隧道长期不均匀沉降预测的蚁群算法

利用上海地区地铁盾构隧道运营期的大量实测沉降资料,找出其长期不均匀沉降规律与蚁群算法思想的对应关系,提出了一种基于蚁群算法的隧道长期不均匀沉降的新预测模型.为了检验新模型的预测效果,运用预测与实测数据对比分析的方式,得出新预测模型的构建及完善思路.研究结果表明,该预测模型利用实测资料直接建模,避免了传统方法计算过程中各种人为因素的干扰,预测精度高,简便易行,适应性强,可供完善的空间也很广,可为今后隧道长期不均匀沉降的预测计算提供一种全新的思路.

Ant Colony Algorithms of Long term Uneven Settlement Prediction in Tunnel

The corresponding relationship between long term uneven settlement law of Shanghai Metro shield tunnel in operation period and the idea of ant colony algorithms is found based on the measured settlement data.Then a new prediction model of long term uneven settlement in tunnel is put forward on the basis of colony algorithms.The forecasting effect of the new model is tested.A comparative study is made of predicted data with the measured data,and the idea of establishing and perfecting the new model is acquired.The results show the prediction model is directly established on real settlement data,which avoids the mistakes due to factitiousness in traditional methods.The established prediction model proves to be accurate,easy to operate and adaptable.Besides,the model can be improved in many fields,which fact is considered as a brand new way to long term uneven settlement prediction calculation of tunnel.