弹性网络正则化移动荷载识别试验研究

Experimental study on moving force identification based on elastic network regularization

针对移动荷载稀疏正则化识别方法存在的不足,提出了弹性网络正则化识别法,构建了弹性网络正则化识别理论和求解方法,分析了单移动荷载工况下3种参数对识别结果的影响。利用MATLAB软件建立了数值模型并进行验证。研究结果表明:该方法在车辆移动速度、测量噪声水平、测点位置组合的相应工况中,识别结果的相对误差分别为6.5%、10%和6.5%,稀疏性较好,能反映移动荷载的特征。试验识别结果与真实荷载吻合度高,且该方法在不同工况中适应性较强,可供工程应用参考。

In view of the shortcoming of moving load regularization identification method,the elastic network regularization identification method was proposed. The theory andsolution strategy of the elastic network regularization identification were established. The influence of three parameters on the identification result was analysed under single moving force. The numerical model was then established using the MATLAB software to verify the effectiveness of the proposed method. The results show that, when the proposed methodis used in the case of vehicle moving speed,measurement noise level and combination measurement point location, the relative errors of identification resultsare 6.5%,10% and 6.5%,respectively. The sparsity is satisfactory.The result can reflect the moving force characteristic. The identification result is in good agreement with the real force. The proposed method has strong adaptability in different working conditions,that can be used as reference for engineering application.