Effectiveness of predicting tunneling-induced ground settlements using machine learning methods with small datasets

Prediction of tunneling-induced ground settlements is an essential task,particularly for tunneling in urban settings.Ground settlements should be limited within a tolerable threshold to avoid damages to aboveground structures.Machine learning(ML)methods are becoming popular in many fields,including tunneling and underground excavations,as a powerful learning and predicting technique.However,the available datasets collected from a tunneling project are usually small from the perspective of applying ML methods.Can ML algorithms effectively predict tunneling-induced ground settlements when the available datasets are small?In this study,seven ML methods are utilized to predict tunneling-induced ground settlement using 14 contributing factors measured before or during tunnel excavation.These methods include multiple linear regression(MLR),decision tree(DT),random forest(RF),gradient boosting(GB),support vector regression(SVR),back-propagation neural network(BPNN),and permutation importancebased BPNN(PI-BPNN)models.All methods except BPNN and PI-BPNN are shallow-structure ML methods.The effectiveness of these seven ML approaches on small datasets is evaluated using model accuracy and stability.The model accuracy is measured by the coefficient of determination(R2)of training and testing datasets,and the stability of a learning algorithm indicates robust predictive performance.Also,the quantile error(QE)criterion is introduced to assess model predictive performance considering underpredictions and overpredictions.Our study reveals that the RF algorithm outperforms all the other models with the highest model prediction accuracy(0.9)and stability(3.0210^(-27)).Deep-structure ML models do not perform well for small datasets with relatively low model accuracy(0.59)and stability(5.76).The PI-BPNN architecture is proposed and designed for small datasets,showing better performance than typical BPNN.Six important contributing factors of ground settlements are identified,including tunnel depth,the distance between tunnel face and surface monitoring points(DTM),weighted average soil compressibility modulus(ACM),grouting pressure,penetrating rate and thrust force.

Accurate machine learning models based on small dataset of energetic materials through spatial matrix featurization methods

A large database is desired for machine learning(ML) technology to make accurate predictions of materials physicochemical properties based on their molecular structure.When a large database is not available,the development of proper featurization method based on physicochemical nature of target proprieties can improve the predictive power of ML models with a smaller database.In this work,we show that two new featurization methods,volume occupation spatial matrix and heat contribution spatial matrix,can improve the accuracy in predicting energetic materials' crystal density(ρ_(crystal)) and solid phase enthalpy of formation(H_(f,solid)) using a database containing 451 energetic molecules.Their mean absolute errors are reduced from 0.048 g/cm~3 and 24.67 kcal/mol to 0.035 g/cm~3 and 9.66 kcal/mol,respectively.By leave-one-out-cross-validation,the newly developed ML models can be used to determine the performance of most kinds of energetic materials except cubanes.Our ML models are applied to predict ρ_(crystal) and H_(f,solid) of CHON-based molecules of the 150 million sized PubChem database,and screened out 56 candidates with competitive detonation performance and reasonable chemical structures.With further improvement in future,spatial matrices have the potential of becoming multifunctional ML simulation tools that could provide even better predictions in wider fields of materials science.

Online tool-wear measurement of small-diameter end mills based on machine vision

The objective of this study was to develop an online tool-wear-measurement scheme for small diameter end-mills based on machine vision to increase tool life and the production efficiency. The geometrical features of wear zone of each end mill were analyzed, and three tool wear criterions of small-diameter end mills were defined. With the uEye camera, macro lens and 3-axis micro milling machine, it was proved the feasibility of measuring flank wear with the milling tests on a 45# steel workpiece. The design of experiment （DOE） showed that Vc was the most remarkable effect factor for the flank wear of small-diameter end mill. The wear curve of the experiments of milling was very similar to the Taylor curve.

Small-time scale network traffic prediction based on a local support vector machine regression model

In this paper we apply the nonlinear time series analysis method to small-time scale traffic measurement data. The prediction-based method is used to determine the embedding dimension of the traffic data. Based on the reconstructed phase space, the local support vector machine prediction method is used to predict the traffic measurement data, and the BIC-based neighbouring point selection method is used to choose the number of the nearest neighbouring points for the local support vector machine regression model. The experimental results show that the local support vector machine prediction method whose neighbouring points are optimized can effectively predict the small-time scale traffic measurement data and can reproduce the statistical features of real traffic measurements.

Micro electrical discharge machining of small hole in TC4 alloy

Aiming at machining deeply small holes in TC4 alloy,a series of experiments were carried out on a self-developed multi-axis micro electrical discharge machining(micro-EDM)machine tool.To improve machining efficiency and decrease relative wear of electrode in machining deeply small hole in TC4 alloy,many factors in micro-EDM,such as polarity,electrical parameters and supplying ways of working fluid were studied.Experimental results show that positive polarity machining is far superior to negative polarity machining;it is more optimal when open-circuit voltage,pulse width and pulse interval are 130 V,5μs and 15μs respectively on the self developed multi-axis micro-EDM machine tool;when flushing method is applied in micro-EDM,the machining efficiency is higher and relative wear of electrode is smaller.

Performance Improvement of the LM Device and Its Application to Precise Measurement of Motion Trajectories within a Small Range with a Machining Centre

In order to apply the LM device previously developed to precisely measuring small motion trajectories located on the different motion planes, three major improvements are successfully performed under the condition of completely maintaining the advantages of the device. These improvements include 1) development of a novel connection mechanism to smoothly attach the device to the spindle of a machining centre;2) employment of a new data sampling method to achieve a high sampling frequency independent of the operating system of the control computer;and 3) proposal of a set-up method to conveniently install the device on the test machining centre with respect to different motion planes. Practical measurement experiment results with the improved device on a machining centre sufficiently demonstrate the effectiveness of the improvements and confirm several features including a very good response to small displacement close to the resolution of the device, high precision, repeatability and reliance. Moreover, based on the measurement results for a number of trajectories for a wide range of motion conditions, the error characteristics of small size motions are systematically discussed and the effect of the movement size and feed rate on the motion accuracy is verified for the machining centre tested.

Design and Performance Analysis of Small Denture Machining Equipment

The research and application on small denture machining equipment are great breakthrough for modern dental restoration technology. In this paper, a small denture machining equipment made of two spindles with four-axis was designed based on machining characteristics and functional analysis. Position accuracy and re-position accuracy were measured by accuracy instrument. In order to test its machining capacity, some typical microstcucture parts, such as straight channel, hemispherical surface, and molars coronal, were selected for high speed milling. It was obtained that the denture machining equipment met the machining requirements with high quality and efficiency, according to the acquisition and analysis of form and position errors, surface roughness, and 3-D profile.

Structural Reliability Analysis Based on Support Vector Machine and Dual Neural Network Direct Integration Method

Aiming at the reliability analysis of small sample data or implicit structural function,a novel structural reliability analysis model based on support vector machine(SVM)and neural network direct integration method(DNN)is proposed.Firstly,SVM with good small sample learning ability is used to train small sample data,fit structural performance functions and establish regular integration regions.Secondly,DNN is approximated the integral function to achieve multiple integration in the integration region.Finally,structural reliability was obtained by DNN.Numerical examples are investigated to demonstrate the effectiveness of the present method,which provides a feasible way for the structural reliability analysis.

Applications of Localized Phase Compensation Method to Design a Stabilizer in a Multi-machine Power System

为了演示和验证稳定器设计的就地相位补偿法在多机电力系统中的应用,介绍在多机电力系统中,就地补偿设计稳定器的2个应用实例。第1个实例是在多机电力系统中就地补偿设计电力系统稳定器（power system stabilizer,PSS）,阻尼电力系统局部模振荡。第2个实例是就地补偿设计附加在静态同步补偿器（static synchronous compensator,STATCOM）上的稳定器,抑制多机电力系统中的区域模振荡,并给出在一个16机电力系统中的应用计算和仿真结果。

Analysis and correction of the machining errors of small plastic helical gears by ball-end milling

Many small-size precise plastic helical involutes gears are used in electrical appliances to transmit rotary movements con- tinuously and smoothly. Ball-end milling is an effective method for trial manufacture or small batch production of this type of gear, but the precision of the gear is usually low. In this research, the main sources of the errors of the gear, machining errors of the tooth profile and trace of the gear obtained were analyzed. The correction amounts for these errors are then determined by using a CNC gear tester. They are used to generate a new 3D-CAD model for gear machining with better nrecision.

基于机器学习构建肺腺癌骨转移自动化模型

目的采用机器学习算法对关键变量进行识别,并对肺腺癌(LUAD)患者骨转移风险进行预测。方法回顾性分析2019年1月至2022年6月淮南东方医院集团附属肿瘤医院收治的132例确诊非小细胞肺癌(NSCLC)患者的临床资料,包括是否发生骨转移、年龄、性别、病理类型、吸烟状况、T分期、N分期、骨转移前是否有其他部位的转移,以及癌胚抗原(CEA)、碱性磷酸酶(ALP)、鳞状细胞癌抗原(SCCA)、糖类抗原125(CA125)、细胞角蛋白19片段抗原21-1(CYFRA21-1)、神经元特异性烯醇化酶(NSE)、钙(CA)水平。采用LASSO回归分析方法来筛选与骨转移相关的关键特征,并将其用于构建6种机器学习模型,另收集63例NSCLC患者的临床数据用于模型的外部验证。不同模型的预测性能通过受试者工作特征曲线(ROC曲线)来评估。校准曲线和DCA曲线用于验证所建模型的准确性和获益能力。使用SHAP包对logistic模型进行解释。结果LASSO回归分析最终筛选了4个重要变量,包括性别、N分期、CEA水平和糖类抗原CA125水平。在6种机器学习模型中,logistic模型在训练集(AUC=0.710)、测试集(AUC=0.705)和外部验证集(AUC=0.655)均具有最佳的预测效能和稳定性。SHAP图显示在logistic模型中4个变量的权重从高到低依次为CEA、性别、T分期和CA125。成功构建了LUAD骨转移的机器学习模型和网页计算器。结论logistic预测模型可以识别LUAD骨转移高风险患者,这有助于临床医生指导高危患者做出适当预防措施。

基于多元统计分析的小样本数据预测模型设计

若小样本数据预测误差较大,会直接影响数据应用效果,为提升小样本数据预测精度,提出基于多元统计分析的小样本数据预测模型设计方法。将小样本数据放入SPSS软件中,结合自助法完成小样本数据的经验分布分析。基于样本数据经验分布特征,结合具备学习能力的Fisherface算法对小样本上数据实施预分类,建立测试样本类别标签,实现小样本数据的特征提取。通过多元统计分析数据特征的主元成分,确定模型回归函数,结合支持向量机构建数据预测模型,通过上述模型完成小样本数据的精准预测。实验结果表明,使用上述方法开展小样本数据预测时,预测误差较低,效率较高,说明其预测效果较好。

基于机器学习的寒区渠道冰情的遥感监测方法

寒区渠道冬季运行时常出现冰情,控制平封的封冻过程会大幅降低渠道输水能力,调控不当甚至可能产生冰塞、冰坝等灾害。国内外开展了大量渠道冰情研究,以期提升渠道冰期输水能力,但受限于观测资料的时空密度,数值模拟结果难以验证,调度决策缺少依据。遥感技术因其具有监测范围大、时效性高的特性,在渠道冰情监测中具有较大的应用潜力。为探索适用于寒区渠道冰情遥感监测的方法,该研究以南水北调中线京石段明渠段为研究区,基于Sentinel-2影像的11个波段反射率构建了完全特征、优选特征和组合特征3类特征空间数据集,作为支持向量机(support vector machine,SVM)、最大似然估计(maximum likelihood estimation,MLE)、随机森林(random forest,RF)分类算法输入,训练得到了9个地物分类器,用于渠道结冰范围识别,并采用北拒马闸前影像渠道结冰范围提取试验,对比不同分类算法和输入特征组合下的分类性能。结果表明:在渠道结冰范围识别中,近红外、可见光和短波红外是关键波段。在样本数量有限的条件下,SVM算法结冰范围识别精度最高,不同特征输入下制图精度(producer’s accuracy,PA)可达85.10%~87.91%,错分误差(commission error,CE)为10.84%~16.08%;RF算法在完全特征和优选特征输入下分类精度与SVM接近,PA为84.67%~86.61%,CE为13.76%~14.41%,但其在组合特征下分类结果严重偏离实际;MLE算法在3类特征下的分类精度均较低,不适宜作为渠道结冰范围识别算法。综合来看,SVM算法对特征空间敏感性较低,在不同的特征输入下均能实现渠道结冰范围的高精度提取;RF算法对特征空间敏感性较高,当输入特征发生变化时,结冰范围识别精度不稳定。最后以完全特征下的SVM算法为例,进行了分类器的时空泛化性验证,结果表明模型在不同时间、不同渠段下,制图精度不低于82.09%,错分误差不高于13.82%,分类模型精度均较好,能有效识别渠道结冰范围。该研究方法可为寒区输水工程冰情监测提供新思路,亦可为类似工作提供参考。

混合增强型机器学习算法在稀土供应链金融中评价中小企业信用风险的研究

稀土是支撑高端技术创新和新能源产业发展的关键原材料之一,研究解决稀土供应链中小企业融资困难的问题,做强我国稀土产业链,更好地维护国家战略利益是当务之急。供应链金融作为创新型融资方式成为实现中小企业融资授信的一种主要手段,但其中信用风险问题成为融资决策中需解决的最关键问题之一。本文提出了一种混合增强型机器学习算法,首先采用动态透镜成像反向学习改进的海洋捕食者算法(IMPA)对支持向量机算法(SVM)进行优化,再采用AdaBoost算法对优化后的SVM进行集成,建立AdaBoost-IMPA-SVM模型。采用该模型对供应链金融风险进行评价,重新建立供应链金融风险体系指标,通过相关性分析进行特效选取,并从计算机通信及其他制造业选取52家中国上市中小企业2019—2021年期间140个样本作为特征变量输入模型。仿真实验结果验证了该模型相较于其他信用风险评价模型具有更好的分类识别性能。

小转弯曲线隧道TBM选型与掘进姿态调控方法

小转弯隧道施工时全断面隧道掘进机(full-section tunnel boring machine,TBM)的选型设计和掘进姿态控制具有特殊性,目前还缺乏通用的理论依据和指导方法。针对此问题,首先,对传统类型TBM小转弯选型进行研究,结合已有项目数据和几何模拟,确定传统类型TBM能适应的最小转弯半径。然后,对双盾敞开式TBM的推进系统和导向系统进行针对性设计,通过分析双盾敞开式TBM推进系统结构和实际施工,提出转弯时双盾敞开式TBM推进油缸内外侧行程差值的理论计算方法和施工过程中的姿态调控方法。最后,得出如下结论:1)当隧道转弯半径小于200 m时,敞开式TBM适应难度较大,需要采用双盾敞开式TBM;2)结合抚宁抽水蓄能电站项目实际施工情况,提出的双盾敞开式TBM的理论计算方法和姿态调控方法确保了转弯段隧道的轴线偏差在要求范围内。

基于机器学习的水泥基灌浆料强度预测方法

针对采用小直径芯样法准确预测水泥基灌浆料抗压强度的问题,使用压力试验机分别对水泥基灌浆料标准尺寸试块和小直径芯样进行抗压强度试验,并基于试验数据,采用支持向量机回归(SVR)和随机森林回归(RFR)对水泥基灌浆料抗压强度进行回归预测。结果表明:标准尺寸试块均呈正反相接的四角锥体破坏形态,而高径比为0.7和1.0的小直径芯样呈正反相接的圆锥体破坏形态,高径比为1.2的小直径芯样呈斜裂缝剪切破坏形态;标准尺寸试块和小直径芯样的抗压强度值均服从正态分布,且无离群值;随着龄期的增长,标准尺寸试块和小直径芯样的抗压强度提高,且具有早期强度较高的特性;直径46 mm芯样的抗压强度较小,且更易受加工精度的影响;在给定的龄期和直径下,高径比为0.7的芯样抗压强度值最大,抗压强度离散程度最小;RFR预测模型对水泥基灌浆料抗压强度的预测效果更优。所提方法可较准确预测水泥基灌浆料抗压强度,为水泥基灌浆料抗压强度的预测研究提供了参考。

不均衡小样本下多特征优化选择的生命体触电故障识别方法

针对现有的剩余电流保护装置无法有效识别触电事故的问题,该文提出了一种不均衡小样本下多特征优化选择的生命体触电故障识别方法。首先通过变分自编码器(VAE)对实验收集到的生命体触电小样本数据进行增殖以实现正负样本均衡;然后在时域上提取能够反映波形动态变化特性的23个特征量,并利用高斯核Fisher判别分析(GKFDA)与最大信息系数(MIC)法从中选择最优表达特征组;最后,提出基于遗忘因子的在线顺序极限学习机(FOS-ELM)算法实现生命体触电行为的鉴别。实验结果表明,所提方法利用不均衡小样本触电数据集就可以训练出一个优秀的分类模型,诊断准确率可达98.75%,诊断时间仅为1.33 ms。其优良的性能结合在线增量式学习分类器设计,使得模型具备新知识学习能力,具有极好的工程应用前景。

端云协同智能计算的关键问题、方法和应用

端云协同智能计算是大数据、云计算、边缘计算发展的产物,可在保护用户隐私的前提下显著提升数据利用率,实现智能计算实时响应能力与服务鲁棒性的优势互补,而相应技术研发和实践应用具有复杂性。本文剖析了端云协同智能计算的应用价值,凝练了端学习效率优化、端少样本过拟合、端模型定制化、分布差异下虚假关联学习、通信开销与计算效率平衡等方面的技术难题;系统梳理了端云协同智能计算中主流方法研究进展,涉及作为应用基石的高效计算硬件、以端为中心的协同计算、以云为中心的协同计算、端云双向协同计算、可信端云协同智能计算等主要方向;总结了推荐系统、自动驾驶、安防系统、教育模式等端云协同智能计算的垂直领域应用情况。着眼端云协同智能计算的未来发展,需重点研究云资源在端模型个性化中的应用策略、端云协同多目标优化算法、端-端与云协同计算的优化策略。

基于深度学习的自适应苹果图像多缺陷检测

苹果图像检测中,对较小缺陷识别与分类时,检测时间与精度的平衡问题一直是该领域的研究重点。为实现苹果缺陷的快速高精度检测,提出了一种基于深度学习的ACE-YOLO自适应局部图像检测算法,通过深度学习缩小检测区域,利用通道注意力机制把计算机算力集中到局部检测范围以缩短检测时间,采用图像增强算法使检测细节更清晰,通过在模型中增加小目标检测层来提高检测精度。该算法利用深度学习实现局部细节检测,与常规算法相比增加了注意力机制,其检测速度提高了25%;由于引入了局部图像增强算法,并增加了小目标检测层,其在对14类苹果缺陷进行识别时,平均检测精度也由86.1%提高到95.2%。实验表明,该算法更适用于苹果缺陷的检测。

YB45包装机商标纸内舌微量弯曲装置的设计

目的解决YB45硬盒包装机在生产过程中出现的商标纸内舌插盖导致小盒盖打不开的问题。方法根据YB45包装机自身的结构特点,对商标纸内舌涂胶装置及折叠装置进行改进,设计一种小盒商标纸内舌微量弯曲装置。此装置可以对商标纸内舌部位进行微量弯曲,内舌弯曲后的小盒商标纸在折叠过程中能有效防止烟包小盒插盖。结果使用商标纸内舌微量弯曲装置后,小盒插盖的数量由改进前的5.3盒/月降至0盒/月;机车效率由改进前的96.3%增至96.9%;质量得分由改进前的97.9分增长为98.5分。结论商标纸内舌微量弯曲装置能有效解决YB45硬盒包装机生产过程中出现的小盒插盖问题,有效保障了YB45包装机小盒包装的产品质量,同时也提升了机车效率。