基于GPC融合Adam的MIMO变风量暖通空调冷却侧节能控制方法

Energy saving control method of MIMO VAV HVAC cooling side based on GPC and Adam

摘要:

MIMO系统在变风量暖通空调冷却侧表现为多个变量之间存在强耦合关系。传统单变量控制难以处理多变量间的耦合效应，可能导致调节一个变量时引发其他变量波动，增加能耗或牺牲舒适度。为保障变风量暖通空调的稳定运行，提出基于GPC融合Adam的MIMO变风量暖通空调冷却侧节能控制方法。在MIMO系统中通过变风量条件下的机组参数拟合运算，定义空调暖通回路与冷却回路之间的换热关系，以求解空调冷却侧的实时能耗，完成变风量暖通空调系统的冷却能耗计算。基于GPC反馈校正机制，设计线性控制器模型，显式处理MIMO系统的耦合关系，通过预测模型协调多变量调节，避免单变量控制的冲突。在此基础上，根据Adam算法的收敛性原则，实施对空调冷却参数的初始化运算，从而在泛化处理冷却参数的同时，补偿模型误差和工况变化，维持解耦控制的有效性。进而生成节能控制策略，完成MIMO变风量暖通空调冷却侧节能控制。实验结果表明，所提方法能将暖通空调冷却侧的耗电量降低至1100kWh，且在温湿度恒定的情况下，所获得的制冷效果相对较优。

Abstract:

MIMO systems exhibit strong coupling relationships among multiple variables on the cooling side of variable air volume HVAC systems. Traditional univariate control is difficult to handle the coupling effects between multiple variables, which may cause fluctuations in other variables when adjusting one variable, increase energy consumption, or sacrifice comfort. To ensure the stable operation of variable air volume HVAC, a MIMO variable air volume HVAC cooling side energy-saving control method based on GPC and Adam fusion is proposed. By fitting the unit parameters under variable air volume conditions in MIMO systems, the heat transfer relationship between the HVAC circuit and the cooling circuit is defined to solve the real-time energy consumption on the cooling side of the air conditioning system and complete the cooling energy consumption calculation of the variable air volume HVAC system. Based on the GPC feedback correction mechanism, a linear controller model is designed to explicitly handle the coupling relationship of MIMO systems, coordinate multivariable regulation through predictive models, and avoid conflicts in univariate control. On this basis, according to the convergence principle of Adam algorithm, the initialization operation of air conditioning cooling parameters is implemented, so as to compensate for model errors and operating conditions changes while generalizing the cooling parameters, and maintain the effectiveness of decoupling control. Further generate energy-saving control strategies to achieve energy-saving control on the cooling side of MIMO variable air volume HVAC systems. The experimental results show that the proposed method can reduce the power consumption on the cooling side of HVAC to 1100 kWh, and the refrigeration effect obtained is relatively better under constant temperature and humidity.