Takagi-Sugeno fuzzy model identification using improved multiswarm particle swarm optimization in solar photovoltaics

• Дата: 2025
• Автори: Zdiri, S., Moulahi, M., Messaoudi, F., Zaafouri, A.
• Формат: Стаття
• Мова: English
• Опубліковано: National Technical University "Kharkiv Polytechnic Institute" and Аnatolii Pidhornyi Institute of Power Machines and Systems of NAS of Ukraine 2025
• Теми: improved multiswarm particle swarm optimization; particle swarm optimization; specific Takagi-Sugeno modeling
• Онлайн доступ: http://eie.khpi.edu.ua/article/view/327924
• Назва журналу: Electrical Engineering & Electromechanics

Introduction. The particle swarm optimization (PSO) algorithm has proven effective across various domains due to its efficient search space exploration, ease of implementation, and capability to handle high-dimensional problems. However, it is often prone to premature convergence, which limits its performance. Problem. This issue becomes critical in identifying Takagi-Sugeno (T-S) fuzzy models, especially in complex systems like solar photovoltaic (PV) applications, where model accuracy is vital for tasks such as maximum power point tracking (MPPT) and shading compensation. Goal. This manuscript introduces an improved multiswarm PSO (I-MsPSO), designed to enhance search performance and robustness in identifying T-S fuzzy systems. The method is particularly suited to nonlinear modeling challenges in renewable energy systems. Methodology. I-MsPSO divides the swarm into 4 independent subswarms, each operating in a local region with specific inertia weights and acceleration coefficients. Periodic information sharing between subswarms allows the algorithm to converge collectively toward optimal solutions. A new modeling approach, specific Takagi-Sugeno modeling (STaSuM), is introduced, using I-MsPSO to determine both the structure and parameters of T-S fuzzy systems. Results. The I-MsPSO’s performance was tested on benchmark optimization problems and real-world engineering cases. Results show that STaSuM produces highly accurate and generalizable fuzzy models, outperforming existing techniques. Scientific novelty lies in the development of I-MsPSO, which enhances the traditional PSO by using 4 interactive subswarms with customized parameters, and the creation of STaSuM for advanced T-S fuzzy system identification. Practical value. I-MsPSO and STaSuM provide a powerful optimization and modeling framework, offering robust and accurate solutions for nonlinear and dynamic environments. Their structure makes them especially valuable for future applications in MPPT control, fault-tolerant modeling, and real-time optimization in PV energy systems. References 39, table 5, figures 8.