Article
Computer Science, Artificial Intelligence
Benyamin Abdollahzadeh, Farhad Soleimanian Gharehchopogh, Seyedali Mirjalili
Summary: Metaheuristics play a crucial role in solving optimization problems, often inspired by the collective intelligence of natural organisms. This paper introduces a new metaheuristic algorithm, GTO, inspired by gorilla troops' social intelligence in nature. Results show that the GTO outperforms existing metaheuristics on most benchmark functions and engineering problems, especially in high-dimensional scenarios.
INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Mahdis Banaie-Dezfouli, Mohammad H. Nadimi-Shahraki, Zahra Beheshti
Summary: The paper proposes a representative-based grey wolf optimizer (R-GWO) to address the weaknesses of GWO, introducing a search strategy that enhances exploration and diversity. Experimental results demonstrate that R-GWO outperforms competitor algorithms on most benchmark functions and engineering design problems, with an overall effectiveness of 95.4%.
APPLIED SOFT COMPUTING
(2021)
Article
Energy & Fuels
Hai Tao, Ahmed A. Ewees, Ali Omran Al-Sulttani, Ufuk Beyaztas, Mohammed Majeed Hameed, Sinan Q. Salih, Asaad M. Armanuos, Nadhir Al-Ansari, Cyril Voyant, Shamsuddin Shahid, Zaher Mundher Yaseen
Summary: The study introduced a novel intelligence model by hybridizing ANFIS with two metaheuristic optimization algorithms for accurate global solar radiation prediction. The proposed model outperformed other models by 25.7%-54.8% in terms of accuracy, showing potential for improvement in prediction accuracy through hybridization.
Article
Computer Science, Artificial Intelligence
Qingke Zhang, Hao Gao, Zhi-Hui Zhan, Junqing Li, Huaxiang Zhang
Summary: This paper proposes a novel metaheuristic optimizer called the growth optimizer (GO), which is inspired by the learning and reflection mechanisms of individuals in their growth processes. The algorithm is tested on the 2017 IEEE Congress on Evolutionary Computation test suite and shows competitive performance compared to 50 state-of-the-art metaheuristic algorithms. It also demonstrates promising results in solving real-world optimization problems. The source code of the GO algorithm is publicly available for access.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Mathematics
Juan Li, Qing An, Hong Lei, Qian Deng, Gai-Ge Wang
Summary: Levy flight is a random walk mechanism used in metaheuristic algorithms to solve NP-hard problems. It exhibits a movement pattern of large and small jumps in local areas, allowing it to escape local optima and expand the search area. Research shows the superiority of Levy flight-based metaheuristic algorithms in various fields.
Article
Engineering, Multidisciplinary
Alma Rodriguez, Octavio Camarena, Erik Cuevas, Itzel Aranguren, Arturo Valdivia-G, Bernardo Morales-Castaneda, Daniel Zaldivar, Marco Perez-Cisneros
Summary: This paper introduces a modified version of Grey Wolf Optimizer, which achieves a better balance, increased accuracy, and avoids convergence at local minima through synchronous-asynchronous processing and increasing diversity operations.
APPLIED MATHEMATICAL MODELLING
(2021)
Article
Computer Science, Artificial Intelligence
Changting Zhong, Gang Li, Zeng Meng, Wanxin He
Summary: In this work, the EOOBLE algorithm is proposed to solve high-dimensional global optimization problems. It combines the opposition-based learning strategy, the Levy flight strategy, and the evolutionary population dynamics strategy to improve the convergence capacity and performance. Experimental results show that the EOOBLE algorithm outperforms other state-of-the-art metaheuristic algorithms and variants of EO.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Qian Yang, Jinchuan Liu, Zezhong Wu, Shengyu He
Summary: This paper proposes a new hybrid algorithm called LMWOAGWO, which combines Levy flight with modified WOA and GWO. The algorithm improves the convergence accuracy and speed by introducing dynamic weighting strategy and greedy strategy. Extensive numerical experiments and comparisons demonstrate the effectiveness and superior performance of LMWOAGWO algorithm in solving standard benchmark functions and real-world optimization problems.
APPLIED SOFT COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Zenglin Qiao, Weifeng Shan, Nan Jiang, Ali Asghar Heidari, Huiling Chen, Yuntian Teng, Hamza Turabieh, Majdi Mafarja
Summary: The GOMGBO algorithm is an improved version of the traditional GBO algorithm, incorporating various mechanisms to enhance performance. Experimental results show faster convergence speed and higher precision, as well as superior effectiveness in engineering problems.
INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Yingxue Chen, Linfeng Gou, Huihui Li
Summary: Pressure retarded osmosis (PRO) is a promising renewable energy resource that is developing rapidly. To ensure stable operation and higher energy extraction, it is important to minimize oscillations and increase convergence speeds in fluctuating environmental conditions. The proposed algorithm, based on a combination of differential evolution (DE) and levy flight technique, enhances the reliability and effectiveness of the PRO system. The algorithm has been evaluated and validated in complex operational environments, showing significant improvement in specific energy extraction compared to the classic method.
APPLIED SOFT COMPUTING
(2022)
Article
Computer Science, Information Systems
Haiyang Xiao, Jia Guo, Binghua Shi, Yi Di, Chao Pan, Ke Yan, Yuji Sato
Summary: This paper introduces a new particle swarm optimization algorithm called TMBPSO, which enhances the search ability and the ability to escape local minimums by using a twinning memory storage mechanism and a multiple memory retrieval strategy. Experimental results with CEC2017 benchmark functions and five other population-based optimization algorithms confirm that TMBPSO can achieve high accuracy results for non-linear functions.
Article
Mathematics, Applied
Mehdi Rashidi Meybodi, Arash Bahar, Francesc Pozo
Summary: This study proposes a new optimization algorithm, named Iteration Dependent Optimizer (IDO), based on swarm intelligence theory. By controlling the movement speed of each individual, the algorithm shows advantages and capabilities in finding better solutions.
MATHEMATICAL METHODS IN THE APPLIED SCIENCES
(2022)
Article
Mathematics
Ahmed A. Ewees
Summary: This paper proposes an improved optimization method called EGBO, which enhances the local search capability of the standard gradient-based optimizer (GBO) by introducing expanded and narrowed exploration behaviors. The EGBO is evaluated using global optimization functions and benchmark feature selection datasets, and is compared to other optimization methods. The results demonstrate that EGBO outperforms the compared methods and the standard GBO, achieving high accuracy in selecting significant features and solving global optimization problems.
Article
Engineering, Chemical
Saleh Masoud Abdallah Altbawi, Saifulnizam Bin Abdul Khalid, Ahmad Safawi Bin Mokhtar, Hussain Shareef, Nusrat Husain, Ashraf Yahya, Syed Aqeel Haider, Lubna Moin, Rayan Hamza Alsisi
Summary: In this paper, a new optimizer called improved gradient-based optimizer (IGBO) is proposed to enhance the performance and accuracy of the algorithm in complex optimization and engineering problems. The proposed IGBO incorporates additional features such as adjusting the best solution with inertia weight, fast convergence rate with modified parameters, and a novel functional operator (G) to avoid local optima. The effectiveness and scalability of IGBO are evaluated through benchmark functions and real-world optimization problems, confirming its competitiveness and superiority in finding optimal solutions.
Article
Mathematics
Muhyaddin Rawa, Abdullah Abusorrah, Yusuf Al-Turki, Martin Calasan, Mihailo Micev, Ziad M. Ali, Saad Mekhilef, Hussain Bassi, Hatem Sindi, Shady H. E. Abdel Aleem
Summary: This paper proposes two hybrid algorithms for estimating solar cell parameters, which minimize the root mean square error between measurement and simulation results. These algorithms improve convergence characteristics and achieve good results on different equivalent circuit models and photovoltaic modules.