为提高埋地管道剩余寿命预测的精确度，对传统灰色马尔科夫预测模型进行优化。将管道腐蚀速率视为一个灰色系统，对灰色模型的原始数据光滑处理后建立等维新信息无偏灰色模型，预测腐蚀速率的宏观值。以优化的灰色模型预测值残差为基础结合马尔科夫链模型，进行二次平滑处理和白化系数寻优，得出残差修正值。最终，结合两种优化模型得出管道腐蚀速率的预测值。实例检验证明，该模型能有效克服系统长期动态预测上的不足，与传统灰色马尔科夫链预测模型相比，预测精度提高了40.33%，预测结果与实测值有更高的拟合程度。

In order to improve the accuracy of the residual life prediction of buried pipelines, the traditional grey Markov prediction model was optimized. The corrosion rate of pipeline was regarded as a gray system, after smooth processing of the original data of the gray model, an unbiased gray model of iso-dimensional new information was established, and the macroscopic value of corrosion rate was predicted. Based on the residual of the predicted grey model predictive value, combined with the Markov chain model, the second smoothing process and whitening coefficient optimization were performed, and the residual correction value was obtained. Finally, a combination of two optimization models yielded a predicted value for the corrosion rate of the pipeline. The example test proved that the model could effectively overcome the shortcomings of long-term dynamic prediction of the system. Compared with the traditional grey Markov chain prediction model, the prediction accuracy was improved by 40.33%, and the prediction result had a higher degree of fitting with the measured value.