Estimating efficient value of controllable variable using an adaptive neural network algorithm: Case of a railway system

Abstract

This study proposes a method, using adaptive neural network (ANN), to predict, estimate and evaluate performancevariables without requiring any restrictive assumptions, taking case of a railway system. Also, by means of this method, it wouldbe possible to compare actual performance data with estimated values and route their assignable causes in future periods. Energyconsumption norm of vehicles in case of energy railway and real data of energy consumption in Iranian railway is considered.