Prediction of Energy Consumption in the First Line of Tehran Metro: GMDH Neural Network Approach

Today, energy and its consumption are the main strategic plan of organizations and also the development of urban transport systems by considering a variety of economic, scientific, industrial, climate and growing urbanization is essential. Analysis of past trends in energy is the key to predict future trends, with regard to the rate of development of metro, for planning and future-oriented macro economic policies.  in this research has been used to predict the energy consumption of Tehran Metro Line 1 from the GMDH Neural Network Model Which is capable of detecting and screening low-input input variables In the course of training the network and removing them during the exam period. and also Was compared To understand the accuracy of the prediction with the ARIMA model. in this research, was detected twelve variables affecting Tehran's metro energy consumptionand is considered as input variables of the model. The results indicate that The GMDH neural network model has a much lower error rate than the ARIMA model and has a higher predictive accuracy.