摘要
传感器最优配置的任务是构建一个最佳的传感器数量、位置,以期获得传感器成本与系统指定监测性能指标之间的最佳平衡。在能够反映结构特性的重要位置布置合理数量的传感器不但可以有效提高经济性,而且由于传感器的不同布置影响结构动力参数的测试结果,优化传感器的布置位置还可以提高结构的测试精度。针对基于结构损伤识别的传感器的最优布置、基于Fisher信息阵行列式值最大的有效独立法和基于模态应变能最大的运动能量法的缺陷而提出的有效独立-驱动点残差法以及基于二重结构编码遗传算法的传感器优化布置法进行了比较分析,指出了这三种方法对减振控制结构传感器优化布置实际应用中存在的问题。通过对比分析可知三种改进方法均能有效地进行传感器的优化布置,基于结构损伤识别的传感器布置能准确有效地得到存在噪声下的结构损伤位置及程度;有效独立-驱动点残差法可以有效地避免信息丢失问题且能较好保存结构的特性;改进遗传算法能够搜索到最优解且识别精度较高。为进一步进行优化布置的研究提供一定借鉴与参考。
The goal of optimizing sensor arrangement is realized by optimal number and location of sensors,and a balance between the cost of sensors and the specified target of monitoring.The cost of a structural system could be improved effectively by placing a proper number of sensors on important locations of the observed structure,and the sensor position affects the test results of structural dynamic parameters.Three effective arrangement methods for sensor optimization were put forward in the paper.The approach for optimal sensor placement was investigated and analyzed among the three methods.the existing problems on the practical application of sensor optimal arrangement were figured out in damping control structures.Results show that: the three different methods can effectively optimize the arrangement of sensors;accurate results indicating the location and degree of structural damages under noisy test data can be obtained efficiently by sensors optimization;the loss of data can be avoided and the characteristics of structures could be preserved effectively by the effective impendence driving-point residue method;the improved genetic algorithm can search for the optimal solution and obtain a higher recognition accuracy.And the results will provide some reference for further research of optimal arrangement.
出处
《建筑结构学报》
EI
CAS
CSCD
北大核心
2010年第S2期100-104,共5页
Journal of Building Structures
基金
陕西省教育厅专项科研基金(09JK550)
江苏省结构重点实验室开放基金(ZD0803)
西安市科技攻关课题(GG04051)
陕西省自然科学基金(2003E230)
关键词
减振控制结构
传感器优化布置
有效独立-驱动点残差法
遗传算法
damping control structure
optimal sensor placement
effective impendence driving-point residue method
genetic algorithms
