QUALITY ASSURANCE METHODOLOGY FOR HIGH-SPEED TRACK INSPECTION SYSTEMS
Railroad tracks provide two main functions, they support and distribute the static weight and dynamic loads into the foundation and they provide the lateral steering guidance to the wheels and axles so the rail vehicles can travel smoothly along the guide way at the desired speed. There are more than 10 commonly used geometry parameters to characterize the quality of the track, including gage, super elevation, profile and alignment. Each parameter is a continuous variable of distance along the track location. The series of measurements of a particular parameter at a fixed sampling rate along the track can be regarded as time-series if we consider distance as time. With some qualifiers, these time series can be modeled by stationary stochastic processes. In this presentation, statistical methods for stationary processes will be used to analyze the repeatability of the measurement system. The measurement requirement and the method of quality assurance presented in this article was developed and applied to distance-based data series; the methodology is applicable equally to data series based on time or other measurement references. Using railroad track geometry as the practical example, the procedure to collect sample data sets, the process to overcome some sampling control issues, checking the repeatability of measurement results are presented to show how statistical hypothesis testing method is applied to assess the quality of the data. Aside from the data repeatability analyses, there are additional quality assurance procedures on a track inspection system, such as controlled laboratory tests and overthe-road endurance tests. These are not addressed in this paper. Two different proposed approaches, one from the system supplier and one from the system purchaser, are presented in this paper. While either approach can be used to get a reasonable assessment of the measurement system, one was shown mathematically to be superior to the other.
Ta-Lun Yang
ENSCO Inc.,5400 Port Royal Road,Springfield,Virginia,22151 USA
国际会议
厦门
英文
238-246
2011-10-28（万方平台首次上网日期，不代表论文的发表时间）