摘要
基于自主研发的超磁致伸缩材料(Giant Magnetostrictive Material,GMM)作动器的主动控制特性,应用T-S(Takagi-Sugeno)型模糊神经网络设计了主动控制系统,该系统以GMM作动器两端节点的相对速度和相对位移作为输入,计算输出控制电流。通过神经网络的自适应学习功能进行模糊划分及生成模糊规则,利用模糊系统的推理能力对空间结构模型进行基于地震响应的主动控制仿真,同时与标准型模糊神经网络系统进行仿真对比。结果表明,二者对空间结构模型的主动控制都能达到良好效果,基于T-S型模糊神经网络推理简单,其仿真速度远快于标准型。因此,采用T-S型模糊神经网络对空间结构进行主动控制更能满足工程应用需求。
Based on independently developed Giant Magnetostrictive Material（ GMM） active control actuator,a Takagi-Sugeno（ T-S） fuzzy neural network control system of a spatial structure was designed,in which the relative displacement and relative speed of two nodes at the end of the active-member were taken as inputs,and the output control current was calculated by the network. Taking advantage of the adaptive neural network learning function to achieve the fuzzy division and to generate fuzzy rules,an active control simulation of the spatial structure model under the action of seismic response by using the fuzzy reasoning capability,was carried out and the results were compared with the results produced by the simulation of a corresponding standard fuzzy neural network model. The results demonstrate that both the models can achieve good control effects,but the simulation speed of the T-S fuzzy neural network is far faster than the standard model. Therefore,the T-S fuzzy neural network controller can better meet the needs of engineering application requirements.
出处
《振动与冲击》
EI
CSCD
北大核心
2015年第24期1-6,11,共7页
Journal of Vibration and Shock
基金
国家自然科学基金(51178388
51108035)
国家重点实验室开放项目(08KF02)
陕西省工业攻关项目(2013K07-07
2014K06-34)
西安建筑科技大学创新团队资助项目
关键词
GMM作动器
模糊神经网络
主动控制
仿真
空间结构
GMM active control actuator
fuzzy neural network
active control
simulation
spatial structure
