基于POA-BP的TP2管材自由弯曲成形结果预测

将壁厚减薄率和椭圆率作为管材自由弯曲成形结果的评价指标,选取弯曲模与管材间隙值、弯曲模圆角半径值、管材弯曲变形区长度、导向机构圆角半径值、导向机构与管材间隙值作为影响因子。利用数值模拟方法对管材自由弯曲成形结果的评价指标和影响因子建立样本库,并随机选取6组作为测试样本,其余的作为训练样本,结合BP神经网络和鹈鹕优化算法对预测模型进行训练,构建POA-BP神经网络预测模型对管材自由弯曲成形结果进行预测。结果表明,POA-BP预测模型的壁厚减薄率和椭圆率的最大预测误差不超过2%,故POA-BP预测模型能够有效预测管材成形结果。     

灰色关联分析的神经网络模型在轴承故障预测中的应用

为提高小样本轴承故障预测的精度,提出了灰色关联分析的神经网络预测模型。利用神经网络训练样本数据,并使用灰色关联分析不断调整其隐含层节点数,以寻找到最优的数据解,完成训练过程;基于已训练好的模型对未来时间点的运行状态进行分析,并根据设备的理论诊断标准实现故障预测。将提出的模型与同样可预测小样本波动数据的灰色马尔科夫预测模型进行实例应用对比,结果显示,灰色关联分析的神经网络预测模型效果明显优于灰色马尔科夫模型。     

切削表面粗糙度的人工神经网络预测

以易切削黄铜的加工表面粗糙度与各种加工参数的关系为对象,将L9( 34)型正交切削试验数据作为训练学习样本,同时以与正交试验参数有关的6个样本作为预测样本,用BP神经网络对其进行了预测。结果表明:经设计的BP神经网络训练1183次,其最大误差不超过5 % ;人工神经网络与正交试验相结合,能大大节省预测时间和费用,效果很好。     

基于神经网络的涡扇发动机中间稳态参数模型

从某型涡扇发动机飞行试验数据出发,基于三层前向神经网络,通过比较训练数据集、神经网络结构以及训练算法对输出结果的影响,确定了该型发动机中间稳态参数预测模型。利用非样本点飞行试验数据模型进行了检验,结果表明,通过神经网络建立的涡扇发动机中间稳态参数预测模型具有可靠的预测精度及模型推广能力,该方法对建立涡扇发动机健康监测模型具有一定的参考作用。     

基于BP神经网络的摩擦材料滑动摩擦性能预测

为了预测不同测试条件下,湿式扭矩限制器中摩擦材料的滑动摩擦性能,建立试验条件与摩擦性能之间的神经网络预测模型。选用BP神经网络进行建模,以不同测试压力和转速为输入变量,相应测试条件下的摩擦扭矩和扭矩稳定性为输出变量,通过一系列参数对比,确定最佳神经网络结构为单隐含层神经元个数为10,并采用L-M算法。采用L-M算法对网络进行训练,利用建立的神经网络对材料的滑动摩擦性能做进一步预测。结果表明:压力对扭矩及其稳定性的影响大于转速;无论是扭矩还是其稳定性,预测结果与实验结果具有相同的变化趋势,最大相对误差小于6%,所建立的神经网络预测模型可以用于滑动摩擦条件下对摩擦材料滑动摩擦性能的预测。     

基于BP神经网络的弯管机回弹量预测

建筑钢结构不同管材的弯曲曲率变化较大,施工时需制作大量胎架,即费时费料又占用场地,故设计了一种冷弯机结构,基于该冷弯机,利用Midas7.8软件对建筑钢结构用管材进行冷弯成形模拟,基于神经网络建立3个输入参数的回弹量数据模型,选择1 000组模拟数据作为训练数据训练神经网络,500组模拟数据作为测试数据测试网络,将预测结果和样本结果进行比较和分析,结果表明,所建立的神经网络预测模型满足误差要求,可以用来预测大管径厚管壁管材冷弯成形后的回弹量。该项研究为开发具有自适应回弹量补偿性能的数控弯管系统提供理论基础。     

基于神经网络的激光熔覆高度预测

激光成形过程中,对熔覆高度进行实时检测,从而实现熔覆高度闭环控制是成形高质量零件的保证.激光成形过程是一个多参数耦合的非线性过程,大量激光参数对成形熔覆表面质量具有重要影响.在分析激光参数对熔覆高度影响的基础上,建立利用激光工艺参数预测熔覆高度的误差反向传播(Back propagation,BP)神经网络模型,完成了网络算法设计.通过激光成形试验采集样本,利用训练样本对所建立的网络进行训练,完成网络输入输出高度映射关系,并利用测试样本对所训练的网络进行检验.仿真试验表明,神经网络熔覆高度预测模型具有很高的精度,验证了该预测模型在理论和实践上的可行性与有效性.神经网络熔覆高度预测模型为实现激光加工过程熔覆高度实时预测与闭环控制打下基础,对提高成形产品质量具有重要意义.     

考虑电控数据的LSTM神经网络进给轴热误差建模

针对进给轴热误差建模中忽略电控数据和时间序列影响的问题,提出一种考虑温度变化与电控数据的长短期记忆(Long-Short Term Memory,LSTM)神经网络热误差预测模型.以三轴立式加工中心为试验对象,首先对进给轴进行热变形分析,再以温度变化、电控数据为输入样本,建立了LSTM神经网络热误差预测模型,随后通过与仅考虑温度变化的LSTM神经网络,以及同时考虑温度变化与电控数据的BP神经网络进行对比分析,试验论证表明,对数控机床进给轴进行热误差建模时,在考虑温度变化的基础上,进一步考虑电控数据可以提高模型的预测精度和鲁棒性,且在同样输入条件下,LSTM神经网络热误差预测模型相较于BP神经网络有更好的预测精度和鲁棒性.     

考虑电控数据的LSTM神经网络进给轴热误差建模

针对进给轴热误差建模中忽略电控数据和时间序列影响的问题,提出一种考虑温度变化与电控数据的长短期记忆(Long-Short Term Memory,LSTM)神经网络热误差预测模型.以三轴立式加工中心为试验对象,首先对进给轴进行热变形分析,再以温度变化、电控数据为输入样本,建立了LSTM神经网络热误差预测模型,随后通过与仅考虑温度变化的LSTM神经网络,以及同时考虑温度变化与电控数据的BP神经网络进行对比分析,试验论证表明,对数控机床进给轴进行热误差建模时,在考虑温度变化的基础上,进一步考虑电控数据可以提高模型的预测精度和鲁棒性,且在同样输入条件下,LSTM神经网络热误差预测模型相较于BP神经网络有更好的预测精度和鲁棒性.     

基于遗传算法优化神经网络的光伏发电预测

针对光伏发电系统输出电能随机性对电力系统的影响问题,对光伏输出功率预测方法进行了研究,提出了基于遗传算法优化BP神经网络的光伏预测模型。对光伏发电系统的历史发电量数据和气象数据进行了数学归纳,分析了天气类型、温度、太阳辐射强度等因素对光伏发电功率的影响。建立了基于遗传算法优化的神经网络光伏发电预测模型,提前一天进行了功率预测,且通过动态修正进一步提高了模型预测精度。运用Matlab为神经网络与遗传算法工具箱,对训练好的模型在不同日类型下进行了测试和评估,并与传统神经网络预测结果进行了对比分析。研究结果表明,利用遗传算法优化方法提高了神经网络模型预测光伏输出功率的精度,预测结果与实测结果之间的平均误差百分比减小,预测方法具有工程应用意义。     

应用神经网络研究煤气鼓风机振动趋势预报

设计煤气鼓风机的振动状态监测方案,介绍神经网络的预测原理,选择4-4-1神经网络拓扑结构,采用改进共轭梯度算法进行网络训练,建立振动趋势预报模型,并运用于鼓风机的振动趋势预报。实验结果表明,基于神经网络的预报技术能准确、有效地预报机组的振动趋势。     

基于神经模糊控制理论的数控机床热误差建模

将基于神经模糊控制理论的建模方法--模糊神经网络建模法应用到数控机床热误差建模当中,讨论了热误差模糊神经网络的结构及建模原理;对大型数控龙门导轨磨床主轴箱系统进行建模试验,采用非接触式红外温度测量仪和千分表分别测量主轴箱系统温度值与主轴热误差,得到两组独立的试验数据,一组用来建立主轴箱系统热误差模糊神经网络预报模型,另一组用来对模型进行验证。试验结果表明,模糊神经网络模型预测精度高,泛化能力强;将模糊神经网络建模方法与径向基函数神经网络建模方法进行综合对比,分析结果表明,模糊神经网络建模方法具有更好的建模效率、建模鲁棒性及预测性能。     

基于改进灰色神经网络模型的板料成形缺陷预测研究

为了准确预测和减少板料成形过程中可能出现的缺陷,提出了一种改进的灰色神经网络预测模型。该模型利用BP神经网络辅助灰色预测模型进行预测。其中,灰色模型进行粗预测,神经网络模型修正其误差,再通过寻找最佳权值以优化灰色模型中微分所对应的背景值,进而得到精度更高的灰色神经网络模型。以国际著名板料成形数值模拟会议NUMISHEET'93的方盒件拉深为例,运用改进的灰色神经网络模型,预测其拉裂和起皱。结果表明,改进的灰色神经网络模型具有很高的预测精度,相比于未改进的灰色神经网络模型,预测结果更加准确和稳定。     

A method for predicting hobbing tool wear based on CNC real-time monitoring data and deep learning

Intelligent monitoring and diagnosis of tool status are of great significance for improving the manufacturing efficiency and accuracy of the workpiece. It is difficult to quickly and accurately predict the wear state of worm gear hob under different working conditions. This paper proposes a novel approach to predict hob wear status based on CNC real-time monitoring data. Based on the open platform communication unified architecture (OPC UA) technology and orthogonal test, the machine data of motor power, current, etc. related to tool wear are collected online in the worm gear machining process. And then, an improved deep belief network (DBN) is used to generate a tool wear model by training data. A growing DBN with transfer learning is introduced to automatically decide its best model structure, which can accelerate its learning process, improve training efficiency and model performance. The experiment results show that the proposed method can effectively predict hob wear status under multi-cutting conditions. To show the advantages of the proposed approach, the performance of the DBN is compared with the traditional back propagation neural network (BP) method in terms of the mean-squared error (MSE). The compared results show that this tool wear prediction method has better prediction accuracy than the traditional BP method during worm gear hobbing.     

一种高速数控机床电主轴表面温度智能预测方法

电主轴是高速数控机床核心功能部件,电主轴损坏基本是电主轴发热引起的.电主轴温度场具有复杂的非线性特征,神经网络在处理非线性系统温度预测方面得到了广泛的研究,神经网络与传统模型相比具有更好的适时预报性和持久性.论文利用遗传算法优化BP神经网络建立电主轴表面温度预测模型.预测结果表明,未优化的BP神经网络与遗传神经网络预测误差相对比,遗传神经网络对电主轴表面温度预测具有更高的预测精度和稳定性.     

Measurement and compensation for volumetric positioning errors of CNC machine tools considering thermal effect

The measurement and compensation of volumetric positioning errors can be used to significantly improve the accuracy of machine tools. In this paper, a sequential step diagonal measurement is introduced to measure nine volumetric positioning errors in a short time. Measurements under various thermal conditions are preformed to understand the relationship between the volumetric positioning errors and the machine temperature field and variations. A radial basis function neural network is used to predict the volumetric positioning errors at all positions based on the temperature distribution of the machine. Compensation experiment is carried out to validate the performance of the measurement and the prediction method. The experimental results show that the volumetric accuracy of the machine tool is significantly improved by the error compensation.     

基于ACO-BP神经网络的光伏系统发电功率预测

为准确预测光伏发电量,减少并网光伏对大电网的影响,引入相似日概念,对夏季预测日的平均温度、最高温度、最低温度以及天气类型进行分析。在历史数据中选取具有相似天气特征的发电功率数据和天气数据作为神经网络的训练样本,建立ACO-BP神经网络光伏发电功率预测模型,并将预测结果与传统BP神经网络和PSO-BP神经网络预测结果相比较。实验结果表明,该模型具有较高的预测精度。     

Prediction of the machined surface quality of ball-end milling of H13 die steel using MLBP method

Abstract

Based on the characteristics of the surface quality prediction system of high-speed milling, the prediction model is used to predict the surface quality of analyzing the advantages of the two methods of using the multilinear and BP neural network model (MLBP) method. This article through the in-depth study of the surface quality, study the surface quality prediction based on the characteristics of multiinput multioutput nonlinear systems, respectively, established a linear regression equation, BP neural network model, and the surface quality of specific conditions to start prediction. The prediction results show that these prediction methods can play a special role as certain conditions. However, owing to the limitations of multiple linear regression and BP neural networks, their generalization ability and robustness cannot meet actual needs. Drawing on the idea of interpolation, and analyzing the advantages and disadvantages of linear regression and BP neural network to solve nonlinear problems, a new prediction method is developed. The main idea are to use interpolation method to insert preprediction under the premise of linear prediction; to process the values and obtain a unified prediction result from linear regression; to combine the experimental results from the pretreatment results; to use these input information as the input content of the BP neural network; to establish a training model based on the BP neural network model self-learning process. This training model predicts the quality of the machined surface. This method is abbreviated as the MLBP method. The experimental results and comparison of model prediction results show that this method can effectively improve the generalization ability and robustness of the prediction model, and further improve the model’s prediction accuracy.     

Study of sample temperature compensation in the measurement of soil moisture content

Since the near-infrared (NIR) spectrum is susceptible to sample temperature fluctuations, we investigate the influence of sample temperature on the predictive power of calibration model for soil moisture content (MC) and propose the multi-source information fusion technology based on back propagation neural network (BPNN) to compensate for sample temperature effect. With the discrete wavelet transform (DWT) as the pre-processing method and the least squares support vector machine (LS-SVM) regression as the modeling method, a model at 20 °C to predict MC of the soil samples at other temperatures was established. The results show that except for 20 °C, the root mean square error of prediction (RMSEP) are large. We analyze the predicted results with the dual-factor analysis of variance without duplication and the result shows that the effect of sample temperature on the prediction model for soil MC is significant. A temperature compensation model was then established with combining of soil MC and sample temperature based on BPNN. The predicted results showed that the prediction precision of the model was improved significantly.     

Formation of iron aluminide coatings on plain carbon steel by TIG process

Micro end-milling is widely used in many industries to produce micro products with complex 3D shapes. The accurate modeling and prediction of surface roughness are important for evaluating the productivity of the machine tools and the surface quality of the machined parts. This paper presents an accurate surface roughness model based on the kinematics of cutting process and tool geometry by considering the effects of tool run-out and minimum chip thickness. The proposed surface roughness model is validated by micro end-milling experiments with the miniaturized machine tool. The results show that the proposed surface roughness model can accurately predict both the trends and magnitude of the surface roughness in micro end-milling.