变速箱齿轮自动生产线的刀具更换决策方法研究

发布时间：2018-05-13 02:41

本文选题：变速箱齿轮 + 自动生产线　；参考：《重庆大学》2016年博士论文

【摘要】：刀具磨损是齿轮加工中普遍存在的客观现象,磨损引发的刀具性能退化是影响齿轮质量的主要因素。但是,频繁的刀具更换不仅会增加刀具成本,还会造成大量生产中断,影响制齿机床的生产效率。在变速箱齿轮自动生产线中,频繁的刀具更换还可能造成整条生产线运行中断。因此,需要对变速箱齿轮自动生产线刀具更换进行合理控制。本文结合国家科技重大专项“轿车变速箱齿轮加工自动生产线”课题,以变速箱齿轮自动生产线为研究对象,将工件加工质量、切削能耗以及刀具失效作为影响刀具更换决策的主要因素,对生产线机床刀具更换决策问题进行研究,旨在提出能够降低生产成本、提高生产效率和生产过程能量效率的刀具更换决策方法。首先,分析专项所研制变速箱齿轮自动生产线的生产特点及工艺流程;分析生产线机床刀具磨损对齿轮生产过程质量失效概率、单位工件切削能耗以及生产过程连续性的影响;提出变速箱齿轮自动生产线刀具更换决策问题,并建立变速箱齿轮加工自动生产线刀具更换决策研究框架,分析刀具更换决策研究的决策目标、影响因素、建模技术等。其次,以单位工件长程平均成本为优化目标对变速箱齿轮自动生产线机床周期性刀具更换决策问题进行研究。分析影响刀具更换决策的成本因素,基于Logistic方程建立齿轮生产过程的质量失效概率模型,进而对刀具磨损引起的齿轮加工过程质量损失进行估计;以刀具更换周期为决策变量,以单位工件长程平均成本最小为目标,基于更新报酬定理建立周期性刀具更换决策模型,研究模型的求解方法并对最优解的特性进行分析。再次,以生产率和能量效率为优化目标对变速箱齿轮自动生产线机床周期性刀具更换决策问题进行研究。用离散时间、离散状态的马尔可夫链描述刀具磨损过程,进而分别建立长程平均生产率模型和长程平均能量效率模型;在此基础上以刀具更换周期为决策变量,以生产率和能量效率为优化目标建立双目标优化的周期性刀具更换决策模型;利用马尔可夫链的长游程性验证模型的准确性,分析长程平均生产率和长程平均能量效率随刀具更换周期的变化特性。然后,基于部分可观察马尔可夫决策过程(Partially Observable Markov Decision Process,POMDP)理论对变速箱齿轮自动生产线机床动态刀具更换决策问题进行研究。依据POMDP基本框架以及刀具磨损对质量失效概率和切削能耗的影响规律,建立刀具更换决策的观察与刀具磨损过程马尔可夫链状态之间的关联;确定刀具更换决策行动集合,分析观察、行动与能量效率之间的联系,建立关于决策阶段能量效率的即时报酬函数;以有限阶段期望总报酬最大为目标建立动态刀具更换决策模型,分析最优刀具更换策略的特性及模型参数变化对其产生的影响。最后,结合重庆蓝黛动力传动机械股份有限公司变速箱齿轮自动生产线机床刀具更换问题,对所提出的周期性刀具更换决策方法的应用情况进行分析,结果表明所提方法能够实现生产性能的改善。
[Abstract]:Tool wear is a common objective phenomenon in gear processing. The deterioration of tool performance caused by wear is the main factor affecting the quality of gear. However, frequent tool replacement will not only increase the cost of cutting tools, but also cause a large number of production interruptions and affect the production efficiency of the machine tool. In the automatic gear production line of the gearbox, frequent knives are used. The replacement of the whole production line may be interrupted. Therefore, it is necessary to control the tool replacement of the gearbox gear automatic production line. In this paper, the automatic production line of the gearbox is taken as the research object with the subject of the automatic production line of the gearbox gear in the national science and technology major special project. The machining quality of the workpiece is cut. Energy consumption and tool failure are the main factors affecting tool replacement decision. The decision problem of tool replacement for machine tool is studied. The purpose of this paper is to put forward a tool replacement decision method which can reduce production cost, improve production efficiency and energy efficiency of production process. The characteristics and process flow of the production line, analyze the effect of tool wear on the quality failure probability of the gear production process, the energy consumption per unit workpiece and the continuity of the production process, put forward the decision problem of tool replacement in the automatic transmission line gear production line, and establish the research framework for the tool replacement decision of the gearbox gear addition automatic production line. The decision goal, influence factor and modeling technology of tool replacement decision research are analyzed. Secondly, the decision problem of periodic tool replacement of machine tool in automatic transmission line gear production line is studied by using the average cost of the unit work work as the optimization goal. The factors affecting the decision of tool replacement are analyzed, and the gear production is established based on the Logistic equation. In order to estimate the quality loss of the gear machining process caused by tool wear, the tool replacement period is taken as the decision variable and the average cost of the unit work work is minimum as the goal. The periodic tool replacement decision model is established based on the renewal reward theorem, and the solution method of the model is studied and the optimal solution is studied. The characteristics are analyzed. Thirdly, the decision problem of the periodical tool replacement for the machine tool of the transmission gear automatic production line is studied with the optimization goal of productivity and energy efficiency. The tool wear process is described by the discrete time and discrete state Markov chain, and then the long range average productivity model and the long range average energy efficiency model are established respectively. On this basis, using the tool replacement cycle as a decision variable, the cycle tool replacement decision model of double objective optimization is established with the goal of productivity and energy efficiency as the optimization goal. The long range average productivity and the long range average energy efficiency vary with the changing cycle of the tool with the accuracy of the long range validation model of Markov chain. Then, based on the partial observable Markov decision process (Partially Observable Markov Decision Process, POMDP), the dynamic tool replacement decision problem of the machine tool is studied. Based on the basic framework of POMDP and the influence of the tool wear on the quality failure probability and the cutting energy consumption, it is built. The relation between the observation of the decision making of the tool replacement and the Markov chain state of the tool wear process, determine the set of decision action of the tool replacement, analyze the relation between the observation, the action and the energy efficiency, establish the instant reward function on the energy efficiency of the decision stage, and establish the dynamic tool with the goal of the maximum total reward of the finite order. By changing the decision model, the characteristics of the optimal tool replacement strategy and the influence of the change of model parameters on the tool change are analyzed. Finally, the application of the new method for the more decision making of the proposed periodic tool is analyzed in combination with the tool replacement of the machine tool of the Limited by Share Ltd transmission gear automatic production line of the Chongqing blue and deah power transmission machinery. The method mentioned in the Ming Dynasty can improve the production performance.

【学位授予单位】：重庆大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TG61

【相似文献】