摘要
针对温度控制的大惯性、大滞后、非线性特点,提出采用基于小波神经网络辨识器的模糊神经自适应控制的中央空调房间温度控制器的设计方案。由于小波神经网络的非线性映射能力比一般神经网络要强,所以基于小波神经网络的辨识器可以获得很高的辨识精度。而且,模糊神经自适应控制器随着系统动态特性的改变可以在线改变其控制规则,从而进行客观准确的控制。与普通模糊控制方法相比较,仿真试验说明了系统设计的有效性。
Aiming at the great inertia,heavy to lag behind,non-linear characteristic of temperature control,proposed a design plan of central air condition room temperature controller through adopting fuzzy neural network adaptive control way using wavelet neural network as identifier.Because the wavelet neural network have the better ability to approximate non-linear functions,the identifier can identify more precisely.Moreover,the controller can alter its rules online along with the change of system dynamic characters to control objectively and accurately.On the base of comparison with general fuzzy controller,the result from simulation shows the validity of the system design.
出处
《仪表技术与传感器》
CSCD
北大核心
2004年第7期37-38,52,共3页
Instrument Technique and Sensor
基金
建设部基金资助项目(03-2-117)
关键词
变风量空调
小波神经网络
模糊神经
自适应控制
Variational Air Volume (VAV)
Wavelet Neural Network
Fuzzy Neural Network
Adaptive Control
