Template-Based Voting Logic for Dual Train Detection Systems

Abstract

A train detection system is crucial in railway signaling and control systems. To improve the reliability and safety of train detection systems, the conventional railway operator in Taiwan, Taiwan Railways Administration (TRA), adopted dual detection systems for each signal block. However, conventional design logics for dual detection systems cannot fully and simultaneously improve reliability and safety. Thus, in this research, an innovative logic is proposed, namely, template-based voting (TBV) logic, to automatically correct possible train detection errors on the basis of preprocessed simulation data. The TBV logic first applies a “template generation module” to generate and store all possible train movement data in a database. Then an “occupation status comparing module” executes the standard or special process to trace passing trains and correct possible detection errors on the basis of the templates. Results of case studies and a sensitivity analysis demonstrate that this logic outperforms other existing logics with respect to integrated performance in reliability and safety. Thus, the application of this logic can substantially improve the reliability and safety of a dual train detection system with fixed blocks.