基于CNN的太阳能热泵—壁挂炉供暖系统约束控制方法

Constraint Control Method of Solar Heat Pump and Wall-Hung Furnace Heating System Based on CNN

为了保障供暖系统的供热效果,提高其能源利用率,研究了一种基于CNN的太阳能热泵-壁挂炉供暖系统约束控制方法。通过梯度下降法优化CNN训练过程,提升其训练与收敛效率,运用优化训练后的CNN引入输入与输出的约束,构建太阳能热泵-壁挂炉供暖系统约束控制模型,对太阳能热泵-壁挂炉供暖系统的约束控制,实例测试结果表明,该方法具有较高的收敛速度及较低的训练函数损失值,可有限控制供暖系统的供水温度,满足供暖系统的实际供水温度需求。同时,水温控制误差的平均值比设定的0.007℃供暖系统控制水温误差约束值低,可约束控制供暖系统的阀门开合度保持在预先设定的合理范围内,且阀门开合度的波动趋势稳定,整体控制性能显著,能够为供暖系统供热质量及能源利用率的提升提供保障。

In order to ensure the heating effect of heating system and improve energy utilization,a constraint control method of solar heat pump wall mounted furnace heating system based on Convolutional Neural Network(CNN) is studied.The gradient descent method is used to optimize the training process of CNN to improve its training and convergence efficiency.The input and output constraints are introduced into the trained CNN to construct the constraint control model of solar heat pump wall mounted furnace heating system,and the constraint control of solar heat pump wall mounted furnace heating system is carried out.This method has higher convergence speed and lower loss value of training function.It can control the water supply temperature of heating system to meet the actual demand of water supply temperature.Meanwhile,the average value of water temperature control error is lower than the set constraint value of 0.007 ℃.The results show that the valve opening and closing degree of the controllable heating system can be kept in a reasonable range,and the fluctuation trend of the valve opening and closing degree is stable,and the overall control performance is significant,which can provide guarantee for the improvement of the heating quality and energy utilization rate of this heating system.