基于遗传神经网络的VAV空调系统预测模型

针对单纯的机理建模方法难以准确预测变风量空调系统（VAV）的参数，利用BP神经网络构建了变风量空调系统的预测模型，并将遗传算法与BP网络相结合，提出运用遗传算法对神经网络的权值和阈值进行遗传搜索，寻优后再进行BP运算，以克服BP算法收敛速度慢、易陷入局部解的缺点。通过实验平台采集了大量数据对所建模型进行训练和验证，结果表明，模型对空调送风参数以及房间温湿度的预测结果与实测数据能很好拟合，精确度高，泛化能力强。

Prediction Model of VAV Air Conditioning Systems Based on GA-ANN

For the mechanism modeling method is difficult to predict the parameters of VAV system accurately, this paper use BP neural network to build a predictive model of VAV system and combine the genetic algorithm with BP network. So that it can overcome the disadvantages of BP algorithm such as low speed converge and easily being subject to the partial minimum, because of using genetic algorithm to search the weights and thresholds of BP network to find the optimal solution before BP operation carried out. After collecting a lot of data for training and verifying model through the experimental platform, the results show that the model predictions can simulate the measured data accurately on the temperature and humidity of the room and the parameters of air conditioning, and with a high generalization capability.