模拟脉动风速场下高层结构调谐质量阻尼器群智能优化控制*
- 摘 要
-
（1 合肥工业大学土木与水利工程学院， 合肥 230009； 2 土木工程结构与材料安徽省级实验室， 合肥 230009）
［摘要］利用抽取时间的快速傅里叶变换技术优化谐波叠加法，根据Davenport脉动风速谱，模拟了风速时程曲线，在生成的脉动风速谱密度曲线中，目标谱与模拟谱走势高度一致，为风振控制计算的精确性提供了有力保证。高层结构采用的是国际通用的第二代Benchmark模型，控制方法为调谐质量阻尼器(TMD)控制；并引入粒子群优化算法，以结构层绝对位移和的无穷范数为目标函数进行了优化研究；为考察此方法的鲁棒性，分别对结构考虑了无刚度不确定性以及具有±15%刚度不确定性的情形，数值计算结果表明了优化TMD控制在高层结构风振控制中的有效性，为人工智能技术在传统结构振动控制领域的应用提供了广阔的思路。［关键词］调谐质量阻尼器; 谐波叠加法; Benchmark结构; 层绝对位移和; 粒子群算法中图分类号：TU112.41 文献标识码：A 文章编号：1002-848X(2015)19-0041-07
Intelligent optimal control of tuned mass damper groups for highrise structure excited by simulated fluctuating wind speed fieldWang Xiaojin1, 2, Zhu Dayong1, 2, Huang Wei1, 2, Liu Cheng1, 2, Lu Kunlin1, 2（1 School of Civil and Hydraulic Engineering, Hefei University of Technology, Hefei 230009, China; 2 Anhui Provincial Laboratory of Civil Engineering and Materials, Hefei 230009, China）Abstract: The Fast Fourier Transform technology based on Decimation In Time (DIT-FFT) was employed to optimize the harmonic wave synthesis method, and the wind speed time-history curves were numerically simulated according to Davenport fluctuating wind speed spectrum. As observed from the generated spectral density function curves, the objective spectrum was highly consistent with the trend of simulated spectrum, which provided a strong guarantee for accuracy of windinduced vibration control computing. The second generation Benchmark structure was universal internationally, and tuned mass damper (TMD) algorithm was served as the controller for high-rise structure. Particle Swarm Optimization (PSO) algorithm was introduced to be the optimization tool, and the infinity norm of Layer Absolute Displacement Sum (LADS) was selected as the fitness function to carry out optimization study. In addition, no uncertainty and ±15% stiffness uncertainty in stiffness were considered respectively for robustness study. The numerical calculation results reveal that optimal control of TMD has an excellent capability in reduction of the wind-induced vibration of the high-rise building, and it can provide a broad view for application of artificial intelligence technology in traditional structural vibration control field.Keywords: tuned mass damper; wave synthesis method; Benchmark structure; Layer Absolute Displacement Sum; Particle Swarm Optimization
*国家自然科学基金资助项目（51179043）。作者简介：王小金,硕士研究生, Email: xiaojin_wang_hfut@126.com。
