Distributed Decision Making Based Risk Management For Virtual Enterprise

With the increasingly fierce competition in the market, the virtual enterprises are becoming a realistic choice of enterprises. As a new business mode, the virtual enterprises bring tremendous opportunities to improve the competitiveness of enterprises to a certain degree. However, without enough practical and proper theoretical experience, the failure probability of virtual enterprises is high. The virtual enterprises have certain advantages, but the complexity of their organization makes the uncertainties increasing, the risk analysis control is difficult and the risk management is complex. In order to achieve smooth operation and expected profits, the virtual enterprises must succeed to avoid the risks. Therefore, the risk management for virtual enterprises is significant in theoretical and practical research.This paper first reviews the present researchs on virtual enterprise, risk management for virtual enterprise and distributed decision making, and then analyses the relationship among virtual enterprise partners and discusses the possible relationship among virtual enterprise partners. The relationship among the members can direct impact on the management of virtual enterprises. Targeting the risk management problems, the virtual enterprises should adopt appropriate strategies for different relationships. While the advantages of distributed decision making lie in solving the problem with distributed characteristics, and providing unified model framework to solve different problems. The distributed decision making thus can visually describe the relationship among virtual enterprise partners and fully reflect the composition of virtual enterprises as well as its operation features. It can also show how risks change in operation. Thus, compared with centralized decision making, distributed decision making can control the risk more effectively. The paper proposes distributed decision making methods to control their risks and carries out following research:(1) An overview of virtual enterprise risk management and distributed decision making related theories is proposed. Firstly, the virtual enterprise risk management objectives, content and management processes are described, the virtual enterprise risks and corresponding induced factors are summarized, the virtual enterprise risk assessment methods and the virtual enterprise risk-sharing principle are described, the virtual enterprise risk management strategies and preventive measures are summed up; Secondly, a distributed hierarchy of decision-making system, information status, system classification, coupled equations, forecasting model and solving the model thinking are introduced, and the basic idea of how to apply the distributed decision making theory to solve the above-mentioned virtual enterprise risk management problems is described briefly, the following researches are based on the above theory.(2) For the virtual enterprises that constitute team relationship among members, the risk management problems are discussed based on symmetric and weak asymmetric risk information. For the symmetric risk information, this paper builds a constructional-DDM risk management model for virtual enterprises. In case the risk information is asymmetric, it deals with how to reserve risk control funds based on forecasted risk, and how to take full advantage of the funds to control the risk. Meanwhile, the organizational-DDM risk management models for virtual enterprises are established. To solve these models, the Taboo Search and Genetic and Serial section algorithms are designed, and the performance analysis of algorithms is analyzed based on changing their parameters. The results show that the model and algorithms are effective.(3) As to the virtual enterprises that form forced team relationship, the paper explores the risk management problems in case the risk information is strict asymmetry, and establishes a corresponding organizational-DDM risk management model for virtual enterprises. Meanwhile, designs GLS algorithm to solve the model. The result shows that the model and algorithm are effective to control the risk.(4) Aiming at the virtual enterprises that constitute non-team relationship, the paper gives discussion on the risk management problems based on symmetric and asymmetric risk information respectively. Seeing that the risk information among each member is symmetric, it discusses the characteristics of risk management, establishes a coordinated-DDM risk management model, analyzes the existence of solution, and compares the different coordination principles. If the risk information among members is asymmetric, the paper sets out several possibilities of risk management, including hidden action, hidden information, hidden action and information agency problems, establishes principal-agent DDM risk management models, and designs related enumeration algorithms and the methods of combining the LINGO software to solve these models. In addition, it proves the distributed decision making for solving risk management problems of virtual enterprises to be effective when comparing with centralized decision making.(5) The paper describes web-based multi-agent system for risk management of virtual enterprises, and analyzes the necessity and advantages of the web-based risk management system, describes the advantages of multi-agent technology applied in risk management system, and gives an account of the key technology and design process of system development and the multi-agent framework of the system.