Özet:

The work-energy relationship based on nature shows us that the condition for movement and ability to do business is energy. Developing technology, factors such as industrialization increase production while causing increased energy consumption. The transport sector, which is a hand of industrialization, has an important place in the energy consumption industry. In this study, energy consumption is studied in the transport sector, especially in Turkey, where the transport potential is high in road transport. In the framework of the study, artificial intelligence techniques, artificial nervous networks (YSA) and adapted nervous fool output system (ANFIS), and regression techniques, using multi-variant linear regression (MLR) methods, the energy consumption forecast in road transport is modeled. The road network length, vehicle-km, annual average daily traffic (YOGT), motor vehicle number and population parameters are studied as independent variables. The comparison of the forecast models takes into account the correlation rate (R), the average quarter of errors (HKO) and the average percentage of errors (OYH) performance criteria. According to performance criteria, the best model is obtained by the YSA method. The best model’s R, HKO, OYH values are 99.5%, 184002.7 and 2.44% respectively. It is aimed at developing the transportation policy.

Anahtar Kelimeler: