Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
8122768 | Sustainable Energy Technologies and Assessments | 2018 | 13 Pages |
Abstract
This paper presents a robust natural-frame based renewable energy source (RES) interfacing control scheme utilizing Neural Network (NN) identified adaptive leaky least mean square (ALLMS). The integration of RES in distributed network leads to quandaries on power quality and stability. These quandaries directly demands precise generation of current references under sundry operating scenarios. The self-learning facility of proposed NN identified ALLMS control model along with a detuned LC filtering circuit can make felicitous shaping of the VSC output under various system perturbance. The proposed methodology utilizes a variable leaky factor to eschew wavering of weights involved in the estimation and extraction process of current references. To achieve most expeditious and precise convergence of the controller, accentuation is made to cull a congruous value for step-size utilized in controlling actions. Emolument of reactive power, load balancing and mitigation of harmonics are taken care of in the proposed control scheme. The comparative study on power quality in-between proposed and conventional adaptive current control techniques, modelled in MATLAB/Simulink, is also presented and it justifies the sound performances of the proposed control scheme under diverse system dynamics.
Related Topics
Physical Sciences and Engineering
Energy
Energy Engineering and Power Technology
Authors
Moushumi Patowary, Gayadhar Panda, Bimal C. Deka,