Modelling of Track Layout for Intelligent Railway Signalling System: A Machine Learning Application

The railway signalling domain is a complex critical infrastructure, linking communication and number of control elements. Ensuring safety in railway signalling system is always considered as a guarantee of intact operation of the railway. Current signalling system composes of centralized controllers which provide a single feature such as interlocking, and level crossing control. The Indian Railways (IR) uses Panel Interlocking (PI), Route Relay Interlocking (RRI), and Solid State Interlocking (SSI) or Electronic Interlocking (EI) for signalling safety, however, permitting movement of the trains lies in the hands of a human. The main challenge is to combine multiple sources of data and define a system which can intensify the functionality of the system. This paper mainly focuses on development of an automated model, beneficial to Intelligent Signalling System (ISS). Assessing its ability to take a decision which authorizes the movement of trains according to the timetable and modify it depending on real-time information using Machine Learning (ML). For modelling, IR standard single line station layout is considered and graphical model-based design techniques are implied. For analysis consider the track sections as nodes, signals as the start point and the end point linked to specific routes and assessing the developed model for various operating scenarios keeping a strict check on completeness and consistency. Implementation of such system in the railway network will not only provide a comprehensive level of safety in railway transportation but also takes a step forward towards systematizing various methods and strategies such as rescheduling system, monitoring performance under one roof using ML.