基于混沌蜂群算法的高耸结构振动信号小波阈值去噪方法研究

A Kent chaos artificial bee colony algorithm based wavelet thresholding for high-rise structure vibration signal denoising

为了有效去除实测振动信号中的噪声,改进了一种基于Kent混沌人工蜂群(KCABC)算法的振动信号小波阈值去噪方法。该算法采用Kent混沌映射初始化蜂群,引入锦标赛选择机制选择食物源,并结合混沌策略搜索最优解。基于广义交叉验证(GCV)阈值构造了目标函数,采用改进的KCABC算法搜索最优阈值,实现了不基于噪声先验知识的振动信号阈值去噪。通过对广州新电视塔4组实测振动信号的处理,比较了改进的KCABC算法与粒子群优化(PSO)算法、标准蜂群(SABC)算法以及Logistic混沌蜂群(LCABC)算法的去噪性能。结果表明:提出的KCABC算法具有较快的收敛速度和较高的搜索精度,能够有效去除高耸结构振动信号中的噪声部分。

A Kent chaos artificial bee colony( KCABC) algorithm based wavelet thresholding has been put forward for eliminating noise in the measured vibration signals. The Kent chaotic mapping is employed to initialize the colony.Besides,the tournament selection mechanism is introduced to select the food source. The chaotic search is also occupied to explore the optimum solution. The KCABC algorithm is then utilized to search the optimal threshold value through minimizing the generalized cross validation( GCV) threshold based objective function. The vibration signal denoising that is not based on the prior knowledge of the noise is thus implemented. Four sets of measured vibration signals of Guangzhou new TV tower are processed. The particle swarm optimization( PSO) algorithm,standard artificial bee colony( SABC) algorithm,and Logistic chaos artificial bee colony( LCABC) algorithm are also taken for contrast experiments. The comparison results indicate that the proposed KCABC algorithm has better enhancement in convergence speed and precision,and can effectively eliminate the noise in the vibration signals of the high-rise structure.