many researchers.One of the effective schemes is the prioritized MO-MPC strategy.However,the terminal constrained conditions imposed in existing prioritized MO-MPC strategy increase the online computational burden of MPC.Moreover,there is no economic performance analysis of objectives in the existing prioritized MO-MPC strategy of systems subject to constraints and nonlinearity.Therefore,study on prioritized MO-MPC schemes of constrained nonlinear systems has the important theoretical significance and potential application values.Based on the existing results on MO-MPC,the dissertation uses the triplet conditions and the concept of lexicographic optimization to study prioritized MO-MPC schemes of constrained nonlinear systems and their srability synthesis methods.The main contributions and innovations of the dissertation are as follows:(1)A stabilized lexicographic MO-MPC strategy is presented for prioritized multi-objective optimization control problems of constrained nonlinear systems.The triplet conditions of terminal penalty functions,terminal constraint sets and local controllers are exploited.Meanwhile,regional assumptions of the highest-priority cost criteria with respect to terminal constraint sets are introduced to define the initial feasible region of the closed-loop system.By the concept of lexicographic optimization,the multi-objective model predictive controller is designed under the initial feasible condition.The online computational burden of the controller is reduced since the optimization problem is not subject to the terminal conditions.By comparing to the conventional lexicographic MO-MPC,the advantages of the strategy of this chapter are validated.(2)A lexicographic economic MPC strategy is proposed for constrained nonlinear systems with multiple economic performance criteria.The theoretical results of the upper bound of the each objective performance are successfully established by combining the concept of lexicographic optimization,three element stability conditions of MPC and value function-based constraints of optimization objectives.Meanwhile,stability of the closed-loop system is guaranteed at the optimal economic equilibrium point.Moreover,feasibility of the multi-objective economic MPC strategy is discussed under no terminal constrained conditions.The validity of the strategy in this chapter is verified by a numerical example.(3)For the combustion processes of circulating fluidized bed boiler with multivariable coupling,nonlinearity,constraints and multiple economic performance criteria optimization requirements,a MO-MPC controller of the combustion processes of circulating fluidized bed boiler is designed by combining the proposed lexicographic multi-objective economic MPC strategy.By comparing to the conventional MO-MPC scheme,the advantages of the proposed strategy are validated.