Reactive power control in decentralized hybrid power system with STATCOM using GA, ANN and ANFIS methods

• Investigation of STATCOM performance in presence of high probabilistic uncertainty for different disturbances in input wind power and reactive power load demand.
• System studies during dynamic conditions with composite load model comprising non linear loads.
• Comparison of voltage control and STATCOM reactive power using various tuning methods.

In this paper, STATCOM performance for voltage–reactive power control is investigated by comparing different tuning methods, used to evaluate gain parameters of STATCOM controller in presence of high probabilistic uncertainty in input wind power and reactive power load demand. To control voltage transient response in least time, reactive power demand is managed by STATCOM. The conventional methods for tuning gain parameters of STATCOM controller do not satisfactorily operate in case of random disturbances and therefore, advanced controllers such as Genetic Algorithm (GA), Artificial Neural Network (ANN) and Adaptive Neuro Fuzzy Inference System (ANFIS) are required. The main contribution of the paper is: (i) Investigation of STATCOM performance in presence of high probabilistic uncertainty with step changes in input wind power and reactive power load demand, (ii) system studies during dynamic conditions with composite load model in lieu of static load model in the system, (iii) comparison of voltage control and STATCOM reactive power using various tuning methods. Results comparison through all tuning methods show that advanced tuning methods are able to preserve optimal performances over wide range of disturbances using Integral of Square of Errors (ISE) criterion.