Profile URL: https://hdl.handle.net/20.500.13091/11994
Job Title:Doç. Dr.
Email Address:hcimen@ktun.edu.tr
Main Affiliation:02.04. Department of Electrical and Electronics Engineering
Status: Current Staff
ORCID:
0000-0003-0104-3005
0000-0003-0104-3005Scopus ID:
57205614115
57205614115YÖK Akademik: 5159C7F4E31CFE92
Google Scholar:
3NID0WEAAAAJ
3NID0WEAAAAJWeb of Science ID:
MFH-0713-2025
MFH-0713-2025Name Variants:
Cimen, H. Ç. Halil
20 results
Scholarly Output Search Results
Now showing 1 - 10 of 20
Article Citation - WoS: 3Citation - Scopus: 4Examining the Influence of Sampling Frequency on State-Of Estimation Accuracy Using Long Short-Term Memory Models(Springer Science and Business Media Deutschland GmbH, 2024-04-16) Arabaci, H.; Ucar, K.; Cimen, H.Lithium-ion batteries’ state-of-charge prediction (SoC) cannot be directly measured due to their chemical structure. Therefore, a prediction can be made using the measurable data of the battery. The limited measurable data (current, voltage, and temperature) and the small changes in charge/discharge curves over time further complicate the prediction process. Recurrent neural network-based deep learning algorithms, capable of making predictions with a small number of input data, have become widely used in this field. Particularly, the use of Long Short-Term Memory (LSTM) has shown successful results in one-dimensional and slowly changing data over time. However, these approaches require high computational power for training and testing processes. The window length of the data used as input is one of the major factors affecting the prediction time. The window length of the data varies depending on the sampling frequency and the length of the lookback period. Reducing the window length to shorten, the prediction time makes feature extraction from the data difficult. In this case, adjusting the sampling frequency and window length properly will improve the prediction accuracy and time. Therefore, this study presents the effects of sampling frequency and window length on the prediction accuracy for LSTM-based deep learning approaches. Prediction results were examined using different metrics such as MAE, MSE, training, and testing time. The study’s results indicate that training and testing times can be shortened when the sampling frequency and window length are properly adjusted. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2024.Conference Object The Effect of Different Open-Circuit Voltage Curves on the Accuracy of Model-Based State Estimation in Lithium-Ion Batteries(IEEE, 2025) Arabaci, Hayri; Ucar, Kursad; Cimen, HalilDue to their unique characteristics, the condition monitoring of lithium-ion batteries in the applications where they are used is inevitable. Most of the condition monitoring approaches employed today are based on electrical equivalent circuit models (ECM). Currently, there are many different structures of ECM in use. However, the common component among all of them is the voltage source that provides the open circuit voltage (OCV), and its accuracy directly affects the results of the analyses. The voltage of this source depends on the state of charge (SoC) of the battery. This dependency varies based on the structure of the battery. Therefore, the relationship between the OCV and the SoC of a battery is determined using data obtained from experiments conducted for this purpose. These experiments involve complex processes, including charging, discharging, and/or operation at low current values. Due to the structure of lithium-ion batteries, the OCV curve obtained for charging differs from the one obtained for discharging. Therefore, in condition monitoring applications, OCV curves derived from these two processes are used in models based on different approaches. These approaches generally include four types of OCV curves derived from the data of charging and discharging experiments conducted at one-wentieth of the nominal current level: the charge curve alone, the discharge curve alone, the curve obtained by averaging the charge and discharge curves, and a curve that aligns with the discharge curve at high states of charge and the charge curve at low states of charge. In addition, OCV curves obtained directly from current and voltage data in Hybrid Pulse Power Characterization (HPPC) tests are also used. In this study, a different approach is examined, where the OCV during charging follows the charge curve and the OCV during discharging follows the discharge curve, based on the direction of the current. The results obtained from this approach are compared with those obtained using other approaches. In the comparisons, the mean absolute error and the root mean square error calculated during the estimation of the terminal voltage are used as criteria. The comparison shows that the lowest error values are achieved when the OCV and SoC curve obtained from low-current discharge experiments is used.Doctoral Thesis Mikro Şebekelerde Derin Öğrenme Destekli Enerji Yönetimi(Konya Teknik Üniversitesi, 2020) Çimen, Halil; Çetinkaya, NurettinGeleneksel elektrik enerjisi sistemlerinin en önemli gereksinimlerinden birisi, üretim ve tüketim dengesinin anlık olarak sağlanmasıdır. Bunu gerçekleştirmek için yüksek kapasiteli güç santralleri kullanılmakla beraber, çok sayıda yenilenebilir enerji santralinin şebekeye entegre olması ve bu santrallerin enerji üretiminin belirsiz olması, bu dengenin sağlanmasını riske atmaktadır. Mikro şebekeler, belirli bir alanda bulunan yüklerin ve dağıtılmış üretim santrallerinin koordineli bir şekilde işletilmesi ile bu dengeyi yerel olarak sağlayabilmekte ve dağıtım şebekesi üzerindeki stresi azaltabilmektedir. Bu tez çalışmasında, konut mikro şebekelerinin optimum şekilde işletilmesi için iki seviyeli Enerji Yönetim Sistemi (EYS) sunulmuştur. Geliştirilen EYS'nin en önemli bileşeni ise, mikro şebekede bulunan tüketicilerin izlenmesini ve kullandıkları cihazların anlık olarak tespitini sağlayabilen Müdahaleci Olmayan Yük İzleme (MOYİ) sistemidir. MOYİ, kullanıcıların ana sayacından okunan toplam tüketim verisinin çeşitli sinyal işleme yöntemleri ile analiz edilerek, cihaz bazında tüketim verilerinin elde edilmesini sağlayan bir yaklaşımdır. Aktif olarak çalışan yüklerin anlık olarak tanınması ve tükettikleri enerjinin tahmini, sunulan iki farklı derin öğrenme modeli ile gerçekleştirilmiştir. Aalborg Üniversitesi, Enerji Teknolojileri Departmanı, IoT Microgrid Laboratuvarı'nda bulunan dokuz farklı cihaz için Geçitlenmiş Özyinelemeli Birimler (GÖB) tabanlı bir derin öğrenme modeli kullanılarak gerçek zamanlı bir yük tanıma analizi gerçekleştirilmiştir. Analiz edilen cihazın tipine bağlı olarak %65 ile %96 arasında bir analiz başarısı elde edilmiştir. Bunun haricinde çevrim dışı ve gerçek zamanlı analiz arasındaki doğruluk farkı test edilmiş ve gerçek zamanlı analizin doğruluk oranının, %5 ila %10 arasında daha düşük olduğu gözlemlenmiştir. Geliştirilecek EYS için sadece yük tanıma analizinin yetersiz olduğu düşünülerek yeni bir GÖB tabanlı derin öğrenme modeli sunulmuş ve cihazların hem açık/kapalı olma durumları hem de tükettikleri güç, çevrim dışı olarak analiz edilmiştir. Sunulan model, iki ayrı analizi aynı anda yapabilecek kapasiteye sahip olduğu için literatürde daha önce kullanılmış derin öğrenme modellerinden daha yüksek performans göstermiştir. Tezin ikinci kısmında ise MOYİ analizinden elde edilen çıktılar kullanılarak, efektif bir EYS mekanizması tasarlanmıştır. MOYİ analizinin EYS'ye dahil edilmesinin temel amacı, müşterilerin tüketim alışkanlıklarını göz önüne alarak, onların elektrik faturalarını azaltmak, şebekenin sunacağı talep tarafı yönetimi uygulamalarından maksimum şekilde faydalanabilmelerini ve ek teşvikler alabilmelerini sağlamaktır. Her tüketici farklı bir yaşam tarzına, dolayısı ile farklı tüketim alışkanlıklarına sahiptir. MOYİ analizi sayesinde, her müşterinin yaşam alışkanlıkları ve tüketim davranışları öğrenilerek tüketiciye özel bir enerji yönetimi tasarlanabilmektedir. Tezde sunulan iki seviyeli EYS'nin birinci seviyesinde, MOYİ analizinden elde edilen sonuçlar istatistiksel olarak incelenmiş ve elde edilen veriler kullanılarak, tüketiciye özgü bir maliyet optimizasyonu gerçekleştirilmiştir. Optimum maliyet, cihazların kullanım zamanlarının, elektrik fiyatlarının yüksek olduğu periyotlardan düşük olduğu periyotlara optimum olarak kaydırılması ile sağlanmıştır. Bu sayede tüketicilerin konforu gözetilerek elektrik faturalarının otomatik olarak azaltılması sağlanmıştır. İkinci seviyede ise, mikro şebekede bulunan üretim ve tüketim birimlerinin kapasiteleri ve kısıtları dikkate alınarak, mikro şebekenin optimum şekilde işletilmesi hedeflenmiştir. Bu kapsamda çoklu bir amaç fonksiyonu tanımlanarak hem mikro şebekenin işletme maliyetinin düşürülmesi, hem de şebekede oluşacak yeni piklerin engellenmesi amaçlanmıştır. Geliştirilen EYS algoritması, Aalborg Üniversitesi, Enerji Teknolojisi Departmanı, AC/DC Microgrid Laboratuvarı'nda gerçek zamanlı olarak test edilmiştir. Elde edilen sonuçlar, geliştirilen EYS algoritmasının uygulanabilirliğini ispatlamıştır. Bununla birlikte farklı optimizasyon periyotlarının, EYS performansı üzerindeki etkisi analiz edilmiştir. Optimizasyon için kullanılan tahminlerin kabul edilebilir doğrulukta olması durumunda, optimizasyon periyodunun uzamasının, EYS performansını arttırdığı gözlemlenmiştir. 6, 12 ve 24 saatlik periyotlar için yapılan deneyler sonucunda, 24 saatlik optimizasyon periyodunun, bataryanın daha efektif bir şekilde kullanılmasını ve mikro şebekenin daha ekonomik olarak işletilmesini sağladığı tespit edilmiştir.Article Citation - WoS: 3Citation - Scopus: 4Distribution, Scale, and Context Sensitive, Convolutional Neural Network-Based Soc Estimation for Li-Ion Batteries(Ieee-inst Electrical Electronics Engineers inc, 2025-02-01) Cimen, HalilLi-ion batteries play a crucial role in green energy goals, but estimating their parameters is challenging due to their nonlinear structure, aging effects, and varying chemistries. In this article, a distribution, scale and context sensitive, convolutional neural network-based state of charge estimation model is proposed. First, the proposed model improves generalization by addressing data distribution shifts in batteries across different temperatures through individual sample handling. Second, by stacking convolutional layers with varied receptive fields, the model captures both local and global dependencies, providing the model with multiscale features and hierarchical representation. Finally, we add a self-attention module to enhance learning of input sequences by focusing on relevant parts and understanding the global context of features. Experiments were performed on single-domain and cross-domain settings to prove the effectiveness of the model. The results obtained demonstrate that the proposed model significantly outperforms state-of-the-art approaches in terms of both accuracy and generalization capability.Conference Object Citation - WoS: 4A Dual-Input Multi-Label Classification Approach for Non-Intrusive Load Monitoring via Deep Learning(IEEE, 2020-05-01) Cetinkaya, Nurettin; Vasquez, Juan C.; Guerrero, Josep M.; Palacios-Garcia, Emilio J.; Cimen, HalilNon-intrusive load monitoring (NILM) is the process of obtaining appliance-level data from users' total electricity consumption data. These data can be of great benefit, especially in demand response applications. In this paper, a multi-label classification for NILM based on a two-input gated recurrent unit (GRU) is presented. Since the presented method is designed with a multi-label approach, great savings in training time are achieved. While a separate model is trained for each appliance in the literature, only one model is trained in the proposed model. Besides, the model was trained using two different inputs. The first is the total active power value consumed by the whole house. The second input is the Spikes obtained by analyzing this active power consumption. Simply put, spikes are obtained by analyzing the instant power changes in active power. Both inputs are evaluated with a convolutional layer and necessary features are extracted. Obtained features are fed into the GRU to be able to analyze time-dependent changes. The simulation results show that an additional input can slightly improve the analysis accuracy. Besides, it was found that the second input is useful especially in the analysis of short-term devices.Article Dinamik Oylama Tabanlı Topluluk Derin Öğrenme ile Benzer Mahsullerin Sınıflandırılması(2025-06-30) Eşme, Engin; Şen, Muhammed Arif; Cimen, HalilÜrün başına ilaçlama, sulama ve hasat gibi akıllı tarım uygulamalarını gerçekleştirebilen otonom makinelerin geliştirilmesinde derin öğrenme tabanlı yaklaşımlar, özellikle görüntü sınıflandırma ve veri analizi gibi görevlerde başarılı uygulamalar sergilemektedir. Tarımsal üretimde verimliliği artırmak ve sürdürülebilirliği sağlamak için bitki türlerinin doğru tanınması ve yabancı otlardan ayırt edilmesi kritik bir öneme sahiptir. Birbirine benzer görünüme sahip tarımsal mahsullerin sınıflandırılması, mevcut yöntemlerle zorlu bir problem olmaya devam etmektedir. Bu araştırma, benzer görünümlü tarımsal ürünlerin otomatik olarak tanınması ve sınıflandırılması konusunda önemli bir adım olmakla birlikte, akıllı tarım teknolojilerinin geliştirilmesine yönelik teorik ve pratik bir temel oluşturmaktadır. Bu çalışma, genellikle birbirine çok benzeyen tarımsal mahsul görüntülerini sınıflandırmayı amaçlayarak 17 farklı derin öğrenme modeli ve dinamik oylama yöntemini kullanmaktadır. Veri seti, kenevir, mısır, pirinç, şeker kamışı ve buğday olmak üzere beş benzer görünümlü mahsul türüne ait toplam 804 görüntüden oluşmaktadır. Güvenilirliği sağlamak amacıyla 10 katlı çapraz doğrulama kullanılmış ve tüm modellerde aynı örnek setleriyle tutarlı deneyler gerçekleştirilmiştir. Elde edilen sonuçlar modellerin birbirine çok benzer tarımsal mahsül görüntülerini sınıflandırma performanslarını doğruluk, eğitim süresi ve disk alanı açısından rapor etmektedir. Deneysel bulgulara göre, ShuffleNet test setinde %98,63 ile en yüksek bireysel doğruluğa ulaşmış, ancak topluluk yaklaşımı bu değeri %99,75'e yükseltmiştir. Önerilen topluluk yaklaşımı doğruluğu artırmakla kalmayıp daha fazla sağlamlık ve kararlılık sağlamaktadır. Ayrıca elde edilen teorik bilgi ve sonuçlar, bu alanda geliştirilecek akıllı tarım makinelerine entegre edilebilecek ve daha verimli şekilde çalışmasını sağlayacaktır.Conference Object Citation - WoS: 3Citation - Scopus: 8Mitigation of Voltage Unbalance in Microgrids Using Thermostatically Controlled Loads(IEEE, 2018) Çimen, Halil; Çetinkaya, NurettinWith the development of technology, some important developments occur in electrical energy systems. Especially smart grids and micro grids will be an important part of energy systems in the following years. With the increasing use of micro grids, more renewable power plants will be included in the system. However, the production of renewable energy sources varies according to the weather conditions. Therefore, the generated energy is not known beforehand. Increasing single-phase loads (such as electric vehicle charging stations), unpredictable renewable generation and uncertainty of consumption can lead to problems of voltage unbalance in the micro grid. This problem can damage the network in many ways. In this study, the problem of voltage unbalance occurring in the microgrid was tried to be reduced by demand side management. The comfort of the customer should not be affected when demand side management is performed. For this reason, the most suitable load for demand side management is thermostatically controlled loads (TCL). An algorithm has been proposed to reduce voltage unbalance by using these loads. This algorithm was tested with the help of PSCAD/EMTDC program.Article Citation - WoS: 17Citation - Scopus: 19Voltage Sensitivity-Based Demand-Side Management To Reduce Voltage Unbalance in Islanded Microgrids(INST ENGINEERING TECHNOLOGY-IET, 2019-07-19) Çimen, Halil; Çetinkaya, NurettinMicrogrids (MG) provide advantages such as providing energy supply to areas far from the distribution grid, efficient use of resources by supporting demand management and having a more dynamic grid. However, if an advanced control system is not implemented in the islanded MG, problems of power quality may arise. One of these problems is the voltage unbalance. Increasing the number of single-phase roof-mounted PV plants and the number of electric vehicle charging stations in recent times may negatively affect voltage unbalance. One of the methods used to mitigate this problem is the demand-side management (DSM). Here, a solution method based on DSM is presented for the voltage unbalance problem that may occur in an islanded MG. Thermostatically controlled loads, which are often used for DSM, are preferred as controllable loads. A new and novel control algorithm based on voltage sensitivity have been developed. Effects of TCLs on different buses and phases are determined with voltage sensitivity matrix including neutral components. The proposed control algorithm reduces successfully both the voltage unbalance factor and the number of controlled TCLs. The algorithm was tested with the PSCAD/EMTDC analysis software.Conference Object Multi-Scale Self-Attention Convolutional Neural Network for Energy Storage State-Of Estimation(Institute of Electrical and Electronics Engineers Inc., 2024-12-06) Çimen, H.State-of-Charge (SOC) estimation in Li-ion batteries increases the efficiency of energy management systems, extending battery life. Accurate SOC prediction helps users better meet their energy needs while optimizing the energy consumption of devices. However, the use of batteries under different operating conditions makes SOC estimation a challenge. In this paper, a deep learning-based model that can accurately predict the SOC of li-ion battery cells under different temperatures has been proposed. The model is able to extract different feature maps by analyzing the input sequence data at different scales. In this way, correlations between different time periods can be revealed. Another feature of the model is that it can detect long-term patterns in each feature map by analyzing the obtained multi-scale features with a self-attention mechanism. In this way, SOC prediction will be improved not only based on previous data but also based on the relevant points in the sequence. In the experiments for positive temperatures, an average of 29.1 % success increase was achieved for LA92 drive cycle and 12.3 % for UDDS. In the experiments for all temperatures, an average of 61.9% success increase was observed for LA92 drive cycle and 46.2 % for UDDS. © 2024 IEEE.Article Enhancing Generalization Performance of CNN-Based State-Of Estimation for Lithium-Ion Batteries(Elsevier, 2025-11-01) Cimen, Halil; Ucar, Kursad; Arabaci, HayriLithium-ion batteries are the most important component of electric vehicles. Since it has a chemical structure, the State-of-Charge (SOC) of the batteries cannot be determined precisely, so it is estimated by various methods. However, the generalization capability of the methods to data obtained from experiments at different temperatures and different batteries is still a major challenge. Moreover, the distribution shifting occurring in time series may also reduce the generalization ability. In this paper, the generalization capacity problem has been addressed, and a SOC estimator based on a convolutional neural network framework is proposed. The proposed method reduces the internal covariate shift during training by using batch normalization and improves the generalization performance by normalizing each instance independently by using instance normalization. The results have been compared with state-of-the-art SOC estimation methods and increased accuracy has been observed. Tests were carried out by creating different scenarios. In the experimental results, the benchmark models were outperformed by achieving a 57.1 % increase in MAE accuracy for tests with data obtained at all temperatures (-20 degrees C, -10 degrees C, 0 degrees C, 10 degrees C, 25 degrees C), 15.5 % for positive temperatures (0 degrees C, 10 degrees C, 25 degrees C) and 24.9 % for negative temperatures (-20 degrees C, -10 degrees C, 0 degrees C).
Research Topics
Domains
Physical Sciences
Fields
Engineering
Subfields
Electrical and Electronic EngineeringAutomotive EngineeringControl and Systems Engineering
Specific Research Areas
Smart Grid Energy Management
Advanced Battery Technologies Research
Microgrid Control and Optimization
Advancements in Battery Materials
Optimal Power Flow Distribution
Sustainable Development Goals
7AFFORDABLE AND CLEAN ENERGY
8
Research Products
13CLIMATE ACTION
2
Research Products
9INDUSTRY, INNOVATION AND INFRASTRUCTURE
2
Research Products
15LIFE ON LAND
1
Research Products
2ZERO HUNGER
1
Research Products
12RESPONSIBLE CONSUMPTION AND PRODUCTION
1
Research Products
4QUALITY EDUCATION
1
Research Products
8DECENT WORK AND ECONOMIC GROWTH
1
Research Products
Documents
19
Citations
593
h-index
8
Documents
19
Citations
439
Publication Collaboration
| Affiliation Name | Count |
|---|---|
| Konya Technical University | 17 |
| Aalborg University | 10 |
| Selçuk University | 4 |
| National Kaohsiung University of Science and Technology | 1 |
| Istanbul Technical University | 1 |
1 / 2
Data obtained from OpenAlex
| Journal | Count |
|---|---|
| Journal of Energy Storage | 2 |
| Applied Energy | 2 |
| 2018 2ND INTERNATIONAL SYMPOSIUM ON MULTIDISCIPLINARY STUDIES AND INNOVATIVE TECHNOLOGIES (ISMSIT) | 1 |
| 2020 ZOOMING INNOVATION IN CONSUMER TECHNOLOGIES CONFERENCE (ZINC) | 1 |
| 20th Ieee Mediterranean Eletrotechnical Conference (Ieee Melecon 2020) | 1 |
Current Page: 1 / 4
Scholarly Output
20
Articles
11
Views / Downloads
33/18
Supervised MSc Theses
0
Supervised PhD Theses
1
WoS Citation Count
335
Scopus Citation Count
445
Patents
0
Projects
0
WoS Citations per Publication
16.75
Scopus Citations per Publication
22.25
Open Access Source
3
Supervised Theses
1
Scopus Quartile Distribution
Competency Cloud
