Estimating Future Electric Vehicle and Charging Station Numbers Using LSTM and GRU Deep Learning Methods: Kocaeli Province Example

Öz

In this study, the number of electric vehicles and charging stations in Kocaeli in 2030 was estimated with two different deep learning models.Data obtained from TURKSTAT and EMRA were used for training LSTM and GRU models.In addition to the prediction results, a statistical prediction was made for 2030 with the support of an analysis company using existing data.Thus, three different prediction results were obtained. To understand which of the LSTM and GRU models provides higher prediction accuracy, low error rate and high success scores were used.LSTM and GRU achieved the same score with a value of 0.99 in the R2 success metric.While LSTM provided a value of 0.5007 and GRU a value of 0.38 in the MAE error metric, LSTM provided a value of 3.05 and GRU a value of 2.92 in the MSE error metric.The DTW metric scores are 126.97 in LSTM and 125.35 in GRU. According to the metric scores, it was determined that the GRU model gave the best result.2030 Kocaeli province charging station GRU model predictions were positioned on a neighborhood basis by preserving the existing charging stations.

Anahtar Kelimeler

• Electric vehicle
• charging station
• deep learning
• artifical intelligence
• python.

Geleceğe Yönelik Elektrikli Araç ve Şarj istasyonu Sayılarının LSTM VE GRU Derin Öğrenme Yöntemleri Kullanılarak Tahmin Edilmesi: Kocaeli İli Örneği

Öz

Yapılan bu çalışmada iki farklı derin öğrenme modeli ile 2030 yılı Kocaeli ili elektrikli araç ve şarj istasyonu sayılarının tahmini yapılmıştır. LSTM ve GRU modellerinin eğitimi için TÜİK ve EPDK’dan alınan veriler kullanılmıştır. Yapılan tahmin sonuçlarına ek olarak mevcut veriler ile bir analiz firmasından alınan destek ile 2030 yılı için istatiksel bir tahmin yapılmıştır. Böylece üç farklı tahmin sonucu elde edilmiştir. LSTM ve GRU modellerinden hangisinin daha yüksek tahmin doğruluğunu sağladığı, düşük hata oranı ve yüksek başarı skorları ile belirlenmiştir. LSTM ve GRUR2 başarı metriğinde 0,99 değeri ile aynı skoru sağlamışlardır.MAE hata metriğinde LSTM 0.5007, GRU ise 0,38 değerini sağlarken, MSE hata metriğinde LSTM 3,05 ve GRU 2,92 değerini sağlamıştır. DTW metriği skorları ise LSTM’de 126,97, GRU ‘da ise 125,35’tir. Metrik skorlarına göre GRU modelinin en iyi sonucu verdiği belirlenmiştir. 2030 yılı Kocaeli ili şarj istasyonu GRU modeli tahminlerinin, mevcut şarj istasyonları muhafaza edilerek mahalle bazında konumlandırması yapılmıştır.

Anahtar Kelimeler

• Elektrikli araç
• şarj istasyonu
• derin öğrenme
• yapay zeka
• python.

Kaynakça

1. [1] Alanazi F.,“Electricvehicles: Benefits, challenges, and potential solutions for wide spread adaptation”, Applied Sciences, 13:1-23, (2023).
2. [2] Mastoi M. S.,Zhuang S., Munir H. M., Haris M., Hassan M., Usman M., Bukhari S. S.H., Ro J. S.,“An in-depthanalysis of electric vehicle charging station infrastructure, policyimplications, and future trends”., Energy Reports, 8: 11504-11529, (2022).
3. [3] Al-Hanahi B., Ahmad I., Habibi D., Masoum M. A., “Charging infrastructure for commercial electric vehicles: Challenges and future Works”, Ieee Access, 9: 121476-121492, (2021).
4. [4] Ahmad F.,Iqbal A., Ashraf I., MarzbandM.,“Optimal location of electric vehicle charging station and its impact on distribution network: A review”, Energy Reports, 8: 2314-2333, (2022).
5. [5] Powell S.,Cezar G. V., Min L., Azevedo I. M., Rajagopal R., “Charging infrastructure Access and operation to reduce the grid impacts of deep electric vehicle adoption”, Nature Energy, 7: 932-945, (2022).
6. [6] Majhi R. C.,Ranjitkar P., Sheng M., Covic G. A., Wilson D. J.,“A systematic review of chargin ginfrastructure location problem for electric vehicles” Transport Reviews, 41: 432-455, (2021).
7. [7] Hassan Q.,Viktor P., Al-Musawi T. J., Ali B. M., Algburi S., Alzoubi H. M., Jaszczur M.,“There newable energy role in the global energy Transformations”, Renewable Energy Focus, 48:1-1,(2024).
8. [8] Hou F.,Chen X., Yang F., Ma Z., Zhang S., Guo, F., “Comprehensive analysis method of determining global long-term GHG mitigation potential of passenger battery electric vehicles”, Journal of Cleaner Production, 289: 1-1, (2021).

9. [9] Ternel C.,Bouter A., Melgar J., “Life cycle assessment of mid-range passenger cars powered by liquid and gase ousbio fuels: Comparison with gren house gas emissions of electric vehicles and forecast to 2030”, Transportation Research Part D: Transport and Environment, 97:1-1, (2021).
10. [10] Okoh A. S.,Onuoha M. C.,“Immediate and future challenges of using electric vehicles for promoting energy efficiency in Africa’scleanenergytransition”, Global Environmental Change, 84: 1-1, (2024).
11. [11] Çelikkaya A., “Türkiye’nin Karbon Fiyatlandırma Politikasının Yeniden Gözden Geçirilmesi”, Maliye Çalışmaları Dergisi, 71: 15-27, (2024).
12. [12] Demir A.,“Hibrid, Tam Elektrikli ve Yakit Hücreli Araçlar Trendi ve Emniyet Yükümlülüklerinin Değerlendirilmesi”, İstanbul Ticaret Üniversitesi Fen Bilimleri Dergisi, 21: 36-155, (2022).
13. [13] Özdemir E.,ErcanB.,“Avrupa Yeşil Mutabakatının Enerji Sektörüne ve Otomotiv Endüstrisine Etkileri ve Sonuçları”, Avrupa Bilim ve Teknoloji Dergisi, 51: 190-202, (2023).
14. [14] Koohfar S.,Woldemariam W., Kumar A.,“Prediction of electric vehicle scharging demand: A transformer-based deep learning approach”, Sustainability, 15: 1-17, (2023).
15. [15] Yi Z.,LiuX. C.,Wei R., Chen X., Dai J.,“Electric vehicle charging demand forecasting using deep learning model”, Journal of Intelligent Transportation Systems, 26: 690-703, (2022).
16. [16] Shanmuganathan J.,Victoire A. A., Balraj G., Victoire A.,“Deep learning LSTM recurrent neural network model for prediction of electric vehicle charging demand”, Sustainability, 14: 1-28, (2022).
17. [17] Sasidharan M., Kinattingal S., Simon S.P., “Comparative Analysis of Deep Learning Models for ElectricVehicle Charging Load Forecasting”, J. Inst. Eng. India Ser. B, 104: 105–113, (2023).
18. [18] Rasheed T.,Bhatti A. R., Farhan M., Rasool A., El-Fouly T. H.,“Improving the efficiency of deep learning model susing supervised approach for load forecasting of electric vehicles”, IEEE Access, 11: 91604-91619, (2023).
19. [19] Chung D. W.,Ko J. H., Yoon K. Y.,“State-of-charge estimation of lithium-ion batterie susing LSTM deep learning method”, Journal of Electrical Engineering&Technology, 17: 1931-1945, (2022).
20. [20] Wei M., Ye M., Li J. B., Wang Q., Xu X.,“State of charge estimation of lithium-ion batteries using LSTM and NARX neural Networks”, Ieee Access, 8: 189236-189245, (2020).
21. [21] Shahriar S. M.,Bhuiyan E. A., Nahiduzzaman M., Ahsan M., Haider J.,“State of charg eestimation for electric vehicle battery management systems using the hybrid recurrent learning approach with explain able artificial intelligence”. Energies, 15: 1-1, (2022).
22. [22] Yang F.,Zhang S., Li W., Miao Q.,“State-of-charge estimation of lithium-ion batteries using LSTM and UKF”, Energy, 201: 1-1, (2020).
23. [23] Alansari M., Al‐Sumaiti A. S., AbughaliA.,“Optimal placement of electric vehicle charging infrastructures utilizing deep learning”, IET Intelligent Transport Systems, 18: 1529-1544, (2024).
24. [24] Adil M.,Mahmud M. P., Kouzani A. Z., Khoo S. Y.,“Optimal location an dpricing of electric vehicle charging stations using machine learning and stackelberggame”, IEEE Transactions on Industry Applications, 60: 4708-4722, (2024).
25. [25] Zhao Z., Lee C. K., Ren J., Tsang Y. P.,“Optimal EV fast charging station deployment based on a reinforcemen tlearning framework”, IEEE Transactions on Intelligent Transportation Systems, 24: 8053-8065, (2023).
26. [26] Hameed B. Z.,Shah M., Naik N., Singh Khanuja H., Paul R., Somani B. K., “Application of artificial intelligence-based classifier stop redict the out come measures and stone-free status following percutaneous nephrolithotomy for staghorn calculi: cross-validation of data and estimation of accuracy”, Journal of Endourology, 35: 1307-1313, (2021).
27. [27] Medvedeva M.,Vols M., WielingM.,“Using machine learning to predict decisions of theEuropean Court of Human Rights”, Artificial Intelligence and Law, 28: 237-266, (2020).
28. [28] Mangipinto A.,Lombardi F., Sanvito F. D., Pavičević M., Quoilin S., Colombo E., “Impact of mass-scale deployment of electric vehicles and benefits of smart charging across all European countries”, Applied Energy, 312: 1-1, (2022).
29. [29] Fischer A. M., Eid M., De Cecco C. N., Gulsun M. A., Van Assen M., Nance J. W. and Schoepf U. J., “Accuracy of an artificial intelligence deep learning algorithm implementing a recurrent neural network with longshort-term memory for the automated detection of calci fiedplaques from coronary computed tomography angiography”, Journal of Thoracic Imaging, 35: 49-57, (2020).
30. [30] Castán-Lascorz M. A., Jiménez-Herrera P., Troncoso A. and Asencio-Cortés G.,“A new hybrid method for predicting univariate and multivariate time series based on pattern forecasting” , Information Sciences, 586: 611-627, (2022).
31. [31] Han S.,Meng Z., Zhang X., Yan Y.,“Hybrid deep recurrent neural Networks for noise reduction of MEMS-IMU with static and dynamic conditions”, Micromachines, 12: 1-1,(2021).
32. [32] Elsayed N.,Maida A. S., Bayoumi M.,“Reduced gate convolutiona long short term memory using predictive coding for spatiotemporal prediction”, Computational Intelligence, 36: 910-939, (2020).
33. [33] Wang Q.,Peng R. Q., Wang J. Q., Li Z., Qu H. B.,“NEWLSTM: An optimized long short-term memory language model for sequence prediction”, IEEE Access, 8: 65395-65401, (2020).
34. [34] BhandariH. N.,Rimal B., Pokhrel N. R., Rimal R., Dahal K. R., Khatri R. K.,“Predictingstock market indexusing LSTM”, Machine Learning with Applications, 9: 1-1, (2022).
35. [35] ArunKumar K. E.,Kalaga D. V., Kumar C. M. S., Kawaji M., Brenza T. M.,“Forecasting of COVID-19 using deep layer recurrent neural networks (RNNs) with gated recurrent units (GRUs) and long short-term memory (LSTM) cells”, Chaos, Solitons & Fractals, 146:1-1, (2021).
36. [36] Bouktif S.,Fiaz A., Ouni A., Serhani M. A.,“Multi-sequence LSTM-RNN deep learning and metaheuristics for electric load forecasting”, Energies, 13: 1-1, (2020).
37. [37] Zarzycki K.,Ławryńczuk M.,“LSTM and GRU neural networks as models of dynamical processes used in predictive control: A comparison of models developed for two chemical reactors”, Sensors, 21:1-1, (2021).
38. [38] Wright L. G.,Onodera T., Stein M. M., Wang T., Schachter D. T., Hu Z., McMahon P. L.,“Deep physical neural Networks trained with back propagation”, Nature, 601: 549-555, (2022).
39. [39] Astawa I. N. G. A.,Pradnyana I. P. B. A., Suwintana I. K.,“Comparison of rnn, lstm, andgrumethods on forecasting website visitors”, J. Comput. Sci. Technol. Stud, 4: 11-18, (2022).
40. [40] Xia M.,Shao H., Ma X., De Silva C. W.,“A stacked GRU-RNN-based approach for predicting renewable energy and electricity load fo rsmart grid operation”, IEEE Transactions on Industrial Informatics, 17: 7050-7059, (2021).
41. [41] Zarzycki K.,Ławryńczuk M.,“LSTM and GRU neural networks as models of dynamical processe sused in predictive control: A comparison of models developed for two chemical reactors”, Sensors, 21:1-1, (2021).
42. [42] Öktem, K., Tekerek, A., “Bitcoin Price Direction Prediction Using Machine Learning on a Very Small Dataset.” Politeknik Dergisi 1-1, (2025). https://doi.org/10.2339/politeknik.1667403
43. [43] Alsaideen, M., Ertem, Z., “A Comprehensive Analysis of Resilient Multivariate Forecasting Models for Steel Plate Price Prediction.” Politeknik Dergisi, 28(2), 627-637, (2025).
44. [44] Kabakçı, D. ve Akbaş, E., “Automated learning rate search using batch-level cross-validation”. Sakarya University Journal of Computer and Information Sciences, 4(3), 312-325, (2023).