Deformation Monitoring Data Processing Study Of The Foundation Pit

Along with the advance of modern civilization and rapid development of economy, emerging tall and super-tall building projects begin to popularize. The high-rise buildings will produce distortion influenced by various kinds of factors. If the distortion exceeded certain limit, which can affect the building itself and the user’s safety. Therefore, monitoring the engineering structures within the deformation influenced range timely, and accurately deal with the analysis of monitoring data is very important.This paper firstly introduces how to draw up a construction about deep foundation pit monitoring scheme reasonably,main contents include compilation principles, compilation basis,monitoring methods and setting monitoring alarm value. Considering the uncertainty of net shape change, and depending on the difference of reference frame, we can choose three different adjustment methods dealing with monitoring network data, which is respectively classical free adjustment and rank defect free network adjustment and quasi-stable adjustment.Treatment analyze two phase foundation pit monitoring network with these three methods, the results found that in the case of coordi nates for the parameter, the observed value corrections are the same.To determine network reference frame, we firstly need to analyze the fixed point and fixed point in the net, this paper introduces the mean spacing method of hypothetical test to inspect overall and analyze the stability. of network point. For the gross error in monitoring data,this paper uses the theory of robust estimation, properly adjust the equivalent weight factor. With resistance, such as residual method of absolute minimum and estimation, Denmark method converges faster and with less number of iterations required in the process of adjustment. So, Denmark after correction factor regression method of estimation, both to keep the original poor resistance to estimate the method features and functions, and further improve the detection of gross error. with calculating monitoring network data.It is proved that Denmark robust estimation can effectively resist the influence of gross error.For traditional data processing model, mostly static model is established for simulating data. Due to the complex diversity of deformation mechanism,we need to introduce changes factors such as geology, temperature, mechanics, hydrology, the traditional static model have been unable to meet the actual demand. The kalman filter is a good method to solve this difficult problem. This method combines model data forecast with parameter estimation and understands target state timely. But in practice it also has some limitations, such as determining the state equation and observation equation is difficult. Adaptive kalman filtering method is hand in hand with the recursive calculation of filtering noise and relevant parameters of the initial model correction, in order to reduce the influence of model error, make the filtering result in more accordance with the actual results.This paper introduces the basic theory of kalman filter system, analysis of how to determine the model of a series of initial value, and focus on using the adaptive kalman filtering theory to establish a monitoring model. Analysis of engineering example of a foundation pit, using matlab standard kalman filtering and adaptive kalman filtering calculation program, calculation and analysis of multiphase deformation monitoring data, and comparing wi th monitoring of the actual value, to calculate the result of the adaptive kalman filtering method of data processing is better, can meet the needs of the project.