Land Subsidence Prediction System and Method Based on Support Vector Machine

Abstract

The invention discloses a ground subsidence prediction system and method based on a support vector machine, which is mainly used to solve the analysis requirement of massive data in the process of shield tunneling that the matching system cannot cope with. The system includes a Hadoop big data analysis platform, data collection module, data preprocessing module, model building module, and settlement prediction module. In the prediction method, a clustering based on confidence interval estimation and Raida criterion is proposed for preprocessing denoising The algorithm detects abnormal points; when building a prediction model, the iterative calculation process of the weight vector is updated with the average value of N partitions. This strategy is combined with stochastic gradient descent to improve the training speed. The present invention has massive data storage and analysis capabilities and high-performance redundancy capabilities. The data analysis is real-time and efficient, and it is used for the prediction of ground subsidence in the shield construction process, providing reference and information for engineering managers and construction operators. Adjustment basis.

Description

technical field [0001] The invention belongs to the technical field of industrial big data, and in particular relates to ground subsidence prediction in the field of subway shield tunneling construction, specifically a support vector machine-based ground subsidence prediction system and method, which can be used to predict the amount of ground subsidence during shield tunneling construction . Background technique [0002] Urbanization is an important part of my country's infrastructure. In recent years, with the development of the economy, the process of urbanization in my country has been accelerating, the development and utilization of urban underground space has developed rapidly, and a large number of tunnel engineering projects have been produced. Underground engineering represented by the shield construction method has become the mainstay of urban underground construction. Way. [0003] The so-called ground subsidence prediction refers to the prediction of the degree ...

AI Technical Summary

Problems solved by technology

The traditional land subsidence analysis method is to train the model in stand-alone and serial mode. When faced with massive data, problems such as slow speed, low efficiency and poor fault tolerance are exposed, which directly affect the prediction of land subsidence and bring construction risks.

Images