Research on predicting the growth trend of natural gas reserves will help provide theoretical guidance for natural gas exploration in Sichuan Basin. The growth trend of natural gas reserves in Sichuan Basin is multi-cycle and complex. The multi-cyclic peak is screened by the original multi-cyclic peak judgment standard. Metabolically modified GM(1,3) gray prediction method is used to predict the multi-cycle model parameters. The multi-cycle Hubbert model and Gauss model are used to predict the growth trend of natural gas reserves. The research results show that: (1) The number of cycles of natural gas reserves curve during 1956–2018 is 13. Natural gas reserves will maintain the trend of rapid growth in the short term. (2) Metabolism modified GM(1,3) gray prediction model can improve the accuracy of model prediction. The prediction accuracy of Hubbert model is higher than that of Gauss model. By 2030, the cumulative proven level of natural gas will reach 52.34%. The Sichuan Basin will reach its peak of proven lifetime reserves in the next few years.

天然气储量增长趋势预测研究将为四川盆地天然气勘探提供理论指导。四川盆地天然气储量增长趋势多周期复杂。多循环峰由原始多循环峰判断标准筛选。代谢修正GM(1,3)灰色预测方法用于预测多循环模型参数。多周期Hubbert模型和Gauss模型用于预测天然气储量的增长趋势。研究结果表明：（1）1956-2018年天然气储量曲线循环次数为13次，天然气储量短期内将保持快速增长的趋势。(2) 代谢修正GM(1,3)灰色预测模型可以提高模型预测的准确性。Hubbert 模型的预测精度高于 Gauss 模型。到2030年，天然气累计探明水平将达到52.34%。四川盆地将在未来几年内达到探明终生储量的高峰。