Efficiency improvement for small-scale single-phase grid-connected inverters

Abstract

This Ph.D. research focuses on efficiency improvement methods for small-scale single-phase grid-connected inverters. Many attempts applied to increase inverters’ efficiency found in the recent literature can be primarily summarized into two categories: specific topologies and complicated control algorithms, which result in a considerable increase in system cost and control complexity. Meanwhile, the power quality is another important issue for grid-connected inverters to comply with the utility standards. To overcome the drawbacks of the traditional methods and satisfy the utility requirement, a novel control algorithm called “variable switching frequency control (VSFC)” is developed as a major part of this thesis work to increase the overall efficiency of the inverter through selecting optimal switching frequencies of pulse width modulation (PWM) in real time while meeting requirements of grid interconnection standards. According to the inverter loss analysis and current harmonic estimation model presented in this dissertation, the selection of switching frequencies for the grid-connected inverter optimized by the proposed “VSFC” ensures that the inverter operates with maximum efficiencies at different output levels. In addition, as the operational switching frequency applied to the inverter under “VSFC” changes along with variations of work conditions, when the inverter operates at a low or medium switching frequency, the time-delay effect caused by sampling distribution, computation of the control program in DSP and inherent PWM generator update is amplified and can severely degrade the system stability and performance. Thus, a robust current control scheme featuring high adaptability to time delays and system uncertainties and high robustness to parameter mismatch is developed in this research. The proposed scheme is built on a structure of the predictive current controller and developed with an improved time-delay compensation technique which greatly reduces the current tracking errors through a simple weighted filter predictor (WFP) and completely eliminates static voltage errors introduced by uncertain system disturbances through a robust adaptive voltage compensator (AVC). The developed new control methods for grid-connected inverters have been verified through computer simulations and laboratory experiments. The results of simulation and experiment investigation have demonstrated the improvements of these methods in overall inverter efficiency and performance.