基于BP神经网络的面向STEP-NC加工工步序列生成

针对STEP-NC非线性工艺规划中加工工步序列的生成问题,提出一种基于BP神经网络的工步序列智能化决策方法。在分析STEP-NC基于加工特征数据模型的基础上,提出面向STEP-NC加工工步序列的表示方案,并对其约束条件进行了分析;构建了基于BP神经网络的加工操作方法选择模型、机床选择模型和刀具选择模型的智能化决策策略,并对模型的有效性进行了训练与测试,充分融合了零件加工经验的定性化与神经网络的定量化决策方法,采用加工工步选择概率大小来实现STEP-NC工步序列的智能化生成;基于Solidworks三维CAD平台,开发了相应的原型系统,并以孔加工工步序列的生成为例,进一步验证了所提方法的正确性和有效性。     

基于遗传算法的工步优化排序方法

针对数控加工中心上零件加工工步的排序问题,以辅助加工时间最短为优化目标,使用遗传算法对零件在一次装夹情况下的加工工步进行优化排序。提出了使用特征关系图和特征高度描述待加工特征之间加工的优先顺序、采用工步优先关系矩阵校验工步序列合理性的方法。论述了初始群体的生成、遗传算子以及工步优化排序的过程和算法。实际应用表明,该方法可有效提高工艺规划系统中工步的优化排序能力。     

航空发动机整体机匣铣-车复合加工工艺优化

针对加工过程中的多次工件转位和刀具更换对加工质量和生产效率的重要影响,在分析工步排序和合理确定辅助时间的基础上,建立了以最短辅助时间为目标函数的工步排序优化模型,利用自适应遗传算法完成工艺规划的全局优化决策,获得了最佳工步排序和最短辅助时间.优化过程中,为了使工步序列满足工艺约束规则,首先根据工艺约束规则确定工步优先关系矩阵,然后利用优先关系矩阵对工步序列进行校验和调整.通过工步优化,有效地降低了发动机机匣铣-车复合加工过程中的辅助时间,达到了提高生产效率和加工质量的目的.     

面向能效的多工步数控铣削工艺参数多目标优化模型

为实现数控平面铣削工艺参数和工步数的节能优化,对多工步数控平面铣削加工过程能量构成进行分析,引入机床、刀具和加工质量约束,综合考虑工艺参数和工步数的关系,建立了以主轴转速、进给量、背吃刀量、铣削宽度和工步数为优化变量,以能量效率和加工成本为优化目标的多工步数控平面铣削工艺参数多目标优化模型。应用基于自适应网格的多目标粒子群算法对模型进行求解。以某夹具多工步数控平面铣削工艺参数优化为例,验证了模型的有效性和实用性。     

基于遗传算法与有向图拓扑排序的工艺路线优化

针对工艺设计中的工艺路线优化问题,归纳了工步间的基本优先级约束关系。基于约束关系,将整个工艺活动过程转化为工步有向图,工步节点间的拓扑关系以约束矩阵的形式存储。提出了约束矩阵判错的检测方法;建立了工步图的拓扑排序模型。设计了一种随机的深度优先搜索算法对工步图进行拓扑排序,得到全部可行的一定数目初始工艺计划作为遗传算法的初始种群。算法中,提出了基于车间动态资源的加工序列编码策略;定量分析了工艺计划评价准则,采用罚函数的方法将目标函数和约束条件建立成一个无约束的优化目标函数,由此确定了染色体的适应度函数;设计了遗传操作算子(选择、交叉、变异),并通过基于模拟退火机制的精英策略加速算法收敛。最后,通过实例证明了该算法的有效性。     

四主轴头五轴机床多工步加工路径优化

针对四主轴头多工步加工空行程耗时多的问题，基于空行程路径最短原则研究了无碰撞多工步加工路径优化方法。以四主轴头机床为对象确定了关节坐标系，在单面加工空行程规划基础上，建立了多面空行程计算模型和防碰撞多工步路径优化模型。基于同一面优先排序原则，在内层混合粒子群算法(GA-PSO)基础上融合了贪心算法，提出了双层混合粒子群优化算法(GA-PSO)。其中内层算法用于优化单一面上路径轨迹，外层算法用于优化多面间加工顺序。通过内外层优化算法间的不断迭代分层优化空行程，实现了整体多工步路径优化。对双层GA-PSO算法进行了算例分析，与遗传算法相比具有更快的收敛速度和更短的加工路径，最后通过虚拟仿真进一步验证了该路径优化算法的可行性。     

基于遗传算法的零件工步优化研究

对于复杂零件的加工,其工步顺序直接影响加工质量和生产效率.研究了工步排序原则和典型的工艺路线,以辅助加工时间最短为优化目标,建立了加工中心中零件加工工步排序数需模型.采用遗传算法对工步顺序进行优化,并在隔代映射遗传算法基础上引入自适应策略,使IP_GA中的交叉、变异概率根据适应度大小自动调节,提高了收敛速度及解的质量.实际应用表明,该方法可有效提高工艺规划系统中工步的优化排序能力.     

加工中心上基于模拟退火和遗传算法的工步优化问题研究

系统在地研究了工步优化问题的数学模型,给出基于模拟退火和遗传算法的工步优化的求解步骤,提出可替代工种链,自适应变异概率的概念;采用知识库推理,ＯＤＢＣ技术建立了零件工步优化的过程模型,将之应用于以加工中心为主要加工设备的创成式ＣＡＰＰ之中。     

低碳低成本约束下箱体零件加工路线优化方法

针对箱体零件工艺设计过程中工艺路线的低碳低成本优化问题,分析了复杂箱体零件的加工特征。在加工工步排序原则和机械加工工艺路线的基础上,根据多色集理论构建了工步排序问题的“特征-工步”和“特征-面孔”围道布尔矩阵,提出了以最低总碳排放（低碳）和最低加工成本（低成本）为优化目标的机械加工工艺路线多目标优化方法,并将改进的遗传算法应用到求解工步排序的问题中,从而得到最优解。以某型号箱体零件为例,验证了所提优化方法的有效性和可行性。     

基于多目标优化的制造资源分配研究

为了实现智能制造环境下企业制造资源的有效利用,提出了一种基于多目标优化的制造资源分配优化方法。构建了考虑加工成本、加工时间、加工质量和制造单元适合度的多目标优化模型,采用一种基于时变速度和位置加权更新的改进的粒子群算法,对该优化模型进行求解。最后通过仿真验证了所提方法的可行性。     

基于遗传算法的加工中心工步排序优化方法

在加工中心加工复杂零件时，工步多，其顺序直接影响加工质量和生产效率。在分析工步排序原则和典型工艺路线的基础上，建立了一个加工中心工步排序优化模型；将遗传算法应用到工步排序优化中，对算法的应用机理进行了详细讨论，提出了一种启发式基因重组方法；并示例验证了工步排序优化方法。     

Influence of the machining sequence on the residual stress redistribution and machining quality: analysis and improvement using numerical simulations

The manufacturing of aluminium alloy structural aerospace parts involves multiple steps, the principal ones being the forming (rolling, forging etc.), the heat treatments and the machining. During this last step, the final geometry of the part is obtained. Before machining, the workpiece has therefore undergone several manufacturing steps resulting in unequal plastic deformation and metallurgical changes which are both sources of residual stresses. On large and complex aluminium alloy aeronautical parts, up to 90 % of the initial workpiece volume can be removed by machining. During machining, the mechanical equilibrium of the part is in constant evolution due to the redistribution of the initial residual stresses.The residual stress redistribution is the main cause of workpiece deflections during machining as well as of post-machining distortion (after unclamping). Both can lead to the non-conformity of the part with the geometrical and dimensional tolerance specifications and therefore to a rejection of the part or to additional conforming steps. In order to improve the machining accuracy and the robustness of the process, the effect of the residual stresses has to be considered for the definition of the machining process plan. In this paper, a specific numerical tool [2] allowing to predict workpiece deflections during machining and post-machining distortion is used to study the influence of the machining sequence on the machining quality in taking into consideration the initial residual stresses. A first machining process plan defined as the reference case is simulated. Simulation results are then compared with experimental ones showing the feasibility to use the developed tool to predict the machining quality depending on the initial residual stresses, the fixture layout and the machining sequence. Using the computational tool, a method to optimise the machining quality depending on the initial workpiece and on the machining sequence is presented. A machining process plan allowing to respect the tolerance specifications is then defined. This demonstrates the feasibility to adapt and to optimise the machining process plan to ensure conformity of the part with the tolerance specifications.     

基于多色集合理论和遗传算法的加工中心工步排序研究

针对加工中心上一次装夹下复杂的工步排序问题,通过实例介绍一种基于多色集合理论和遗传算法的工步排序优化算法。首先,在分析工步排序原则和典型工艺路线的基础上,根据多色集合理论建立加工中心上工步排序问题的约束模型。然后,以辅助时间最短为优化目标,建立其数学优化模型。最后,将遗传算法应用到工步排序中从而得出最优解。实例证明,在多色集合约束模型约束下的遗传算法能够很好地求解加工中心上的工步排序问题,排序结果接近最优且可以大幅提高加工中心的效率。     

基于仿真的数控铣削加工参数优化研究

针对数控铣削加工参数优化问题,通过加工仿真,计算每一走刀步的切削深度和切削宽度并划分区间。在每一个由组合划分区间内所有刀步构成的一个加工特征组合段上建立了多目标优化模型,模型采用遗传算法对每一加工特征组合段的加工参数进行优化,自动修改NC程序中的加工参数并反映优化结果。应用实例证明,提出的多目标优化模型和优化求解算法正确、有效。     

Multi-criteria end milling parameters optimization of AISI D2 steel using genetic algorithm

This paper focuses on using multi-criteria optimization approach in the end milling machining process of AISI D2 steel. It aims to minimize the cost caused by a poor surface roughness and the electrical energy consumption during machining. A multi-objective cost function was derived based on the energy consumption during machining, and the extra machining needed to improve the surface finish. Three machining parameters have been used to derive the cost function: feed, speed, and depth of cut. Regression analysis was used to model the surface roughness and energy consumption, and the cost function was optimized using a genetic algorithm. The optimal solutions for the feed and speed are found and presented in graphs as functions of extra machining and electrical energy cost. Machine operators can use these graphs to run the milling process under optimal conditions. It is found that the optimal values of the feed and speed decrease as the cost of extra machining increases and the optimal machining condition is achieved at a low value of depth of cut. The multi-criteria optimization approach can be applied to investigate the optimal machining parameters of conventional manufacturing processes such as turning, drilling, grinding, and advanced manufacturing processes such as electrical discharge machining.     

190柴油机曲轴数控铣削加工的分析探讨

曲轴是发动机的主要零件,其加工质量和加工效率直接影响发动机的工作性能和制造。曲轴粗加工采用CNC外铣加工工艺,使工件的加工精度、适应多品种生产的柔性和质量的稳定性都有了明显的改善。我厂引进了两台数控曲轴外铣铣床,特别适合于那些对平衡块侧面需要加工、轴颈没有沈割槽的曲轴。其较之车削、内铣等工艺,具有更高的生产效率。以济柴190型柴油机曲轴加工为实例,简要讲述了用数控曲轴铣床进行曲轴粗加工、半精加工的铣切工艺过程步骤和优点。     

基于自适应蚁群算法的工艺路线优化

针对计算机辅助工艺设计中最优方案选择,提出一种以制造资源更换率最低为目标的自适应蚁群优化方法(Adaptive ant colony algorithm,AACA)。通过分析零件特征,根据精度要求对制造特征进行分解,提出加工元概念。加工元被定义为特定的制造特征、加工阶段、加工方法、制造资源、装夹位置的集合,工艺路线的确定被转换为对加工元的优化顺序安排问题。以制造特征之间的几何位置约束,各加工阶段的先后顺序约束为基本元素,构造加工元优选约束矩阵,给出基于优选约束矩阵的加工元优选原则。在加工元优先顺序约束和可用制造资源的共同约束下,将缩短加工周期、提高加工质量和降低加工成本的综合目标表达为制造资源更换率最低,进行优化目标函数的数学建模。指出加工元优化排序可类比旅行商问题,并选择AACA进行优化求解。实例分析表明提出的方法可以可靠和有效地得到符合生产实际的工艺路线。     

Empirical modeling and optimization of wire electrical discharge machining

Wire electrical discharge machining is a widely used process in manufacturing industries to machine complex profiles. The performance of any machining process is based on choosing the right combination of input parameters. Metal removal rate and surface roughness are the most important output parameters, which decide the performance of a machining process. The selection of optimal parameters in wire electrical discharge machining is difficult as it is a complex process and involves a large number of variables. The present work models the metal removal rate and the surface roughness in terms of the input variables using the response surface methodology and, consequently, the developed mathematical models are utilized for optimization. Since the influences of machining parameters on the metal removal rate and the surface roughness are opposite, the problem is formulated as a multiobjective optimization problem. Non-dominated sorting genetic algorithm is then applied to obtain the Pareto-optimal set of solutions.     

面向低碳制造的机械加工系统工艺优化模型及方法研究

低碳制造在现代化企业中是一种可持续发展的模式,在实际应用中具有明显的经济效益与社会效益,在当前的机械制造行业成为其显著的特点.低碳制造的实行主要以科学技术为支撑,在机械加工过程中对资源消耗和废弃物的排放具有重要的辅助作用.面对低碳制造的机械加工系统工艺优化模型及方法进行研究,对整个加工过程中的各要素以及各要素之间的关系进行了描述及分析,与机械加工系统在运行过程中产生的资源消耗与环境影响相结合,对工艺的优化进行了具体的分析,同时对优化的方法进行了简单的分析,希望可以为有关低碳制造方面的研究提供一些理论性的参考与借鉴.     

Abrasive water and slurry jet micro-machining techniques for fabrication of molds containing raised free-standing micro-features

The demand for metallic micro-molds that can be used for inexpensive mass production of polymeric microfluidic chips is increasing. Existing manufacturing techniques such as soft-lithography and photolithography can require multiple time-consuming steps, especially when the aim is to create three-dimensional features. In this study, the feasibility of using abrasive water jet machining (AWJM) and abrasive slurry jet machining (ASJM) to fabricate such micro-molds in Al6061-T6 and SS316 was studied. Jet raster scans under various combinations of process parameters were used in order to machine micro-pockets containing free-standing structures, representing molds for casting microfluidic chips with channel networks. As expected, for both materials and using both ASJM and AWJM, the pocket roughness decreased as the distance between adjacent raster scans (step size) decreased, but the lowest waviness occurred at an intermediate step size. The best quality pockets were achieved on SS316 using ASJM with the intermediate step size and the highest possible slurry mass flow rate. Unmasked machining could not be used to fabricate molds with sharp-edged intersecting features, and a novel hybrid AWJM/ASJM masked machining technique was thus introduced. An undercut and an undesirable erosion near the edges of the mask formed if the position of the last raster scan closest to the mask was not carefully controlled. Possible reasons for these phenomena were discussed in terms of the likelihood of jet deflection off the machining kerf and mask, and the resulting erosion due to secondary slurry flow. By careful selection of the process parameters, it was demonstrated that high quality molds with both single and intersecting free-standing structures at multiple heights could be fabricated, thus making three-dimensional microfluidic chip mold fabrication feasible.