| High voltage pulse dust removal power supply is a common method of dust removal of waste gas.With the increasing attention of the country to environmental protection,its important role is becoming more and more prominent.Domestic coal-fired power plants,cement,metallurgy and other scenes use electric dust removal equipment,has a wide range of transformation and use basis,and high pressure dust removal power supply as its energy supply unit,with the corresponding control system,can greatly improve the dust removal efficiency,boost the last kilometer of the electric dust remover.Furthermore,as far as the characteristics of the power supply itself are concerned,it can avoid the voltage pole limitation of the basic dc power supply,greatly reduce the frequency of the reverse corona and the suppression of spark discharge phenomenon,so as to maintain a larger dust removal output power and break through the efficiency bottleneck.Based on this,this thesis is based on the original parallel symmetric boost pulse power supply,designed step by step on the basis of60 k V dc voltage,superimposed with 80 k V pulse voltage,and built a parallel pulse power supply prototype,and conducted a series of experiments to verify.The main circuit of the superposition power supply is analyzed,the working principle is analyzed,and the simplified calculation model is given.The main parameters are calculated and all the parameters and part of the device selection are given.Then,the transition state of flashover is analyzed in detail.In particular,according to different flashover states,the type of spark flashover is divided,and the simulation verification is given.The structure of the main circuit and the arrangement of some devices are given.From the control parameter setting,combined with the control system hardware platform based on DSP and FPGA,the control system is described as a whole.The main regulation mechanism is introduced,and the interaction between pulse and dc is highlighted.And to undertake the classification of spark flashover,highlighting the identification and processing of spark types;The hardware architecture and software realization of the control system are introduced briefly.In order to solve the problem of bus boost phenomenon,a solution strategy is proposed and verified by simulation.Based on BP neural network,the prediction mechanism of low-frequency,low-frequency and high-order harmful pulse spark phenomenon is proposed,and the corresponding BP model is built.Network training,simulation verification and network continuous optimization are carried out to verify the feasibility of this method.On the prototype platform built,three typical experiments were completed by means of step-down experimental verification,that is,operation parameter verification,zero turn-off and spark flashover experiment.In addition,for the phenomenon of busbar boost,additional experiments of busbar voltage stabilization are added to achieve the purpose of verification and optimization,and some references are provided for subsequent studies. |
