Emergency Decision-making Based On The Improved Group Decision-making Method

In recent years, along with frequent emergencies, related departments have greatly increased the attention to the problem of emergency management. According to the properties of complexity, polytropy and urgency, emergency events often involves various factors with complicated relationships. If we can make accurate decisions timely according to the development of emergency, loss could be reduced and the situation could be brought under control.With the rapid development of the society, economy and science, the decision environment to decision makers has become more and more complex. Collective wisdom plays an important role in making accurate decisions. Group decision is to let professionals from different department and specialty to participate in decision-making process. Based on the shortcomings of traditional group decision making method including the blindness of making decisons, the accuracy of historical decisions and the probability of collective mistake from choosing experts, three recommendations for improvement are put forward:(1) The traditional voting rule of group decision is adjusted;(2) The decision-making process is improved;(3) The event type of making decisons is taken into account. We can make full use of the information after implemention of scheme and extract the accuracy rate of the experts’ historical decision, which could be considerd as the feedback weight to adjust the objective weights of decision-makers. Then an emergency decision model could be built based on the improved group decision-making method. This article utilized the proposed algorithm into the choice and evaluation of emergency plans from the wind power plant. Comparing the expert weight, program score and ranking of alternative schemes and explaining the reasons for differences, the necessity of using the accuracy rate of the experts’ historical decision as a factor for adjusting expert weight was demonstrated. Then more valuable reference information for group decision research could be provided.