On-line evaluation of loadability limit for pool model with TCSC using back propagation neural network

This paper presents an approach for on-line evaluation of loadability limit for pool model with Thyristor Controlled Series Compensator (TCSC) using Back Propagation Neural Network (BPNN). The optimal location, setting of TCSC and loadability limit for various load patterns in off-line are determined using Differential Evolution (DE) algorithm. This approach uses AC load flow equations with constraints on real and reactive power generations, transmission line flows, magnitude of bus voltages and TCSC setting. The input parameters to BPNN are real and reactive power loads at all buses. Data for training the BPNN is generated through Optimal Power Flow (OPF) solution using DE and the trained BPNN is tested with unseen load patterns. Sequential Forward Selection (SFS) belonging to greedy wrapper method is used for the selection of best optimal input features. Simulations are performed on 39 bus New England test system and IEEE 118 bus system. Solution accuracy and computation time are analyzed. The results obtained illustrate that, for on-line evaluation of loadability limit of pool model with TCSC, BPNN is accurate with minimal Mean Squared Error (MSE) and less computation time.

► Loadability limit for pool model with TCSC is evaluated using BPNN. ► With TCSC, conventional methods are not suitable for data generation. ► Hence, DE is proposed for data generation through OPF solution to train the BPNN.