کنفرانس مهندسی برق ایران

صفحه اصلی / سی و چهارمین کنفرانس بین المللی مهندسی برق

Enhancing the Accuracy of Long-Term Load Forecasting using a Hybrid PSO-ANFIS Model

نویسندگان :

Saman Mahmoodi¹ Sayed Taher Sharifi²

1- دانشگاه کردستان 2- اداره برق شهرستان بانه

کلمات کلیدی :

ANFIS،Hybrid-approach،Load forecasting،PSO algorithm،Real distribution system

چکیده :

Accurate load forecasting is a cornerstone of efficient operation and planning in modern distribution systems, especially as grids evolve with the integration of distributed energy resources (DERs), electric vehicles (EVs), and smart metering technologies. This research introduces an intelligent hybrid model for load forecasting, in which Particle Swarm Optimization (PSO) is employed to fine-tune the parameters of an adaptive network-based fuzzy inference system (ANFIS), thereby significantly improving the model forecasting accuracy and flexibility. Given the high volume of input data and the convergence of the estimation process, three key features are introduced to establish a ratio among the inputs of load, temperature, and population growth. These features are defined as potential inputs for the proposed model. The model has been evaluated using a real-world of Baneh city distribution system and simulations were conducted using the DIgSILENT PowerFactory 15.1, while the result analysis was carried out in MATLAB 2021a software. Across the evaluated forecasting, the model achieved RMSE of 1.86% and MAE 1.43%. The findings underscore the robustness and scalability of the proposed PSO-ANFIS framework, positioning it as a viable solution for intelligent energy management.