摘要
工程质量问题具有隐蔽性及突发性特点,已有的质量管理研究往往从管理者的视角探讨质量管理思想或方法,鲜有从工程项目自身质量特征入手,将工程项目视为有机系统进行研究。受生物免疫系统的启发,本文提出工程质量缺陷免疫系统概念,以期实现工程质量预警和缺陷原因识别。首先,构建了工程质量缺陷免疫体系并论述了各部分的运行机理。然后,研究了具有三层质量缺陷预警与防御机制的免疫算法。通过规则识别、适应性识别和协同识别三层次算法,实现质量缺陷免疫体系的运行。同时,将MOBP神经网络算法引入协同识别模式,实现了质量缺陷预警与质量缺陷原因识别。该质量缺陷免疫体系能够通过学习完成抗体池的进化,提高工程质量系统的免疫能力。最后,运用Matlab编写程序进行算例计算,所得结果展示了质量缺陷免疫系统对质量管理实践的有效性。
The project quality issue is characterized by concealment and abruptness. Poor quality can trigger a chain reaction. Quality defects, poor management, and other issues still exist in the engineering practice. Therefore, the project quality of engineering remains a hot and difficult issue in theoretical and practical studies. We believe that the quality immune system should conclude the functions of immune identification, immune system evolutionary, and early warning. Project quality is a complex system which is affected by many factors. Some mutations of the quality will appear and make a difference when the internal and external environmental of the quality system change. The previous studies often discussed quality management theories or techniques from the perspective of superintendent, not from the essential characteristics of quality defects or based on the complex organic system. If we regard the project as an engineering system which can quickly find out quality threats and clear them, we may propose a new intelligent management. Based on the ideas mentioned above, we explore a quality immune system for quickly discovering and analyzing quality issues in order to maintain the stability of project quality. First of all, we build up the engineering quality defects immune system and discuss its functions by mapping the biological immune system to the quality immune system. Furthermore, the running mechanism and the basic application process of quality immune systems are claimed.Then, the three-layer quality immune algorithm is studied for defects, early-warnings, and defense mechanisms. The organism immune levels, such as Phagocytes identification mode, Lymphocyte identification mode, and T Helper identification mode, are applied to a quality immune system. The system can help form the rule identification, adaptive identification, and synergetic identification. These three layers of quality immune algorithm constitute the quality immune system which can identify the quality defects and give the early-warning. Rules identification model is to set a warning threshold value of quality according to quality specification or special quality requirements. When the date exceeds the threshold value, the quality immune system will start warning. Adaptability identification model refers to an early warning method of the quality that takes the occurrence time and frequency of the quality defects into consideration, which could achieve the targeted early warning in different project phases effectively. This study focuses on the prevention and control of high frequency quality accidents. The synergy identification model is used to further recognize the reasons of the quality defects based on the early warnings in order to implement the intelligent management of the quality. Meanwhile, we introduce the MOBP neural network algorithm into the synergetic identification layer to identify the causes of quality defects. The quality defect date with the characteristics of the project is entered to train the neural network. The trained neural network can accurately identify the quality defects when the new quality data（antigen） is produced. The synergetic identification is able to realize the evolution of antibody pool to keep the system stable. When the neural network cannot judge or identify the antigen when a new one appears, it is necessary to confirm it manually. The confirmed data will be added to the antibody pool as a new antibody. Finally, the application method of quality immune system is elaborated in order to guide the practices of the engineering quality management. We use a case study to prove the effectiveness of this method with the program development using the MATLAB software. The results provide recommendations on how to carry on the early warning and identify the reasons of the defects.
出处
《管理工程学报》
CSSCI
北大核心
2015年第4期205-212,共8页
Journal of Industrial Engineering and Engineering Management
基金
国家自然科学基金资助项目(71271085)
北京社科"十二五"规划项目(12JGB044)
关键词
工程质量
免疫系统
质量缺陷预警
project quality
immune system
quality defects warning
作者简介
乌云娜（1956-），女，蒙古族，吉林省前郭旗人，华北电力大学经济与管理学院教授，博士生导师，研究方向：工程与项目管理。
