基于模糊逻辑与神经网络的高层结构半主动控制

Study on semi active control of high rise structuresbased on fuzzy logic and artificial neural network

根据剪切型结构动力特性提出AVSD开关控制律。把开关控制律作为专家知识,利用模糊逻辑转化为模糊控制规则。为了进行时滞和结构动力特性时变的控制补偿,采用神经网络在线自适应跟踪辨识方法进行在线辨识和响应预测。最后以某框架结构为例进行仿真分析,结果表明这一方法控制效果及鲁棒性好、实际应用方便可靠。

According to the dynamic characteristics of high rise shear type structure, this paper presents a semi active control method for engineering structures based on Fuzzy Logic and Artificial Neural Networks. In this method, the on off switch law applied to AVSD system is used as the expert knowledge and the Fuzzy Logic theory is transformed to fuzzy control law. Based on this control, the AVSD fuzzy control system is formed. In order to improve the control resulting from time delay and solve the time varying problem of structural dynamic characteristics existing in the vibration course, neural network on line self adaptive tracing identification method is used to predict the structural dynamic responses and identify the structural dynamic characteristics. At last, computer simulation analysis for a frame structure is carried out. The results indicate that this method has remarkable control effect and robustness, and is convenient and reliable to be used in practice.