论文部分内容阅读
在复合式地源热泵系统中,控制策略存在着极大的优化空间。本文提出了直接比较冷却塔和土壤换热器出口水温的方法,并建立了土壤换热器预测模型。首先,为了提高预测精度,将所建立模型分为2类,一类为利用当前时刻变量进行预测,称为静态模型;另一类为加入上一个时刻的变量进行预测,称为动态模型。而后,为了研究复合式地源热泵系统不同运行模式下神经网络预测土壤换热器出口水温的准确性与可行性,根据样本特点建立了一系列模型,并与复合式地源热泵系统的动态数值模型进行了比较。结果表明,无论处于何种运行模式下,神经网络都能够准确预测土壤换热器的出口水温,且动态模型具有较高的精度,绝对误差不超过0.2℃。
In compound ground source heat pump systems, there is a great room for optimization of control strategies. This paper presents a direct comparison of cooling tower and soil heat exchanger outlet water temperature method, and established a soil heat exchanger prediction model. First of all, in order to improve the prediction accuracy, the established model is divided into two categories, one for the prediction of the current moment variables, called the static model; the other for adding the variables of the previous moment to predict, known as the dynamic model. Then, in order to study the veracity and feasibility of neural network predicting the outlet water temperature of soil heat exchanger under different operation modes, a series of models are established according to the characteristics of the sample, and are compared with the dynamic value of the composite ground source heat pump system The model is compared. The results show that the neural network can accurately predict the outlet water temperature of the soil heat exchanger in any operating mode, and the dynamic model has high accuracy with an absolute error of less than 0.2 ℃.