摘要
针对金属填料蒸发型空调系统内部传热传质过程的复杂性,利用人工神经网络的非线性映射功能、学习功能和记忆功能,通过对数据样本的学习,建立起描述金属填料蒸发型空调系统性能与其影响因素间关系的神经网络模型,再根据此模型来预测影响因素变化引起的蒸发型空调系统性能变化。利用试验数据对网络模型进行训练和验证,检验模型预测的准确性。结果表明:网络模型输出结果与实验数据比较吻合,相对误差大多在6%以内,证实该网络模型对金属填料蒸发型空调系统性能预测的可行性。
Based on characters of Artificial Neural Network （ANN） such as non -liner mapping, self-learning, the ANN model was set up to predict the relationship between effective factors and performance of metal-packed evaporative air conditioning system. In order to reflect the performance of the predictive model, experimental data were used to test and verify the model. The outcome of the ANN model is in good agreement with experimental data and relative difference is within 6 percent. So it is feasible to predict the performance of metal-packed evaporative air conditioning system.
出处
《重庆大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2007年第5期103-107,共5页
Journal of Chongqing University
关键词
金属填料
蒸发型空调系统
人工神经网络
模型
metallic packing
model
evaporative air conditioning system
artificial neural network
