基于神经网络的暖通空调控制系统延迟时间的确定

Hvac System Delay Time Determination Using an Artificial Neural Network

本文提出一种人工神经网络(ANN)方法，确定暖通空调(HVAC)系统对控制信号的响应延迟时间。在研究中选用了四层网络，在网络学习过程中采用了一种加速方法，以改善梯度下降法。对制热和制冷两种工作方式的实验数据，采用ANN和传统算法两种方法确定延迟时间，并进行比较。结果表明ANN可以有效地用于确定HVAC系统的延迟时间。

This paper presents an artificial neural network (ANN) approach to determine the delay time for a heating,ventialting and air－conditioning (HVAC) plant to respond to control actions.The ANN is a fully connected four－layer network.An acceleration technique was used to improve the General Delta Rule for the learning process.Experimental data for heating and cooling modes were used with both the ANN and a traditional mathematical method to determine the delay time.The results show that ANNs can be used effectively determining the delay time for AVAC systems.