Multivariate grey disaster prediction model

Abstract

Grey catastrophe prediction can predict the date of catastrophe, so that people can take precautions in advance to reduce losses. For example, the use of catastrophe forecasting methods can predict the occurrence of disasters such as droughts, floods, landslides, typhoons and coal mine accidents, which is conducive to the government, relevant departments and the public to take corresponding actions to reduce the adverse effects of disasters on the society. The catastrophe prediction model is mainly to study the catastrophe date sequence and to find its regularity, and through the establishment of GM (1,1) model to predict the date of the subsequent catastrophe. As we can see, the occurrence of catastrophe is caused by the influence of multiple factors. Therefore, it is not enough to take only one factor into consideration when forecasting, combining it with other factors is necessary. Multivariate grey prediction model MGM(1, m) can describe the variables from the dimension of system, and can better reflect the relationship between the variables in the system, so as to predict multiple variables. Therefore, the study combines multivariate grey prediction with catastrophe prediction to propose a novel multivariate grey disaster prediction model. The new model comprehensively considers the relationship between the catastrophic dates of variables in the catastrophic system, avoids the shortcomings of the traditional single factor grey disaster prediction and predicts the date of the next catastrophe from the dimension of system. Multivariate grey disaster prediction model is a combination of catastrophe prediction model and multivariate grey prediction model. First, find n catastrophe dates which belong to the first variable. Then, based on the n catastrophe dates of the first variable, find the n catastrophe dates of other variables respectively and get the catastrophe date matrix. Finally, the prediction can be achieved by establishing the MGM (1, m) model for the catastrophic date matrix. In order to compare the simulation accuracy and prediction accuracy of the traditional grey catastrophe prediction model with the multivariate grey disaster prediction model, we use an example to simulate and analyze. It can be seen from the result that the multivariate grey disaster prediction model has higher simulation accuracy and prediction accuracy than the traditional grey disaster prediction model. The research indicates that the multivariate grey disaster prediction model complements and improves the grey prediction theory, improving the precision of simulation and prediction. It is worth promoting to other similar multi-variable disaster system.