Tunicate Swarm Algorithm-Neural Network for Adaptive Power System Stabilizer Parameter

Abstract

Tunicate Swarm Algorithm (TSA) is a metaheuristic method that imitates the life of the tunicate. It occurs during navigation and foraging using jet propulsion and swarm behavior. A feed-forward neural network ( FFNN) is a neural network that is often used, and applied. computational methods have been widely used to optimize FFNN weights in order to produce better output. This paper proposes a compound algorithm based on a tunicate swarm algorithm to optimize an FFNN. It is applied to power system stabilizers. The proposed method is compared with other algorithm methods such as the feed-forward (FFNN), cascade forward backpropagation (CFBNN), focused time delay (FTDNN), and distributed time delay (DTDNN). The proposed method has the ability to improve the output of FFNN methods. The proposed method has the average ability to reduce the overshoot of the speed by 35.17% and the undershoot of the rotor angle by 15.36% . In addition, the proposed method has better capabilities than the comparison method. The results of the experiment show that the use of the submitted algorithm has preferable adaptability and performance than the other methods.