基于神经网络的建筑物沉降原因分析

在我们应用回归方法对建筑物沉降原因分析及沉降趋势的预测中，由于实际情况的复杂性及主观认识的局限性，这样所得的结果含有较多的人为因素，可能会与实际情况有所差异。针对这种情况，讨论了应用神经网络方法来发现和验证引起建筑物沉降的因素及对沉降趋势的预测。实例表明，该方法能取得较好的效果。

Reason Analysis of Building Sedimentation Based on Neural Networks

During the tendency forecast and the reason analysis of the building sedimentation with regression method, the result may be include artificial factors and is different from the reality due to the complicacy of the reality and the limitation of the subjectivity knowledge. To deal with this, the author discusses how to find and verify the factors leading building sedimentation and how to forecast the sedimentation tendency with neural networks. Demonstrated by the practice of one project, good effects can be received with this method.