复杂产品开发过程仿真与预测研究  

Research on Complex Product Development Process Simulation and Prediction

作　　者：赵韩[1] 武照云[2] 韦艳丽[1] 宋晖[1] 

机构地区：[1]合肥工业大学安徽省数字化设计与制造重点实验室,合肥230009 [2]河南工业大学机电工程学院,郑州450007

出　　处：《系统仿真学报》2009年第22期7350-7354,7358,共6页Journal of System Simulation

基　　金：安徽省科技攻关计划项目(07010202108)

摘　　要：为了更有效的对复杂产品开发过程进行预测分析,提出一种两层设计结构矩阵模型作为仿真的结构关系模型。建立了服从三角分布的时间与成本模型,采用拉丁超立方抽样法对该模型进行抽样并生成样本点。采用任务返工概率、返工影响与学习效应曲线来描述开发过程的迭代特征,对任务返工概率的变化调整、返工需求判断策略、多任务返工执行约束等问题给出了相应的仿真模拟方法。在此基础上提出了改进的产品开发过程仿真算法,并设计开发了基于Matlab的复杂产品开发过程仿真平台。最后,通过一个实例证明了该仿真平台的有效性和实用性。In order to predict and analyze complex product development process （PDP） effectively, a two-layer design structure matrix model was proposed as structural relation model in simulation. The schedule and cost model obeyed triangular distribution and was sampled by Latin Hypercube Sampling Method to generate initial data. The iterative characteristics of PDP were described by rework probability, rework impact, and the learning curve. The simulation methods of some key problems were proposed, including the rework probability adjustment, the rework judgment policy, and the multi-task rework execution constraints. Then, the improved simulation algorithm of PDP was proposed. The simulation platform of PDP was developed based on Matlab. Finally, an example was used to illuminate the validity and practicability of the simulation platform.

关 键 词：复杂产品开发过程 设计结构矩阵 过程仿真 迭代 风险预测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]