【作者】 赵文燕；

【导师】 何桢；

【作者基本信息】 天津大学 ， 管理科学与工程， 2010， 博士

【摘要】 基于产品平台的产品开发模式通过充分挖掘已有系统资源,以尽可能少的内部多样化满足外部需求的不断变化,实现成本和风险的降低。复杂产品因其需求、结构、技术、制造和管理等方面的复杂性,导致产品设计参数和约束繁多,设计结果多变,定制生产周期长、风险大、成本高。在对复杂产品设计时只有充分利用现有成熟技术和零部件才能够提高复杂产品设计的可靠性,达到预期设计效果。因此为复杂产品建立产品平台的重要性相对于简单产品而言更加凸显。同时,复杂产品多变的使用环境,要求复杂产品的产品平台具有较好的稳健性。本文根据产品平台设计的任务,与稳健设计思想相比较,提出产品平台设计过程的本质是稳健设计。因为传统稳健设计不针对产品族设计问题,本文对稳健设计的概念进行扩展,并基于此提出产品平台的稳健设计模型,指出建立稳健产品平台的过程就是分析目标产品族结构变量对设计变量变化的敏感性。在稳健设计模型的基础上,本文提出面向复杂产品建立稳健产品平台的流程,包括设计变量提取和分析、系列划分、模块划分以及可调节模块的稳健参数设计等。该流程针对复杂产品的特点,应用实验设计方法中分区组的方法,提出建立系列产品平台以提高通用率;在流程的各个步骤中均充分考虑产品平台的稳健性。本文针对复杂产品稳健产品平台建立流程中的两个关键问题进行了研究:(1)对复杂产品设计变量识别和筛选的方法进行了研究。采用鱼骨图分析导致产品结构变量变化的因素,剔除人为原因造成的不必要变型;改进粗数方法,用粗数方法对客户需求进行重要度评价,并约简不重要客户需求;引入敏感性分析方法研究客户需求与技术参数之间的关系以及技术参数和产品主要部件之间的关系,进而得到划分平台系列的关键参数。(2)研究产品平台结构变量获取与表达方法,并将平台结构变量定义为实体模块和结构参数。在每个平台系列内利用反向鱼骨图对产品实体进行分解,基于敏感性分析将实体模块划分为标准模块、虚拟模块、需分解模块、柔性模块和可调节模块;建立可调节模块的参数化平台,给出了基于敏感性分析确定单值平台参数、多值平台参数和可调节参数的确定方法。更多还原

【Abstract】 The mode of product development based on product platform can satisfy the ever-changing external demands with as little as possible internal diversity, and reduce cost and risk, by making full use of the existing system resources. Complex products show numerous complexities in the customer demands, structure, technique, manufacture, management and other aspects, which results in multifarious design parameters and constraints, uncertained design results, long lead time, high risk and cost. Then it is rather important for complex product design to use existing mature technique and components fully, so as to enhance the reliability of the design and achieve the anticipated design effect. In brief, the product platform building for complex product has more importance than that for simple product. At the same time the product platform built for complex product should be robust to adapt greatly uncertain application environment.By comparing the aim of product platform design with that of Robust Design, it indicates that the process of product platform design in nature is robust design. Traditional Robust Design fails to deal with the product family design, so the concept of Robust Design is extended here, on which the robust design model for product platform design is put forward. The model indicates that the process of robust product platform design is a process of performing sensitivity analysis to analyze the impact of the change of design variables on the structure variables of the target product family.Based on the model of robust design, a flowchart of robust product platform design for complex products is presented, which includes design variables drawing and analysis, series division, modularization, and robust parameter design of scaled module. According to the characteristic of complex products, by using block design in Experimental Design for reference, product platforms could be serialized for increasing commonality. Additionally, the demands of robustness are carefully considered in each step of product platform design.Two key issues in the process of robust product platform building for complex products are emphasized on: (1) the way of design variables identification and selection is researched on. First, fishbone diagram is introduced to identify and analyze the factors result in the changes of the product structure parameters, and to reject some artificial reason result in unnecessary modification of design. Second, the method of rough number is improved to perform importance evaluation for customer demands and reduce unimportant demands. Third, to obtain the key parameters as the reference index for series division, a method based on sensitivity analysis from customer requirements to technical parameters and further to product components is advised. (2) The approach of acquisition and description of the structure variable of product platform is studied on. The structure variables are defined as entity modules or structure parameters. The entity modules are divided into standard modules, flexible modules, scalable modules, virtual modules and decomposable modules through sensitivity analysis. And the initial modules are obtained by decomposing the product with reverse fishbone diagram in the interior of each series. Additionally, the parametric product platform is built for scaled modules, and the way of parameter identification is put forwards for single-valued, multi-valued and scaled value parametric platform.更多还原