Multi Objective Optimization Using Non-Dominated Sorting Approach For Load Frequency Control

• Main Author: Salam, Saifullah
• Format: Thesis
• Language: English
• Published: UTeM 2016
• Subjects: T Technology (General); TJ Mechanical engineering and machinery
• Online Access: http://eprints.utem.edu.my/id/eprint/18877/1/Multi%20Objective%20Optimization%20Using%20Non-Dominated%20Sorting%20Approach%20For%20Load%20Frequency%20Control%2024%20Pages.pdf; http://eprints.utem.edu.my/id/eprint/18877/; http://library.utem.edu.my:8000/elmu/index.jsp?module=webopac-d&action=fullDisplayRetriever.jsp&szMaterialNo=0000102300
• Institution: Universiti Teknikal Malaysia Melaka

This work provides an in-depth analysis on the implementation of multi objective optimization based on non-dominated sorting for load frequency control by using proportional-integral-derivative (PID) controller for interconnected reheat thermal power system. The load frequency control is use to control the frequency and inter-area (tie line) oscillation. When the load demand slightly changes, the frequency of the system will be affected and the control action need be react as soon as possible to prevent system instability. The real power is related with frequency which will slightly change due to any changing in frequency. The two area power system model will be considered in this investigation. The optimization technique which has been implemented is the Evolutionary Programming (EP). The weighted sum approach of EP is considered in order to provide multiple solution point by varying the weighted. The non-dominated point has been selected to optimize the PID gains to provide least overshoot value and fast frequency response of the system. Two objectives function will be considered for this investigation and has been characterized by the performance criterions which are Integral of Time Multiplied Absolute Error (ITAE) and Integral of Time Weighted Squared Error (ITSE). Optimum PID parameters which able to provide the best performance in terms of lower settling time and less overshoot value in the frequency deviation response will be determined.