<p>Carbon emission trading system is one of the important means for China to tackle climate warming and achieve the dual-carbon goals. However, due to the characteristics of the carbon emission trading prices, such as nonlinearity, instability, and complexity of predictor variables, the performance of existing prediction models is relatively poor. In order to solve the above problems, a prediction model considering the predictor variables of carbon emission trading prices is proposed in this paper. The proposed method first analyzes 18 predictor variables in six categories and Lasso is used to select predictor variables as inputs for the prediction model. Secondly, sliding time window technique is adopted to analyze the dynamics and trends in the carbon market. Then, differential evolution (DE) algorithm is utilized to improve the performance of support vector regression (SVR), where SVR is used as the prediction model for carbon emission trading prices. Finally, the carbon emission trading price is taken as a sample set in Hubei Province, together with the selected predictor variables, as input to the prediction model. The experimental results show that the proposed model obtains optimal results in three performance evaluation metrics, including root mean square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE). Compared with the basic model, the results decreased by 1.6081, 1.4753 and 5.5418% respectively, which improved the prediction accuracy of the model. And the proposed model has also achieved better performance in predicting the prices of China’s carbon emissions trading market.</p>