Applications of strapdown inertial systems in curvature detection problems

Described in this thesis is the successful application of a strapdown inertial system to curvature detection problems, for the purpose of the analysis of perturbations in the trajectory of vehicles. It is shown that the principles of differential geometry can be applied directly to curvature analysis of vehicle routes, utilizing the high relative accuracy of measurement data output from strapdown reference units. Conventional methodology utilizes coordinate geometry in reconstruction of curvature values. In this thesis, it is shown that acceleration and angular rate data can also be used to compute curvature parameters. The software module SINS was developed for strapdown navigation and curvature computations. Field tests were performed on a Light Rail Transit (LRT) vehicle and on a deep mine shaft hoist system. The LRT survey proved that curvature determination, using a combination of both typical coordinate methods and the rate data methods developed for this thesis, can be a useful tool in the detection of rail track misalignment, while the mine tests showed conclusively the utility of the strapdown system in dynamic analysis of trajectory information.