Application of Machine Learning for Tool Condition Monitoring in Turning

The machining process is primarily used to remove material using cutting tools. Any variation in tool state affects the quality of a finished job and causes disturbances. So, a tool monitoring scheme (TMS) for categorization and supervision of failures has become the utmost priority. To respond, traditional TMS followed by the machine learning (ML) analysis is advocated in this paper. Classification in ML is supervised based learning method wherein the ML algorithm learn from the training data input fed to it and then employ this model to categorize the new datasets for precise prediction of a class and observation. In the current study, investigation on the single point cutting tool is carried out while turning a stainless steel (SS) workpeice on the manual lathe trainer. The vibrations developed during this activity are examined for failure-free and various failure states of a tool. The statistical modeling is then incorporated to trace vital signs from vibration signals. The multiple-binary-rule-based model for categorization is designed using the decision tree. Lastly, various tree-based algorithms are used for the categorization of tool conditions. The Random Forest offered the highest classification accuracy, i.e., 92.6%.

     

Quantum algorithm for neighborhood preserving embedding

Neighborhood preserving embedding (NPE) is an important linear dimensionality reduction technique that aims at preserving the local manifold structure. NPE contains three steps, i.e., finding the nearest neighbors of each data point, constructing the weight matrix, and obtaining the transformation matrix. Liang et al. proposed a variational quantum algorithm (VQA) for NPE [Phys. Rev. A 101 032323 (2020)]. The algorithm consists of three quantum sub-algorithms, corresponding to the three steps of NPE, and was expected to have an exponential speedup on the dimensionality n. However, the algorithm has two disadvantages: (i) It is not known how to efficiently obtain the input of the third sub-algorithm from the output of the second one. (ii) Its complexity cannot be rigorously analyzed because the third sub-algorithm in it is a VQA. In this paper, we propose a complete quantum algorithm for NPE, in which we redesign the three sub-algorithms and give a rigorous complexity analysis. It is shown that our algorithm can achieve a polynomial speedup on the number of data points m and an exponential speedup on the dimensionality n under certain conditions over the classical NPE algorithm, and achieve a significant speedup compared to Liang et al.'s algorithm even without considering the complexity of the VQA.     

PSO-LSSVM对LIBS定量分析精度的提高

针对土壤定量分析受基体效应影响大，LIBS定量分析精度不佳等问题，采用粒子群算法对LSSVM进行优化，提高模型的精确度。选取Pb Ⅰ 405.78 nm和Cr Ⅰ 425.44 nm作为分析谱线进行分析。采集十二个不同浓度样品的特征光谱，每个浓度样品在不同点采集20组数据，将其中17组数据设为训练集，3组数据设为预测集，用LSSVM和PSO-LSSVM两种方法建立定标模型。对比两种模型的拟合相关系数（R2）、训练集均方根误差（RMSEC）和预测集均方根误差(RMSEP)。由于自吸收效应的影响，随着浓度的增加，预测值逐渐低于实际值，LSSVM定标模型的拟合程度较低，无法达到实验要求，模型性能有待提高。利用粒子群算法对LSSVM的模型参数惩罚系数和核函数参数进行优化，得到最佳的参数组合，Pb元素为(8 096.8, 138.865 7)，Cr元素为(4 908.6, 393.563 5)，用最佳的参数组合构建LSSVM的定标模型。相比于LSSVM，PSO-LSSVM定标模型的精确度更高，Pb和Cr元素的R2提高到了0.982 8和0.985 0，拟合效果明显提升。Pb和Cr元素的训练集均方根误差由0.026 0 Wt%和0.027 2 Wt%下降到0.022 4 Wt%和0.019 1 Wt%，预测集均方根误差由0.101 8 Wt%和0.078 8 Wt% 下降到0.045 8 Wt%和0.042 0 Wt%，模型的稳定性进一步提高。说明PSO-LSSVM算法能够更好地降低土壤基体效应和自吸收效应带来的影响，提高分析结果的精确度与稳定性。     

Co-Training for Visual Object Recognition Based on Self-Supervised Models Using a Cross-Entropy Regularization

Automatic recognition of visual objects using a deep learning approach has been successfully applied to multiple areas. However, deep learning techniques require a large amount of labeled data, which is usually expensive to obtain. An alternative is to use semi-supervised models, such as co-training, where multiple complementary views are combined using a small amount of labeled data. A simple way to associate views to visual objects is through the application of a degree of rotation or a type of filter. In this work, we propose a co-training model for visual object recognition using deep neural networks by adding layers of self-supervised neural networks as intermediate inputs to the views, where the views are diversified through the cross-entropy regularization of their outputs. Since the model merges the concepts of co-training and self-supervised learning by considering the differentiation of outputs, we called it Differential Self-Supervised Co-Training (DSSCo-Training). This paper presents some experiments using the DSSCo-Training model to well-known image datasets such as MNIST, CIFAR-100, and SVHN. The results indicate that the proposed model is competitive with the state-of-art models and shows an average relative improvement of 5% in accuracy for several datasets, despite its greater simplicity with respect to more recent approaches.     

Dynamic Modeling of the Feed Drive System of a CNC Metal Cutting Machine

Studying the vibrational behavior of feed drive systems is important for enhancing the structural performance of computer numerical control (CNC) machines. The preload on the screw and nut position have a great influence on the vibration characteristics of the feed drive as two very important operational conditions. Rotational acceleration of the screw also affects the performance of the CNC feed drive when machining small parts. This paper investigates the influence of preload and nut position on the vibration characteristics of the feed drive system of a CNC metal cutting machine in order to be able to eliminate an observed resonance occurred at high rotational speeds of the screw, corresponding to high feed rates. Additionally, rational structural parameters of the feed drive system are selected in order to increase the rotational acceleration for improving the performance of the CNC machine. Experiments and analyses showed that by selecting specific parameters of feed drive system and simultaneously applying a certain value of preload, a 97% increase in rotational acceleration and 30% time reduction considering the vibration resistance at high rotational speeds can be achieved.     

紫色球杆菌视紫红质光谱特性的机器学习研究

近年来，机器学习等人工智能技术被应用于蛋白质工程，其在蛋白质结构、功能预测、催化活性等研究中具有独特优势。在未知蛋白质结构的情况下，将蛋白质序列和功能特性与机器学习相结合，基于序列-活性关系（innovative sequence-activity relationship，ISAR）算法，将蛋白质氨基酸序列数字化，用快速傅里叶变换（fast four transform，FFT）进行预处理，再进行偏最小二乘回归建模，可在数据集较少情况下拟合得到最佳模型。通过机器学习对紫色球杆菌视紫红质（gloeobacter violaceus rhodopsin，GR）的突变体蛋白质氨基酸序列与光谱最大吸收波长进行建模，获得了最佳模型。用最佳索引LEVM760106建模得到的确定系数R² 为0.944，均方误差E为11.64。用小波变换进行的预处理，其R² 虽也约为0.944，但E大于11.64，不及FFT进行的预处理。方法较好地解决了蛋白质序列与功能特性之间的数学建模问题，在蛋白质工程中可为预测更优的突变体提供支持。     

Optimal time delays in a class of reaction-diffusion equations

ABSTRACT

A class of semilinear parabolic reaction diffusion equations with multiple time delays is considered. These time delays and corresponding weights are to be optimized such that the associated solution of the delay equation is the best approximation of a desired state function. The differentiability of the mapping is proved that associates the solution of the delay equation to the vector of weights and delays. Based on an adjoint calculus, first-order necessary optimality conditions are derived. Numerical test examples show the applicability of the concept of optimizing time delays.     

利用带无标签数据的双支持向量机对恒星光谱分类

恒星光谱分类是天文技术与方法领域一直关注的热点问题之一。随着观测设备持续运行和不断改进，人类获得的光谱数量与日俱增。这些海量光谱为人工处理带来了极大挑战。鉴于此，研究人员开始关注数据挖掘算法，并尝试对这些光谱进行数据挖掘。近年来，神经网络、自组织映射、关联规则等数据挖掘方法广泛应用于恒星光谱分类。在这些方法中，支持向量机(SVM)以其强大的学习能力和高效的分类性能而备受推崇。SVM的基本思想是试图在两类样本之间找到一个最优分类面将两类分开。SVM在求解时，通过将其最优化问题转化为具有(QP)形式的凸问题，进而得到全局最优解。尽管该方法在实际应用中表现优良，但为了进一步提高其分类能力，有的学者提出双支持向量机(TSVM)。该方法通过构造两个非平行的分类面将两类分开，每一类靠近某个分类面，而远离另一个分类面。TSVM的计算效率较之传统SVM提高近4倍，因此，自TSVM提出后便受到研究人员的持续关注，并出现若干改进算法。在恒星光谱分类中，一般分类算法都是根据历史观测光谱来建立分类模型，其中最关键的是对光谱进行人工标注，这项工作极为繁琐，且容易犯错。如何利用已标记的光谱以及部分无标签的光谱来建立分类模型显得尤为重要。因此，提出带无标签数据的双支持向量机(TSVMUD)用以实现对恒星光谱智能分类的目的。该方法首先将光谱分为训练数据集和测试数据集两部分；然后，在训练集上进行学习，得到分类依据；最后利用分类依据对测试集上的光谱进行验证。继承了双支持向量机的优势，更重要的是，在训练集上学习分类模型过程中，不仅考虑有标记的训练样本，也考虑部分未标记的样本。一方面提高了学习效率，另一方面得到更优的分类模型。在SDSS DR8恒星光谱数据集上的比较实验表明，与支持向量机SVM、双支持向量机TSVM以及K近邻(KNN)等传统分类方法相比，带无标签数据的双支持向量机TSVMUD具有更优的分类能力。然而，该方法亦存在一定的局限性，其中一大难题是其无法处理海量光谱数据。该工作将借鉴海量数据随机采样思想，利用大数据处理技术，来对所提方法在大数据环境下的适应性展开进一步研究。     

Ensemble-SISPLS近红外光谱变量选择方法

近红外光谱具有高维小样本的特点，变量选择是提高定量分析模型稳健性和可解释性的一种有效方法。确定独立筛选（SIS）是一种基于边际相关性的超高维数据变量选择方法，广泛用于基因微阵列数据的变量选择。SIS具有将数据维度降低至样本大小规模的能力，其降维能力与LASSO相当，在相当宽泛的近似条件下，由于具有安全筛选性质，所有重要变量被保留的概率趋于1。基于确定独立筛选偏最小二乘（SIS-SPLS）的变量选择是一种迭代式的SIS变量选择方法，首先利用SIS方法完成光谱重要变量的初选；然后根据重要变量的边际相关性大小进行逐步前向选择：建立偏最小二乘回归模型，依据贝叶斯信息准则（BIC）确定最终的变量选择结果。SIS-SPLS以逐步前向选择的方式实现对重要变量的增量式筛选，随着潜变量个数的增加及因变量残差的逐步减小，SIS-SPLS方法选择的变量个数将趋于稳定。然而仅以边际相关性对变量重要性进行评价，当光谱变量个数远大于样本数时，该方法也存在选择的变量过多、变量选择结果不够稳健等问题。为进一步提高小样本情况下变量选择的稳健性，将集成学习引入SIS-SPLS方法之中，提出了一种集成SIS-SPLS变量选择方法（Ensemble-SISPLS）。该方法首先对校正集样本进行自助重采样，对采样得到的每一个校正子集分别使用SIS-SPLS方法进行变量筛选，通过投票机制并设置频次阈值对所有校正子集的变量选择结果进行集成，选择出现频次大于给定阈值的变量并建立偏最小二乘回归模型，计算5折交叉验证均方根误差。对频次阈值和潜变量个数两个关键参数使用网格搜索法进行优选，根据子模型的交叉验证均方根误差和变量个数对子模型性能进行综合评价，以最优子模型包含的变量作为最终的变量选择结果。分别在Corn数据集和当归数据集上进行变量选择实验，比较Ensemble-SISPLS，SIS-SPLS和UVE-PLS三种变量选择方法的性能。其中当归数据集共77个样本，样本采自甘肃岷县和渭源县，使用Nicolet-6700型近红外光谱仪扫描得到所有样本的近红外光谱并对当归中的阿魏酸含量进行预测。Ensemble-SISPLS方法在Corn数据集上选择的变量个数、RMSEP和决定系数分别为22，0.000 8和0.999 8；SIS-SPLS方法在Corn数据集上选择的变量个数、RMSEP和决定系数分别为97，0.007 3和0.998 8。Ensemble-SISPLS方法在当归数据集上选择的变量个数、RMSEP和决定系数分别为24，0.018 1和0.996 3；SIS-SPLS方法在当归数据集上选择的变量个数、RMSEP和决定系数分别为38，0.022 6和0.994 3。结果表明，该方法进一步提高了变量选择结果的稳健性和预测能力。Ensemble-SISPLS变量选择方法有效结合了SIS-SPLS较强的变量选择能力和集成学习良好的泛化能力，提高了变量选择的稳健性。此外，由于在子模型的预测能力和变量个数之间进行了折中，一定程度上减少了选择变量的个数，提高了模型的可解释性。     

高光谱成像的土壤剖面水分含量反演及制图

传统土壤水分的获取方法仅可获得离散的土壤水分点位数据，难以获得剖面上精细且连续的水分含量分布图。研究了野外条件下利用近红外高光谱（882～1 709 nm）成像反演剖面土壤水分含量（SMC），并实现精细制图的可行性。研究剖面位于江苏省东台市，我们利用近红外高光谱成像仪对剖面进行了5天原位连续观测，共采集了280个土样用于烘干法测定SMC。原始高光谱图像经数字量化值（DN）校正、黑白校正、拼接、几何校正、剪切和掩膜等一系列预处理后，提取各采样点的平均光谱反射率。提取光谱（Raw）经吸光度[LOG₁₀(1/R)]，Savitzky-Golay平滑（SG）、一阶微分（FD）、二阶微分（SD）、多元散射校正（MSC）和标准正态变量（SNV）转换后，采用偏最小二乘回归（PLSR）和最小二乘支持向量机（LS-SVM）方法建立SMC预测模型，并对比分析不同光谱预处理方法与建模方法组合条件下SMC的预测精度。结果表明，光谱反射率随SMC增加逐渐降低，不同光谱预处理方法的预测精度有所差异，除MSC方法外，同一光谱预处理方法的LS-SVM模型预测精度均高于PLSR模型，并且基于LOG₁₀(1/R)光谱的LS-SVM模型对SMC预测精度最高，其建模集的决定系数（R2_c）和均方根误差（RMSE_c）分别为0.96和0.65%，预测集的决定系数（R2_p）、均方根误差（RMSE_p）和相对分析误差（RPD_p）分别为0.88，1.05%和2.88。利用最优模型进行剖面SMC的高空间分辨率精细制图，通过比较SMC反演图中提取的预测值与实测值关系发现预测精度较高（R2: 0.85～0.95, RMSE: 0.94%～1.02%），且两者在剖面中的变化趋势基本一致，说明SMC反演图不仅能很好地反映出土壤水分在整个剖面中毫米级的含量分布信息，也可反映出同一位置处不同天数间的含量差异。因此，利用近红外高光谱成像结合优化的预测模型，能够实现土壤剖面SMC的定量预测及精细制图，有助于快速、有效监测田间剖面土壤水分状况。