Revisiting the 7/23 train accident using computer reconstruction simulation for causation and prevention analysis

Abstract

Accident reconstruction by means of computer simulation facilitates the analysis of the causes of accidents and deriving suggestions for preventing an accident. A train movement simulation model based on cellular automata is developed in order to reconstruct the train collision process. An event-driven accident reconstruction procedure is proposed to describe the aberrant train movement processes. The selection of accelerations and decelerations is based on the statistically average tractive, resistive, and braking forces related to train movement speeds. The reconstruction results regarding the collision accident between multiple-unit trains D3115 and D301 on July 23, 2011 in China (the 7/23 accident for short) demonstrate that the proposed reconstruction methodology can achieve remarkable accuracy. The 7/23 accident reconstruction reveals that train D301 ran according to illogical track circuit code signals, indicating the disappearance of train D3115 running ahead, which is consistent with the official report. It is suggested that online analysis of train movement situations is conducive to quantitatively justifying train operation mechanisms and safety. A further finding is that the marginal emergency braking positions for preventing such an accident are only located beyond approximately 0.555 km. The lesson drawn from this accident reconstruction is that advanced automatic train control cannot paralyze the driver’s alertness, particularly in an environment of atrocious weather, which often induces failure of the train control equipment and endangers the safety of train operations.