Selçuk Kürşat İşLeyen

This paper addresses the coordinated task assignment and route planning for a fleet of heterogeneous unmanned air vehicles (UAVs) operating in a stochastic threat environment. Each UAV is capable of engaging and destroying designated targets, while enemy air defenses pose risks of UAV losses based on specified survival probabilities. The study aims to minimize mission tardiness within a fixed time horizon and maximize cumulative mission rewards by neutralizing the highest number of targets while minimizing UAV losses. We first formulate the problem as a stochastic, multi-objective mixed-integer linear program defined on a graph, integrating task scheduling, timing, and connectivity constraints to simultaneously optimize UAV routes and target assignments. For small-scale instances, this formulation can be solved optimally. However, the problem’s NP-hard, combinatorial nature renders exact methods impractical for larger instances. To address this, we develop a novel two-stage heuristic, combining simulated annealing and genetic algorithms. In the first stage, simulated annealing generates a robust initial task allocation common to all scenarios. In the second stage, a genetic algorithm refines the routes and assignments independently for each scenario. A custom scenario generator was developed to vary survival probabilities, target densities, distance metrics, and UAV capacities to rigorously test solution quality. Computational experiments demonstrate that the proposed heuristic produces high-quality solutions across various settings within tractable time frames, offering decision makers a practical tool to plan UAV-based airstrike operations under uncertainty.

This study addresses the cooperative multitask assignment and path planning of heterogenous UAVs with aircraft kinematics. The purpose of this study is to minimize the maximum mission time and total distance traveled by UAV fleet which execute multi-tasks in a military theatre composed of targets with air defense threat circle as well as no-fly zones. This study aims to optimize multi-objectives with structural and timing constraints. The study introduces a complex and well-structured mixed integer linear program, which successfully optimize path planning and multitask allocation of UAVs simultaneously. Contrary to common fashion in literature, the proposed model encompasses most of the real-life military requirements for UAV tactical operations, rather than leaving them as assumptions or integrating some of them by the results of commercial simulators. In terms of path planning phase, the study takes into account Dubin’s distances complying with the flight dynamics of attack and surveillance UAVs, and no-fly zones in the theater. Regarding task allocation, the model in the study satisfies task order requirements of targets such as classification, attack and verification in the respective order by avoiding deadlock caused by waiting cycles over target. Targets are threats to UAVs, so the study also considers the probability of UAV loss during the mission scenario by developing chance constraints integrated in to the model. The model is coded in Matlab, and solved by Gurobi global optimal solver successfully under predefined realistic operative scenarios.

This research delves into the dynamic landscape of transportation systems, with a specific focus on the integration of drones and conventional vehicles. The study presents a Mixed Integer Programming (MIP) model for the Capacitated Multi-Drone Assisted Vehicle Routing Problem (mDroneCVRP), aiming to minimize the time of the last vehicle's arrival at the warehouse. It is essential to highlight that the proposed model was effectively solved using the CPLEX algorithm within the GAMS framework, underscoring the sophistication of the solution approach. The integration of multiple drones into the routing process proves to be instrumental in significantly reducing service time, demonstrating the efficacy of synergizing drone and truck operations. As the number of nodes escalates, emphasizing the necessity for heuristic approaches to address larger instances, the study provides valuable insights into the judicious use of drones in synchronized routing operations. Furthermore, the research challenges conventional assumptions by permitting drones to take off from and land on different vehicles, thereby augmenting operational capabilities and adeptly tackling contemporary transportation challenges.

As technology advances, the integration of robots in the assembly line has become widespread. While robots offer numerous benefits, such as increased productivity and improved product quality, they also result in higher energy usage. Finding the optimal line balance while considering energy consumption is a challenging task in a robotic assembly line that produces multiple product models in a mixed sequence. This paper addresses the mixed-model parallel robotic assembly line balancing and model sequencing (MPRALB/S) problem. The objectives of this problem are to minimize cycle time and energy consumption. The authors have not found any existing research on this topic in the literature. To solve the MPRALB/S problem, a modified non-dominated sorting genetic algorithm II (MNSGA-II) is developed. Since there is no existing benchmark data for the MPRALB/S problem, new datasets are generated for this study. The MPRALB/S problem is illustrated through a numerical example. The performance of MNSGA-II is evaluated with non-dominated sorting genetic algorithm II (NSGA-II) and restarted simulated annealing through commonly used performance metrics including hypervolume ratio (HVR), ratio of non-dominated solutions (RP) and generational distance (GD). According to the results of the computational study, MNSGA-II outperforms NSGA-II in approximately 81% of the problem instances for HVR, 71% for RP, and 76% for GD. The results show that MNSGA-II is an effective approach for solving the MPRALB/S problem and achieves competing performance compared to other algorithms.

The Multi-Product Inventory-Location-Routing Problem with heterogeneous fleet considers a supply chain, which comprises multiple producers, potential distribution centers (DCs) with opening capacity levels and geographically scattered retailers each of which has deterministic demand over a discrete planning horizon. The goal is determining a set of DCs with their capacity levels to open, assigning retailers to the opened DCs, finding product quantities to be ordered by and distributed from opened DCs and determining the fleet and routes to satisfy the demands of retailers with minimum cost. A mixed-integer linear programming model is proposed to describe the problem, which is strengthened by two valid inequalities. Since the commercial solver can only solve the very small-sized instances within a reasonable time, two heuristic methods are developed. Results show that the proposed valid inequalities are effective and both methods provide important savings in acceptable run times compared to the commercial solver.

Teknolojik ve ekonomik ilerlemeyle birlikte sınırların ortadan kalkması, müşteri taleplerinin içerikleri ve hacimlerinin sürekli bir şekilde değişmesini ve beraberinde işletmelerin bu değişimlere hızlı cevap verebilmesi gerekliliğini ortaya çıkarmıştır. Bu ihtiyaç, işletmeleri kitlesel üretim sistemlerinin hızını ve esnek üretim sistemlerinin de esneklik kabiliyetini birlikte karşılayacak modern üretim sistemlerine yöneltmiştir. Bu üretim sistemlerinden bir tanesi de Japonya’da 90’lı yıllarda elektronik sektöründe kullanılmaya başlanan seru üretim sistemleridir. Bu sistemler, geleneksel montaj hatlarını seru adı verilen hücrelere bölerek, bir ürünün üretimindeki tüm görevlerin tek ya da birden fazla yüksek yetenekli çalışan tarafından gerçekleştirildiği sistemlerdir. Bu çalışmanın amacı, kitle üretimin hızını ve esnek üretimin ürün çeşitliliği sağlama kapasitesini karşılayan seru üretim sistemi için optimum seru yükleme planları elde etmektir. Bu nedenle, seru üretim sistemlerinde karşılaşılan problemlerden olan seru yükleme problemi ele alınmıştır. Belirli bir seru tasarımına ait optimal işçi ve iş yükleme planını elde etmeye yönelik karışık tamsayılı yeni bir matematiksel model ve bir mat-sezgisel algoritma önerilmiştir. Modelden ve sezgisel yöntemden elde edilen sonuçların etkinliği, literatürden alınan gerçek bir uygulama veri setiyle karşılaştırmalı olarak gösterilmiş karşılaştırma sonucunda elde edilen seru yükleme planlarının etkin olduğu gözlemlenmiştir.

Fishery industry is one of the main sources of income and most important subsectors for national economies. Nevertheless, natural fish sources have unfortunately diminished recently and not every fish species can be grown in every region due to some reasons. Moreover, many countries have limited resources to meet the need for fish. Therefore, aquaculture comes into prominence to eliminate these problems. However, taking right decisions and selecting the right fish breeds become crucial for fish farming. In this context, the aim of this study is to develop a Decision Support System (DSS) with intent to enable decision makers to determine most suitable fish species according to the features of their farms easily. The system was built based on Classification and Regression Trees algorithm, one of the data mining techniques. Sixty-two breeds of fish and thirteen factors affecting their growth was studied to create a database. The results show that the suggested DSS functions successfully in terms of not only determining appropriate and profitable fish species but also using existing resources more efficiently. It is expected that foreign trade volume will be increased with the raising productivity and; hence, countries will create new business branches which will have reflections in employment figures in the long run.

Silahlı İnsansız Hava Araçları, birçok ülkenin ulusal güvenliğini sağlamak adına askeri operasyonlarda yoğun bir şekilde kullandıkları yapay zekâya dayalı savunma sistemleridir. Bu sistemler sayesinde operasyon alanındaki hareketli ve hareketsiz hedefler, zorlu coğrafik koşullar altında pilot kullanılmaksızın kumanda merkezi yardımıyla imha edilebilmektedir. İnsansız hava aracı filosu tarafından, seyir süresi, mühimmat kapasitesi, yakıt maliyeti ve zaman penceresi kısıtlamaları dikkate alınarak sistemdeki hareketli hedeflerin başarılı bir şekilde imha edilmesi gereksinimi, Hareketli Müşterili Araç Rotalama Problemini ortaya çıkarmaktadır. Bu çalışmada Heterojen Filolu-Zaman Pencereli-Kapasite Kısıtlı Hareketli Müşterili Araç Rotalama Problemi, minimum görev süresi ve görev maliyeti amaçları doğrultusunda çözülmesi amaçlanmıştır. Problemin çözümü için sezgisel algoritmalar(ÇARA, RASA) geliştirilmiş ve metasezgisel algoritmalar (Genetik Algoritma, NSGA-II ve Tavlama Benzetimi Algoritması) kullanılmıştır. Önerilen algoritmaların etkinliği vurucu sayısının 5-10, hedef sayısının 10-35 arasında değiştiği 30 farklı deney seti üzerinde test edilmiştir. Algoritmalar için uygun parametre setinin belirlenmesinde Taguchi yönteminden yararlanılmıştır. Analiz sonucunda Genetik Algoritmanın diğer algoritmalara kıyasla daha başarılı sonuçlar ürettiği tespit edilmiştir.

It is a known fact that basic health care services cannot reach the majority of the population due to poor geographical accessibility. Unless quantitative location-allocation models and geographic information systems (GIS) are used, the final decision may be made on pragmatic considerations which can result far from optimal. In this paper, current and possible (or potential) new locations of pharmacy warehouses in Gaziantep are investigated to provide optimal distribution of hospitals and pharmacies. To do so, first of all, geographic information of 10 current and 10 potential pharmacy warehouses, 231 pharmacies and 29 hospitals are gathered using GIS. Second, a set covering mathematical model is handled to determine coverage capability of current and potential pharmacy warehouses and minimize the number of warehouses to be opened. Finally, P-center and P-median mathematical models are applied to open potential warehouses and to assign pharmacies & hospitals to the opened warehouses so that the total distance and the demand’s longest distance to the source are minimized. Developed integer programming (IP) models and GIS software are compared with on a case study. Computational experiments prove that our approach can find new potential pharmacy warehouses which cover wider areas than current warehouses to service pharmacies and hospitals in the city.