A New Hybrid Load Frequency Control Strategy Combining Fuzzy Sets and Differential Evolution

• Published in: Journal of control, automation & electrical systems automation & electrical systems, 2021-12, Vol.32 (6), p.1627-1638
• Main Authors: Osinski, Cristiano, Leandro, Gideon Villar, da Costa Oliveira, Gustavo Henrique
• Format: Article
• Language: English
• Subjects: Algorithms; Closed loops; Control; Control and Systems Theory; Control stability; Controllers; Electric power systems; Electrical Engineering; Engineering; Evolutionary computation; Frequency control; Frequency stability; Fuzzy control; Fuzzy logic; Fuzzy sets; Hybrid systems; Hydroelectric plants; Load fluctuation; Mechatronics; Parameters; Performance evaluation; Performance indices; Proportional integral derivative; Robotics; Robotics and Automation; Tuning
• ISSN: 2195-3880; 2195-3899
• DOI: 10.1007/s40313-021-00767-0

Electrical power system has to operate properly even in the presence of load variations and other disturbances. One of the concepts that define this condition is frequency stability, and to deal with this issue, load frequency controllers (LFCs) have to be well designed and parameterized. The most frequent control approaches for LFC are PID controllers, with automatic tuning strategies. Recently, fractional-order PID has gained much interest due to its ability to improve closed-loop performance. This paper presents a new hybrid load frequency control strategy, called AFOPID. The main property of this algorithm is, based on a FOPID, to combine fuzzy sets and differential evolution (DE) for optimal parameter settings. The AFOPID’s parameters are tuned online in such a way that, in the occurrence of a load disturbance, the fuzzy logic updates its k p , k i , and k d coefficients to adapt the closed loop to the new operating condition. Then, the fractional coefficients λ and μ are updated by using a DE strategy. The hybrid system improves the overall solution since the fuzzy system first sets a good operation point and then the DE refines the solutions acting on the fractional orders. The proposed controller and tuning method are validated on a mathematical model of a hydroelectric plant belonging to the Brazilian National Interconnected System (SIN). The strategy has been compared with other similar algorithms in the same situation presenting better closed-loop performance. For controllers evaluating, some performance indices were used, such as the analysis of overshoots/undershoots, settling times, the integral square error, the integral time square error, and the integral time absolute error. The results have shown that the hybrid method has proved to provide better closed-loop performance than similar solutions or when each method is used alone.