Investigation of rank order centroid method for optimal generation control

Area of Science:

• Electrical Engineering
• Control Systems Engineering
• Optimization Techniques

Background:

• Multi-criteria decision-making (MCDM) is complex, especially in power system control.
• Automatic Generation Control (AGC) in interconnected power systems requires balancing multiple performance objectives.

Purpose of the Study:

• To propose the Rank Order Centroid (ROC) method for weighting sub-objective functions in MCDM for AGC.
• To design and optimize a Proportional-Integral-Derivative (PID) controller for a two-area interconnected power system (TAIPS) using ROC and optimization algorithms.

Main Methods:

• The Rank Order Centroid (ROC) method was used to determine weights for Integral Time Absolute Errors (ITAE) of frequency deviations, control errors, and tie-line power fluctuations.
• A PID controller was designed based on the weighted objective function.
• The Jaya Optimization Algorithm (JOA) was implemented to optimize the objective function, compared against TLBOA, LJA, NMSA, EHOA, and DEA.
• Six case studies were performed with statistical analysis, including the Friedman rank test, to validate performance.

Main Results:

• The JOA-based PID controller demonstrated superior performance in managing frequency and tie-line power fluctuations under various load conditions.
• Systematic weighting of sub-objective functions using ROC improved controller design.
• Comparative analysis confirmed the efficacy of JOA over other tested optimization algorithms.

Conclusions:

• The proposed ROC method combined with JOA offers an effective approach for designing robust PID controllers for AGC in TAIPS.
• This methodology enhances system stability and performance by optimally balancing multiple control criteria.
• The findings provide a valuable framework for future research in intelligent power system control.