Research On Emergency Decision-Making For Unconventional Disasters And Accidents Based On Scenario-response

With the rapid development of economy and the widespread use of new energy and new process, unconventional accidents and disasters, such as hazardous chemical leakage accidents and earthquakes, present the unconventional features in recent years. Usually, unconventional accidents and disasters have the distinct characteristics of low probability of occurrence, complex evolution process and unclear evolvement path. When such unconventional emergency events occur, they lead to unexpected catastrophic consequences. In view of the unique characteristics of unconventional accidents and disasters, emergency decision-making technology and methodology play a crucial role in reduction and mitigation of their effects. In this paper, a key emergency decision-making technology based “scenario-response” is developed to solve unconventional emergency decision-making problems. In this paper, some typical domestic and overseas cases are first studied to explore the occurrence and evolvement rule of unconventional accidents and disasters. Then some new decision-making theories and methods, such as dynamic bayesian networks, case-based reasoning, and cloud theory, are utilized to specified decision-making problems. Finally, a practical unconventional emergency decision case is presented. The main contents can be summarized as follows:(1) Construction of dynamic bayesian networks-based scenario inference model of unconventional accidents and disasters. On the basis of analyzing the characteristics and evolution rule of unconventional accidents and disasters, their evolution mechanism and paths are discussed. Considering the causal relationship among the scenario elements of accidents or disasters, the scenario network structure diagram is developed. Methods of scenario analysis and bayesian network are used to forecast and estimate the development and evolution trend of scenarios. Finally, the scenario inference model of unconventional accidents and disasters is constructed to obtain the synthetic scenario evolution map. Emergency decision-makers could make decision based on the current state and evolution trend of the accident, then adjust and deploy emergency forces according to the possible evolution path of accident. The model could be served as the basis for emergency decision-makers to make scientific decisions with limited time constraint.(2) Formulation of emergency decision-making scheme generation methodology based on knowledge element and case-based reasoning. In this study, the scenario inference result is considered as disposal goal, and the scenario representation methodology based on knowledge element is first presented through analyzing the characteristics element of unconventional accidents and disasters’ scenarios. Then their basic scenario information is extracted to develop the structured representation method of cases. Finally, a scenario matching algorithm based on structural similarity and attribute similarity is designed, and the inference decision-making process is implemented to generate emergency decision-making schemes of unconventional accidents and disasters. The methodology could obtain alternative emergency schemes, and provide feasible reference for emergency decision.(3) Development of cloud-based emergency decision-making scheme optimization model. In view of the fuzzy and uncertain characteristics of scheme optimization, cloud model which could achieve the conversion between qualitative and quantitative variables is applied to the emergency decision-making, and the corresponding arithmetic is presented in the study. Under circumstance of coexistence of fuzziness and randomness, the model could handle expert evaluation and decision problems, and could improve decision-making objectivity and effectiveness.(4) The application of the emergency decision-making technology and theory in practical case. For verification purpose, Da Lian“7.16” Oil depot fire and explosion accident is presented. Experimental results obtained demonstrate the proposed emergency decision method under the mode of “Scenario-response” can generate emergency decision schemes dynamically according to the evolving accident scenarios, and provide the optimal accident disposal strategy, which could reduce the uncertainty effectively and make emergency disposal of unconventional accidents and disasters more rational.This study presented scenario inference and emergency decision-making scheme optimization methodology for unconventional accidents and disasters, which laid a significant foundation for the rationality and accuracy of disposal of unconventional accidents and disasters. Above research result enriched and innovated further emergency decision-making theory of “Scenario-response”-based unconventional accidents and disasters. It can not only provide new idea and feasible reference for scientific decision of unconventional accidents and disasters, but also provide technical support for the developing the emergency assistant decision-making system.