Estimating transportation energy demand in Turkey using the artificial bee colony algorithm
Introduction
Energy demand has been increasing all over the world with the various social and economic developments. Similarly, growing population and urbanization as well as socio-economic development in Turkey have caused a rapid increase in the energy demand for many sectors of the country.
The economy of Turkey expanded by in 2010 and by in 2011, becoming the 18th largest economy in the world and the 7th largest economy among the European Union countries [29]. The improvements in economy have changed the travel habits of the citizens, who are now able to afford a personal vehicle for comfort and preferred to use their private vehicles rather than depending on public transportation. Hence, the developments in economy have also triggered an increase in the transportation energy demand (TED) of the country. According to the data of sustainable development indicators of the Turkish Statistical Institute (TSI), the transportation sector has a share of in the energy consumption of Turkey. The same indicators also showed that energy consumption on highways was nearly in the transportation sector in 2012 [40], [41].
According to the reports of the International Energy Agency (IEA), Turkey will likely see the fastest medium to long-term growth in energy demand among its member countries. On the other hand, the total primary energy demand of the country is estimated to be more than double reaching million tons of oil equivalent (MTOE) by 2020. During the last decade, in terms of the increase in the demand for natural gas and electricity, Turkey was the second country after China. Therefore, Turkey is expected to become one of the most dynamic energy economies of the world in terms of the increase in energy demand [21].
Turkey depends heavily on energy imports because of its steadily increasing energy demand and limited energy resources. Today, the country has to import almost three fourths of its energy needs. Energy dependence is more evident in the transportation sector. Oil and natural gas are two main fuels in the country, a great portion of which is imported every year (approximately and , respectively). On the other hand, Turkey aims to reduce its dependency on imported fossil fuels through gradual commissioning of nuclear power into the Turkish energy mix. Turkey aims to establish a nuclear capacity of more than by 2030. In addition, Turkey intends to increase wind and solar energy sources. It plans to produce of its electricity need from renewable energy sources by 2023 [21].
Energy planning studies require an understanding of the past, present and future energy demands. Future energy demands are estimated using energy demand models. Modeling the energy demand is an important step of planning since governments can develop appropriate strategic plans in the light of realistic projection based on these models. Hence, energy sources can be used more efficiently in various sectors.
Modeling energy demand in the transportation sector is usually dependent on many transportation and socio-economic variables such as income, vehicle ownership, housing size, population increase and number of trips. However, it is difficult to include all these parameters when developing TED models since it usually requires a more detailed study with massive data sources. For time- and cost-efficiency, a TED model should be expressed with a simple mathematical form and based on data that is easily obtained such as population, gross domestic product etc.[33].
There are several direct search algorithms for determining the optimal design such as artificial bee colony (ABC), ant colony optimization (ACO) and particle swarm algorithms. ABC is one of the swarm-based optimization techniques, which was originally proposed by Karaboga [22]. There are two good review papers related to ABC algorithm. The first one presented by Bansal et al. [3] presents the developments and applications and gives a brief comparison of the performance of ABC with some of other heuristic search algorithms. They declared that it has reportedly outperformed other search heuristics and is not sensitive to initial parameter values and not affected by the increasing dimension of the problem. The second paper presented by Karaboga et al. [23] provided a comprehensive survey of the advances on ABC and its applications. They mentioned that ABC algorithm has been extensively researched through the world and the number of publications related to ABC algorithm in the literature increases exponentially. ABC was also compared to other algorithms and a remarkably better performance with a 100% success rate of this algorithm was reported [35].
The ABC algorithm has three main advantages; being easy to implement, requiring fewer control parameters and being robust. Other direct search techniques involve the use of tunable parameters [e.g., crossover rate, mutation rate in the Genetic Algorithm (GA)] and their performance concerning the number of function evaluations heavily depend on the control parameters. Therefore, in this study, the ABC algorithm was used to estimate the TED of Turkey to present an alternative algorithm that can be effectively and easily used for the prediction of future energy demand of Turkey. Linear, exponential and quadratic forms of models were developed based on the parameters of the gross domestic product (GDP), population and total annual vehicle-km.
This paper is organized as follows: Section 2 presents relevant literature. The details of the ABC algorithm and the development of the models are explained in Sections 3 ABC optimization algorithm for transportation energy demand, 4 The development of transportation estimation demand models, respectively. In Section 5, the performance of the models is evaluated and estimation of energy demand of Turkey by 2034 is discussed based on two possible scenarios. The paper is concluded in Section 6.
Section snippets
Literature review
In Turkey, during the last four decades, TED, GDP, population and vehicle-km have increased by approximately , , and times, respectively due to the expanding economy, growing population and developments in the transportation sector [15], [40], [41]. This situation has attracted the attention of many researches to future energy demand projections using different modeling approaches such as artificial intelligence (AI) and regression analysis (RA) techniques. Recently, various AI
ABC optimization algorithm for transportation energy demand
In optimization problems related to the prediction of future demands, the major task is to find an optimal function and coefficients of this selected function. In this process, the objective is to minimize the sum of the square of errors, which is the deviation between observed and calculated or estimated data. This is mathematically expressed as follows:where is the observed values; is the estimated values as a function of the independent variables
The development of transportation estimation demand models
The pattern of transportation energy consumption has shown a rising trend in most of the countries in the world. In this study, TED estimation models were developed utilizing the ABC algorithm to predict the energy demand of Turkey based on the data between 1970 and 2013. The data related to GDP, population and total annual vehicle-km were collected from the World Bank [44], TSI [40], [41] and the General Directorate of Highways [15], respectively, and TED data was obtained from the World
Results and discussion
The ABC algorithm was used to estimate the coefficient for each mathematical model. The number of bees and maximum number of cycles (MNC) were selected as 40 and 1000, respectively in the optimization process. The limit value was taken as ( is the number of design variables) as suggested by Karaboga [22]. The parameters were selected as 500 for all TED models. Twenty independent runs were performed. Table 2 presents the best outputs and the best and average values of
Conclusions
The transportation sector consumes approximately of the total energy sources of Turkey. The economic growth in the past few decades has resulted in a significant increase in the annual TED of the country. Hence, a reasonable estimate is important in terms of planning for future TED. In this study, the ABC algorithm was successfully applied to estimate the future TED of Turkey for the period from 2014 to 2034 using 44-year historical data of GDP, population and total annual vehicle-km
References (50)
- et al.
Policies to reduce energy use and environmental emissions in the transport sector: a case of Delhi city
Energy Policy
(1997) - et al.
Transport energy demand modeling with meta-heuristic harmony search algorithm, an application to Turkey
Energy Policy
(2008) - et al.
Estimating energy demand of Turkey based on economic indicators using genetic algorithm approach
Energy Convers Manag
(2004) Implications of transportation policies on energy and environment in Kathmandu Valley, Nepal
Energy Policy
(2003)- et al.
Genetic algorithm approach to estimate transport energy demand in Turkey
Energy Policy
(2005) - et al.
Another look at measures of forecast accuracy
Int J Forecast
(2006) - et al.
Modelling and prediction of Turkey’s electricity consumption using artificial neural networks
Energy Convers Manag
(2009) - et al.
Projection of future transport energy demand of Thailand
Energy Policy
(2011) - et al.
The forecast of motor vehicle energy demand and CO2 emissions from Taiwan’s road transportation sector
Energy Policy
(2009) Demand forecast for road transportation fuels including gasoline, diesel, LPG, bioethanol and biodiesel for Turkey between 2013 and 2023
Renew Energy
(2014)
Use of artificial neural networks for transport energy demand modelling
Energy Policy
Estimating emissions on vehicular traffic based on projected energy and transport demand on rural roads: policies for reducing air pollutant emissions and energy consumption
Energy Policy
General procedure for long-term energy-environmental planning for transportation sector of developing countries with limited data based on LEAP (long-range energy alternative planning) and Energy PLAN
Energy
Monitoring urban transport air pollution and energy demand in Rawalpindi and Islamabad using LEAP model
Energy
Artificial bee colony algorithm for optimization of truss structures
Appl Soft Comput
Prediction of net energy consumption based on economic indicators (GNP and GDP) in Turkey
Energy Policy
Estimating the net electricity energy generation and demand using the ant colony optimization approach: case of Turkey
Energy Policy
Improvement of energy demand forecasts using swarm intelligence: the case of Turkey with projections to 2025
Energy Policy
Estimates of hydroelectric generation using neural networks with the artificial bee colony algorithm for Turkey
Energy
Modeling and optimization of a chiller plant
Energy
A new hybrid artificial bee colony algorithm for robust optimal design and manufacturing
Appl Soft Comput
Long-term outlook of energy use and CO2 emissions from transport in Central and Eastern Europe
Energy Policy
Forecasting the transport energy demand based on PLSR method in China
Energy
Projection of future transportation energy demand of Jordan using adaptive neuro-fuzzy technique
Energy
Optimal placement of elastic steel diagonal braces using artificial bee colony algorithm
Steel & Compos Struct
Cited by (55)
Realization of Turkey's energy demand forecast with the improved arithmetic optimization algorithm
2022, Energy ReportsCitation Excerpt :According to their study, they stated that the GA based method was found successful results for the estimation process. In another study, Sonmez et al. [8] have proposed a forecasting model by using the artificial bee colony (ABC) technique ın order to estimation the Turkey’s future transportation energy demand.
Prediction of transportation energy demand by novel hybrid meta-heuristic ANN
2022, EnergyCitation Excerpt :The outcomes revealed that the transportation energy requirement is anticipated to raise throughout 2010–2030. Sonmez et al. [60] proposed three distinct mathematical models to calculate transport energy requirement of Turkey utilizing the artificial bee colony algorithm. For the establishment of the models, GDP, overall yearly vehicle-km, and population were used as variables.
A novel method for customer-oriented scheduling with available manufacturing time windows in Cloud Manufacturing
2022, Robotics and Computer-Integrated ManufacturingUsing neural network modelling for estimation and forecasting of transport sector energy demand in developing countries
2022, Energy Conversion and Management