Main Article Content
Given the fact that vicious internal competition and the epidemic have led to a decline in operating income, how to improve the total operating income of container port group has become an important issue for decision makers of container port group. In order to increase the total operating revenue of port group and suppress the vicious price competition within the port group, a second-order forecasting game model is proposed, which is based on Improved GM (1,1) model, logit model and Bertrand Nash equilibrium model. Firstly, the GM (1,1) model is improved by using the three-point smoothing method to predict the future container throughput; secondly, the logit model is used to simulate the choice behavior of container freighter to container port according to the container port utility; finally, the non-cooperative game and cooperative game revenue of container port group are calculated by using the Bertrand Nash equilibrium model. The results show that the prediction accuracy of container throughput is 0.68% higher than before, and the total operating revenue of cooperative game and non-cooperative game is 52.3% and 35.5% higher than the original revenue, respectively. The research shows that the second-order prediction game model can make a more accurate prediction of the container throughput, and can effectively improve the operating income of the container port group, which lays a theoretical foundation for the managers of the port group to formulate a reasonable freight rate.