Abstract: The process management of emergency tasks is a necessary requirement in disposal operation of emergency plan on the computer network platform. Taking the advantage of modular management of business process, in this paper Unified Modeling Language (UML) is introduced to describing the dependence relationships between activities in an emergency plan, and mapped to a design structure matrix (DSM). The four points scale rule is used to measure the degree of information dependence relationships of the flow dependence, as well as the dependences of the role, the equipment and the facilities in the resource dependence between the activities. The indictor of the dependence volume and its calculation method is proposed. Then, clustering optimization model is carried out, in which the minimizing of the total dependence volume between different modules and maximizing of the total dependence volume inner all modules are token as the objectives.Acoordingly, a genetic algorithm is designed to solve this problem. At last, a case study is analyzed, and the calculation result proves that the proposed method is validity and effectively. This method could bitterly reduce the complex degree of organization management, and creat the conditions for fast restructuring and integrating functions of module tasks.

Key words: emergency plan, modularization of process management, cluster analysis, design structure matrix (DSM), genetic algorithms