The impact of traffic flow on fuel consumption in transportation sector: eva‎luation and optimization using artificial intelligence

12/18/2023 12:00:00 AM

With the increasing dependence of human life on energy, the role of energy in the functioning of various economic sectors of countries has become crucial. On the other hand, with the introduction of vehicles in human life, human life faced growth and development. However, this development in the transportation industry, as one of the most significant sources of energy consumption for humans, also brings challenges, such as environmental pollution. Environmental protection is one of the essential aspects of sustainable development, and the first measure of its sustainability is air quality. The amount of greenhouse gas emissions, one of the determining factors of air quality, has a direct relationship with the number of energy carriers consumed in a country. Considering the 25% share of the transportation sector in the total energy demand in the country and a large number of urban vehicles, the importance of gasoline consumption can be realized. The enlargement of cities and the expansion of the use of vehicles and the increasing number of vehicles in the country (more than 20 million vehicles) lead to an increase in the density of cars and, as a result, waste of time, high fuel consumption, and irreparable damage to the environment. Considering the relationship between traffic and fuel consumption, this will cause a significant reduction in fuel consumption. Estimating the traffic volume and its relationship with fuel consumption in the country will play an influential role in reducing fuel consumption and costs and developing economic and social benefits. Regression models and artificial neural networks were discussed as the best options among various methods to eva‎luate the relationship between traffic volume and fuel consumption. The results of the thesis show that the artificial neural network model with a hidden layer had an accuracy of about 82% in learning and 80% in predicting delay time for test data.

انرژي , مصرف سوخت , ترافيك , يادگيري ماشين , شبكه عصبي مصنوعي

Energy , Fuel Consumption , Traffic , Machine Learning , Artificial Neural Network

S.Mohammad Hashemi

Dr. Meysam Farajolahi, Dr. Rohoullah Ahmadi