Linear Quadratic Optimal Control Based Research On Photovoltaiv Grid-connected System

Grid-connected photovoltaic(PV) generation has become central issue among theacademia and industry almost all of the countries. This thesis focuses on controlstrategies of single-phase two-stage grid-connected PV systems, which includes depthanalysis of its working principles, modeling of the Boost converter and the LCL filterbased inverter, which suitable for the application of the linear quadratic optimalcontrol. Furthermore, the comprehensive and systematic research also consists ofmaximum power point tracking(MPPT) methods, inverter DC input voltage stabilitycontrol and the grid-connected current waveform correction.To achieve the maximum power point tracking and inverter DC input voltagestability control, a current continuous mode(CCM) Boost Converter integrated controlstrategie is presented which combines dynamic compensation based linear quadraticoptimal control and fuzzy variable step perturbation and observation(P&O) MPPTmethod. The specific description is: the use of the Gauss-Seidel iterative algorithm isproposed to improve output characteristics of the practical model for the fourparameters PV modules, which can be adjusted according to the external temperatureand light intensity changes; A fuzzy control based a variable step size P&O ispresented and its improvements is the use of fuzzy control for variable step sizetracking which overcoming the contradictory between the tracking accuracy andresponse speed of traditional fixed step P&O and the introduction of the minimumstep formula for calculating the light intensity caused by mutations in tracking failures;The implementation of the inverter DC input voltage stability control is put forwardaccording to the different stage of MPPT algorithm through two sets of controls andboth of the controls are the application of the dynamic compensation based linearquadratic optimal control in the Boost converter to restrain the inverter DC sidevoltage sags caused by insolation falling and to suppress the secondary side of powerdisturbances. The simulation results demonstrate the effectiveness of the proposedcontro.lA hybrid controller incorporating servo dynamic compensator based dynamicdisturbance rejection-optimal pole assignment and repetitive control is presented forthe suppression of the harmonics caused by the LC-filtered rectifiers. The repetitivecontroller is adopted to improve the grid-connected current tracking accuracy and power factor. And a dynamic disturbance rejection-optimal pole assignment isintroduced as follows: a servo compensator based LCL-filtered grid-inverter model isestablished and the dynamic disturbance rejection-optimal pole assignment isintroduced by make use of freedoms of selecting feedback matrix to develop theability of dynamic disturbance rejection. The simulation results verify the perfectperformance of the proposed control which Applied to those grid-connected inverterwith higher requirements current waveform.