摘要
为提高车辆动态称重压电信号的信噪比,要在奇异谱分析(SSA)降噪算法中确定合理的嵌入维数与重构阶次。基于Cao算法,提出一种稳定性方法确定嵌入维数,并以累积能量贡献率变化量确定重构阶次,在Lorenz信号上加入不同方差的高斯白噪声进行实验仿真,验证SSA降噪算法的可行性和有效性,并评价噪声方差对结果的影响。在不同车速下,对五种载重不同的车辆进行工程实测试验。结果表明,车速在10~50 km/h,平均称重误差控制在2.72%~4.72%,利用SSA降噪算法处理车辆动态称重压电信号,达到较好的降噪效果,称重精度及稳定性均能满足实际工程要求。
In order to increase the signal to noise ratio(SNR)of piezoelectric signals in vehicle weigh-in-motion(WIM),it is necessary to determine reasonable embedding dimension and reconstruction order of the singular spectral analysis(SSA)de-noising algorithm.Here,a new stability method based on Cao algorithm was presented to determine embedding dimension,the variation value of cumulative energy contribution rate was used to determine reconstruction order.Gaussian white noise with different variances was added into Lorenz signal to do simulation,verify the feasibility and validity of SSA de-noising algorithm,and evaluate the effect of noise variance on the results.Five vehicles with different loads and at different speeds were used to perform the actual engineering tests.The results showed that when the vehicle speed is within the range of10-50km/h,the average WIM error is controlled within the range of2.72%to4.72%,so using SSA de-noising algorithm to process vehicle WIM piezoelectric signals can reach a better de-noising effect,the accuracy and stability of WIM can meet practical engineering requirements.
作者
刘小锋
冯志敏
陈跃华
张刚
李宏伟
LIU Xiaofeng;FENG Zhimin;CHEN Yuehua;ZHANG Gang;LI Hongwei(School of Marine Science, Ningbo University, Ningbo 315211, China;Ningbo Shan Gong Intelligent Security Technology, Co, Ltd., Ningbo 315100, China)
出处
《振动与冲击》
EI
CSCD
北大核心
2018年第5期180-187,共8页
Journal of Vibration and Shock
基金
国家自然科学基金资助项目(51675286)
浙江省科技计划资助项目(2013C31045)
浙江省科技创新团队项目(2013TD21)
关键词
动态称重
压电信号
嵌入维数
重构阶次
SSA降噪算法
weigh-in-motion (WIM)
piezoelectric signal
embedding dimension
reconstruction order
SSA de-noising algorithm
