Modeling of highways energy consumption with artificial intelligence and regression methods

While developing technology and industrialization factors increase production, they also lead to an increase in energy consumption at the same time. The transportation sector, which is a branch of industrialization, has an important place on the basis of sector in energy consumption. In this study, energy consumption in the transportation sector has been examined, especially in the USA, where freight transport by road has an important place, it has a high potential. Within the scope of the study, energy consumption prediction modeling is made by using artificial neural networks (ANN) adaptive neuro-fuzzy inference system (ANFIS) and Simple Membership Functions and Fuzzy Rules Generation Technique (Fuzzy SMRGT) from artificial intelligence techniques. Artificial intelligence methods were also compared with multivariate linear regressions and multivariate regressions types. Interaction, pure quadratic and quadratic methods were used as multiple nonlinear regression. In the modeling, energy consumption was estimated by taking the highway network length, the number of vehicles and the number of drivers as independent variables. When comparing the prediction models, the determination coefficient (R-2), the root-mean-square error (RMSE) and the average percentage error (APE) performance criteria were taken into consideration. In addition, it was shown that the models performed well based on the metrics in the testing phase. When the performances of the models were compared, it was seen that two models obtained remarkable results. According to performance criteria, the best model is obtained by Fuzzy SMRGT and ANFIS methods. R-2, RMSE, APE values of the best models are Fuzzy SMRGT (0,978; 208,08; % 0,79) and ANFIS (0,969; 282,69; % 1,06), respectively. The Fuzzy SMGRT and ANFIS models have slightly better performance than MLR, MR, ANN models. It is aimed to use the developed models in the evaluation and management of transportation and energy policies.