Research And Design On Single-phase Multiple-step In Grid-connected Photovoltaic Generation System By Data-Driven

In order to promote the photovoltaic technology products which have the adventages of distributed generation into every household, this thesis proposes the PV grid-connected cont-rol strategy by data-driven according to the maximum power-point tracking, the design of power converters, grid performance.The model based control methods are facing the high requirement of practical problems in engineeering and engineering application, the new control method which is the data-driven control theory should meet the requirements of controlled objects, such as imprecise model, excellent adaptiveness and robustness. Model-free adaptive control (MFAC) is a landmark achievement on data-driven control theory. Compared with the tradiational control methods based on models, MFAC is more original in theory, and it is also an advanced control strategy with excellent adaptiveness and robustness.The characteristic of solar cell and the MPPT are analyzed in detail. This thesis deals with the basic model-free adaptive control and proposes its improved methods, namely a method of parameter tuning for model-free controller based on particle swarm optimization, a method of constant voltage by MFAC with disturbance observation,and a method of constant voltage by MFAC with incremental conductance. They are used in the DC/DC link of the PV systems, some general simulators of MFAC are developed and implemented in the Simulink environ-ment. Simulation results show that the proposed control methods have stronger robustnesss, nicer anti-jamming capability and better load adaptability.What is more, the model-free adaptive control and a new method which is data-driven and particle swarm optimization based controller design and its parameters tuning method are used to full-bridge inverters, are compared with the state feed-back lin-earization. Simulation shows that model-free control is obviously better than state feed-back linearization control on the ability of traceability, adaptability, anti-interference. The power compensation is explored as well.Finally, the500W PV grid-connected system is designed,and the forestage multifunction converter of the single phase grid-connected power system and the semi-substantial emulation platform are researched. Based on the building the semi-substantial emulation platform, writing the underlying drivers and related algorithm modules, made the preliminary comm-issioning and testing. The circuit verifies the effectiveness of this method based on the semi-substantial emulation platform,the underlying drivers and related algorithm modules.