Person:
YILDIZ, BETÜL SULTAN

Loading...
Profile Picture

Email Address

Birth Date

Research Projects

Organizational Units

Job Title

Last Name

YILDIZ

First Name

BETÜL SULTAN

Name

Search Results

Now showing 1 - 10 of 17
  • Publication
    Optimum design of a seat bracket using artificial neural networks and dandelion optimization algorithm
    (Walter de Gruyter Gmbh, 2023-10-13) Erdaş, Mehmet Umut; Kopar, Mehmet; Yıldız, Betül Sultan; Yıldız, Ali Rıza; Erdaş, Mehmet Umut; Kopar, Mehmet; YILDIZ, BETÜL SULTAN; YILDIZ, ALİ RIZA; Bursa Uludağ Üniversitesi/Mühendislik Fakültesi/Otomotiv Mühendisliği Bölümü.; Bursa Uludağ Üniversitesi/Mühendislik Fakültesi/Makine Mühendisliği Bölümü.; 0000-0003-1790-6987; AAH-6495-2019; F-7426-2011; CNV-1200-2022; DBQ-9849-2022
    Nature-inspired metaheuristic algorithms are gaining popularity with their easy applicability and ability to avoid local optimum points, and they are spreading to wide application areas. Meta-heuristic optimization algorithms are used to achieve an optimum design in engineering problems aiming to obtain lightweight designs. In this article, structural optimization methods are used in the process of achieving the optimum design of a seat bracket. As a result of topology optimization, a new concept design of the bracket was created and used in shape optimization. In the shape optimization, the mass and stress values obtained depending on the variables, constraint, and objective functions were created by using artificial neural networks. The optimization problem based on mass minimization is solved by applying the dandelion optimization algorithm and verified by finite element analysis.
  • Publication
    Gradient-based optimizer for economic optimization of engineering problems
    (Walter De Gruyter Gmbh, 2022-05-25) Mehta, Pranav; Sait, Sadiq M.; Yıldız, Betül Sultan; YILDIZ, BETÜL SULTAN; Yıldız, Ali Rıza; YILDIZ, ALİ RIZA; Bursa Uludağ Üniversitesi/Mühendislik Fakültesi/Makine Mühendisliği Bölümü.; 0000-0002-4796-0581; 0000-0003-1790-6987; F-7426-2011; AAL-9234-2020; B-3604-2008
    Optimization of the heat recovery devices such as heat exchangers (HEs) and cooling towers is a complex task. In this article, the widely used fin and tube HE (FTHE) is optimized in terms of the total costs by the novel gradient-based optimization (GBO) algorithm. The FTHE s have a cylindrical tube with transverse or longitudinal fin enhanced on it. For this study, various constraints and design variables are considered, with the total cost as the objective function. The study reveals that the GBO provides promising results for the present case study with the highest success rate. Also, the comparative results suggest that GBO is the robust optimizer in terms of the best-optimized values of the fitness function vis-a-vis design variables. This study builds the future implications of the GBO in a wide range of engineering optimization fields.
  • Publication
    A comparative study of state-of-the-art metaheuristics for solving many-objective optimization problems of fixed wing unmanned aerial vehicle conceptual design
    (Springer, 2023-04-11) Anosri, Siwakorn; Panagant, Natee; Champasak, Pakin; Bureerat, Sujin; Thipyopas, Chinnapat; Kumar, Sumit; Pholdee, Nantiwat; Yıldız, Betül Sultan; Yıldız, Ali Riza; YILDIZ, ALİ RIZA; YILDIZ, BETÜL SULTAN; Bursa Uludağ Üniversitesi/Mühendislik Fakültesi/Makine Mühendisliği Bölümü.; 0000-0001-7592-8733 ; AAH-6495-2019; F-7426-2011
    The complexity of aircraft design problems increases with many objectives and diverse constraints, thus necessitating effective optimization techniques. In recent years many new metaheuristics have been developed, but their implementation in the design of the aircraft is limited. In this study, the effectiveness of twelve new algorithms for solving unmanned aerial vehicle design issues is compared. The optimizers included Differential evolution for multi-objective optimization, Many-objective nondominated sorting genetic algorithm, Knee point-driven evolutionary algorithm for many-objective optimization, Reference vector guided evolutionary algorithm, Multi-objective bat algorithm with nondominated sorting, multi-objective flower pollination algorithm, Multi-objective cuckoo search algorithm, Multi-objective multi-verse optimizer, Multi-objective slime mould algorithm, Multi-objective jellyfish search algorithm, Multi-objective evolutionary algorithm based on decomposition and Self-adaptive many-objective meta-heuristic based on decomposition. The design problems include four many-objective conceptual designs of UAV viz. Conventional, Conventional with winglet, Twin boom and Canard, which are solved by all the optimizers employed. Widely used Hypervolume and Inverted Generational Distance metrics are considered to evaluate and compare the performance of examined algorithms. Friedman's rank test based statistical examination manifests the dominance of the DEMO optimization technique over other compared techniques and exhibits its effectiveness in solving aircraft conceptual design problems. The findings of this work assist in not only solving aircraft design problems but also facilitating the development of unique algorithms for such challenging issues.
  • Publication
    A novel hybrid flow direction optimizer-dynamic oppositional based learning algorithm for solving complex constrained mechanical design problems
    (Walter de Gruyter Gmbh, 2023-01-27) Yıldız, Betül Sultan; Pholdee, Nantiwat; Mehta, Pranav; Sait, Sadiq M.; Kumar, Sumit; Bureerat, Sujin; Yıldız, Ali Rıza; YILDIZ, BETÜL SULTAN; YILDIZ, ALİ RIZA; Bursa Uludağ Üniversitesi/Makine Mühendisliği Bölümü; AAL-9234-2020; F-7426-2011
    In this present work, mechanical engineering optimization problems are solved by employing a novel optimizer (HFDO-DOBL) based on a physics-based flow direction optimizer (FDO) and dynamic oppositional-based learning. Five real-world engineering problems, viz. planetary gear train, hydrostatic thrust bearing, robot gripper, rolling bearing, and multiple disc clutch brake, are considered. The computational results obtained by HFDO-DOBL are compared with several newly proposed algorithms. The statistical analysis demonstrates the HFDO-DOBL dominance in finding optimal solutions relatively and competitiveness in solving constraint design optimization problems.
  • Publication
    A novel generalized normal distribution optimizer with elite oppositional based learning for optimization of mechanical engineering problems
    (Gmbh, 2023-02-23) Mehta, Pranav; Yıldız, Betuel Sultan; Pholdee, Nantiwat; Kumar, Sumit; Riza Yıldız, Ali; Sait, Sadiq M. M.; Bureerat, Sujin; YILDIZ, BETÜL SULTAN; YILDIZ, ALİ RIZA; Bursa Uludağ Üniversitesi/Mühendislik Fakültesi/Makine Mühendisliği Bölümü; F-7426-2011; AAH-6495-2019
    Optimization of engineering discipline problems are quite a challenging task as they carry design parameters and various constraints. Metaheuristic algorithms can able to handle those complex problems and realize the global optimum solution for engineering problems. In this article, a novel generalized normal distribution algorithm that is integrated with elite oppositional-based learning (HGNDO-EOBL) is studied and employed to optimize the design of the eight benchmark engineering functions. Moreover, the statistical results obtained from the HGNDO-EOBL are collated with the data obtained from the well-established algorithms such as whale optimizer, salp swarm optimizer, LFD optimizer, manta ray foraging optimization algorithm, hunger games search algorithm, reptile search algorithm, and INFO algorithm. For each of the cases, a comparison of the statistical results suggests that HGNDO-EOBL is superior in terms of realizing the prominent values of the fitness function compared to established algorithms. Accordingly, the HGNDO-EOBL can be adopted for a wide range of engineering optimization problems.
  • Publication
    A new hybrid artificial hummingbird-simulated annealing algorithm to solve constrained mechanical engineering problems
    (Walter de Gruyter Gmbh, 2022-07-26) Yıldız, Betül Sultan; Mehta, Pranav; Sait, Sadiq M.; Panagant, Natee; Kumar, Sumit; Yıldız, Ali Rıza; YILDIZ, BETÜL SULTAN; YILDIZ, ALİ RIZA; Bursa Uludağ Üniversitesi/Mühendislik Fakültesi/Makine Mühendisliği Bölümü.; AAL-9234-2020; F-7426-2011
    Nature-inspired algorithms known as metaheuristics have been significantly adopted by large-scale organizations and the engineering research domain due their several advantages over the classical optimization techniques. In the present article, a novel hybrid metaheuristic algorithm (HAHA-SA) based on the artificial hummingbird algorithm (AHA) and simulated annealing problem is proposed to improve the performance of the AHA. To check the performance of the HAHA-SA, it was applied to solve three constrained engineering design problems. For comparative analysis, the results of all considered cases are compared to the well-known optimizers. The statistical results demonstrate the dominance of the HAHA-SA in solving complex multi-constrained design optimization problems efficiently. Overall study shows the robustness of the adopted algorithm and develops future opportunities to optimize critical engineering problems using the HAHA-SA.
  • Publication
    A novel hybrid fick's law algorithm-quasi oppositional-based learning algorithm for solving constrained mechanical design problems
    (Walter De Gruyter Gmbh, 2023-09-13) Mehta, Pranav; Sait, Sadiq M.; Yıldız, Ali Rıza; YILDIZ, ALİ RIZA; Yıldız, Betül Sultan; YILDIZ, BETÜL SULTAN; Bursa Uludağ Üniversitesi/Mühendislik Fakültesi/Makina Mühendisliği Bölümü.; AAL-9234-2020; F-7426-2011
    In this article, a recently developed physics-based Fick's law optimization algorithm is utilized to solve engineering optimization challenges. The performance of the algorithm is further improved by incorporating quasi-oppositional-based techniques at the programming level. The modified algorithm was applied to optimize the rolling element bearing system, robot gripper, planetary gear system, and hydrostatic thrust bearing, along with shape optimization of the vehicle bracket system. Accordingly, the algorithm realizes promising statistical results compared to the rest of the well-known algorithms. Furthermore, the required number of iterations was comparatively less required to attain the global optimum solution. Moreover, deviations in the results were the least even when other optimizers provided better or more competitive results. This being said that this optimization algorithm can be adopted for a critical and wide range of industrial and real-world challenges optimization.
  • Publication
    A nelder mead-infused info algorithm for optimization of mechanical design problems
    (Walter De Gruyter Gmbh, 2022-08-26) Mehta, Pranav; Yıldız, Betül S.; Kumar, Sumit; Pholdee, Nantiwat; Sait, Sadiq M.; Panagant, Natee; Bureerat, Sujin; Yıldız, Ali Rıza; YILDIZ, BETÜL SULTAN; YILDIZ, ALİ RIZA; Bursa Uludağ Üniversitesi/Mühendislik Fakültesi/Makine Mühendisliği Bölümü.; AAL-9234-2020; F-7426-2011
    Nature-inspired metaheuristic algorithms have wide applications that have greater emphasis over the classical optimization techniques. The INFO algorithm is developed on the basis of the weighted mean of the vectors, which enhances the superior vector position that enables to get the global optimal solution. Moreover, it evaluates the fitness function within the updating stage, vectors combining, and local search stage. Accordingly, in the present article, a population-based algorithm named weighted mean of vectors (INFO) is hybridized with the Nelder-Mead algorithm (HINFO-NM) and adapted to optimize the standard benchmark function structural optimization of the vehicle suspension arm. This provides a superior convergence rate, prevention of trapping in the local search domain, and class balance between the exploration and exploitation phase. The pursued results suggest that the HINFO-NM algorithm is the robust optimizer that provides the best results compared to the rest of the algorithms. Moreover, the scalability of this algorithm can be realized by having the least standard deviation in the results. The HINFO-NM algorithm can be adopted in a wide range of optimization challenges by assuring superior results obtained in the present article.
  • Publication
    Chaotic marine predators algorithm for global optimization of real-world engineering problems
    (Elsevier, 2022-12-24) Kumar, Sumit; Yıldız, Betül Sultan; Mehta, Pranav; Panagant, Natee; Sait, Sadiq M.; Mirjalili, Seyedali; Yıldız, Ali Riza; YILDIZ, BETÜL SULTAN; Bursa Uludağ Üniversitesi/Makine Mühendisliği Bölümü; 0000-0002-7493-2068; AAL-9234-2020
    A novel metaheuristic called Chaotic Marine Predators Algorithm (CMPA) is proposed and investigated for the optimization of engineering problems. CMPA integrates the exploration merits of the recently proposed Marine Predators Algorithm (MPA) with the chaotic maps exploitation capabilities. Several chaotic maps were applied in the proposed CMPA to govern MPA parameters that eventually led to controlled exploration and exploitation of search. This study makes an initial attempt to explore and employ CMPA in decoding complex and challenging design and manufacturing problems. For performance evaluation of the proposed algorithm, CEC 2020 numerical problems having different dimensions and five widely adopted constrained design problems were solved. For all problems, both qualitative and qualitative results are examined and discussed. Moreover, two case studies of multi pass turning were examined by the proposed CMPA algorithm to optimize the cutting operation with a minimum cost of production per unit objective. Furthermore, the suggested CMPA algorithm has been investigated for solving a real-world structural topology optimization problem. Statistical analysis is performed, and the results of CMPA are compared with twelve distinguished algorithms. Outcomes of the proposed variant algorithm on the benchmarks demonstrate its significantly improved performance relative to other optimizers including a variant of MPA and two state-of-the-art IEEE CEC competitions winners algorithms. Findings from the manufacturing process exhibit CMPA proficiency in solving arduous real-world design problems.
  • Publication
    Optimum design of a composite drone component using slime mold algorithm
    (Walter De Gruyter Gmbh, 2023-09-25) YILDIZ, ALİ RIZA; YILDIZ, BETÜL SULTAN; Kopar, Mehmet; Bursa Uludağ Üniversitesi/Mühendislik Fakültesi/Otomotiv Mühendisliği Bölümü.; Bursa Uludağ Üniversitesi/Mühendislik Fakültesi/Makine Mühendisliği Bölümü.; 0000-0003-1790-6987; AAL-9234-2020; F-7426-2011
    Composite materials have a wide range of applications in many industries due to their manufacturability, high strength values, and light filling. The sector where composite materials are mostly used is the aviation industry. Today, as a result of the development of aviation systems, drones have started to be actively used, and many studies have started to be carried out to mitigate them. In this study, the subcarrier part, which is part of the drone, was designed using glass and carbon fiber-reinforced composite materials. Using the data obtained at the end of the analysis, the stacking angle with the optimal displacement and stress value was determined by using the genetic algorithm (GA), gray wolf algorithm (GWO), and slime mold optimization (SMO) techniques in order to develop a carrier with a minimum displacement and stress value of more than 60 MPa. As a result of the optimization, it was determined that artificial intelligence algorithms could be used effectively in determining the stacking angle of composite materials, and the optimum values were determined in the slime mold algorithm.