Modelling, Optimisation and Control of Photovoltaic Energy Conversion Systems

Abstract

With growing energy demands of the society and depleting reserves of conventional energy sources, renewable energy is the need of the hour. Out of the various renewable energy sources available, solar photovoltaic (PV) is emerging as one of the most promising energy solution. To utilise the vast potential of solar PV energy using, various aspects of solar photovoltaic energy conversion are considered in this thesis, starting from input and output measurement of PV system to the system modelling and characterisation, partial shading analysis and the maximum power point tracking. Irradiance and temperature form the input to the photovoltaic converters and are measured by using cost-e ective and high performance solar irradiance meter. Output current-voltage measurement for the PV system is made possible by the development of a closed-loop controlled SMPS based characterisation setup designed with a high bandwidth input lter providing high speed characterisation and stable operation under varying ambient conditions. Steady state modelling of PV system is achieved using a novel sequential parameter estimation method that utilises the measured system response and overcomes the numerical convergence problem. To study the dynamic behaviour of PV modules, their capacitance is evaluated experimentally. The e ect of non-uniform irradiance and temperature on PV modules is further studied with the introduction of a novel subcell modelling approach. This subcell model further facilitates the understanding of hotspot formation on healthy PV modules and their e ect on output characteristics. Non-uniform irradiance and temperature input to the series connected PV modules gives rise to multiple peak power-voltage characteristics, which makes the maximum power point tracking (MPPT) a challenge. A majority of the commercial converters are equipped with local MPPT, due to which their energy yields reduce tremendously under shading conditions. In this thesis, a shading factor based global MPPT (GMPPT) is proposed which incorporates the novel approach of module voltage sensing for GMPPT implementation. The proposed method is seen to be hundreds of times faster than the popular scanning based GMPPT and improves the power capture many-folds as compared to the local MPPT algorithms.