Prediction Model And Analysis Of Ground Surface Settlement In Foundation Pit Of Subway Wind Tunnel

Along with the acceleration of urbanization,the urban population has expanded,and the number of motor vehicles has also exploded,which has brought tremendous pressure on ground transportation.The development of Metro traffic has greatly alleviated the traffic pressure,but the key components of Metro traffic is generally located in the bustling area or large business district of the city,where buildings are dense,traffic is busy,and underground pipelines are intricate.Once the foundation pit problems appear in Metro traffic construction process,will have tremendous impact on the surrounding area.Therefore,it is necessary to regularly monitor the foundation pit and its surroundings,and analyze the monitoring data to predict the future settlement trend of the foundation pit to guide the construction and prevent the occurrence of foundation pit accidents.Taking the foundation pit engineering of a certain section of Shanghai Metro Traffic Line as an example,this paper selects the surface monitoring point D4-1 with great research value for analysis and research,and its settlement is-19.20mm.The data after wavelet noise reduction through two different kinds of equally dimensional and unequal interval prediction models to simulate predition.The results show that the model has the highest accuracy in 5 dimensions.Two kinds of equal-dimensional unequal interval prediction models with three kinds of(average,weighted average and geometric mean)weakened buffer operators are modified.The simulation prediction results show that the average weakening buffer operator has the best effect,and its modified 5-dimensional unequal interval gray Verhulst model has the highest accuracy.Experiments were carried out with another settlement data of monitoring point D1-2,which has a large research value.The settlement of this point is-18.97mm.The same method is used to test and shows that the model has the highest accuracy in 5 dimensions.The simulation results show that the 5-dimensional unequal interval gray Verhulst model modified by the average weakening buffer operator has the highest accuracy,and the weakening buffer operator plays a certain role in improving the prediction accuracy.Figure[27]table[36]reference[95]...