Font Size: a A A

Control Stratigies For High Efficiency And High Power Factor Three Phase Power Supply

Posted on:2017-06-21Degree:DoctorType:Dissertation
Country:ChinaCandidate:X GuoFull Text:PDF
GTID:1318330536976615Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
Power electronic converter is an important conversion device for electric energy utilization.Three-phase power supply is a kind of electronic converter that transforms the commercial three-phase alternating current(AC)power into direct current(DC)power with the required specification.The traditional three-phase power supply uses the diode or thyristor rectifier,which causes many shortcomings including the harmonic pollution,reactive power loss and electromagnetic interference,etc..The soft switching active power factor correction PWM converter with high power factor and high efficiency becomes an important way to solve the above problems,thus,it has broad application prospects to replace the conventional equipment in the fields of high power uninterruptible power system(UPS),reactive power compensation,solar power generation,AC/DC drive system and so on.In this thesis,based on the technical requirements of heating power supply for monocrystalline silicon furnace,the advanced topology structures are proposed.The power factor correction technology,soft switching technology and the advanced control strategies are integrated together to design a set of three phase power supply prototype with high efficiency,high power factor and high performance.The advanced control strategies are the main investagion focus.The main research contents of the thesis are given as follows:(1)Based on the analysis of traditional three phase power supply topology,the two stage converter configuration with the compound active clamp zero voltage soft switching three phase power factor correction converter(CACZVS three phase PFC converter)as the first stage converter and the full bridage phase shifted zero voltage and zero current(FB-ZVZCS)DC/DC converter as the second stage converter are proposed for the power supply topology.To deal with the technical requirements of high power factor and high efficiency of the heating power supply,the different topologies,soft switching conditions and related modulation methods are analyzed contrastively.The selected CACZVS three phase PFC converter can achieve unity power factor and output DC bus voltage regulation,at the same time,the FB-ZVZCS DC/DC converter can achieve the output voltage and output power regulation.The working principle and equivalent circuit of the proposed three phase power supply are analyzed,and the corresponding mathematical models are built,which facilites the subsequent control method research.(1)To deal with the difficulty of multiple parameters selection problems for strong coupling multi-controllers,a multi-objective chaotic particle swarm optimization(PSO)algorithm is proposed to improve the performance of the converter.The traditional feedforward decoupling proportional integral(PI)control algorithm base on the two-phase rotating coordinate system of the first stage CACZVS three phase PFC converter is used as an example for illustrating the systematical way and the procedure of the parameters optimization.Three phase PFC converter contains three control loops including quadrature-axis current loop,direct-axis current loop and output voltage loop,therefore,6 PI parameters for 3 PI controllers need to be designed,and the control parameters are mutual influenced.The optimization objectives include the fast responding speed and zero steady state error of output DC voltage,as well as the unit power factor.this is a complicated multi-parameter(multi-dimensional)and multi-objective optimization problem.To solve the problem,in this article(1)a multi-objective PSO method based on Pareto optimization theory is proposed to get balance among multi-objectives.(2)an improved chaotic particle swarm optimization(CPSO)method is proposed using the unilateral coupled map lattices spatiotemporal chaos model to product the uniform distribution of the initial particles.The proposed CPSO method generates the multi-dimensional uniform distributed initial particles to solve the problem in the traditional CPSO method that can only produce one-dimensional initial particle,thus improves the searching speed and the global optimization performance.(2)Three different nonlinear control methods are proposed to reduce the influence of the converter parameters uncertainties(variation),and to improve the control performance of three phase power supply.The first stage CACZVS three phase PFC converter is nonlinear system with multi-varible strong coupling,circuit parameter uncertainties,and load variation according to the operating conditions,in order to achieve the high control performance of three phase power supply,the following control strategies are proposed,including:a)a feedback linearization control method based on sliding mode load observer.The proposed method uses the sliding mode observer to estimate the load parameter online,at the same time,to deal with the nonlinear and strong coupling features of CACZVS three phase PFC converter,the feedback linearization controllers are designed using the load estimated valuegiven by the observer to improve the adaptiveness of the controller with repect to the time-varying load;b)Considering the load and the output filter capacitor uncertainties,a voltage loop model reference adaptive controller is designed and current loop feedback linearization controller is used.The proposed control scheme has less control parameters and the control parameters are easy to select,the controller has better adaptation to load change and output filter capacitor uncertainty;c)the double closed loop robust sliding mode control method is proposed to deal with the input voltage fluctuation,the uncertainties of input filter inductance,filter inductance equivalent resistance,output filter capacitor and load parameters.Compared with the traditional feedback linearization method,the proposed control methodimproves the system robustness to parameter uncertainties and parameter changes by introducing the robust items into the controller,in addition,the usage of shrinking boundary layer decreases the chattering of control force and reduces the distortion of the input current.As a result,the three phase input current waveforms are smooth and the voltage fluctuations are less than some existing method when the load changes.(3)A robust constant frequency model predictive control(RCF-MPC)method is proposed to realize the current loop constant frequency model predictive control for CACZVS three-phase PFC converter,and the sliding mode variable structure controller is used for the voltage loop.The proposed control scheme improves the performance of the first stage converter under the conditions of unbalanced three-phase input voltage and circuit parameters uncertainty(variation).The practical working condition of the monocrystalline silicon furnace heating power supply are usually affected by unbalanced three-phase input voltage and parameter uncertainty(variation)at the same time,the traditional methods to deal with the unbalancedthree phase input voltage need to consider the three phase positive sequence currentand and negative sequence current,respectively,thus,many control loops are involved and the control structure is complex,in this thesis,the predictive control method is used to solve the problem.The traditional finite state model predictive control method(FS-MPC)for three-phase hard switching PWM converter,because of its variable switching frequency,cannot be directly applied in the three phase PFC soft switching converter in which the switch state have to change at least one time to create soft switching conditions.To solve the above problems,the RCF-MPC method is proposed,the propose method based on the optimal time series constant frequency model predictive control(CF-MPC),the negative sequence current component,as well as the robust item to deal with the prediction model error,is added into the cost function to improve the performance,the voltage loop controller still uses the sliding mode variable structure control method.The propose control scheme improves the performance of the converter under the condition even close to the practical working condition where the three phase input voltages unbalance and the converter parameters uncertainty or variation exist simuteniously.(4)The variable universe fuzzy control method is proposed to realize the output voltage regulation for the second stage FB-ZVZCS converter,compare with the traditional PI control method,the propose method has better performance when the converter model is unknown.The traditional controller of second stage FB-ZVZCS converter are designed based on the small signal linearized model and the linear system theory,however,the accuracy of the small signal linearized model depends on the working point,when the load change with the working state,the control performance of the traditional method will reduce.The variable universe fuzzy control method is proposed to improve the performance of the converter,the propose method does not depend on the accurate model of the converter,and the usage of the universe shrinking factor solves the conflict between the response speed and the control accuracy when the number of fuzzy rules and fuzzy variables are limited.The steady-state tracking accuracy is improved under the premise of ensuring the responding speed.(5)Base on the TMS320F28335 digital signal processing,the control circuit is designed and two-stage converter circuits are built.The experimental platform of 1.2kW three phase power supply is made to verify the control methods and pave the way to practical application of the heating power.The experimental results obtain from the two-stage converter experiment platform shows the expected performance.
Keywords/Search Tags:Three phase power supply, Compound active clamp, Full bridage phase shifted ZVZCS, Three phase unbalance, Parameter uncertainty, PI parameters optimization, Particle swarm optimization, Feedback linearization control, Sliding mode observer
PDF Full Text Request
Related items