Load shedding scheme based on meta-heuristic techniques and multi-criteria decision making for islanded distribution systems / Hazwani Mohd Rosli

• Main Author: Hazwani , Mohd Rosli
• Format: Thesis
• Published: 2024
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://studentsrepo.um.edu.my/15427/2/Hazwani.pdf; http://studentsrepo.um.edu.my/15427/1/Hazwani_Mohd_Rosli.pdf; http://studentsrepo.um.edu.my/15427/

In modern power systems, ensuring stability during contingencies is essential for maintaining a reliable and secure operations. This study introduces a new Load Shedding Scheme (LSS) that incorporates advanced techniques to enhance system stability under adverse conditions. The proposed LSS utilizes discrete meta-heuristic techniques to intelligently select critical loads for shedding. It considers both voltage stability index and power mismatch as fitness functions. Among the techniques employed, Binary Zebra Optimization Algorithm (BZOA) demonstrates the best performance, as evidenced by its capability to achieve both minimum fitness values and minimum standard deviation. Multi-criteria decision-making approaches, including the Analytic Hierarchy Process (AHP) and the Weighted Sum Model (WSM), rank loads based on its priority, stability index, and sizing. The proposed LSS has undergone comprehensive testing on the 11 kV Malaysian distribution system for both 30-bus and 102-bus systems. The testing involved the utilization of various stability indices for assessing performance in both islanding and overloading events. Results consistently demonstrate that the proposed LSS outperforms adaptive techniques across all stability indices, showcasing superior frequency response characteristics. A significant improvement was also observed in both stability index values and voltage profiles for buses that underwent load shedding. Diverse scenarios were also investigated to assess the versatility of the proposed scheme. In an increased load scenario, the LSS adeptly prioritizes nonvital loads for shedding, effectively ensuring stable system operation with the system voltage approaching to 1.0 p.u. Additionally, the proposed scheme exhibits remarkable frequency recovery in scenarios involving both loss of synchronous and induction generator. Comparative analyses with previous techniques highlight the superiority of the proposed LSS in terms of rapid frequency recovery by as much as 73.75% and 28.51% for the 30-bus and 102-bus systems, respectively, compared to the adaptive technique.