The Risk Measurement And Evaluation Of The Agent Unit Under The Whole Process Agent Construction

Under the traditional construction mode, the construction of the government investing project may cause "three super" and other drawbacks. In order to solve these problems, the agent construction comes up. As a new thing, the relative laws and measures of the agent construction is not perfect and it may bring certain risks to every interest part of the agent construction, especially for the important executive main part, that is the agent unit, the risk magnitude of which is significant. So it is necessary to set the agent construction risks of the agent unit as the research object, recognize and appraise the risk magnitude systematically to help the agent unit to strengthen the sense of risks and the ability to manage them, increase the benefit and social position. Furthermore, it helps to cultivate the agent construction market and it is favorable to the operation of national economy.Firstly, this thesis starts from the relative concept and basic modes of the agent construction, reveals the important role and responsibility of the agent unit in the process of agent construction. Under the background of the whole process agent construction, analyze and sum up every type pf risks systematically and build up the breakdown structure figure of the agent construction risk.Secondly, based on the pairwise comparison, considering the fuzziness of the experts' judgment owing to the insufficient data and other elements in the pairwise comparison of risk indexes, introduce the triangle fuzzy number to deal with this fuzziness, combining with the extent of divergence of experts' opinion and the weight difference to determine the synthetic weight of risk index, which is in conformity with the nature of experts' thought.Thirdly, from the angle of risk severity, risk occurrence possibility and risk control ability to measure the risk magnitude of risk index. Adopt the credibility measurement theory to deal with the risk occurrence possibility to avoid the defection that the probability measurement may need much high-quality data. The risk severity considers the influence of experts' risk preference to the severity extent and it may suit the actual circumstance of judgment. Finally, based on the fuzzy inference, output the control rules to acquire the fuzzy number of the risk magnitude of risk index. After the process of defuzzification, combining with the weight of risk index, acquire the entire risk level of the agent construction. Then clarify the application process through an example and make this method operationally.The improved method is suitable for the agent construction risk identification and appraisal by phase and makes up the shortages of traditional risk evaluation method pertaining with the determination of risk index, the comprehensiveness of information treatment and the prediction of risk occurrence possibility. This improves the correctness and reliability of the evaluation. In the application, it only needs to make some modifications of the breakdown structure, and this can be used as the general thread of the agent construction risk measurement and appraisal.