节点文献
基于三维几何模型的计算机辅助工艺规划技术
Research on Computer Aided Process Planning Based on Three Dimensional Geometric Models
【作者】 陈永府;
【作者基本信息】 华中科技大学 , 机械设计及理论, 2006, 博士
【摘要】 随着三维CAD系统在国内企业迅速普及,如何有效利用三维几何模型进行工艺规划是亟待解决的问题。本文研究基于三维几何模型的计算机辅助工艺规划技术,旨在充分发挥三维模型定量分析上的优势,增强CAPP的自动化程度,从而满足快速工艺规划的现实需要。论文围绕着三个主题展开研究:一是面向变形产品的参变式工艺规划;二是基于轻量化的可视化装配工艺;三是可定制的柔性工艺统计报表。作为这三个主题的基础,本文提出中性工艺信息模型,将工艺信息分为工艺管理信息和工艺规程信息,其中工艺规程信息用工艺操作集及其偏序关系表达,该模型为本文后续研究提供数据组织和算法支持。面向变形产品的参变式工艺规划包含三个部分:加工特征识别,耦合约束建模与求解和参变式工艺规程生成。对于加工特征识别,本文提出集成自动识别和交互特征定义的混合式特征识别方法,以提高特征识别方法的健壮性和实用性。对于耦合约束建模与求解,首先将数值型约束与符号知识型约束进行融合,建立统一的表示模型;然后提出耦合约束求解算法,采用有向图表达出约束之间的依赖关系,通过逐级子图分解、“孪生变量法”一阶解耦、求解、再验算等步骤实现耦合约束求解。在此基础上,提出参变工艺规程模型,该模型由相对固定的项目( Fi )和变化的项目( Vi )组成,后者用一系列的参变量表示,并对这些参变量采用耦合约束建模理论进行建模,然后通过耦合约束求解实现参变式工艺规划。为适应大规模装配和快速显示的需要,研究三维模型的轻量化技术。在此基础上,研究基于三为轻量化模型的可视化装配工艺:以直观的方式展现零部件的装配拓扑关系和位置关系;用动画的方式模拟装配过程与装配顺序;进行装配过程中的冲突和干涉检查。为实现可定制的柔性工艺统计报表,提出以零部件作为数据重构核心,用“行匹配列合并运算”将企业零散的数据有序地组织起来,形成立体数据的重构模型。在数据重构模型支持下,将统计报表过程划分为:数据收集、数据处理和格式化输出三个步骤,并提炼出“共性”和“特性”规则,从而通过配置实现可定制的复杂工艺报表。最后,简要阐述了系统架构和实现方式,结合应用实例验证本文提出的基于三维几何模型的计算机辅助工艺规划的相关理论和方法是可行和有效的。
【Abstract】 As three-dimensional CAD systems are more and more popular in enterprises, it becomes a key problem to effectively use three-dimensional geometric models in process planning, which needs to be solved immediately. Therefore, several techniques for CAPP based on three-dimensional (3D) CAD models are discussed here. Advantages of 3D digital models are fully exploited to enhance automation degree of CAPP; as a result, requirements of rapid response of process planning are met.Generally, this dissertation focuses on three subjects as following: parametric process planning for variant parts, visualization of assembly process based on 3D lightweight models and custom-built flexible process reporting. As the foundation of the subjects, a neutral process model, which is composed of management information and process plans expressed as operations with its precedence relationships, is proposed. The model supports for data managing and algorithms discussed below.Parametric process planning for variant parts includes recognizing machining features, modeling and solving numerical and symbolic coupling constraints, and generating parametric process plans. To recognize machining features, a hybrid approach integrating automatic recognition and interactive definition is presented here, which greatly enhances practicability and robustness. Numerical and symbolic constraints are incorporated to establish a uniform model. And the dependence of constraints is expressed by a digraph. Then an algorithm for solving coupling constraints is proposed, which contains the following steps: dividing the digraph into sub-graphs gradually at first; breaking the coupling constraints with first order by twin-variables; solving them respectively and checking computations at last. On the basis of the model a parametric process model is proposed, in which processes are expressed as fixed items ( Fi ) and variable items ( Vi ). The latter, denoted by serial variables, is constrained by coupling constraints. And then process plans are generated by solving numerical and symbolic coupling constraints.Lightweight technique for 3D models is discussed to improve the speed of displaying large-scale assemblies. And based on it, visualization of assembly process that includes showing topological structure and positions, animating assembly procedure and sequence, and checking conflict and interposition of parts is accomplished.To achieve custom-built flexible process reporting, a cube data reconstruction model based on part information is presented to put the heterogeneous data from different sources in order through row-matching and column-merging. Based on the reconstruction model, reporting procedure is regarded as three steps: collecting data, handling data and outputting data as format. And common rules and special ones for each step are abstracted. Then, custom-built complex process reports are realized through interactive definition and configuration with the rules.Structure and implementation of the system are discussed at last. At the same time, some examples in real industrial enterprises are given to show that the theories and methods for CAPP based on 3D models presented above are valid and practicable.