Research On DC/DC Converter Applied In Photovoltaic System

It is known that photovoltaic system is one of the cleanest, most practicable and most promising resources in all kinds of renewable energy resources nowadays. In recent years, the main drawbacks are that the efficiency of conversion is low and the initial installation cost is considerably high. This problem can be solved by applying an MPPT(maximum power point tracking) controller between the PV (Photovoltaic) array and load, which obtains the maximum power from PV array. This paper researches on DC/DC converter applied in PV system, aiming at making full use of the solar photovoltaic energy, and carries out theory and experimental researches on MPPT controlling in PV system.In Chapter two of the paper, the principle of the solar panel is introduced first. Second, MATLAB is used to build the simulation of the output characteristics for the solar cell under various temperature and isolation. The simulation results confirm that the solar cell is non-linear.Conventional perturb and observation (P&O) MPPT strategy will make the final operation point of the solar module oscillate at the MPP, thus results in the power loss of the solar energy. The thesis proposes an improvement of P&O MPPT algorithm that can reduce signal noise and sample error without delay commonly related to the P&O method. Then two DC/DC converters are compared and Boost converter is presented to realize conversion.Finally, experimental results for a Boost controlled with a TMS320LF2407 DSP system are presented to validate the analysis and verify the performance of the proposed control scheme. By comparing algorithm based on two observation objects: input parameters and output parameters, conclusion comes as follows: P&O based on input parameters can reduce system vibration in steady state effectively ,and keep fast transient response speed as well. On the other hand, P&O based on output parameters is easier to realize, because either parameter of the two is enough and there is no need to do multiplication. However, the transient response is times slower than that of the algorithm based on input parameters.