Research On The Multi-disaster Integrated Predition & Pre-warning And Decision Support System

The long-established improper emergency management system and some functional disadvantages result in the low overall level of informatization in the disaster management field, intellectualization in particular, which seriously restricts the efficiency of disaster prevention and mitigation. With the development of information technology and its application in the field of disaster management, together with the concept of integrated disater management and integrated disaster reduction and mitigation, disater prediction and information management is no longer the functional area or research scope of a single department or a single disaster, which instead needs multisectoral and multidisciplinary collaboration and the establishment of comprehensive prediction model and unified disaster information management mechanism. However, the complexity of natural disasters determines that it will be very diffcult to achieve its comprehensive prediction, while the diversity of disasters and the extensiveness of disaster bearing bodies makes the information transfer and sharing a very difficult task.This paper leads the natural disaster chain theory into the research project and constructs the multi-disater coupling prediction computer model based on the theroy through the abstraction and decomposition of the physical process of disaster chain combined with the agent and event chain modeling technology, which solves the problem of various isomeric models coupling and integrated prediction. Geared toward the disaster information management and consistence, the research analyses the characteristic and management modes of engineering disaster bearing bodies and puts forward and develops the information model of engineering disaster bearing bodies with digitization method oneness applying the theory and method of information model. Based on this model, the engineering services information management platform supporting mul-disaster dynamic emergency comprehensive management has been developed. Using historical disaster cases and the knowledge reuse technolgoy such as data mining and case-based reasoning, a new modelling theory, technologies and methodologies of disaster prediction is presented and inference mechanism upon cases and integration mechanisms is put forward. Based on the above research results, the project develops MD-PWDSS, a prototype system of multi-disaster comprehensive prediction and early warning and decision support system based on disaster chain. It provides such functions as single disaster prediction, multi-disaster coupling prediction based on disaster chain, information management of engineering disaster bearing bodies, prediction model support upon cases and disaster relief decision support, which can realize varigrained disaster prediction and multi-disaster dynamic emergency comprehensive management. It provides a integrated environment of disaster prediction, information sharing and relief decision. In addition, the open system architecture based on intelligent agent technology provides good conditions for further improvement and expansion of disaster management function. Finally, the feasibility and effectiveness of MD-PWDSS is verified instancing on seismic data of Shantou City, Guangdong Province.The research results show that the multi-disaster coupling prediction model and its system based on disaster chain can solve the problem of comprehensive prediction and emergency management in the process of natural disaster management and will improve the efficiency and quality of management. The research provides new ways of thought and methods for improving the information and intellectualization of natural disaster comprehensive management in our country, which is significant to the theoretical research and practical applications.