The Applied Research On Adaptive Kalman Filter In Data Processing Of Deformation Monitoring

There are some systematic errors in the process of deformationmonitoring, To be more accurate representation of the deformation of thedeformable body, we put forward a higher demand for field monitoringdata, therefore, deformation monitoring data processing is a deformationmonitoring process an important part.At present, for the deformation monitoring data processing method.There are Regression analysis method, Time series analysis method, GreySystem method, Kalman filter method and so on. The Kalman filtermethod is widely used method of data processing, it can not only reducethe measurement error, but also to predict the next deformation trend, theKalman filter is a dynamic data processing method, it does not requirelarge amounts of data, only today observational data can be achieved the filtering and prediction. Traditional Kalman filter is prone to divergenceto handle a large number of observational data, so the model is modifiedto maintain smooth data, this is the adaptive Kalman filter.At present, the commonly adaptive Kalman filter method is thevariance compensation adaptive Kalman filtering, the maximum aposterior estimate adaptive Kalman filter and the estimating variancecomponent adaptive Kalman filter, they are able to eliminate the randomdisturbance error and get close to the real information, they have manyincomparable advantages with traditional adjustment methods, especiallyfor the large amount of data, but the effect of the different adaptivemethods is not the same, this article aims to investigate the optimallyadaptive Kalman filtering method. for the deformation monitoring dataprocessing.Based on lots of the documents and materials, firstly this articledescribes the purpose and significance of the deformation monitoring,the methods and significance of the monitoring technology of dataprocessing, Secondly, the research expound the basic principles ofKalman filtering systematically, and research the theory, methods andmodels of the Variance-compensation adaptive Kalman filtering, themaximum a posterior (MAP) estimation adaptive Kalman filtering andthe estimating variance components adaptive Kalman filtering; Finally,based on the deformation monitoring data of the Taiyuan Wanda Plaza Project, using the general Kalman filtering, Variance-compensationadaptive Kalman filtering, maximum a posterior (MAP) estimationadaptive Kalman filtering and estimating variance components adaptiveKalman filtering four methods to carry on processing, and with theMatlab software establishment data reduction program, to verify theresults of data processing for the four methods, Through the comparativeanalysis, we gain the optimal adaptive Kalman filtering method.