Study On The Emergency Decision-making Of Emergency Environmental Accidents Based On CBR

Emergency environmental accidents is characterized by high uncertainty,complicated evolution model and severe destruction,which exerts great pressure and difficulties on emergency work of enterprises and government departments.How to improve the quality of emergency decision is the focus of current emergency works.The previous cases in the emergency decision possess important reference value,which are the descriptions of emergency incidents and disposal measures and the essential basis of emergency decision.As the technology increasingly advances,emergency is becoming more and more complex,and it also requires higher standard of emergency decision.At present,the defects of domestic emergency cases is featured by lower utilization of informatization,decentralized expression structure and incomplete contents.With the accumulation of cases and the increase of storage difficulty,it is very difficult for decision makers to find appropriate reference cases in a very short period of time.The environmental emergency is urgent,uncertain and destructive,which requires decision-making body must respond in time and immediately.However,the conventional emergency decision theory and technology can hardly meet the requirements of dynamics.With analysis of the unpredictability and non-structural features of environmental emergency,the thesis adopt the situation-oriented disposal method to make the emergency decision.Emergency decision is one of the core contents and key links of emergency management,which depends the efficiency of emergency management.Environmental emergency decision-making not only faces the difficulties of shorter time,changeable environment and insufficient information but also needs to give full consideration to the emerging,developing and dying of emergency as well as the uncertainty of resource consumption and implementation effect in meeting any emergencies.The thesis focus on the exploration of a more scientific and efficient decision-making method by grasping the key factors that influence the emergency decision making by scenario analysis.In order to enhance the usability of previous experience,this thesis,based on the analysis of the features of environmental emergency,emergency decision mechanism and demand analysis,combines the scenario analysis method,case-based reasoning technology and system dynamics theory.It analyzes the scenario structured cases,scenario interference,scenario similarity matching and methods for finding suitable solutions from emergency cases.The thesis presents a situational reasoning method for environmental emergencies?Through analyzing the evolution mechanism of environmental emergency from different dimensions according to the methods of disaster science and system dynamics,the thesis concludes the rule of incidents development mechanism.In accordance with the decision-making process of the emergency body,it analyzes the mechanism of emergency decision-making and presents the key points of emergency decisions.On the basis of in-depth analysis of the composition features of the situation,this thesis proposes the extraction method of elements and constructs the spatial model of the situation through the multi-dimensional situational space method.By using ontology to create knowledge of top-level domain tasks and other aspects,the information standard in the same field is unified,and information sharing and integration is more convenient.According to the characteristics of scenario factors,this thesis proposes a multiattribute scenario matching algorithm which is applicable to emergency environmental events,and improves the retrieval efficiency.By setting the threshold value in advance,the utility value of the emergency effect of each retrieved case can be calculated,and the second screening of the source case can be completed which improves the quality of case reasoning.Finally,the decision model presented in this thesis is verified by examples of environmental emergencies.The application of the emergency environmental accidents showed that the method not only can establish the relationship between the instance and the ontology model quickly,but also improve reliability and achieve knowledge sharing.Furthermore,the method can realize the multi-dimensional expression of the scene and meet the "scene-response" emergency decision requirements.