Page 318 - Distributed model predictive control for plant-wide systems

292 Distributed Model Predictive Control for Plant-Wide Systems

13.5 Conclusion
This chapter proposes a hierarchical structure of DMPC economic algorithm to schedule the
convention electric refrigerator and ice storage tank system. The algorithm includes two levels:
the upper level algorithm is a mixed integer programming economic optimization; and the
bottom level algorithm is dynamic performance optimization of electric refrigerators based on
the DMPC. The upper algorithm supplies the power references of the electric refrigerators in
the bottom level, and then the lower level uses the DMPC algorithm to track the references
above quickly to speed up the dynamic response. The chapter applies the scheduling algorithm
in the low-district cooling system of a super-high building in Shanghai, China. Results show
that the algorithm is able to reflect basic characteristics of cold sources in such cooling system
and put forward optimum strategy under different conditions of load and electricity price so
as to realize the goal of the lowest cost.