Abstract
Numerous disruptive incidents including technical issues as well as natural disasters have affected Railway Networks (RN) in the past few decades. For optimizing the RN, the work created a novel RN architecture with four separate modules. The infrastructure module allots disruption to the RN. The control system module defines whether the RN is disrupted or not utilizing the Gaussian Membership Function-based Fuzzy Decision Making (GMF-FDM) process. If it is interrupted, it is evaluated at the train service together with the passenger network, and the operational module is triggered. Two main operations are being executed by the operational module: (i) Improved Cockroach Swarm Optimization (ICSO) reroute the trains by means of ascertaining the shortest route, as well as (ii) Mixed Integer Linear Programming (MILP) reschedule the train's timings in relation to the passenger. Based on train, passenger, along with station data, significant constraints were extracted, which were inputted to the Vulnerability Assessment (VA) module. In which, the chosen paths vulnerability is assessed by the proposed Enhanced Elman Deep Learning Neural Network (EEDLNN). An accuracy of 96.61% is obtained by the experimental outcome, which is better when weighed against the conventional methods.