<p>The precise prediction of carbon emission trading prices is of utmost importance as it contributes to developing a stable and scientifically robust carbon pricing mechanism. However, it's challenging to predict carbon prices accurately due to the inherent randomness and instability associated with time series data. In this study, we propose a novel approach for forecasting the time series of carbon emission trading prices based on ensemble AI-driven intelligent optimization method. Initially, our model employs the Ensemble Empirical Mode Decomposition (EEMD) technique to decompose the carbon price series into Intrinsic Mode Functions (IMFs). Subsequently, each IMF is individually forecasted using the Support Vector Regression (SVR). Finally, the obtained forecasts are combined to generate the final prediction. We develop an intelligent algorithm based on Differential Evolution (DE) and Grey Wolf Optimizer (GWO) to optimize the hyperparameters of SVR model. To validate the effectiveness of the proposed model and algorithm, three sets of experiments are conducted. In terms of prediction accuracy, the model significantly improves the prediction accuracy, with a Mean Absolute Error (MAE) of 0.0924 and a Mean Squared Error (MSE) of 0.0158 on the shanghai carbon emission trading dataset. The results have revealed the proposed forecasting model demonstrates stable and accurate predictions, presenting advantages over alternative models.</p>