【作者】 潘瑞林；

【导师】 刘晓冰；

【作者基本信息】 大连理工大学 ， 企业管理， 2010， 博士

【摘要】 成本管理的重要内容是实现对生产资源成本要素的合理配置。本文针对特钢企业成本管理中生产资源成本要素配置不合理的问题,在分析国内外特钢企业生产资源成本要素配置方法研究与应用现状的基础上,以东北特钢集团大连基地生产资源成本要素配置实践为背景,应用系统、集成、优化与控制的管理思想,从生产过程和投入两个阶段研究了特钢企业生产资源成本要素配置方法和体系,主要研究内容如下：针对目前生产资源成本要素分配方式选择不当和效率低的问题,根据特钢企业生产资源成本要素分配方式选择的特点,提出了一种基于支持向量机的生产资源成本要素分配方式自学习选择模型,建立了资源的成本项目、作业与分配方式之间的映射关系,实现了特钢企业生产资源成本要素分配方式的自动选择和匹配。并采用决策树算法对选择规则进一步提取,使隐性的知识显性化,通过应用实例给出了该方法的具体实现过程。针对以往采用的线性分配系数不能满足非线性、复杂生产工况的问题,提出了一种生产资源成本要素分配系数确定新方法。利用邻域粗糙集理论对影响分配系数的众多属性进行属性约减和重要度计算。对于批量生产的产品,随机属性需要按照一定规则到实例库中抽样分析得到其分布规律,根据概率密度函数与所消耗资源的经验数据将其取值范围划分为小区间并赋予分配系数；对于小批量或者新产品,以粗糙集约减属性集、属性重要度作为检索条件到实例库中检索相近的实例,完成实例修正与保存等操作,得到产品的资源分配系数。通过与原有方法的对比分析验证了该方法的有效性。针对目前投入阶段生产资源成本要素配置不合理的问题,综合考虑了产品成本、质量等核心目标,研究了投入资源成本要素优化控制机理。基于特钢企业ERP／MES／PCS／CMS信息化平台,结合面向成本设计(DFC)的先进设计理念和资源配置功能,构建了特钢产品基于DFC的投入资源成本要素配比优化控制模型,并采用蚁群算法对特钢产品的配比进行了优化设计,实现了投入阶段生产资源成本要素的实时控制和合理配置,并给出了企业应用的计算实例。将生产资源成本要素配置方法与信息技术集成运用,提出了基于生产资源成本要素配置方法的成本管理系统的体系结构,并结合案例企业成本管理现状,给出了面向对象的系统分析、设计方法。该方法及其信息支持系统的研究和实际应用,有利于推动特钢成本管理理论与实践的发展,并在一定程度上完善了作业成本法的相关理论,对企业提高成本信息决策相关性、合理配置生产资源具有现实意义。更多还原

【Abstract】 The main function of cost management is to achieve the reasonable configuration of cost factor of the production resources (CFPR). On the basis of analyzing the research and application status of CFPR configuration in the special steel enterprises of both domestic and oversea, under the background of the application in Dalian Steel Plant of Dongbei Special Steel Group Co. Ltd., the management thoughts of system, integration, optimization and control are applied to study the configuration methods system of CFPR for special steel enterprise from preparation stage and production stage. The concrete researches include the following.In order to solve the improper selection and inefficiency of allocation modes in the process of allotting CFPR, a self-learning selection model of allocation modes of CFPR based on Support Vector Machine is proposed through analyzing features of allocation modes selection in special steel enterprise. The mapping among cost items, activities and allocation modes is established, and CFPR allocation modes automatically match is achieved in special steel enterprise. Rules of choosing modes are extracted from the sample sets by using decision tree algorithm, which makes the implicit knowledge explicit and makes it easy for people to understand and operate. Finally, the realization process is showed through an application example.To fulfill the requirements of nonlinear and complex operating conditions, a new determined method of resource allocation coefficient is proposed. Firstly, neighborhood rough set theory is applied to reduce the numerous attributes and the weighting value of each attribute is calculated. For the mass product, the random features need to sampling analysis in case library following a certain rules and its regularities of distribution is figured out. The domain is divided into short intervals according to probability density function and empirical data of resources consumed. A new allocation coefficient for every short interval is obtained in the same time. For the small-batch or new product, resource allocation coefficient of the products are obtained by using important attributes and its weighting values to match similar cases in case library, adapting a new case with similar cases, and saving to case library. Moreover, the effectiveness of the proposed method is demonstrated through comparison with original method.With a systematic consideration to factors of cost and quality, the optimal control mechanism that could enhance the degree of rational configuration of CFPR in the preparation stage is proposed. Integration with advanced design idea and resource configuration function of Design For Cost, the optimal control model of CFPR recipe is established based on ERP/MES/PCS/CMS information environment in the special steel enterprise. The recipe of special steel product is optimized based on Ant Colony algorithm in order to achieve real-time control and the rational configuration of CFPR in the preparation stage. Finally, the computing example is showed through an application example.The integrated use of configuration methods of CFPR and information technology, the Cost Management System architecture is proposed based on the configuration methods of CFPR. The object-oriented system analysis and system design method combined with cost management situation of case enterprise are given. The research and practical application of configuration methods of CFPR and its information support system are in favor of promoting cost management theory and practice in special steel enterprise, and improving Activity Based Costing to some extent. At the same time, it has practical significance to improve the enterprise decision-making relevance and allocate production resources felicitously.更多还原