摘要
基于变限积分理论,构造了响应函数最小二乘意义下的加权变限积分。通过适当次数的积分滑动平均,有效过滤测量噪声中的高频噪声。针对测量响应中残留的低频噪声,使用L_∞范数拟合正则化方法识别载荷,提出了一种选取L_∞范数拟合正则化方法最优正则化参数的单调性检验方法。数值仿真及试验验证说明单调性检验方法可以有效确定L_∞范数拟合正则化方法的最优正则化参数,得到比传统L_2范数正则化方法拟合性质更好精度更高的识别载荷;针对遥测数据中噪声特点,使用模拟遥测数据利用L_∞范数拟合正则化方法对冲击载荷进行了有效识别。
Based on the variable limit integral theory, the weighted variable limit integral of response function was constructed with the least squares estimation technique. By performing the integral moving average method several times, the high-frequency noise in the measured noise was filtered effectively. Aiming at the residual low-frequency noise in the measured response, the L∞norm fitting regularization method was used for load identification. In order to realize this regularization method, a monotonicity validation method was proposed to select the optimal regularization parameters. The numerical simulation and the test verification showed that the proposed monotonicity validation method can determine the optimal regularization parameters of the L∞norm fitting regularization method effectively; in addition, the identified load is more accurate and more reliable than that using the traditional L2norm regularization method; aiming at the characteristics of noise in telemetry data, the impact load is identified effectively by using the L∞norm fitting regularization method and simulated telemetry data. © 2017, Editorial Office of Journal of Vibration and Shock. All right reserved.
出处
《振动与冲击》
EI
CSCD
北大核心
2017年第9期101-107,共7页
Journal of Vibration and Shock
基金
国家自然科学基金资助(11372084)
