关键词: GM（1, 1）模型, 发展系数, 参数辨识, 改进模型

Abstract: The least square method is used to solve the parameters in GM（1, 1） model. When there are outliers in the data, this method will increase the prediction error of the model. From the viewpoint of optimizing parameters, the fourth-order Runge-Kutta method based on Simpson integral formula is used to modify the parameter identification of GM（1, 1） model, and a new improved GM（1, 1） model is proposed to reduce the prediction error of the model. At the same time, the scope of application of the improved model is further analyzed from two cases of different values of development coefficient and different prediction steps. Finally, an example is given to verify the effectiveness of the improved model.

Key words: GM（1, 1） model, development coefficient, parameter identification, improved model