steel annealing processes as hybrid systems【DOC精选】.docVIP

Neural-optimal control algorithm for real-time regulation of in-line storage in combined sewer systems??Original Research ArticleEnvironmental Modelling Software Optimization of hybrid systems??Original Research ArticleComputers Chemical Engineering An overview of dynamic optimization problems with hybrid systems embedded is presented. The control parameterization approach is examined in detail, where the possibility of mode switching within the hybrid system causes nonsmoothness in the Master problem, thus invalidating gradient based solvers. To handle this, a decomposition approach is proposed that divides the general problem into smooth subproblems. The subproblem of determining the optimal mode sequence given fixed transition times is considered for continuous time linear time varying hybrid systems. A mixed-integer reformulation of the problem is proposed that retains the linear structure of the embedded hybrid system. A deterministic branch-and-cut framework incorporating outer approximation cuts and a dynamic bounds tightening heuristic is then employed to solve the nonconvex mixed-integer problem to guaranteed global optimality with a finite number of iterations. It is shown that the dynamic bounds tightening heuristic can have a dramatic effect on accelerating convergence of the algorithm. Future directions for research are also discussed, including the application of recent developments in semi-infinite programming for the formal safety verification of processes under uncertainty. Article Outline 1. Introduction 2. Model 3. Simulation and Sensitivity Analysis 4. Optimization 5. Decomposition Roadmap 6. Optimal Mode Sequences 7. Optimal Transition Times 8. Summary and Future Directions 8.1. Safety Verification Under Uncertainty 8.2. Control Parameterization using Nonsmooth Optimization Acknowledgements References participation and optimal control in water resources planning and management??Original Research ArticleApplied Mathematics and Computa