One of the most important factors for successful agricultural production is the irrigation system in place. In this study, a precision irrigation system, which takes advantage of the various phases of plant growth, was developed and implemented using the sensor network technology integrated with IOS/Android. The amount of water in the soil was measured via sensors that were placed on certain points of the area to be irrigated. These sensors were placed near the root of the product. Data from sensors was transmitted via Wi-Fi in real-time to a mobile phone based on IOS/Android. In the light of obtained data, the seasonal precision irrigation system was created depending on the amount of water required by the plants at each stage of their growth stage. The required energy of the system was provided by solar energy. The system can be controlled by smart phones, which increases the usability of the system. When design performance was analyzed, it was observed that some important advantages such as obtaining high efficiency with water, time and energy saving and reducing the workforce were ensured. Five separate laterals were used for the irrigation system. There were valves on each lateral, which realized the opening and closing process depending on the water need. A total of 16 humidity sensors were used in the irrigation system and the data from these sensors was transferred to the IOS/Android server via the programmable controller (PLC). The basic electrical equipment in the irrigation system was monitored and controlled via mobile devices. Control parameters were obtained by comparing the real values and reference values by a closed-loop system and determine the new working status of the irrigation system.

At present, active detectors are commonly used for detection of land mines. Land mines can be detected with high precision through active detectors. However, the operating principle of active detectors can also lead to vital dangers. When detecting mines in the field, electrical signals sent to the environment from active detectors sometimes trigger the mine blasting mechanism and cause mine explosion. Another way to detect land mines without triggering the blasting mechanisms is to use passive detectors. The biggest handicap of passive detectors is that they cannot detect mines as much as active detectors. This causes that passive detectors are as dangerous as at least active detectors. In this case, passive detectors can cause dangerous results like active detectors. In this paper, we have developed solutions that eliminate the handicaps of passive mine detectors. For this purpose, a new approach, which is established on artificial intelligence based on the magnetic anomaly, measurement height, and soil type, is suggested. The experimental setup is designed to verify and test the proposed approach. In this respect, the actual data measured under different conditions were recorded and processed with modern and effective artificial intelligence techniques; and alternative models were developed. With the proposed approach, the mines are detected with a success rate of 98.2%. This success rate in detection is promising for the passive mine detectors. A significant contribution of the developed model in terms of literature is the successful classification as well as the detection of mines. In experimental studies conducted with real data, five different types of mines are classified as 85.8% success rate. The proposed model has been a pioneering study on mine classification in the literature. Moreover, the realization of this paper with a passive mine detector proves the success of the proposed approach.

The use of Unmanned Aerial Vehicles (UAVs) has rapidly expanded across various applications in recent years. The autonomy requirements for UAVs include obstacle recognition, obstacle avoidance, and Safe Landing Zone (SLZ) detection. A fundamental challenge common to these vehicles is path planning, which determines where they should navigate and how they should find their routes. Although advancements in technologies utilizing computer vision have accelerated developmental progress, achieving a fully autonomous system necessitates solving multiple complex problems. During flight, UAVs must be capable of recognizing and assessing unexpected situations, generating a new path to continue operation, and ultimately completing the return journey and landing safely. This review provides a comprehensive perspective on the applicability of methods used to address autonomy-related challenges in UAVs and examines trends in recent studies. A total of 211 studies conducted between 2000 and 2025 were analyzed, revealing a three-phase growth pattern: 2000–2007 exhibited a low number of publications and a stagnant period; 2007–2014 showed a moderate increase in publications with rising awareness; and 2014–2025 experienced a significant surge in published articles. The results indicate that SLZ detection is the most frequently addressed issue, while visual-based control techniques are the most commonly employed methods. This study enables an evaluation of which methods are predominantly used in different application areas, the types of analyses conducted to validate proposed solutions, and which UAV types are more suitable for specific applications.

In this study, breast cancer classification as benign or malignant was made using images obtained by histopathological procedures, one of the medical imaging techniques. First of all, different noise types and several intensities were added to the images in the used data set. Then, the noise in images was removed by applying the Wavelet Transform (WT) process to noisy images. The performance rates in the denoising process were found out by evaluating Peak Signal to Noise Rate (PSNR) values of the images. The Gaussian noise type gave better results than other noise types considering PSNR values. The best PSNR values were carried out with the Gaussian noise type. After that, the denoised images were classified by Convolution Neural Network (CNN), one of the deep learning techniques. In this classification process, the proposed CNN model and the VggNet-16 model were used. According to the classification result, better results were obtained with the proposed CNN model than VggNet-16. The best performance (86.9%) was obtained from the data set created Gaussian noise with 0.3 noise intensity.

Runge Kutta (RUN) is an up-to-date and well-founded metaheuristic algorithm. The RUN algorithm aims to find the global best in solving problems by going beyond the traps of metaphors. For this purpose, enhanced solution quality mechanism is used to avoid local optimum solutions and increase the convergence speed. Although the RUN algorithm offers promising solutions, it is seen that this algorithm has shortcomings, especially in solving high dimensional multimodal problems. In this study, the solution candidates that guide the search process in the RUN algorithm are developed using the Fitness-Distance Balance (FDB) method. Thus, using the FDB-based RUN algorithm, the global optimum value of many optimization problems will be obtained in the future. CEC 2020 which has current benchmark problems was used to test the performance of the developed FDB-RUN algorithm. 10 different unconstrained benchmark problems taken from CEC 2020 were designed by arranging them in 30/50/100 dimensions. Experimental studies were carried out using the designed benchmark problems and analyzed with Friedman and Wilcoxon statistical test methods. According to the results of the analysis, it was seen that the FDB-RUN variations showed a superior performance compared to the base algorithm (RUN) in all experimental studies. In particular, it has been shown to provide more effective results for the continuous optimization of high-dimensional problems.

This study aims to improve the energy performance of residential buildings. heating load (HL) and cooling load (CL) are considered as a measure of heating ventilation and air conditioning (HVAC) system in this process. In order to achive an effective estimation, hybrid machine learning algorithms including, artificial bee colony-based k-nearest neighbor (abc-knn), genetic algorithm-based knn (ga-knn), adaptive artificial neural network with genetic algorithm (ga-ann) and adaptive ann with artificial bee colony (abc-ann) are used. Results are compared classical knn and ann methods. Thence, relations between input and target parameters are defined and performance of well-known classical knn and ann is improved substantialy.