It is difficult for the commander and decision-maker to obtain the related information from complicated battlefield situational elements for judging the trend of battlefield.Aiming at this problem,we regarded the battlefield situation as a grey system,used the grey system theory to classify the acquisition of situational elements,and analyzed the degree of correlation between the elements to provide the basis for the commander to make decision.The data of battlefield situational elements was predicted by establishing a combined model of GM(1,1)-RBF model and DGM(2,1)-RBF models.To increase the prediction performance, the improved algorithm was used to optimize the single model of the combined models.Simulation result shows that the optimized combined models has much higher prediction accuracy and efficiency.