Research On Key Technologies Of Shunt Active Power Filter

The development of science and technology promotes the large-scale application of new power electronic devices,which brings convenient life to people.At the same time,the nonlinear characteristics of various components themselves also aggravate the harmonic pollution in the power system.At present,how to effectively reduce or even eliminate harmonics in power system and improve power quality has become an urgent problem in power system.As a traditional harmonic control device,passive power filter can only eliminate the fixed frequency harmonics,and its application scope is greatly limited.Active Power Filter(APF),as a kind of new Power electronic equipment,can effectively control the dynamic harmonics in the power grid,which has rapidly developed into the main measure of current harmonic control.The research object of this thesis is the shunt three-phase three-wire system voltage source active power filter.The research of this thesis mainly focuses on three aspects:harmonic current detection,DC side capacitance voltage control and current tracking compensation control.After the analysis and comparison of several commonly used harmonic detection methods,this paper proposed to optimize the initial weight threshold of BP neural network with Beetle Antennae Search(BAS),and applied the trained BAS-BP neural network to harmonic current detection.DC capacitive voltage control is an important part of the research of active power filter.Firstly,the reason of DC capacitive voltage fluctuation is analyzed in this thesis.Aiming at the shortcomings of traditional PI control,a fuzzy adaptive PI+PI piece-wise control method is proposed.In the selection of harmonic current tracking compensation control strategy,this thesis introduces the advantages and disadvantages of the three commonly used control methods,and uses mathematical modeling to explain the fundamental reason of the system delay when the conventional SVPWM control is used for harmonic current compensation.In order to solve this problem,this thesis proposes to use BP neural network to predict the instruction current,so as to realize the application of deadbeat SVPWM current tracking compensation control in active power filter.In order to make the conclusion more convincing,the current commonly used control methods were introduced in the simulation process for comparison,and analyzed and explained one by one according to the simulation results.The simulation results show that the optimization schemes proposed in this thesis have achieved the expected results and can effectively improve the overall performance of the active power filter.In the experiment,the TMS320F28335 DSP is used as the main control chip of the active power filter.Aiming at the hardware design of the active power filter,this thesis explains the selection of its main components,and gives the circuit diagram of the core function module of the active power filter.In the aspect of software programming,this thesis introduces the programming ideas of the main functional modules,and explains them in the way of flow chart.