Research on Accident Early Warning Model Based on Entropy Weight and Extension Theory

The occurrence of accident is influenced by multiple factors, so in the process of accident warning model construction, we should take full account of the relationships between all of the factors and the accident and the roles that these factors played in the accident. This paper constructs an accident warning model by employing the entropy weight and extension theory. On one hand, this model employs entropy weight theory to determine the weights of various factors, which will avoid the subjectivity of weights determination and the complexity of computing. On the other hand, it constructs accident warning classic and partial unit matter-elements with applying extension theory and method, and this step can avoid the possible subjectivity in this evaluation model by using comprehensive correlation degree as evaluation criteria. Finally, take coal and gas outburst accident as an example, the validity of the model is verified from the weights of various factors which cause outburst and outburst risk size. This is a new methodology in the process of accident warning model construction.