基于热湿负荷与自适应预测时域微网优化调度  

Optimal scheduling of microgrid based on heat and humidity load with adaptive prediction horizon length

作　　者：林俊光 周雅敏 冯彦皓 马聪 吴凡 郑梦莲 俞自涛[2] LIN Jun-guang;ZHOU Ya-min;FENG Yan-hao;MA Cong;WU Fan;ZHENG Meng-lian;YU Zi-tao(Zhejiang Energy Group Research Institute Limited Company,Hangzhou 311100,China;Institute of Thermal Science and Power Systems,Zhejiang University,Hangzhou 310027,China)

机构地区：[1]浙江浙能技术研究院有限公司,浙江杭州311100 [2]浙江大学热工与动力系统研究所,浙江杭州310027

出　　处：《浙江大学学报（工学版）》2023年第9期1832-1842,共11页Journal of Zhejiang University：Engineering Science

基　　金：国家重点研发计划项目(2019YFE0126000);浙江浙能技术研究院有限公司科技项目(No.ZNKJ-2019-087)。

摘　　要：为了提升建筑冷热电联供(CCHP)微网灵活性并减少负荷波动,开展建筑湿负荷参与日前需求响应的效果和日内预测时域的自适应调节方法研究.在日前阶段,构建包含建筑湿负荷的冷负荷需求响应模型;在日内阶段,提出基于预测负荷方差的自适应预测时域模型预测控制(MPC)方法.在日前阶段,分析分时电价下包含湿负荷的冷负荷需求响应对总经济成本和蓄能水罐蓄冷量的影响;在日内阶段,分析采用自适应预测时域MPC方法对计算时间、成本和各类设备工作状态的影响.结果表明,考虑湿负荷的冷负荷需求响应降低了日前阶段成本7.75%;在日内阶段,自适应预测时域MPC方法不仅能够平衡计算时间和成本,还能够增加蓄能量和平滑燃气内燃机出力.In order to improve the flexibility of combined cooling,heating and power(CCHP)microgrid and to reduce load fluctuation,the effect of building humidity load participation in day-ahead demand response and the adaptive intra-day prediction horizon length were studied.In the day-ahead time domain,a cooling load demand response model including the building humidity load was constructed.In the intra-day time domain,a model prediction control(MPC)method with adaptive prediction horizon lengths based on the variance of the predicted load was proposed.The effects of the demand response of the cold load containing humidity load on the total economic cost and the storage capacity of the storage tank under the time-of-day tariff were analyzed during the day-ahead time domain.The effects of the impact of using the MPC method with adaptive prediction horizon lengths on the calculation time,the cost and the working status of various types of equipment were analyzed during the intra-day time domain.Results showed that the cooling load demand response considering the humidity load reduced the cost of the day-ahead scheduling by 7.75%.The MPC method with adaptive prediction horizon lengths not only balances the calculation time and the cost,but also increases the storage and smoothes the output of the gas-fired internal combustion engine in the intra-day time domain.

关 键 词：冷热电联供(CCHP) 热湿负荷 优化调度 模型预测控制(MPC) 自适应预测时域 

分 类 号：TM73[电气工程—电力系统及自动化]