In the GM (1,1) prediction model, the development coefficient a and grey action b have a direct impact on the prediction accuracy of the model. Based on the analysis of GM modeling principle and the influence of parameters on model accuracy, an elite ACO algorithm with adaptive pheromone concentration adjustment and GM (1,1) fusion prediction model are proposed. Without changing the expression of GM (1,1) model, the improved ACO algorithm is used to solve the optimal parameters of the model. The experimental results show that compared with the traditional GM (1,1) model, the prediction accuracy of the improved ACO algorithm and GM (1,1) fusion model can also get better prediction effect under the condition of large error of the traditional GM model, and has advantages over the traditional model in applicability. It is a new idea to improve the accuracy of the model. It also shows that it is reasonable and scientific to use the adaptive elite strategy to improve the ant colony algorithm and improve the global optimization ability of the algorithm.