摘要
研究了京基100大厦AMD控制系统在线计算的关键技术。为实施在线实时控制,采用集中质量法将原结构简化为二维串联多自由度模型,且所建模型能较好地反应原结构动力特性,为结构主动控制在线计算提供了兼顾计算精度和效率的数值模型;分别设计了H2/LQR控制算法、极点配置算法、模糊神经网络控制算法等三种主动控制算法,并进行了对比分析,得到了合适的算法及相关参数;设计了基于加速度的状态观测器,基于此观测器确定了加速度传感器的安装位置。
The key techniques for implementation of online algorithms for AMD control system used in KK100 are studied in this paper.For the feasibility to execute real-time control,the structure was simplified to a reduced order model,which almost has the same dynamic properties as the original structure,and can meet the precision and efficiency requirements.Based on the reduced order model,three online active control algorithms including H2/LQR,pole assignment,and fuzzy control algorithm were designed and parameters are optimized and analyzed.The pole assignment algorithm is selected for the real AMD control application of KK100 with the three algorithms comparison.Meanwhile,an acceleration-based state observer is designed,with which the locations of accelerometers are determined.
出处
《振动与冲击》
EI
CSCD
北大核心
2011年第11期230-235,共6页
Journal of Vibration and Shock
基金
国家自然科学基金重点项目(50938001)
关键词
AMD控制
降阶模型
在线计算
观测器设计
AMD control
reduced-order model
online algorithm
acceleration-based state observer