DESIGN AND ANALYSIS OF A FUZZY LOGIC CONTROLLED PHOTOVOLTAIC SYSTEM WITH A BIDIRECTIONAL DC-DC CONVERTER

Аннотация

This paper presents a detailed model of a photovoltaic (PV) system integrated with a bidirectional DC-DC converter, employing the Adaptive Neuro-Fuzzy Inference System (ANFIS) for intelligent control. The model considers two key input variables: time and ambient temperature. By incorporating ANFIS, the system adapts dynamically to environmental fluctuations, enhancing its overall performance. Key performance indicators such as voltage stability, current fluctuation mitigation, and battery charge optimization are analyzed to assess system effectiveness. Simulations are carried out in the MATLAB/Simulink environment, providing a robust framework for evaluating system behavior under varying operating conditions. Results indicate that the integration of ANFIS significantly improves energy flow management, enhances stability, and ensures a higher quality of power output. Furthermore, the model demonstrates adaptability to changing external conditions, making it a viable solution for real-world renewable energy applications and intelligent PVpower management. Keywords:photovoltaic panel, fuzzy logic, fuzzy controller, ANFIS, MATLAB/Simulink modeling, bidirectional converter, MPPT,renewable energy. 1.Introduction The accurate modeling of photovoltaic (PV) systems is vital in advancing renewable energy technologies. These systems are inherently nonlinear and highly sensitive to external parameters such as solar irradiance and ambient temperature. Achieving optimal performance requires not only the physical modeling of components but also the integration of advanced control strategies and power converters. Recent advancements in PV system research span various domains, including component-level electrical modeling, intelligent control algorithm development, and forecasting techniques. Several studies have addressed these aspects comprehensively. For example, [1] examines how shading affects electrical mismatches in PV systems, particularly highlighting the role of bypass diodes in reducing power loss and reverse voltage. However, it notes inefficiencies under light shading, where current splits between the module and diode. In [2], a support vector machine (SVM)-based algorithm predicts PV output based on weather classification into four categories (clear, cloudy, foggy, and rainy). This model, validated on a 20 kW system, achieved high accuracy in power forecasting, showing promise for grid-connected applications. DC-DC converters are integral to PV efficiency. Study [3] introduces a converter design utilizing a coupled inductor in both series and parallel modes, yielding greater voltage gain than traditional configurations. In [4], an improved PI-controlled boost converter model simplifies control circuit design while enhancing voltage regulation. page2390 INTERNATIONAL JOURNAL OF ARTIFICIAL INTELLIGENCE ISSN:2692-5206,ImpactFactor:12,23 AmericanAcademicpublishers,volume05,issue05,2025 Journal: https://www.academicpublishers.org/journals/index.php/ijai Maximum Power Point Tracking (MPPT) remains a cornerstone of PV optimization. A fuzzy logic-based MPPT method in [5] demonstrated 94.8%–99.4% accuracy across varying irradiance (700–1000 W/m²) and temperature (25–60°C). Similarly, [6] presents a PV model combined with a bidirectional energy storage system, validating the PI algorithm for effective energy flow control. Study [7] explores solar radiation modeling using clearness index and time-based diurnal patterns to refine converter duty cycles. In [8], bifurcation analysis uncovers nonlinear instability phenomena, emphasizing the necessity for robust control schemes. These findings affirm the centrality of converter control, intelligent methods (e.g., ANFIS, fuzzy logic controllers, SVM), and advanced MPPT techniques in PV research [9]. Recent developments further reveal the superiority of AI-based MPPT algorithms. Deep learning models, such as LSTM, achieve up to 30% higher efficiency in variable light conditions compared to traditional perturb and observe(P&O) and ANN methods [10]. Innovations in converter design are also noteworthy. For example, [11]introduces a three-phase interleaved boost converter directly interfacing with Li-ion batteries, eliminating the need for separate charging circuits. Adaptive strategies like ASGAO-RBFN not only improve MPPT accuracy but also enhance converter efficiency [12]. Moreover, integrating PV and wind systems into the grid calls for advanced energy management techniques. Study [13] shows that battery systems with controlled reverse energy flows can stabilize operation and reduce losses, especially in regions with limited infrastructure. Building upon this body of research, the present study proposes a MATLAB/Simulink-based model of a PV system equipped with a bidirectional DC-DC converter. The model employs the ANFIS algorithm to achieve high- precision control inthe face of variable external conditions, making it a promising candidate for furtherdevelopment and real-world