基于贝叶斯和决策树方法的汽车电子研发项目变更决策模型研究

发布时间：2018-09-12 07:15

【摘要】：随着移动互联网时代的到来，新科技如雨后春笋，以前所未有的速度在各个领域涌现出来。根据相关统计，在当前的汽车产业中，70%的创新来自于汽车电子技术。同样，从PC、手机的发展情况来看，我们有足够的信心预测，汽车电子将很快成为汽车行业的兵家必争之地，汽车电子技术产品也成为各个整车厂的兵戎相见之地。 YFVE公司，作为上汽集团的子公司，虽然在传统的汽车内外饰，低端的汽车电子产品领域占据着国内的头把交椅，并在国际上也占据一定的市场份额，但是对于带有高科技的创新技术的高端汽车电子产品，却还在起步阶段。而经过前两年的尝试，虽然公司的汽车电子研发跌了跟头，最终以失败收场。但是在这个研发过程中，暴露出来的是首要问题是对于带有多变性的高科技的研发项目的变更管理决策的能力的不足。对于带有创新性质的高科技汽车电子项目，由于鲜有可以借鉴的经验，并且在技术上存在着很大的未知性，所以传统的项目管理方式，尤其是对于变更的管理决策方式并不能满足这样的汽车电子研发项目的需求。针对汽车电子研发项目所发生过的项目变更管理问题，我们需要找到一种科学的，系统的方法来辅助我们管理，尤其是对于不确定性的技术变更的分析和决策的管理。本文将首先介绍汽车电子技术的发展历史和现状，然后阐述YFVE公司在进行研发项目的变革管理决策中所发生的问题。然后在第三章将着力建立一个决策模型。主要是基于贝叶斯决策方法对变更的先验概率进行分析，并得到更加合理的准确的后验概率。之后采用决策树的方法，对于技术变更的可能性作一个整体的分析比较，并通过收益分析，，最终选择出可能会带来最大收益的决定。最后，结合今年YFVE公司拿到的新高端研发项目的实际使用案例，阐述该模型在工作中的实际运用情况及效果，从而有效的证明该模型的科学性及可行性。
[Abstract]:With the arrival of the mobile Internet era, new technologies are springing up and emerging in various fields at an unprecedented speed. According to relevant statistics, 70% of the innovation in the current automotive industry comes from automotive electronics technology. Similarly, judging from the development of PC, mobile phones, we are confident enough to predict that automotive electronics will soon become a war taker in the automotive industry. YFVE, a subsidiary of SAIC, although in the traditional automotive interior and exterior decoration, the low-end automotive electronics field occupies the top position in the domestic market. It also occupies a certain market share in the world, but it is still in its infancy for the high-end automotive electronics with high-tech innovation technology. And after two years of trial, although the company's automotive electronics research and development fell, and finally failed. However, in this research and development process, it is revealed that the most important problem is the lack of the ability to make change management decisions for hi-tech R & D projects with variability. For high-tech automotive electronics projects with innovative nature, because there are few experiences to learn from, and there is a great uncertainty in technology, so the traditional project management methods, Especially for the change management decision-making method can not meet the needs of such automotive electronics R & D projects. In order to solve the problem of project change management in automotive electronics R & D projects, we need to find a scientific and systematic method to assist us in management, especially in the analysis and decision management of uncertain technical changes. This paper first introduces the development history and current situation of automotive electronics technology, and then expounds the problems occurred in the decision making of change management for R & D projects in YFVE Company. Then in the third chapter will focus on the establishment of a decision model. The prior probability of change is analyzed based on Bayesian decision method, and a more reasonable and accurate posterior probability is obtained. Then the decision tree is used to analyze and compare the possibility of technology change, and through the income analysis, the decision that may bring the maximum profit is selected. Finally, combined with the actual use case of the new high-end R & D project obtained by YFVE Company this year, this paper expounds the actual application and effect of the model in the work, so as to effectively prove the scientific nature and feasibility of the model.
【学位授予单位】：上海交通大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP18;F426.471

【参考文献】