Study Of Current Control Strategy Of Three-phase Grid-connected Inverter Based On Sliding Mode Prediction

Over the past few decades, due to the increasingly serious environmental pollution and energy crisis, it has become the major task of current social development to transform the development pattern and optimize the energy structure.The use of new energy technologies, such as wind power generation and photovoltaic power generation, can convert clean energy into electrical energy, which contributes to reducing fossil-fuel energy consumption. So the new energy technology has been the emphasis expansive direction for optimizing energy structure. The three-phase grid-connected inverter is the core section of the convertion between clean energy and electrical energy, and its performance directly affects the application and the development of the new energy technology. Thus control strategies of inverters have been becoming a research focus in the new energy technology field.In actual operation, many influence factors,such as environment condition,service life, etc., can cause parameter drifts of elements, especially filter inductance.As a result,ideal parameters mismatch actual parameters, which reduces the modeling and controlling accuracy and affects the power quality adversely. It is necessary for inverters to possess strong robustness to eliminate negative effects of parameters mismatch. Through accessing to a large number of references, it is found that sliding model prediction has excellent dynamic performance and strong robustness. According to this theory, the prediction model is builded by the switching function of sliding mode control. Through feedback correction and rolling optimization, the influence of parameter fluctuation and external disturbance can be eliminated, so prediction results can get corrected immediately. Thus the real-time performance index , which covers all kinds of elements including control constraints and state constraints, can be solved optimally. System state will reach and hold the ideal sliding mode along reference trajectory quickly. Thus this thesis has studied the current control strategy of inverters base on sliding mode prediction to improve response rate and robustness under parametric mismatch condition.The two-level three-phase grid-connected inverter is the main research object,and its models in common coordinate systems are builded. On the basis of the feature analysis of its models, space vector PWM technique is selected. Traditional current control strategies base on PI control and deadbeat control are analysed. Their weaknesses for dynamic performance or anti-disturbance capability are also indicated. Sliding mode predictive control is used for the current control strategy of inverters. Its control principle and design method of control law are analysed. Its robust stability is also proved. Simulation results, which are compared with results of traditional current control strategies, indicate the improvement of dynamic performance and robustness.