|
In recent years, the rapid development of China's urban underground rail transit, security has become a widespread concern about the problem with the development of the underground rail transit, subway tunnels, tunnel deformation monitoring is more and more attention. Laser scanning technology for its high-precision, high-density, real-time and initiative characteristics, is to receive widespread attention, mobile laser scanning technology used in the subway tunnel deformation monitoring is a new method of tunnel deformation monitoring. This is a new non-contact real-time security detection technology can make up for the deficiency of traditional detection methods, which uses a non-contact, real-time online, comprehensive access to multi-sensor data fusion structure information, based on the computer network communication technology, cost low, the process can be repeated, with the characteristics of real-time, dynamic, high-precision, can improve monitoring accuracy and spatial and temporal analysis of the overall monitoring ability. The technology also avoid a the monitoring operations contradiction and the conventional method in which the poor working conditions. Massive, redundant, strong noise characteristics due to laser scan data, data processing challenges. This paper analyzes the characteristics of the tunnel laser scanning data errors, and pointed out that the laser scan data error exists in the radial direction, rather than simply exist in the vertical direction. Features of this error, the proposed space conversion Cartesian space converted to polar coordinate space, restore error direction and Gross Error Elimination become relatively simple. Since the laser scanning data contains a lot of gross errors, it is necessary to adopt a robust estimation method, this article describes the principle of robust estimation and robust estimation methods, the last symbol constrained least squares and the method is proved reasonable. The main research results are as follows: (1) by the the Laplace distribution M estimated model proved symbolic constraints Robust Least Squares reasonableness, for application of the method to lay the foundation. (2) the shape of the tunnel, the sign constraints Robust Least Squares algorithm has been improved greatly improve the efficiency of the algorithm. (3) analysis of the direction of a the tunnel laser scanning data error with space conversion method restore data error direction and use of the sign constraints Improved Robust Least Squares successfully removed the gross errors in the the tunnel laser scanning data .
|