Abstrak
Chaos dan voltage collapse muncul pada sistem tenaga listrik akibat gangguan energi pada beban kritis. Untuk mengatasinya maka dirancang PID-SVC yang berbasis-layered recurrent network (LRN).PID berbasis-LRN
IntroductionThe Loading on electric power systems (EPS) was growth rapidly. On the other hand, the built of power plants and transmision line were very slow due to economical and environtment reasons. Therefore, the existing EPS must be operated in critical condition on boundary of stability regions. Meanwhile, voltage collapse was found at the highest loading parameter existed on saddle node bifurcation (SNB) and voltage collapseanalysis was done by using center manifold with static and dynamic approach [1]. The appeared of Hopf bifurcation (HB), chaos, crisis and voltage collapse before SNB were identified and classified in [2][3]. The appeared of HB in critical EPS was suppressed by using linear and nonlinear controllers in [4][5].Chaos appeared in rotor speed, angle and voltage magnitude due to disturbingof energy (DE) [6]. Furthermore, chaotic behavior was observed in power systems and modelled using Elman recurrent neural networks (RNN). To validate the RNN model was compared to mathematical model, where the results of the both models were identical [7]. Controlling chaos and voltage collapse using continuation technique [8], ANFIS-based composite controller (CC)-SVC [9] and an additional PID-loop [10]. Artificial neural network(NN) was used in indentification system and control application [11][12][13]. SVC controlled by NN was used to enhanched dynamic stability of EPS [14] andRNN was used as power system stabilizer in single machine [15].Controlling chaos and voltage collapse using layered recurrent network (LRN)-based PID-SVCwere focused in this research. This controller was chosen because it was trainedin offline modeby train data and easy to be implemented.This paper is organized as follows: Power system model is explained in Section 2. LRN-based PID-SVC controller design is detailed in