Design and Real-time Implementation of Adaptive Grid Synchronization Control Schemes for Photovoltaic System

Abstract

In order to meet the increasing demand of electric power and to reduce the environmental pollution, a lot of research works have been directed for generating power from renewable energy sources such as Photovoltaic (PV) and wind energy systems. Power generation from PV system is considered to be a promising approach owing abundantly availability of solar irradiances. PV systems can be operated either in standalone (Off- Grid) or in grid connected (On-Grid) mode. Depending upon the power conversion stages involved, a Grid Connected PV (GCPV) system can be further classified as single stage or double stage systems. Single stage grid connected PV system is very much popular owing to its simple configuration and less power loss. The thesis first reviews the works reported on designing control algorithms for grid synchronization of a three phase PV system. Subsequently, a number of filtering algorithms such as Improved Linear Sinusoidal Tracer (ILST), damped Second Order Generalized Integrator (d-SOGI), Lattice Wave Digital Filter (LWDF), Least Means Square (LMS), Least Means Absolute Third (LMAT) and Least Means Fourth (LMF) have been employed to design adaptive grid synchronization control schemes for achieving effective grid synchronization of a three phase PV system. Then a new adaptive control scheme employing Normalized Least Means Absolute Third (NLMAT) filtering has been developed for grid synchronization of a PV system. The aforesaid control algorithms together with the proposed NLMAT algorithm are implemented in grid connected PV system. For achieving reliable operation, all the parameters of the GCPV system such as DC link voltage, DC link capacitor, selection of rating of Insulated Gate Bipolar Transistor (IGBT) in the inverter, series-parallel combination of PV panels etc. for a GCPV system have been designed judiciously. An LCL filter has also been designed instead of a conventional LC filter for obtaining improved performance. To achieve maximum efficiency of the proposed controller, an adaptive hysteresis band current controller (AHBCC) has been implemented such that appropriate gate pulses are generated for the VSI. The performances of all these controllers are compared by first pursuing simulation studies by using MATLAB/Simulink 2017a. The outcome of the comparative assessment of their effectiveness in accomplishing control functions such as (i) extraction of fundamental active component of load current, (ii) maximum power extraction (iii) harmonics mitigation and (iv) injecting active power to the grid is analyzed. The assessment is performed under sudden load change and change in solar irradiance conditions. In order to verify the effectiveness of the proposed NLMAT controller, further viii a prototype PV system has been developed in the laboratory. All these control schemes are then implemented in real-time on the prototype PV system. After comparing the performances of the proposed controller with the other controllers mentioned above in real-time, it is found that the proposed NLMAT controller exhibits superior performance amongst all the aforesaid grid synchronizing control schemes. Also it is observed that THD yielded in the grid currents with newly developed NLMAT control algorithm is the least amongst all the above controllers and remaining well within the limit prescribed by of IEEE-1547 and IEEE-519 standards. The thesis also describes the design of a 90kW roof top grid connected PV system installed at Electrical Science Building at National Institute of Technology Rourkela, Odisha, India. It then presents an economic evaluation the aforesaid PV system pursued by using PVsyst6.70 software