Application Of Wavelet Analysis And Combination Model In Settlement Prediction

Along with urban modernization and enlargement, space development has nowdays become one of the mainstreams in city construction. At the same time, large-scale bridge building and water projects have also increased, which brings considerable benefits to our daily life and social development, but on the other hand raises our concerns for the hidden dangers. The significant increment of size and height for high-rise buildings entails high-level stability and safety. In other words, supporting capacity of the foundation needs to be strengthened, in order to guarantee the stability and safety of neighbor buildings and pipelines. It is therefore necessary to monitor the settlement of these high-rise buildings on a regular basis according to the design requirement and precisely estimate their deformation trends, so as to take immediate actions, avoid potential accidents and ensure personal security.Wavelet-based threshold de-noising is utilized in this paper to analyze the monitoring data of building deformation and finally make reasonable predictions. This paper selects typical observation points from a real example, called Celebrity Impression Phase 2. First of all, prediction results were separately calculated based on different models in combination with the observation statistics to identify their strengths and weaknesses. Then, a pretreatment of the original observation data was conducted in line with wavelet-based threshold de-noising approach. Last but not least, the writer of this paper made a result comparison between original data and wavelet de-noised data by using Grey Combination Prediction, Simple Grey Prediction and Time Series Prediction models. It is concluded that results from combined prediction models are superior than those from single prediction models, that is, wavelet de-noised data improves smoothness of the original data, which, to a certain degree, would enhance the prediction accuracy of these models.