基于Fast-ICA算法的改进EEMD算法在桥梁工程中的运用

Application of Improved EEMD Algorithm in Bridge Engineering Based on Fast-ICA Algorithm

目前,以振动分析为基础的桥梁结构健康状态评估方法逐渐受到人们的重视,模态参数识别作为振动分析的关键问题之一,还需进一步完善。模态参数识别即是对结构的振动信号进行动力特性参数的识别,以得到结构的频率、振型以及阻尼比。为了精确地识别出桥梁结构的模态参数,需先对传感器采集的结构响应信号进行分解和重构以保留结构的真实信息,再对重构信号进行模态参数辨识。现阶段,集合经验模态分解算法作为常用的信号分解算法之一,存在两大缺陷,即所得本征模态函数间存在模态混叠现象和有效本征模态函数的定义没有统一的标准。基于此提出对应的改进算法,首先将夹角余弦法用于筛选有效IMF分量;其次将盲源分离算法中的Fast-ICA算法嵌入EEMD分解算法中以避免模态混叠现象的发生;接着将所提算法运用于模拟信号以验证其可行性;最后将所提改进算法运用于实际桥梁工程的响应信号中,并对比分析所得模态参数值以验证该算法的可靠性。结果表明,所提算法较现有EEMD算法具有更好的信号分解效率,且能够有效地剔除结构的虚假信息,保留实际桥梁结构的真实信息。

At present,the method for evaluating the health status of bridge structures based on vibration analysis has gradually attracted people’s attention.As one of the key issues in vibration analysis,modal parameter identification needs to be further improved.Modal parameter identification is the identification of dynamic characteristic parameters of structural vibration signal to obtain the frequency,mode and damping ratio of the structure.In order to accurately identify the modal parameters of bridge structure,it is necessary to decompose and reconstruct the structure response signals collected by sensor to retain the real information of the structure,and then to identify the modal parameters on the reconstructed signals.Currently,as one of the commonly used signal decomposition algorithms,the ensemble empirical modal decomposition algorithm still has 2 weaknesses:one is mode aliasing between the obtained intrinsic mode functions,the other is no universal criterion in the definition of effectiveness intrinsic mode functions.Based on this,the corresponding improved algorithm is proposed.First,the effective IMF components are selected using the angle cosine method.Second,the Fast-ICA algorithm in the blind source separation algorithm is embedded in the EEMD decomposition algorithm to avoid the occurrence of modal aliasing.Then,the proposed algorithm is applied to analog signals to verify its feasibility.Finally,the proposed improved algorithm is applied to the response signals of actual bridge engineering,and the obtained modal parameter values are compared and analyzed to verify the reliability of the algorithm.The result shows that the proposed algorithm has better signal decomposition efficiency than the existing EEMD algorithm,it can effectively eliminate the false information of the structure and keep the real information of the actual bridge structure.