基于CPSO-BP的刀具寿命预测算法

为了改善刀具寿命预测的精准度,文章在已有的PSO-BP神经网络算法中引入混沌理论,提出了一种基于混沌粒子群算法优化BP神经网络(CPSO-BP神经网络)的刀具寿命预测方法。该方法采用粒子群算法优化网络权值和阈值,通过混沌扰动更新粒子的位置。CPSO-BP神经网络算法既避免了BP神经网络存在的收敛速度慢、易陷入局部最优的缺点,又改善了全局搜索能力,同时,降低了粒子群优化算法造成早熟收敛或停滞的可能性。仿真结果表明:与已有的PSO-BP神经网络算法相比,该文的CPSO-BP神经网络算法用于刀具寿命预测时收敛速度和预测精度均更胜一筹。     

BP神经网络在立铣刀结构参数优化中的应用

钛合金薄壁件的铣削加工过程中,刀具磨损速度快,并且工件容易变形,其主要因素是加工过程中切削力大,切削温度高。文章利用有限元仿真软件Advant Edge FEM铣削仿真数据,建立整体式立铣刀结构参数与切削力和切削温度的BP神经网络预测模型,并对切削预测模型进行了切削实验验证。在此基础上,利用BP神经网络模型的预测结果对整体式立铣刀的结构参数进行了优化,切削实验证明,优化后的刀具参数可以有效地降低切削力和切削温度,从而有效地改善过程中刀具的切削性能和工件的加工质量。     

超声滚挤压轴承套圈表面残余应力预测模型建立

为实现对超声滚挤压轴承套圈表面残余应力的合理预测,以轴承套圈为研究对象,进行正交试验设计,分别建立了残余应力与各加工参数之间的传统BP神经网络、改进BP神经网络和遗传算法优化BP神经网络(GA-BP)数学预测模型。通过对三种预测模型进行对比分析,结果表明,轴承套圈表面残余应力预测模型预测精度由高到低依次是GA-BP、改进BP神经网络预测模型、传统BP神经网络预测模型;且GA-BP所建立的轴承套圈表面残余应力的预测模型的平均误差控制在2.58%,因此该预测模型可以进行不同参数下电轴承套圈表面残余应力的预测。     

基于进化神经网络的磁粒研磨表面粗糙度预测方法研究

目的 为实现磁粒研磨光整加工的表面粗糙度精准预测,提出一种遗传算法优化表面粗糙度BP神经网络的预测模型.方法 将表面粗糙度作为预测的目标,影响磁粒研磨5052铝合金管内表面粗糙度的五个主要工艺参数作为神经网络的输入.合理设计正交试验,得到不同工艺参数配置下的表面粗糙度值,将其作为神经网络的输出.通过建立非线性预测模型,对比遗传算法优化后和传统BP神经网络的均方差与仿真时间,分析优化前后表面粗糙度的预测效果.结果 通过试验数据建立了结构为5-11-1的神经网络,进化BP神经网络预测模型均方差为0.044,建模仿真时间为0.187 s,其平均相对误差率为13.2％.传统的BP神经网络预测模型均方差为0.231,建模仿真时间为1.840 s.结论 通过遗传算法优化后的BP神经网络均方差更小,建模仿真时间更短,进化BP神经网络可以实现更为精准的预测,同时能够极大地避免传统BP神经网络易陷入局部极小值的弊端.     

基于BAS-BP神经网络的钻削力预测

针对传统预测深孔加工中钻削力精度不高的问题以及BP神经网络本身存在的缺陷,提出了BAS-BP神经网络预测模型。文章基于天牛须算法与BP神经网络相互结合,利用天牛须算法计算优化BP神经网络中的初始权值与阀值,从而建立BAS-BP神经网络的预测模型。并与传统BP神经网络预测模型进行对比。结果表明BAS-BP神经网络克服了训练时间长、收敛速度慢的缺点,预测精度明显提高。     

基于神经网络的喷丸25CrMo合金疲劳寿命及残余应力松弛行为预测研究

首先,本文采用BP神经网络建立了喷丸25CrMo车轴钢疲劳寿命预测模型。然后,在此基础上采用遗传算法（GA）对BP神经网络的预测精度进行了优化。此外,还采用了径向基神经网络（RBF）进行建模分析,并与以上两种模型的预测结果进行对比,结果表明：遗传算法优化的BP神经网络（GA-BP）相比于BP和RBF神经网络具有更高的预测精度,其中训练集和测试集的平均预测精度分别为91.5%和85.4%。然后,基于GA-BP神经网络模型的连接权值矩阵和Garson方程进行了灵敏度分析,从而进一步量化了输入影响因素对喷丸25CrMo车轴钢疲劳寿命的相对影响比重;最后,还采用GA-BP神经网络预测了喷丸25CrMo车轴钢表面残余压应力的松弛行为,结果表明：测试集的平均预测误差仅为3.4%,表明了该神经网络预测性能良好。综上所述,本文采用神经网络建模分析了喷丸25CrMo车轴钢的疲劳性能和残余压应力松弛行为,显著降低了传统疲劳试验所需的成本,并且还保证了较高的准确性。     

基于灰色理论和神经网络的轧制力预测

针对热轧带钢轧制力预测精度不准确问题,建立了灰色轧制力预测模型,通过对比灰色轧制力预测模型和BP神经网络预测模型的优缺点,进一步提出将灰色理论和BP神经网络组合应用到热轧带钢轧制力预测中,并分析比较了轧制力预测模型的相对误差。同时,在相同的轧制条件下,对灰色神经网络轧制力预测模型的预测值和在线平整轧制的实测轧制力值作了比较,所得轧制力预测误差小于±5%。由此可以看出,灰色理论和BP神经网络组合应用的方法能够较准确地实现平整轧制力的预测。     

基于BP人工神经网络的钢轨交流闪光焊焊接接头质量预测

对刘国东等提出的BP(误差反向传播)神经网络归一化模型进行了改进,得到了适合钢轨交流闪光焊落锤质量预测的BP神经网络归一化模型。基于LabView开发软件编制了高速采集软件。采集了U71Mn钢轨焊接工艺正交试验的焊接电流、焊接电压和动立柱的位移,并从中提取加速烧化前一阶段的闪光率、能量输入、焊接时间和烧化量等质量特征量作为BP神经网络预测模型的输入量。建立了输入层单元数为5、隐含层单元数为14的BP神经网络焊接接头落锤质量的预测模型;以正交设计工艺试验的27个焊接接头中的17个作为训练样本,对预测模型进行训练。以余下的lO个作为检验样本,采用将训练后的预测模型进行预测,预测准确率达到90％。     

基于BP神经网络的钢轨闪光对焊接头灰斑面积预测

针对钢轨闪光对焊的特点,根据GAAS80/580焊机记录的压力、电流和动端位移随时间而变化的曲线,从中提取了10个主要影响接头灰斑面积的特征参数作为BP神经网络预测模型的输入量,建立了钢轨闪光对焊接头的灰斑面积预测模型.采用粒子群算法优化了BP神经网络的权值和阈值,并利用优化后的BP网络模型对接头灰斑面积进行了预测.结果表明,提取的特征参数能较好地反映焊接接头灰斑情况,粒子群算法优化的BP神经网络预测模型能较准确地预测出焊接接头灰斑面积.     

基于BP神经网络的表面硬度预测模型

运用MATLAB中BP神经网络强大的数据预测功能,并结合VB编写的用户操作界面,建立了基于BP神经网络的表面硬度预测模型.对该模型进行训练和仿真后,其所得到的结果与实际测量值相差较小,能够达到模型预测的要求,有较好的工程实际意义.     

铝合金切削过程刀具磨损预测研究

为更精确地研究刀具磨损,建立刀具磨损模型至关重要。目前刀具磨损的模型主要是经典的刀具磨损模型和刀具磨损预测模型,刀具磨损预测模型主要为人工神经网络、隐马尔可夫模型和支持向量机模型。分析铝合金切削过程中的刀具磨损机制,总结经典的刀具磨损模型,梳理刀具磨损预测模型。铝合金切削过程中刀具主要的磨损机制为黏着磨损、扩散磨损和磨粒磨损。结果表明：在黏着磨损和磨粒磨损的基础上考虑扩散磨损的刀具磨损理论模型最接近实际加工。     

Simulation based model for tool life prediction in bevel gear cutting

Due to unpredictable tool life behavior in bevel gear cutting, unexpected production stops for tool changes occur. These lead to additional manufacturing costs. Because of its complexity, it is currently not possible to analyze the bevel gear cutting process sufficiently regarding tool life. This restriction leads to an iterative process design and determination of the ideal process parameters by using a trial-and-error approach. As a matter of fact, there is no concept to predict tool life in bevel gear cutting. Thus, a project has been initiated in order to develop a tool life model based on cutting simulation. This report presents the tool life model which combines a manufacturing simulation for bevel gear cutting with a regression model. The data of the regression model are determined by analogy trials. The combination of manufacturing simulation and regression model allows a local tool life prediction along the cutting edge.     

Modelling and prediction of cut shape for plasma arc cutting based on artificial neural network

Abstract

An artificial neural network approach for the modelling of plasma arc cutting processes is introduced. Neural network models have been proposed for predicting the cut shape and estimating the special cutting variables. The implementation of artificial neural networks in the modelling of cutting processes is discussed in detail. The performance of the neural networks in modelling is presented and evaluated using actual cutting data. Moreover, prediction applications of the above neural network models are described for various cutting conditions. It is shown that estimated results based on the proposed models agree well with experimental data; the neural network models yield good prediction results over the entire range of cutting process parameters spanned by the training data. The testing and prediction results show the effectiveness and satisfactory prediction accuracy of the artificial neural network modelling. The developed models are applicable to carbon steel.     

Prediction of cutting force for self-propelled rotary tool using artificial neural networks

In this paper, a cutting force model for self-propelled rotary tool (SPRT) cutting force prediction using artificial neural networks (ANN) has been introduced. The basis of this approach is to train and test the ANN model with cutting force samples of SPRT, from which their neurons relations are gradually extracted out. Then, ANN cutting force model is achieved by obtaining all weights for each layer. The inputs to the model consist of cutting velocity V, feed rate f, depth of cut a_p and tool inclination angle λ, while the outputs are composed of thrust force F_x, radial force F_y and main cutting force F_z. It significantly reduces the complexity of modeling for SPRT cutting force, and employs non-structure operator parameters more conveniently. Considering the disadvantages of back propagation (BP) such as the convergence to local minima in the error space, developments have been achieved by applying hybrid of genetic algorithm (GA) and BP algorithm hence improve the performance of the ANN model. Validity and efficiency of the model were verified through a variety of SPRT cutting samples from our experiment tested in the cutting force model. The performance of the hybrid of GA–BP cutting force model is fairly satisfactory.     

Comparison of tool life between ceramic and cubic boron nitride (CBN) cutting tools when machining hardened steels

This paper describes a comparison of tool life between ceramics and cubic boron nitride (CBN) cutting tools when machining hardened bearing steels using the Taguchi method. An orthogonal design, signal-to-noise ratio (S/N) and analysis of variance (ANOVA) were employed to determine the effective cutting parameters on the tool life. First order linear and exponential models were carried out to find out the correlation between cutting time and independent variables. Second order regression model was also extended from the first order model when considering the effect of cutting speed (V), feed rate (f), hardness of cutting tool (TH) and two-way of interactions amongst V, f, TH variables. The results indicated that the V was found to be a dominant factor on the tool life, followed by the TH, lastly the f. The CBN cutting tool showed the best performance than that of ceramic based cutting tool. In addition, optimal testing parameter for cutting times was determined. The confirmation of experiment was conducted to verify the optimal testing parameter. Furthermore, the second order regression model and exponential model supported the first order model regarding the prediction capability. Improvements of the S/N ratio from initial testing parameters to optimal cutting parameters or prediction capability depended on the S/N ratio and ANOVA results. Moreover, the ANOVA indicated that the cutting speed was a higher significant but other parameters were also significant effects on the tool lives at 90% confidence level. The percentage contributions of the cutting speed, tool’s hardness, and feed rate were about 41.63, 32.68, and 25.22 on the tool life, respectively.     

高温合金高速切削性能分析及参数优化

对铁基高温合金切削过程进行有限元建模和动态数值模拟,并结合高速切削试验平台,通过多元正交试验分析切削参数对切削力及刀具磨损的影响规律,得到基于切削用量的切削力和刀具寿命预测模型。最后应用遗传算法对切削参数进行合理选择与优化,试验所得数据为镍基合金高速切削工艺参数的选择提供了参考和依据。     

Tool flank wear prediction in CNC turning of 7075 AL alloy SiC composite

Flank wear occurs on the relief face of the tool and the life of a tool used in a machining process depends upon the amount of flank wear; so predicting of flank wear is an important requirement for higher productivity and product quality. In the present work, the effects of feed, depth of cut and cutting speed on flank wear of tungsten carbide and polycrystalline diamond (PCD) inserts in CNC turning of 7075 AL alloy with 10 wt% SiC composite are studied; also artificial neural network (ANN) and co-active neuro fuzzy inference system (CANFIS) are used to predict the flank wear of tungsten carbide and PCD inserts. The feed, depth of cut and cutting speed are selected as the input variables and artificial neural network and co-active neuro fuzzy inference system model are designed with two output variables. The comparison between the results of the presented models shows that the artificial neural network with the average relative prediction error of 1.03% for flank wear values of tungsten carbide inserts and 1.7% for flank wear values of PCD inserts is more accurate and can be utilized effectively for the prediction of flank wear in CNC turning of 7075 AL alloy SiC composite. It is also found that the tungsten carbide insert flank wear can be predicted with less error than PCD flank wear insert using ANN. With Regard to the effect of the cutting parameters on the flank wear, it is found that the increase of the feed, depth of cut and cutting speed increases the flank wear. Also the feed and depth of cut are the most effective parameters on the flank wear and the cutting speed has lesser effect.     

Cutting forces modeling considering the effect of tool thermal property—application to CBN hard turning

Force modeling in metal cutting is important for a multitude of purposes, including thermal analysis, tool life estimation, chatter prediction, and tool condition monitoring. Numerous approaches have been proposed to model metal cutting forces with various degrees of success. In addition to the effect of workpiece materials, cutting parameters, and process configurations, cutting tool thermal properties can also contribute to the level of cutting forces. For example, a difference has been observed for cutting forces between the use of high and low CBN content tools under identical cutting conditions. Unfortunately, among documented approaches, the effect of tool thermal property on cutting forces has not been addressed systemically and analytically. To model the effect of tool thermal property on cutting forces, this study modifies Oxley’s predictive machining theory by analytically modeling the thermal behaviors of the primary and the secondary heat sources. Furthermore, to generalize the modeling approach, a modified Johnson–Cook equation is applied in the modified Oxley’s approach to represent the workpiece material property as a function of strain, strain rate, and temperature. The model prediction is compared to the published experimental process data of hard turning AISI H13 steel (52 HRc) using either low CBN content or high CBN content tools. The proposed model and finite element method (FEM) both predict lower thrust and tangential cutting forces and higher tool–chip interface temperature when the lower CBN content tool is used, but the model predicts a temperature higher than that of the FEM.     

有限元仿真预测高温合金切槽刀具寿命的初步探讨

通过分析切槽刀具在切槽过程中的加工难点,结合难加工材料高温合金Inconel 718的材料特性,提出了一种新的刀具设计思路,设计了一种新断屑槽结构的切槽刀具,该断屑槽的特点是凸起部位由平滑过渡的曲面构成。对比刀片为市售国外同类产品是采用尖锐棱角构成的断屑槽。采用DEFORM软件对两种刀片进行有限元切削仿真和切削试验结果表明,新的切槽刀具的综合性能优于国外同类产品,且仿真评价与切削实验结果基本一致,说明采用有限元仿真预测不同方案的高温合金切槽刀具寿命的高低是基本可行的,可以初步应用到切槽刀具的结构设计中来。     

铝合金高速切削刀具磨损研究进展

由于铝合金高速切削过程中切削温度高,导致刀具严重磨损,降低了刀具寿命和零件加工精度,因此准确预测刀具磨损和分析刀具磨损规律至关重要。分别从刀具寿命模型和刀具磨损速率模型概述刀具磨损理论模型研究进展,基于切削用量、刀具性质和冷却方式分析刀具磨损规律。从已有研究来看,在铝合金高速切削过程中刀具磨损随切削速度和进给量增大而增大,切削深度无明显规律;常见刀具磨损有黏结磨损、磨粒磨损和扩散磨损,其中黏结磨损为主要刀具磨损机制。