1.Key Laboratory of Modern Measurement & Control Technology,Ministry of Education,Beijing Information Science & Technology University,Beijing 100192,China
2.China North Vehicle Research Institute,Beijing 100072,China
To achieve the extrapolation of the full-cycle load spectrum from limited load data, this paper proposes a load extrapolation method based on local adaptive bandwidth diffusion kernel density estimation, in view of the limitations of traditional adaptive bandwidth optimization algorithms in parameter selection. This method first reduces the two-dimensional rainflow matrix to one-dimensional equivalent amplitude, then optimizes the local bandwidth based on the local integral mean square error, and uses the local optimal bandwidth to construct a probability density distribution model through diffusion kernel density estimation. Finally, the target frequency load is extrapolated by combining the Monte Carlo simulation method. The preprocessed load data of a certain special vehicle's comprehensive transmission device are compared and verified. The results show that compared with the traditional method, the probability density distribution curve and cumulative frequency curve obtained by the method proposed in this paper are closer to the actual equivalent amplitude. The correlation coefficients and determination coefficients are both closer to 1, with the correlation coefficients being 0.983 8 and 0.999 6 respectively, and the determination coefficients being 0.967 9 and 0.999 1 respectively. The root mean square errors are also smaller, being 5.05×10-5 and 15.9 respectively.
通过式(17),将合成工况的二维均幅值雨流矩阵进行降维,得到一维等效幅值。为验证本文提出的基于局部自适应带宽扩散核密度估计方法的准确性,将其与以下方法进行对比分析:基于式(7)固定带宽扩散核密度估计的载荷外推法(Diffusion kernel density estimation-fixed bandwidth-experience,DKDE-FB-EXP)、基于改进Sheather-Jones法(Improved Sheather-Jones,ISJ)[22]固定带宽扩散核密度估计的载荷外推法(Diffusion kernel density estimation-fixed bandwidth-improved Sheather-Jones,DKDE-FB-ISJ)、结合上述2种固定带宽与式(8)的自适应带宽扩散核密度估计载荷外推法((Diffusion kernel density estimation-adaptive bandwidth-experience,DKDE-AB-EXP)与(Diffusion kernel density estimation-adaptive bandwidth-improved Sheather-Jones,DKDE-AB-ISJ))、基于局部自适应带宽核密度估计的载荷外推法(Kernel density estimation-local adaptive bandwidth,KDE-LAB),并以等效幅值数据作为参照。所有方法均选择高斯核函数,见式(3)。为提高计算效率,设定128个估计点。对于KDE-LAB方法与本文提出的方法,通过归一化等效幅值的最大差值、最小差值及分块数,设置相等的初始带宽h0与初始间隔长度H0,表达式为:
由图5可知,DKDE-FB-EXP方法与DKDE-AB-EXP方法得到的密度分布曲线重合,DKDE-FB-ISJ方法与DKDE-AB-ISJ方法得到的密度分布曲线同样重合,表明2种带宽求取方法的固定带宽与自适应带宽估计结果一致。由于自适应带宽为固定带宽与自适应系数的乘积,因此估计结果相同时,采用固定带宽较为便捷。与其他外推方法相比,本文所提方法更贴近等效幅值的密度分布情况。为了进一步对比密度分布拟合效果,计算等效幅值与上述外推方法间的相关系数R、决定系数R2及均方根误差(Root mean square error,RMSE)。其中,相关系数R与决定系数R2越接近1、RMSE越小,说明拟合效果越好。等效幅值分布的拟合结果如表2所示。
XuBao-rong, WangTao, LiangZi. Generating method and process of load spectra for tracked vehicles based on prediction of manipulating action[J]. Acta Armamentarii, 2022, 43(2): 252-259.
LuoQing-guo, WangXu-dong, ZhangGeng-yun, et al. Loading spectrum simulation and experiment study of the armored vehicle engine[J]. Acta Armamentarii, 2012, 33(12): 1442-1447.
NiuWen-tie, CaiFu-you, FuJing-jing. Load extrapolation method based on adaptive bandwidth kernel density estimation[J].Transactions of the Chinese Society for Agricultural Machinery, 2021, 52(1): 375-384.
YuLi-juan, AnYang, HeJia-long, et al. Research progress and development trend of extrapolation method in electromechanical equipment load spectrum[J].Journal of Jilin University (Engineering and Technology Edition), 2021, 52(1): 375-384.
ZhangWen-sheng. Research on load spectrum compiling for transmission shaft in transmission system of tracked vehicle[D]. Beijing: School of Mechanical Engineering, Beijing Institute of Technology, 2016.
LiuYan, ZhangXi-feng, WangZhen-yu, et al. Contrast of extrapolations in compiling load spectrum[J]. Modern Manufacturing Engineering, 2011(11): 8-11.
[15]
ChenC H, YangZ J, HeJ L, et al. Load spectrum generation of machining center based on rainflow counting method[J]. Journal of Vibroengineering, 2017, 19(8): 5767-5779.
[16]
HeJ L, WangS X, LiG F,et al. Compilation of NC lathe dynamic cutting force spectrum based on two-dimensional mixture models[J]. The International Journal of Advanced Manufacturing Technology, 2018, 98(1/4): 251-262.
[17]
ZhangJ, LuJ W, HanW D, et al. Program load spectrum compilation for accelerated life test of parabolic leaf spring[J]. International Journal of Automotive Technology, 2019, 20(2): 337-347.
YuJia-wei, ZhengSong-lin, ZhaoLi-hui, et al. Extrapolation method of vehicle big-data service-loads based on non-parametric KDE method[J]. China Mechanical Engineering, 2021, 32(11): 1307-1314.
[20]
WangJ X, LiuY, ZengX H, et al. Selection method for kernel function in nonparametric extrapolation based on multicriteria decision-making technology[J]. Mathematical Problems in Engineering, 2013, 2013(8): 1-11.
GaoTian-yu, LiHuan-liang, ZhengZheng, et al.Compiling method of load spectrum of ZLK50 type loader hydraulic cylinder based on rain flow frequency extrapolation[J]. Journal of Mechanical Strength, 2017, 39(4): 951-956.
[23]
WangS X, LiG F, YangH J, et al.Research on extreme value distribution of cutting force in CNC lathe based on generalized extreme value distribution[C]∥2019 4th International Conference on System Reliability and Safety(ICSRS). Piscataway, N J: IEEE, 2019:198-202.
[24]
FogleJ, AgarwalP, ManuelL. Towards an improved understanding of statistical extrapolation for wind turbine extreme loads[J]. Wind Energy, 2008, 11(6): 613-635.
[25]
YuanK, ChenX Q, GuiZ P, et al.A quad-tree-based fast and adaptive kernel density estimation algorithm for heat-map generation[J]. International Journal of Geographical Information Science, 2019, 33(12): 2455-2476.
[26]
RosenblattM. Remarks on some nonparametric estimates of a density function[J]. The Annals of Mathematical Statistics, 1956, 27(3): 832-837.
[27]
ParzenE. On estimation of probability density function and mode[J]. The Annals of Mathematical Statistics, 1962, 33(3): 1065-1076.
LiYan. Research on non-parametric extrapolation method in compiling load spectra of wheel loader[D]. Changchun:School of Mechanical Science and Engineering, Jilin University, 2016.
[30]
AbramsonI S. On bandwidth variation in kernel estimates-a square root law[J]. The Annals of Statistics,1982, 10(4): 1217-1223.
ZhangLong-sheng. Fatigue damage analysis of automobile rear subframe based on kernel density estimation via diffusion[D]. Qinhuangdao:School of Vehicles and Energy, Yanshan University, 2021.
[33]
BotevZ I, GrotowskiJ F, KroeseD P.Kernel density estimation via diffusion[J].The Annals of Stats,2010, 38(5): 2916-2957.
[34]
XuX Y, YanZ, XuS L. Estimating wind speed probability distribution by diffusion based kernel density method[J]. Electric Power Systems Research,2015, 121: 28-37.
YanChu-liang, WangGong-quan. Investigation of rain flow counting method and its program of statistical treatment[J]. Transactions of the Chinese Society for Agriculture Machinery, 1982(4): 90-103.
[37]
SmithA C, BrownE N. Estimating a state-space model from point process observations[J]. Neural Computation, 2003, 15(5): 965-991.