摘要
基于经典的Kelvin线热源模型和Kavanaugh圆柱源理论模型,运用负荷叠加、负荷阶跃、负荷累积的思想,结合人工神经网络建立了土壤源热泵系统地下埋管换热器仿真优化模型。与实验和理论已验证过的圆柱源理论模型的比较分析表明,地下埋管神经网络优化模型相对于传统的地下埋管解析解模型具有良好的计算精度和泛化能力,该模型应用于土壤源热泵系统的长期运行可以大大缩短计算时间,为土壤源热泵系统的优化及设计提供更有效率的模拟计算方法。
Based on Kelvin linear heat source theory and Kavanaugh cylindrical source theory of classic constant heat flux, the artificial neural network simulation model of vertical U tube was built, quoting the idea of superposition principal, step load and load aggregation, then the optimum model was built. Comparing the optimum model with the cylindrical source theory which had already been verified by both experiment and theory, the results indicated that the artificial neural network simulation model of vertical U tube compared with the traditional underground analytic solution model had better computational accuracy and generalization capability. The model is developed to simulation long-term operation performance of GSHPS operation that can reduce the computing time of simulation model and provide more efficiency numerical simulation method for optimize design of GSHPS.
出处
《太阳能学报》
EI
CAS
CSCD
北大核心
2010年第6期738-742,共5页
Acta Energiae Solaris Sinica
基金
国家自然科学基金(50776050)
