摘要
通过物理学电磁理论分析建立大跨桥梁结构电磁驱动AMD振动控制系统的力电计算模型,根据大跨桥梁结构抗震设计特点,提出了基于遗传-BP神经网络的大跨桥梁结构电磁驱动AMD系统输入电压的在线实时控制方法,由传感器采集到的结构状态信号和外界地震激励信号瞬时做出系统输入电压选择,解决传统振动控制计算量繁重、运行慢导致时滞等问题,从而实现地震激励下桥梁结构在线较为精确的振动控制。
The force-electricity relationship of an electromagnet-drived active mass damper(AMD) working for a long span bridge is established based on the electromagnetic theory. Then, according to the earthquake resistance design characteristic of the long span bridge, an online control method of AMD’s input voltage based on GA-BP network is put forword. The input voltage can be instantly chosen as soon as the dynamic response and earthquake acceleration are gathered by the sensors. The big slow computation and time delay problem are both avoided. Thus, online vibration control of a long span bridge can be realized precisely.
出处
《振动与冲击》
EI
CSCD
北大核心
2008年第9期135-138,共4页
Journal of Vibration and Shock
基金
教育部博士点基金资助项目(20030359005)
合肥工业大学科学研究发展基金资助(060801F)
安徽省自然科学基金项目(050440504)
关键词
电磁驱动AMD系统
斜拉桥
BP神经网络
遗传算法
electromagnet-drived active mass damper (AMD) system
cable-stayed bridge
BP neural network
genetic algorithm(GA)
