Research On Control Strategy And Fault Diagnosis Of Four-leg Active Power Filter

With the improvement of people's living standard,more and more non-linear load appears in three-phase four-wire system,which brings harmonic pollution,unbalanced three-phase and inadequate reactive power to low-voltage distribution network.To solve these problems,in order to make the four-leg APF operate reliably and efficiently,this thesis rsearches three aspects of the four-leg APF,including harmonic detection,current tracking control and IGBT open-circuit fault diagnosis.The traditional phase-locked loop can not accurately lock the phase under the condition of three-phase unbalance,harmonic component and DC component,a frequency-locked loop(CFCSOGI-FLL)combining complex filter and improved cascaded second-order generalized integrator is proposed and applied to dq0 harmonic detection.It analysises dq0 harmonic detection based on instantaneous reactive power theory,designes a first-order complex filter phase-locked loop(CF-PLL).Combined with the improved cascaded second-order generalized integrator,this combination has the second-order separation and filtering,which be carried out to eliminate the influence of harmonic component,negative sequence component and DC component.It can correctly lock phase of the power grid under complex operating conditions.The performance of phase-locked and frequency-locked of three phase-locked loop(CFCSOGI-FLL,CF-PLL,DDSRF-PLL)is compared and analyzed under three conditions,including normal power grid,unbalanced power grid voltage,and complex conditionsu combined with unbalanced voltage,DC component and harmonic component.It is validated that the frequency-locked loop proposed in this thesis can accurately detect frequency and phase.In order to improve the rapidity and accuracy of current tracking control for fourleg APF,the nonlinear control strategy based on passive sliding mode control in current tracking is studied.Firstly,the mathematical model of four-leg APF in dq0 rotating coordinate system is established,and a conventional passive controller based on EulerLagrange(EL)model is designed,this chapter uses fuzzy control to realise the selftuning of damping parameters to improve the dynamic and static performance of passive controller.Then terminal sliding mode control is introduced into passive control,and stability analysis is also carried out,which weakens the influence of system parameters on controller performance.The simulation results show that the proposed control strategy has better rapidity,accuracy and robustness than the conventional passive control.The IGBT open-circuit fault diagnosis method based on wavelet packet decomposition and particle swarm optimization BP neural network is studied.This charter analysis the characteristics of output current waveform from network side and line,after single tube failed.Low frequency energy extracted by wavelet packet decomposition and waveform coefficients are used as fault eigenvalues,and particle swarm optimization BP neural network is used for fault diagnosis.The simulation results show the effectiveness of the proposed diagnosis strategy.Finally,the effectiveness and feasibility of the proposed harmonic detection method and passive control strategy are verified on a two-level four-leg experimental platform.A fault diagnosis system based on GUI interface is designed in MATLAB.