Study. To Oriented Hwme The Risk Decision

Professor Qian Xuesen proposed the Hall for Workshop of Metasynthetic Engineering (HWME), which is a new methodology to solve complex problems related to the Open Complex Giant System (OCGS). Decision making is a process that chooses the best program from some alternative programs after analysis and comparison. With the rapid development of the society, there are more and more complex problems. Analyzing the phenomenon, recognizing the essence and choosing the suitable solution are the main tasks in Decision-making research.Firstly, with the guidance of the HWME theory, a well-developed indexes system of risk evaluation for new cars research and development is proposed after the analysis of the flow and the risk of the new cars research and development. Then the indexes, which are contained in the indexes system of risk evaluation, are expounded in detail.Secondly, the methods which are used in the traditional exposure rating are studied in this paper. With the directions of these methods, two methods used in exposure rating are raised. The first method is raised based on the idea of risk- influence tree. The second method is raised based on the idea of BP-ANN.Thirdly, with the research of risk multi-attribute decision making, the method, which is used to get the attribute value is proposed if the distillation function of the attribute is unknown. When the distillation functions of the same attribute are dissimilar in different programs, the method of standardizing this kind of attribute is proposed in this paper. Based on the idea of deviation, the method of comprehensive granting the attribute weight is brought up in this paper. At last, the method—RM-GrayTOPSIS, which is based on TOPSIS and grey correlation analysis(GCA), is brought in to solve the risky decision making problems. The model of this method is also proposed in this paper.At last, with the background of new cars research, the application of the methods and models which are summarized in this paper is studied. These methods and models are proved effectual and banausic.