Cemil Altin

Energy system optimization seeks to utilize multiple energy sources efficiently under technical, economic, and environmental constraints. The increasing integration of renewable energy and the need for sustainable operation have made the optimal planning and management of hybrid energy systems crucial. Classical optimization methods, including Linear Programming, Nonlinear Programming, and simulation-based models, often face limitations when addressing high-dimensional and nonlinear problems. This study introduces a deep learning–based surrogate modeling framework for sizing the components of hybrid renewable energy systems. Initially, Particle Swarm Optimization (PSO) is employed to determine the optimal component sizes for a large number of synthetically generated hourly solar irradiance and load profiles. These optimal solutions are then used as target labels. The associated annual time-series data are transformed into multi-channel Data Map (DMAP) images, which serve as inputs for convolutional neural networks (CNNs). After training, the CNN models are capable of directly estimating the required number of photovoltaic (PV) panels, inverter capacity, and battery units from the DMAP images, eliminating the need to perform the iterative PSO optimization during the prediction stage. Various convolutional neural network architectures, including ResNet, DenseNet121, RegNet, ConvNeXt, EfficientNet, SqueezeNet, MobileNet, and InceptionV3, were evaluated for this multi-output regression task. The results indicate that ResNet and DenseNet121 achieve the best performance, while ConvNeXt provides strong results with a modern architectural design. Among the evaluated models, DenseNet121 achieved coefficients of determination (R²) of 0.934, 0.988, and 0.947 for predicting the sizes of the PV array, inverter, and battery bank, respectively. These results correspond to an average prediction accuracy of approximately 90.6%. ResNet produced similar performance, with its highest R² value reaching 0.983 for inverter sizing. Lightweight networks such as SqueezeNet and MobileNet demonstrate notable effectiveness for resource-constrained systems, whereas InceptionV3 underperforms in leveraging its multi-scale architecture. These results demonstrate that, once the models have been trained, deep learning–based surrogate models can generate sizing decisions comparable to those obtained using PSO with only a fraction of the computational effort. As a result, they provide a fast and practical alternative to conventional iterative optimization methods for component sizing in smart grid and sustainable energy planning applications.

Solar power plants are among the most widely used renewable energy sources today. Varying radiation levels from region to region, and similarly varying consumption depending on the user within a given region, make the optimal sizing of these plants challenging. In this study, a machine learning-based surrogate model for the real-time sizing optimization of solar power plants, trained with a completely original dataset, has been developed. In the first stage, 500 different solar power plant installation scenarios were synthetically generated and evaluated in HOMER, and the obtained optimal sizing outputs were used as training targets for the proposed surrogate model rather than real operational data. The results obtained by applying various machine learning methods to the generated dataset are presented comparatively. Among 7 different machine learning models, XGBoost, Gradient Boosting, and LightGBM demonstrated the best performance. The developed model achieved an average R² score of 0.9425 for a total of 3 targets, while target-specific performance showed R² scores of 0.9747 for inverters, 0.9365 for PV panels, and 0.9165 for batteries. This model serves as a computationally efficient surrogate of the HOMER optimization process, enabling high-accuracy real-time predictions while significantly reducing the computational burden associated with intensive mathematical calculations, iterative procedures, and complex search spaces.

In radial distribution networks with many separate branches and sections, it is very difficult to calculate the optimum conductor cross-section. This work introduces a new tool for optimizing conductor cross-sectional area in electrical distribution systems by utilizing the Teaching-Learning-Based Optimization (TLBO) algorithm. The tool can optimize the conductor crossection of the multisection, branching distribution systems. The maximum current carrying capacity constraint is taken into consideration when formulating the objective function to choose the ideal conductor size for each network segment. The optimal conductor sizes are determined by both desired percent voltage drop and current carrying capacity of the conductor. By calculating the currents drawn from the line segments in advance, the search space of the optimization algorithm is narrowed. MATLAB and Excel were used to determine the ideal conductor size. The conductor, which is chosen using the suggested method, will preserve appropriate voltage levels in radial distribution systems while optimizing the overall savings in conducting material and energy loss costs. The outcomes show that the optimal selection of conductor problem can be solved by the TLBO algorithm in a practical and effective manner. Results of testing the suggested tool on a radial distribution system are noteworthy.

Bu çalışmada, Hibrit Yenilenebilr Enerji Siatemleri optimizasyonunda kullanılmak üzere bu alanda en sık kullanılan HOMER programının olumsuz yönlerini bertaraf eden metasezgisel Parçacık Sürü Optimizasyonu tabanlı optimizasyon aracı tasarımı yapılmıştır. Güvenilirliğinin ortaya konması nedeniyle aynı yükü beslemek için gerekli hibrit sistem, hem geliştirilen araç ile hemde HOMER optimizasyon yazılımı ile optimize edilerek sonuçlar karşılaştırılmıştır. İşlem hızı bakımından yapılan karşılaştırmada aynı sistemi optimize etmede, HOMER yazılımının 936 saniyede elde etmiş olduğu sonucu 17 saniyede elde ederek tasarlanan sistemin daha hızlı sonuca ulaştığı görülmüştür. Ekonomik yönden yapılan karşılaştırmada ise iki ayrı aracın yapmış olduğu optimizasyon sonucunda; Enerji maliyetinde 1.737 % , Toplam Net Bugün ki Maliyette 0.85 % ve Başlangıç Sermayesinde ise 1.895 % lik fark bulunmakta olup sonuçlar arasında önemli bir farkın bulunmadığı gözlenmiştir. Elektriksel sonuçların kıyasında ise Beslenen yüklerde 0.031 %, Beslenemeyen yüklerde 1.071 % lik fark çıkmış olup elektriksel anlamda da sonuçların güvenilirliği ortaya konmuştur. Sonuçlar bu alandaki detaylı optimizasyon çalışmaları için umut vericidir. Ayrıca HOMER yazılımının sunmakta olduğu, ekonomik ve elektriksel detaylar, gerçeği temsil eden özgün bir sevk stratejisi sayesinde tasarlanan araç ile de kullanıcıya sunulmaktadır. Güvenilirlik parametresi olarak HOMER yazılımının içerdiği fakat daha önce metasezgisel algoritmalarla yapılan optimizasyon süreçlerinde hiç kullanılmamış olan kapasite eksikliği parametresi kullanılmaktadır. Kapasite eksikliği parametresi ilk defa metasezgisel algoritma ile yenilenebilir enerji kaynaklarının optimizasyonunda kullanılmıştır. Maliyet fonksiyonu olarak Enerji Maliyeti kullanılmıştır. Kısacası HOMER programına alternatif çok daha hızlı sonuca ulaşabilen ve HOMER programındaki hassasiyet, hantallık ve zorlu arama uzayı oluşturma süreçlerini bertaraf eden özgün, güvenilir bir optimizasyon aracı tasarlanmıştır. Bu araç sayesinde ayrıca, vekil modelleri, makine öğrenmesi veya derin öğrenme tabanlı optimizasyon sistemlerini eğitmek için gerekli optimizasyon çıktıları çok hızlı bir şekilde elde edilerek fazla sayıda veri üretilmesi kolaylaşacaktır.

In this study, an optimisation tool that uses the differential evolution algorithm with a special distribution strategy is designed for the first time to be used in the optimisation of hybrid renewable energy systems. The developed tool and the hybrid optimisation model for multiple energy resources (HOMER) optimisation programme were compared. The tool is much faster than the HOMER programme and can produce almost the same results as HOMER. In addition, a heuristic-based optimisation technique was used for the first time to generate extremely comprehensive findings. The capacity shortage parameter, which is not used much in the literature, is used as a reliability parameter. The cost of energy (COE) was used as the cost function. The results are promising for detailed optimisation studies in this area.

Caudal epidural anaesthesia is usually the most well-known technique in obstetrics to deal with chronic back pain. Due to variations in the shape and size of the sacral hiatus (SH), its classification is a crucial and challenging task. Clinically, it is required in trauma, where surgeons must make fast and correct selections. Past studies have focused on morphometric and statistical analysis to classify it. Therefore, it is vital to automatically and accurately classify SH types through deep learning methods. To this end, we proposed the Multi-Task Process (MTP), a novel classification approach to classify the SH MTP that initially uses a small medical tabular data set obtained by manual feature extraction on computed tomography scans of the sacrums. Second, it augments the data set synthetically through a Generative Adversarial Network (GAN). In addition, it adapts a two-dimensional (2D) embedding algorithm to convert tabular features into images. Finally, it feeds images into Convolutional Neural Networks (CNNs). The application of MTP to six CNN models achieved remarkable classification success rates of approximately 90 % to 93 %. The proposed MTP approach eliminates the small medical tabular data problem that results in bone classification on deep models.

In this study, the wing profile, which is difficult to calculate and determine, has been optimized with the help of Foilsim data and optimization algorithms. Foilsim data provided by NASA (National Aeronautics and Space Administration) and used by many researchers, especially in developing model airplanes, has been provided to use in aircraft wing shape optimization. Although Foilsim is a very useful simulation program for designers, it cannot be used effectively in optimization processes due to its web environment. Lift coefficient is needed for Lift equation in airfoil shape optimization. Lift coefficient depends on angle, camber, and thickness of airfoil Calculation of Lift coefficient is difficult and needs heavy mathematical equations or real experiments. By using Foilsim data and optimization algorithm (Artificial Neural Networks: ANN, Artificial Bee Colony: ABC), wing angle, camber and thickness values, which are difficult to determine and calculate, were estimated and comparative experiments of the values were made. (Fixed Lift, Fixed Speed, Fixed Wing Area). Experimental results have shown that it is a useful study for airfoil shape optimization. In short, in this study, by using the Foilsim data and the optimization algorithm to provide the lifting force determined by the designer, the most suitable angle, camber, thickness values of the wing, which are difficult to determine and calculate, are determined to enable the production of efficient aircraft. The user enters the desired lift value into the ABC optimization algorithm and finds the required wing properties for the desired lift value.

Internet of things and Drones are two new promising innovative technologies which is inevitable in the internet era. These technologies provide modern solutions for many fields. One of these fields is agriculture. Agriculture plays pivot role for humankind, because more than half of the World’s population depends on agriculture. In this study internet of things technology is applied to a drone which is capable for doing agricultural works like spraying, carrying and real time monitoring. An on board android device which is mount on the drone is used to manage the drone over internet by a graphical user interface software designed within the study. The farmer communicates with on board android device over internet by remote desktop application in order to manage drone and get data. The drone will help farmers by getting live data from the farm and do necessary works remotely. The aim of this study is to enable farmers to do remote farming. Agricultural activities have declined in recent years with the increase in migration from the village to the city. Thus, farmers will be able to make remote farming.