Implementation of Neural network based I cos φ Controller for DSTATCOM in Three Phase Four Wire Distribution System under Varying Source and Load Conditions for Power Quality Improvement

The application of power semiconductor devices in the area of power electronics and power system are the key factors which kindle the power quality problems particularly the harmonics, burden of reactive power and neutral current due to unbalance. The above citied power quality problems are predominant due to wide usage of 1-φ / 3-φ , linear and nonlinear loads which are fed from three phase four wire (3P4W) distribution system. The performance of the 3P4W distribution system can be enriched by mitigating the above mentioned problems with proper selection of a power quality compensator and control strategy. In this paper, DSTATCOM is chosen as power quality compensator which comprises of a three phase three leg Voltage source converter (VSC), a single phase active power filter (APF) and a DC bus capacitor. The proposed neural network based control strategy generates the reference supply current, so as to make the source supply only the real fundamental load current component (I cos φ, in which “I” refers the fundamental load current magnitude and cos φ refers the load displacement power factor). The DSTATCOM with neural network based control strategy mitigates harmonics, compensates reactive part of load current, ensures balanced and sinusoidal source current from the supply mains that are nearly in phase with the supply voltage even though there is an unbalance in the three phase load currents and source voltage conditions, compensates neutral current even under unbalanced linear and non-linear load conditions in 3P4W distribution systems. Artificial Neural Network (ANN) controller is also executed to keep up the voltage of the capacitor at the reference value under varying load and source conditions. The DSTATCOM performance is validated under unbalanced, linear/nonlinear loads for all possible conditions of the source like balanced, unbalanced, balanced/distorted and unbalanced/distorted source voltage conditions. The propounded neural network based control algorithm for DSTATCOM presented in this paper is simulated using MATLAB software. The simulation results prove the efficacy of the proposed neural network based control strategy under varying source and load conditions. Key-Words: Neural network, DSTATCOM, Neutral current mitigation, Three phase four wire distribution system, unbalanced and/or distorted source, Power quality.