基于模糊前馈解耦的变风量空调控制方法研究

Control Method based on Fuzzy Feedforward Decoupling for VAV Air Conditioning system

变风量(VAV)空调系统具有多变量、多回路耦合、大滞后、强非线性特点;采用数学建模和仿真实验法,针对送风温度、送风管道静压、室内温度和室内二氧化碳浓度4个相互耦合的主要被控量,通过建立热量平衡和物质传输机理模型和传递函数矩阵,设计基于神经网络的前馈模糊解耦方法;所提出的方法既利用了数学模型可清晰描述被控量之间耦合关系的优点,又避免了传统解耦算法控制效果过于依赖建模精度等问题,仿真和实验结果表明系统获得了良好的控制性能。

Variable Air Volume （VAV） air conditioning system possesses many characteristics such as multi-variable,coupling between multi-control loops,and strong non-linear.In the paper,according to thermal balance and mass transfer processes,modeling and simulation are adopted to establish the mechanism model and transfer function matrix for main control variables,and a fuzzy feedforward-decoupling method based on neural network is developed.The method not only gets the advantage which mathematics model is able to describe the complex coupling relation,but also avoids the defect of traditional decoupling algorithm,which control effect rely on the accurate model seriously.Finally,experiment results indicate that the method is reliable and effective.