Research On Model Predictive Control And System Design Of Static Var Generator

As the society is in rapid progress,the power grid equipment installed capacity grows and the same holds for the nonlinear load and reactive load.They will induce large amounts of reactive current and harmonic current,causing lots of damage to power network,affecting the safety and stability in the operation of power system and impairing the quality of power grid.The method of installing static reactive power generator in the critical grid nodes can not only improve the power factor in power system and reduce the loss of power grid,but also can improve the stability of power system and suppress its harmonic current.In recent years,model predictive control has been widely recognized in the electric power electronic control due to its simple process,superior performance and good robustness.It has achieved a good promoted application in many areas,such as in the inverter,active filter,PWM rectifier and motor drive,but seldom do researches in the field of static reactive power generator.This thesis studies the design and improvement of model predictive controller in static reactive power generator,designing the hardware and software platform of the experimental prototype to verify the performance of control algorithm.The main work is summarized as follows:First,the research background,significance of reactive power compensation devices,its development history and current research status are intruoduced.The topology structure,working principle and reactive power detection method of static reactive power generator are also analyzed,following with the establishment of mathematical modesl under the static coordinate system and rotating coordinate system respectively.Second,the realization of model predictive control based on finite state element in static reactive power generator is studied;the predictive error feedback is used to correct the problem of model mismatch,and the multi step prediction is adopted to solve the control delay.Multi objective optimization control is realized via compensating the performance index evaluation function.It follows with the full range simulation of the control performance of model predictive control algorithm and improved strategy,using the MATLAB SimPowerSystems module.The findings showed that the static var generator could have a better performance with model predictive control algorithm and improved strategy.Then,the paper gives a detailed design to the experimental prototype hardware system separately from the main circuit,sampling circuit,drive circuit,protection circuit and auxiliary circuit these five parts,with TMS320F2812 as the main control chip,presenting main parameters calculation process,main parameters of the components and the specific circuit schematic.Last,this paper performs the system design of the experimental prototype software system according to DSP program,PC and control board procedures these three parts respectively.The system structure of software is introduced and module functions of the main modules,design ideas,implementation processes and main parameters are illustrated as well.