Robust and Adaptive Grid Synchronization Control of a Two Stage Grid-Connected Photovoltaic System

Abstract

In view of supplementing power generation to meet increasing load demand while minimizing environmental pollution, more attention is currently given to use renewable power extraction for Photovoltaic (PV) and Wind energy conversion system. Amongst all the renewable power generation options, PV power generation is being considered as the most suitable one owing to the abundant availability of solar irradiance with pollution¬free operation. A PV system can be operated as standalone or grid connected modes. The thesis focuses on design, development, and practical realization of robust and adaptive control schemes for effective synchronization of a two¬stage three¬phase Grid Connected Photovoltaic System (GCPVS) to the utility grid. The PV power varies continuously during the day, and PV current and voltage characteristics depend upon the irradiation and ambient temperature, respectively. Therefore, to track this random variation in the Maximum Power Point Tracker (MPPT), adaptive MPPT algorithms is necessary on the PV side. On the grid side, the grid voltage distortions, corrupts the grid synchronization controllers. To resolve these issues, a robust controller needs to be designed. The thesis first designs a Adaptive MPPT controller to improve the Maximum Power Point (MPP) tracking performance. The classical MPPT controllers such as P&O has fixed step size, and its MPP tracking performance is influenced by predefined parameters such as perturbation size, sampling time and initial duty ratio. Therefore, fast tracking of the MPP through these MPPT algorithms is difficult. To resolve these issues, a PI controller is employed to generate a variable step size duty ratio, to improve the tracking performance of this proposed Improved Adaptive Perturbed and Observed (IAPO) MPPT algorithm. In this algorithm, the step size is varied based on the PV power variation, which is embedded as an adaptive feature to a PI controller. This adaptive feature results in fast tracking of the MPP even during transients, less oscillations of DC link voltage during steady state, and least dependence on predefined parameters such as initial duty ratio. Subsequently, to achieve current control for synchronization, an Integral Sliding Mode Controller (ISMC) is designed in which issues regarding the harmonics in the grid injected current is reduced. The classical PI current controller performance deteriorates due to grid voltage distortions and hence there is a need to employ harmonic suppression schemes. But, this slows down the current controller dynamics. Whereas, the ISMC aids in achieving faster dynamics in face of modeling and parametric uncertainties. It also suppresses the harmonics in the injected grid current despite the presence of high Total Harmonic Distrotion (THD) vii content in the grid voltage. It is also found to have excellent decoupling of the cross coupling terms and provide independent control of active and reactive current. The ISMC based current controller is then integrated with the IAPO MPPT algorithm for a two stage GCPVS. Its performance is then compared with other control schemes, namely, SMC¬IAPO and PI¬IAPO control schemes. The comparison envisages that with ISMC¬IAPO control approach, the performance of the overall system improves, and this control strategy does provide faster dynamics despite rapid variations in the PV power, uncertainties owing to modeling and parametric variations in the PV system and grid disturbances. It is also found that this ISMC¬IAPO approach gives the best quality current in comparison to ISMC¬P&O, SMC¬IAPO, and PI¬IAPO. The robust performance of the ISMC is then evaluated in grid fault situations where a Vector Current Control with Feedforward (VCCF) control strategy is employed in the control of two stage GCPVS to deliver balanced three phase currents to the gird. The control of two stage GCPVS in the Low Voltage Ride Through (LVRT) situation becomes a real challenge. It is because the control operation needs to regulate the DC link voltage at its nominal value. Therefore, to transfer the reactive power, the PV panel is operated in the de¬rated mode of operation. This is only possible if MPPT operation is halted and boost converter is operated in constant duty ratio mode. This new duty ratio and reactive current reference are calculated based on the percentage of voltage dips. The ISMC based VCCF strategy is evaluated in the three phase symmetrical and two phase unsymmetrical fault. The ISMC provides both active and reactive current to the grid. The ISMC is compared with a PI controller based VCCF strategy and found to have good performance in damping the inrush currents that arise in the event of fault occurrence and fault clearance. All these control algorithms are simulated in MATLAB/Simulink environment followed real-time implementations on a prototype PV system. A prototype of 2 kW PV system is developed in the laboratory. The aforesaid grid synchronization algorithms were then implemented and verified in real-time on the prototype GCPVS, and results obtained are analyzed.