登入選單

返回Google圖書搜尋

Ray-tracing Propagation Models of Complex Guideway Geometries for the Deployment of Train Control Systems

Neeraj Sood

出版	University of Toronto, 2019
URL	http://books.google.com.hk/books?id=_eROzwEACAAJ&hl=&source=gbs_api

註釋

This thesis presents the work carried out in modeling propagation in complex guideway geometries. The primary motivation of this work is to develop high-fidelity models that can enable improvements in the deployment and performance of train control systems. To that end, a 3-D image-based ray tracing tool is developed. It is validated in a variety of scenarios, including an indoor office hallway, an outdoor open environment with an undulating terrain, and an arched road tunnel. The limitations of the basic image-based algorithm are thoroughly investigated, especially with respect to their failings in modeling segmented approximations of circular and arched tunnels. Due to a lack of any rigorous solution that addresses these issues, a heuristic approach is developed to modify the Fresnel reflection coefficients to systematically weaken the contribution of reflected ray-paths that lead to an overestimation of RSS. This approach is inspired by the scattering of an incident wave by a perfectly conducting finite sized rectangular plate. Enabled with these modifications, the ray tracing tool is used to study propagation in a 3.5 km long real, complex tunnel environment in the London Underground. The estimated RSS is within a mean absolute deviation of 4-6 dB compared to the measurements. The input geometries required for these simulations are derived from point cloud data obtained from a laser survey conducted by Thales Canada, by developing a semi-automated error driven framework which enables a systematic study of the relationship between the accuracy of the simulated RSS and the accuracy of the estimation of the modeled geometry. Finally, physics-based models such as the developed ray tracing tool and others are used to characterize and enhance the performance of train control systems. It is shown that such modeling tools can be quite useful as they enable optimization of access points driven by the estimated network level performance indicators, thereby integrating propagation modeling with network-level system design leading to potential efficiency gains, instead of carrying out these tasks sequentially as is done conventionally.