A CERLMDAN-PE-NLMS based denoising method for MEMS gyro signals
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
万方数据
随着武器技术的不断发展,常规弹药的制导化改造已成为一种必然趋势.通过应用精确制导技术,可以显著提高弹药的打击精度和效率.而在实现弹药制导化改造的过程中,精准测量角速度是一项关键核心技术.微机电系统(Micro-Electro Mechanical System,MEMS)陀螺仪存在输出信号噪声大、精度低的问题,针对上述问题,提出了一种自适应噪声完备集合鲁棒局部均值分解(CERLMDAN)和归一化 LMS算法(Normalized Least Mean Square,NLMS)结合的滤波模型.该模型通过在鲁棒局部均值分解(Robust Local Mean Decomposition,RLMD)过程中添加白噪声将原始数据分解为多个乘积函数(Product Functions,PF),并根据排列熵(Permutation Entropy,PE)将PF分为混合PF和有用PF;其次对混合PF使用NLMS去噪;最后,把处理后的PF和有用PF进行重构,得到去噪后的信号.试验表明,本文提出的去噪模型对信号均值与方差有显著提升,信号均值由 0.589 1 提升至 0.539 6,信号方差由 44.473 提升至 5.269 2.
With the continuous development of weapons technology,the guided transformation of conventional munitions has become an inevitable trend.Through the application of precision guidance technology,the striking accuracy and efficiency of munitions can be significantly improved.In the process of realizing the guided transformation of munitions,accurate measurement of angular velocity is a key core technology.The Micro-Electro-Mechanical System(MEMS)gyroscope has the problems of large output signal noise and low accuracy,and an adaptive noise-complete ensemble robust local mean decomposition(CERLMDAN)and Normalized LMS algorithm(NLMS)combined filtering model is proposed to address the above problems.The model decomposes the original data into multiple Product Functions(PFs)by adding white noise to the Robust Local Mean Decomposition(RLMD)process and divides the PFs into mixed PFs and useful PFs according to the Permutation Entropy(PE).Secondly,the mixed PFs are denoised by using the NLMS.Finally,the processed PFs and the useful PFs are reconstructed to obtain the denoised signal.The experiments show that the denoising model proposed in the paper has significantly improved the signal mean and variance,the signal mean is improved from 0.589 1 to 0.539 6,and the signal variance is increased from 44.473 to 5.269 2.
MEMS gyroscopeadaptive noise-complete ensemble robust local mean decompositionnonlocal mean noise reductionarrangement entropy
NETL
NSTL
万方数据
