多冷源系统的递阶协调分布式预测控制

Hierarchical Cooperative Distributed Model Predictive Control for Multi-Types Cold Sources Systems

在现代高层建筑中,从降低运行成本考虑,由多种互补冷源组成的供冷系统应用得越来越广泛。然而要做到高效节能,如何优化运行是关键问题。为此提出一种基于分布式预测控制的递阶优化策略,首先根据各冷源经济模型和分时电价策略,采用混合整数规划方法,优化各冷源的开启状态及最优功率设定值,同时考虑到各冷源的动态调节特性以及物理约束,提出一种目标耦合的协调分布式预测控制方法,在各子系统MPC目标函数中加入了总体目标,进一步优化各冷源的设定值,使得各冷源在动态过程中保证总负荷的同时跟踪最优制冷量,进一步提高经济性能。该策略以上海某超高层建筑低区供冷系统为对象,通过数值仿真验证了方法的有效性。

In modem high-rise buildings, the cooling system composed of multi-type complementary cold sources is more and more widely applied in order to reduce the operation cost, so the operation optimization of such system is the key factor of realizing energy saving. To solve the optimization problem, a kind of hierarchical optimal strategy is presented in this paper. At first, the on-off state and the best setting value of each cold source are optimized with the method of mixed integer programming according to the economic model and time-of-use power price. Then considering dynamic regulation features and physical constraints of each cold source, the coordinated method with coupling targets based on distributed model predictive control is proposed so as to achieve better economic performance. The setting value of each cold source is re-optimized under the distributed framework to guarantee each source tracks the best setting value and the total load is satisfied as well. The strategy is put into use in the cooling system of the lower district in a super-high building at Shanghai and simulation results verify its validity.