Optimized design to adverse transportation conditions for railway freight system.

This study proposed a method for transportation agencies to efficiently and accurately formulate or revise rules for the movement of railway transport while ensuring safety under adverse conditions. Determining such a method in general requires trial-and-error experimentation, which consumes large amounts of time and money. We used the uniform experiment (UE) and generalized linear autoregression (GLAR) to establish our method. Based on it, a series of numerical models were proposed to examine the association between the operational indices of safety (derailment coefficient and rate of wheel unloading) and such factors as the type of wagon, cargo weight, partial loading (covering longitudinal and lateral offset), line condition, and operating speed. The models were used to determine the worst transportation conditions. The results of analysis showed the following: 1) the effect of the speed of operation on the safety indices followed a parabolic law, those of cargo weight and part loading followed a linear law, the type of wagon and line condition exhibited no clear regularity, and some of these factors have an interactive influence. 2) A combination of the UE and GLAR helped deal with the complex multivariate process using the fewest multilevel experiments to accurately determine the most adverse conditions for railway freight transportation. The proposed method provided reference schemes for governmental agencies to study and revise freight management regulations.