Font Size: a A A

Research On Rotor Position Estimation Methods Of The Brushless Synchronous Starter/Generator Based On The Excitation System

Posted on:2020-12-20Degree:DoctorType:Dissertation
Country:ChinaCandidate:S MaoFull Text:PDF
GTID:1522307100473934Subject:Electrical engineering
Abstract/Summary:PDF Full Text Request
The aviation integrated starter/generator(ISG)system is an important technical characteristic of more electric aircraft(MEA).Since three-stage brushless synchronous machines have been widely adopted as main generators on aircraft,the research on ISG system based on three-stage brushless synchronous machines plays a vital role in growing popularity and facilitating development of MEA technologies.Accurate rotor position knowledge of the main generator(MG),essential for the starting control of the brushless synchronous starter/generator(BSSG),is usually provided by mechanical position sensors,which increase system volume,weight,and cost and decrease system reliability.Therefore,rotor position estimation methods of BSSG have important research significance.The traditional high-frequency signal injection methods rely on the fixed motor magnetic saliency to estimate the rotor position.Due to the variable magnetic saliency of MG,traditional methods have risks of failure when applied to MG.What’s more,persistent high-frequency signal injection leads to losses,noises,torque ripples,and reduction of the starting torque of MG.Thus,rotor position estimation methods of MG based on the brushless excitation system(BES)were proposed in this thesis to solve these problems.Both rotor current characteristics and mechanical features of BES are exploited in this method.As a result,persistent highfrequency signal injection is no longer needed and effects of the variable magnetic saliency of MG are avoided.This thesis revolves around four parts of this method,and they are characteristics of BES,rotation angle estimation methods of MG,initial position estimation methods of MG,and the rotor position estimation method of MG at medium-and high-speed ranges.Firstly,characteristics of four kinds of rotor currents of BES under two common commutation modes were analyzed by using a three-phase bridge rectifier circuit model.Because rotor currents of BES cannot be measured directly,their estimation methods were described.Then,error sources of the estimation process were discussed,and errors caused by the pure integrator and exciter parameters were analyzed specifically.Their influence on estimation results of rotor currents of BES was also studied.Secondly,the research on the rotation angle estimation method of MG was carried out.The correspondence between the rotor position estimation error of the main exciter(ME)and characteristics of rotor current estimation results of BES was analyzed.Due to different effects of the rotor current estimation error of BES on this correspondence,two kinds of position estimation indicators of ME,as well as the corresponding rotation angle estimation methods of ME and estimation error processing methods of rotor currents of BES,were proposed.According to the mechanical characteristics of BES,mechanical rotation angles of MG and ME are the same.Therefore,two rotation angle estimation methods of ME corresponded to two rotation angle estimation methods of MG.The effectiveness of the two methods was verified by simulation,and further verified by experiment on BSSG starting control experimental platform built in this thesis.Besides,the comparison analysis of the two proposed methods was given.Then,research on initial position estimation methods of MG was performed.Two methods were proposed according to different signal transmission directions.In the first method,whose feasibility is based on the estimable field current of MG,transmission direction of injected signals is from the stator of MG to the rotor of MG.This method has two steps.The first step is to pre-estimate the initial position of MG with the sinusoidal injection signal,and the second step is to determine the polarity with the step injection signal.In the second method,transmission direction of injected signals is from the rotor of MG to the stator of MG.In this method,initial position of MG is detected with the electromotive force(EMF)of MG,induced by the gradually increasing field current of MG when ME is powered on.The simulation and experimental results showed the effectiveness of the two proposed methods.Besides,the comparison analysis of the two proposed methods was given.Then,the rotor position estimation of MG in the low-speed range was realized by a combination of initial position estimation methods of MG and rotation angle estimation methods of MG,and its validity was also verified by simulation and experiments.Finally,the research on the rotor position estimation method of MG in medium-and high-speed ranges was carried out.The mathematical model of MG for parameter estimation and basic principles and characteristics of the least square algorithm were analysed.Then,the scheme to embed the parameter estimation method in the rotor position estimation method without coupling was proposed by utilizing the rotor position estimation results of MG in zero-and lowspeed ranges and the strong robustness of the sliding mode observer.After the effectiveness analysis of the parameter estimation algorithm itself,the parameter estimation method of MG based on two independent least square estimators was proposed,which avoided the coupling problem between the least square estimators in the traditional methods.With estimated parameters,rotor position estimation method of MG in medium-and high-speed ranges,based on the extended EMF which is observed by the sliding mode observer,was analyzed.Then,the parameter estimation method and rotor position estimation method of MG in medium-and highspeed ranges were verified by simulation and experiments,and the results illustrated the effectiveness of the proposed method.
Keywords/Search Tags:Aviation integrated starter/generator, Three-stage brushless synchronous machine, Brushless excitation system, Rotor position estimation, Parameter estimation
PDF Full Text Request
Related items