【作者】 刘长清；

【导师】 姚英学；

【作者基本信息】 哈尔滨工业大学 ， 机械制造及其自动化， 2007， 博士

【摘要】 数控加工技术在制造业中占有重要地位,是机械加工现代化的主要基础和关键技术,也是发展军事工业的重要战略技术。在数控加工中采用加工过程优化技术,不仅有利于提升数控加工设备的制造能力,而且能产生重大经济效益。传统的加工过程优化方法是采用自适应控制技术,尤其是约束型自适应控制技术。随着计算机技术的发展,国内外研究者开展了基于仿真技术的加工过程优化研究。相对于自适应控制的在线优化,这是一种具有离线性质的优化方法,具有成本低、符合绿色制造、无实时性要求的特点,拥有良好的发展前景。虚拟加工技术是当前加工仿真技术发展的一个新阶段,不仅能提供逼真的加工环境仿真,而且在几何仿真和物理仿真过程中考虑了机床系统的运动特性和动态特性。本文开展了基于虚拟加工仿真的切削过程离线优化技术研究,对离线优化的原理和优化流程进行了分析,并对优化数学模型的建立、约束规则的选择进行了分析与研究。加工过程离线优化的前提是实现准确的加工过程仿真,本文开展了利用实体仿真技术进行加工过程仿真的研究。分析了刀具扫描体生成过程中出现自相交现象的原因,并提出了解决方案。针对实体仿真中容易出现的错误,提出了基于空间扫描轮廓生成刀具扫描体的方法,保证了加工仿真结果的正确性。在加工过程仿真与优化系统中,切削参数的获取对铣削力模型在加工过程中的应用、刀具寿命的预测以及切削参数的优化都具有重要意义。本文提出了采用“剖切法”获取切削参数,其实质是将刀具在进给方向上的半圆柱面代替刀具实体并与材料去除体进行布尔交运算,然后对布尔交结果进行处理来获取切削参数的过程。利用“剖切法”可正确地获得立铣加工中的切削深度、切削宽度以及切入角和切出角。相对于利用切屑几何体获取切削参数的方法,“剖切法”获取切削参数的过程不会增加工件复杂度,但显著降低了切削参数获取的计算量。铣削力预测值与测量值的较好吻合间接验证了切削参数获取方法的正确性。针对加工过程优化的特点,本文选择粒子群优化算法优化切削参数。在标准粒子群优化算法基础上,引入粒子群平均距离对优化过程进行评价与控制。通过选择合适的粒子群平均距离阈值,改进型粒子群算法实现了优化成功率与优化效率的统一。数控程序中的进给速度和主轴转速通常都未得到优化,针对切削参数优化研究较集中于优化进给速度而较少对主轴进行优化的现象,本文对同时实现进给速度和主轴转速优化进行了探索与研究。将主轴转速和进给速度与粒子群优化算法相结合,在粒子寻优过程中实现了对数控程序中的主轴转速和进给速度的同时优化。在ACIS实体造型系统基础上,开发了具有实体仿真功能的铣削加工过程离线优化软件系统。利用该系统根据恒切削力和效率的组合目标对粗加工数控铣削程序进行了优化,并采用优化后的数控程序进行了加工实验。铣削实验验证了离线优化系统的有效性和可靠性。更多还原

【Abstract】 Numerical control (NC) machining occupies an important place in the manufacturing industry. As one of the key technologies, NC machining is the basement for manufacturing modernization, and is strategically important to develop military industry. Using existing NC machining equipments, machining process optimization can increase efficiency and guarantee machining quality, which is important to improve the machineability of the equipments and produce great economic benefits. Traditionally, adaptive control (AC) technology, especially adaptive control constraints (ACC) was applied to optimize and control the machining process. With the development of computer technology, optimizing machining process based on machining process simulation technique becomes a more potential method, which is an off-line optimization method relative to the on-line optimization in AC, and shows the advantage of low cost, agreement with the trend of green manufacturing and without real-time requirement.As a new stage of machining process simulation, virtual machining can supply vivid machining scene, and take full consideration of machine performance such as kinematic characteristics and dynamic characteristics in geometry simulation and physical simulation. In this paper, off-line machining process optimization technology, which based on virtual machining technique, is studied including off-line optimization principle and process, establishment of optimization mathematics model, selection for constraint rules.Because machining simulation is the basement to the off-line optimization, this paper focus on the research on machining process simulation based on solid simulation technology. Depending on the analysis to the reasons of self-intersection in constructing tool swept volume, the solved method is proposed. To eliminate the wrong result in solid simulation, a sweeping method based on a 3-dimision swept profile is introduced.In simulation and optimization to machining process, getting milling parameters is very important to apply cutting force model, to estimate tool life and to optimize cutting parameters. This paper presents a new method which is named“sliced method”, which replaces the tool by its half cylindrical face in feed direction, and treats the intersection geometry between the half cylindrical face and the removed material body of tool path to calculate the milling parameters including radial width, axial depth, cut-in angle and cut-out angle.“Sliced method”does not increase the complexity of workpiece, but is more efficient to calculate the parameters relative to the method according to the chip geometry. The validation of the“sliced method”is test by the good agreement between the measured forces and the predicted forces indirectly.To fit the characteristic of machining process optimization, particle swarm optimization (PSO) algorithm was improved by introduced the particle average distance to the standard PSO algorithm. By controlling the threshold of particle average distance, improved PSO algorithm can achieve the optimal result with smaller iterations.Because spindle revolutions and feed rates are not optimized in most NC file,but most optimization researches on federate optimization, this paper makes study on optimizing federates and spindle revolutions together. Combined with the PSO, feed rates and spindle revolutions in NC file are optimized during each“particle”seeking its optimal value.Based on ACIS solid model system, an off-line optimization system of NC end milling process is developed which has the function of solid simulation. Using the optimization system, spindle revolutions and feed rates in NC file for rough machining are optimized according to the multiple objectives: constant machining forces and increasing machining efficiency. Using the optimal NC file, milling experiments are carried out and the experiment result shows the validity and reliability of off-line milling process optimization system.更多还原