遗传神经网络在地面沉降趋势预测中的应用

Application of Genetic Neural Network in Land Subsidence Prediction

摘要:

受地下流体资源和固体矿物开采、地面动静荷载、地下空间开发等人为因素的影响，区域性地面沉降问题越来越突出；构建区域地面沉降预测预报模型，及时发现潜在隐患并拟定合理的防治措施对保障经济社会与生态环境可持续发展方面具有重要意义。该文基于遗传算法的神经网络在地面沉降趋势预测中应用，对几种算法从训练速度、拟合水平以及预测能力方面作了比较分析。

Abstract:

Accompanying with the development of social economy, due to the effects of exploitation of underground fluid and solid mineral resources, static and mobile ground load, the development of underground space and other factors, regional land subsidence problems have becoming more and more serious. It has an important significance to build a regional land subsidence prediction model to detect potential pitfalls, and formulate reasonable control measures to ensure sustainable development of social economic and ecological environment. In this paper, based on application of neural network by using genetic algorithm in ground settlement prediction, a comparative analysis of several algorithms from training speed, fitting and predictive ability has been carried out.