基于混沌表示和特征注意力机制的机床两轴动态误差预测

针对传统方法难以揭示机床多轴插补动态误差的序列产生机制,各时间维度上的误差时序特征存在相互关联的问题,提出一种融合混沌表示(Chaotic representation, CR)和特征注意力机制(Feature attention mechanism, FA)的级联动态误差预测模型。首先,在证明多元动态误差时变演化具有混沌特性的基础上,对其进行相空间重构,将动态误差参数时间序列背后隐藏的信息在相空间中进行表达。然后,融合特征注意力机制在时间维度上动态分配相点特征权重的同时降低高维演化相空间信息冗余,进一步重塑原系统的动力学状态向量空间。最后,考虑到混沌时变演化具有长程相关性,采用双向长短期记忆(Bi-directional long short-term memory, Bi-LSTM)网络模型逼近混沌相空间内的动力学特性,实现动态误差混沌时间序列信息的有效预测。通过XK-L540型数控铣床实测数据的算例表明,相较于CRFA-LSTM模型,以及单一级联模型CR-Bi-LSTM、FA-Bi-LSTM,本文算法的均方根误差分别降低约35%、16%和43%。

Time series prediction of mining subsidence based on a SVM

In order to study dynamic laws of surface movements over coal mines due to mining activities,a dynamic prediction model of surface movements was established,based on the theory of support vector machines(SVM) and times-series analysis.An engineering application was used to verify the correctness of the model.Measurements from observation stations were analyzed and processed to obtain equal-time interval surface movement data and subjected to tests of stationary,zero means and normality.Then the data were used to train the SVM model.A time series model was established to predict mining subsidence by rational choices of embedding dimensions and SVM parameters.MAPE and WIA were used as indicators to evaluate the accuracy of the model and for generalization performance.In the end,the model was used to predict future surface movements.Data from observation stations in Huaibei coal mining area were used as an example.The results show that the maximum absolute error of subsidence is 9 mm,the maximum relative error 1.5%,the maximum absolute error of displacement 7 mm and the maximum relative error 1.8%.The accuracy and reliability of the model meet the requirements of on-site engineering.The results of the study provide a new approach to investigate the dynamics of surface movements.

The predictability of atmospheric and oceanic motions:Retrospect and prospects

This paper reviews the historic understanding of the predictability of atmospheric and oceanic motions, and interprets it in a general framework. On this basis, the existing challenges and unsolved problems in the study of the intrinsic predictability limit(IPL) of weather and climate events of different spatio-temporal scales are summarized. Emphasis is also placed on the structure of the initial error and model parameter errors as well as the associated targeting observation issue. Finally, the predictability of atmospheric and oceanic motion in the ensemble-probabilistic methods widely used in current operational forecasts are discussed.The necessity of considering IPLs in the framework of stochastic dynamic systems is also addressed.

Dynamic-statistics combined forecast scheme based on the abrupt decadal change component of summer precipitation in East Asia

Based on the 1983~2011 CMAP data,the precipitation anomaly in East Asia and its nearby sea regions（hereafter called East Asia for short） demonstrates the ＂＋-＋＂ pattern before 1999 and the ＂-＋-＂ pattern afterwards; this decadal change is contained principally in the corresponding EOF3 component.However,the NCC_CGCM forecast results are quite different,which reveal the ＂＋-＋-＂ pattern before 1999 and the ＂-＋-＋＂ pattern afterwards.Meanwhile,the probability of improving NCC_CGCM＇s forecast accuracy based on these key SST areas is discussed,and the dynamic-statistics combined forecast scheme is constructed for increasing the information of decadal change contained in the summer precipitation in East Asia.The independent sample forecast results indicate that this forecasting scheme can effectively modify the NCC_CGCM＇s decadal change information contained in the summer precipitation in East Asia（especially in the area of 30°N–55°N）.The ACC is 0.25 and ACR is 61% for the forecasting result based on the V SST area,and the mean ACC is 0.03 and ACR is 51% for the seven key areas,which are better than NCC_CGCM＇s system error correction results（ACC is -0.01 and ACR is 49%）.Besides,the modified forecast results also provide the information that the precipitation anomaly in East Asia mainly shows the ＂＋-＋＂ pattern before 1999 and the ＂-＋-＂ pattern afterwards.