Filtering Technology In Shenyang Subway For The Application Of Deformation Monitoring Data Processing

Based on the background of Shenyang street in the air defense project, deformation monitoring of civil air defense engineering and subway overlapped section. Because of the subway in the operational phase, the real-time monitoring and automatic monitoring system-static level monitoring system. But the system output data, compared with the manual two leveling results, beating larger error data,17%--26%, the monitoring data denoising.This paper prepared filtering software on VS2010, first uses the classical Calman filtering, adaptive Calman filter and wavelet filter three methods on the monitoring data, the results found that the error of classical Calman filter can effectively eliminate the error is 38%, the adaptive Calman filter can effectively eliminate the error is 55%, the wavelet filter can effectively remove the 52%, the three methods when used alone can not meet the requirements of more than 60% error elimination.Aiming at this problem, this paper proposes the use of filtering model combination techniques, according to the characteristics, found that the use of a filtering method alone is:1, the classic Caiman filter is not suitable for long-term use; 2, adaptive Calman filter has obvious advantage in the low frequency part; 3, the wavelet filter in the high frequency part of a remarkable noise reduction. Filtering model combination. According to the filter sequence identified four filtering model combined BCal, BCa2, Ca1B and Ca2B, and by the theoretical knowledge that BCal noise reduction effect is the best. Through an engineering example, filtering model combination method is better than the damping effect using a filtering technology alone, error filtering model combination method can effectively remove the 65%--75%. The filter model combined BCal noise reduction effect is best, can effectively eliminate the error of 75%, greatly improve the accuracy and reliability of the automation of monitoring data, the error is expected to get rid of 60% of the objective, the optimal treatment method of metro Automatic Monitoring of civil air defense projects in the data is combined BCal filtering model.