摘要
当前地震动速度时程时域特性分析方法,仅能分析岩溶区地震动速度时程的振动周期、强弱程度与时间变动的问题,但未能准确计算岩溶区砌体建筑物自振频率,导致砌体建筑环境振动特性分析结果存在误差。深入研究岩溶区砌体建筑环境振动特性分析方法,构建混凝土损伤塑性模型,分析岩溶区砌体建筑材料的屈服(受压)应力-非弹性应变关系、开裂(受拉)应力-非弹性应变关系和损伤因子;采用贝叶斯方法检测岩溶区砌体建筑受压受拉时的自振频率,通过L-M神经网络法消除自振频率后,使用振动特性分析方法准确分析岩溶区砌体建筑环境振动特性。实验结果表明,所提方法分析准确率高达0.99,分析16栋岩溶区砌体建筑环境振动特性耗时仅有5 ms,具有较高的分析精度和效率。
In the past,using the method of analyzing the time-domain characteristics of ground vibration speed,only the problem of the vibration period,the degree of strength,and the time change of ground vibration speed time in the Karst area can be analyzed,but the self-vibration frequency of masonry buildings in the Karst area can not be accurately calculated. The results of analysis on the vibration characteristics of masonry building environment have some errors. To study the new analysis method of vibration characteristics of masonry environment in Karst area,the paper builds concrete damage plastic model,analyzes the yield(compression) stress-inelastic strain relation,crack(tensile)stress-inelastic strain relation and damage factor of masonry building materials in Karst area;Bayesian method was used to detect the vibration frequency of masonry buildings under pressure and pull in the Karst area. After eliminating the vibration frequency by the L-M neural network method,the vibration characteristics analysis method was used to accurately analyze the vibration characteristics of masonry buildings in the Karst area. The experimental results show that the analysis accuracy of the proposed method is as high as 0.99,and the analysis of the vibration characteristics of 16 masonry buildings in Karst areas takes only 5 ms,which has high analysis accuracy and efficiency.
作者
周磊
ZHOU Lei(Yangling Vocational and Technical College,yangling 712100,China)
出处
《华南地震》
2019年第1期97-103,共7页
South China Journal of Seismology
关键词
岩溶区
砌体建筑
环境振动特性
贝叶斯
L-M神经网络
振型
Karst area
Masonry building
Environmental vibration characteristics
Bayesian
L-M neural network
Vibration mode