Novel Neural Network-Based Load Frequency Control Scheme: A Case Study of Restructured Power System

Abstract

Nowadays, neural networks (NN) are being utilized in different control problems because of their excellent ability to model any nonlinear process. NN is suitable for the process having a wide range of operating conditions. In this work, the neural network-based internal model control (NN-IMC) scheme has been considered as a secondary controller for the load frequency control (LFC) problem in the restructured electricity market in order to meet Poolco and bilateral transactions. The proposed control scheme has been implemented on a 75-bus, 15-generator power system. The test system is divided into four areas. It is seen that area frequency errors have been eliminated at a steady state in all cases, and Gencos/Discos shared the increase in demand as per their involvement in the frequency regulation market. The results show that the NN-IMC control scheme has good performance and improves system responses effectively. Further, the performance of the NN-IMC control scheme has also been compared with the fractional-order proportional-integral-derivative (FO-PID) control scheme It is observed that the performance of the FO-PID controller is superior to the NN-IMC scheme in terms of settling time and similar to the NN-IMC control scheme in terms of maximum overshoots/undershoots. The compliance of the NN-IMC scheme has also been checked with NERC standards. It is seen that the NN-IMC scheme also satisfied the CPS1 and CPS2 control standards.