Grey multivariable convolution model with new information priority accumulation

The GM(1,N) model with convolution integral appeals considerable interest in recent researches due to its effectiveness in multivariate time series forecasting. However, the failure of incorporation new information priority principle will cause large errors. To improve the simulation and prediction accuracy, GMC(1,N) model with new information priority accumulation is put forward. A parameter is added to adjust the weight of data. By giving a large weight to the new information, the accuracy of the prediction is improved in theoretical. The priority of new GMC(1,N) model is verified through some cases.