System Identification and Adaptive Control for Grid-Connected Single-Phase AC Inverters

Abstract

This thesis presents adaptive control techniques to improve the stability and performance of grid-tied single-phase inverter systems. There are three main control systems involved in the conversion of DC to AC power: a synchronizer, a current-controller, and a DC-bus controller. The focus of this thesis is to create a control system which responds to the changing operating conditions. In order to adapt to changing operating conditions, estimations of the variables which influence stability and performance are necessary. A synchronization system is proposed, which delivers very accurate estimation of the grid frequency and voltage amplitude across all operating conditions. The synchronization system also provides an accurate and normalized set of orthogonal vectors for the AC output current reference phase. An adaptive parameter estimation technique is proposed to passively estimate the LCL-filter parameters and DC input current. Using the parameter estimations of the output filter, an adaptive state-feedback system is proposed. Next, a frequency-adaptive current controller is proposed, which dynamically tunes to the frequency provided by the synchronizer. Finally, the DC-bus controller is enhanced with the grid-voltage estimate and DC current estimator. The proposed techniques are verified for performance and stability though simulation and experimental results.