基于DDPG算法的冰蓄冷空调系统运行策略优化

Operation Strategy Optimization of Ice Storage Air Conditioning System Based on DDPG Algorithm

空调的不合理运行会增加建筑物总能耗,加大电网峰谷负荷差,影响电网的安全运行。针对此问题,提出一种基于深度确定性策略梯度(deepdeterministicpolicygradient,DDPG)算法的冰蓄冷空调系统运行策略优化方法,以平滑冰蓄冷空调系统的功率曲线和降低系统运行费用为目标,求解各个时刻制冷主机和蓄冰槽的运行功率。通过仿真实验分析DDPG算法的性能,结果表明,该方法不仅平滑了空调系统的功率曲线,实现了电网调峰的目标,同时降低了空调系统的运行费用,具有良好的经济效益。与传统的主机优先和融冰优先方法相比,该方法的功率平滑效果和经济效益都更好。

The unreasonable operation of air conditioning will increase the total energy consumption of buildings and the load difference between peak and valley of power grid,then affect the safety of power grid operation.To solve this problem,an optimization method of operation strategy for ice storage air conditioning system based on deep deterministic policy gradient(DDPG)algorithm is proposed.With the goals of smoothing the power curve of ice storage air conditioning system and minimizing the operation cost of the system,the operation power of chiller and ice storage tank at each time is calculated.The simulation results show that the proposed method is an efficient method.On the one hand,it smooths the power curve of the air conditioning system,and achieves the goal of peak regulation of power grid.On the other hand,it reduces the operating cost of the air conditioning system,and has good economic benefits.Compared with the traditional methods of chillers priority and ice melting priority,this method has better power smoothing effect and economic benefits.