Power Conditioning With Intelligent Control Using a Novel Recursive Stochastic Optimization

This paper presents a novel smart-grid power-conditioning system with intelligent control for solving a critical issue affecting the performance and life of three-phase asynchronous machines that are highly sensitive to voltage and phase drifts power disturbances of the electric grid. The present work of controlling the power supply of an electric-grid is based on an original power conditioning control algorithm; named recursive stochastic optimization (RSO) algorithm. An asynchronous machine steady-state model is derived based on symmetrical components analysis, permitting the evaluation of the effect of both voltage and phase unbalance using voltage and current unbalance factors. The three-phase power supply is conveyed through a power conditioning and intelligent control system, driven by a digital signal processor (DSP) controller, which measures the input currents, estimates the unbalance factors, and searches for voltage levels that minimize the electric-grid unbalance using the RSO, in order to act on a feedback power-electronics stage. The RSO algorithm shows excellent results along with very fast response time of the order of 20 ms allowing real-time control. Results obtained are validated with a Simulink dynamic model. The intelligent control system architecture based on TMS320C6748 DSP board with signal acquisition and control microcontroller is proposed.