Research On The Method Of Harmonic Analysis For The Current Signal Of Locomotive Driving System Power Line

Harmonic detection is an important part in the process of the realization of harmonic diagnosis technology.It plays a role in providing a diagnostic basis for diagnosis of electrical equipment status,and has a great influence on the correctness.In this paper,the characteristics of current in the electric wire of locomotive and vehicle driving system and the requirements of the state monitoring of the drive system are concerned,problems such as the method of harmonic analysis and the design of the harmonic detection system are studied.Main research contents and research achievements,in this paper,include:(1)Two common methods of harmonic analysis are studied.First,the basic theory of Fourier transform is briefly introduced.Then,the shortage of it when calculating the harmonic parameters is explained and therefore introduces an improved algorithm of it windowed interpolation algorithm.Second,concerned window and interpolation problem of windowed interpolation algorithm,spectrum characteristics of several common window functions are studied,and with the method of polynomial approximation,their practical interpolation formula is calculated.Finally,the results of simulation explain that,windowed interpolation algorithm can effectively restrain the synchronous sampling of the Fourier transform of Zou Bo effect and fence effect,improve the accuracy of harmonic analysis under asynchronous sampling condition.Firstly,the basic principle of wavelet packet decomposition and reconstruction is briefly introduced.Secondly,according to the change of the spectrum of the signal after the wavelet packet decomposition process,the cause of the wrong frequency band order of the wavelet packet decomposition is deduced,and the correct frequency bands order are given.At last,according to the characteristics of the harmonic signal,the db45 wavelet is chosen as the wavelet function of the harmonic analysis.The simulation results show that the harmonic analysis method based on db45 can detect the harmonic parameters accurately,and it also has a good performance on the detection of the Non-stationary signal.(2)A new method of harmonic analysis is proposed.When giving a summary on the the research of wavelet packet analysis and Fourier transform analysis,that the two kinds of harmonic analysis methods are not very good to meet the requirements of current harmonics analysis in the electric wire is revealed.So a new method of harmonic analysis,which is based on the harmonic wavelet filter,is proposed in this paper.First,the special characteristics of the harmonic wavelet filter as a filter is analyzed.In view of the shortcomings of the slow Time domain decay rate and large fluctuations of discrete spectrum of the harmonic wavelet,a new method to speed up the harmonic wavelet is obtained by using the Gauss function in the time domain.Then,the detection of the harmonic detection method based on harmonic wavelet filter is presented.Simulation results show that the harmonic detection method based on harmonic wavelet filter does not need the condition of synchronous sampling,and has a good performance on both steady state harmonic detection and non-steady state harmonic detection.The results not only include the time domain waveform of each harmonic,but also can export instantaneous amplitude of each harmonic,which can be used directly by harmonic diagnosis technique.(3)a harmonic detection system is designed.Firstly,the overall framework of the harmonic detection system is introduced,and the working process of the system is briefly analyzed as well.Then the hardware and software design of the harmonic detection system are introduced.In the aspect of hardware,the main circuit of data acquisition module based on SIM3U167 and magnetic probe circuit is presented.In the aspect of software,the working process of singlechip and PC is briefly introduced.Finally,a set of experiments are designed to test the accuracy of the harmonic detection system using the harmonic analysis method presented in the third chapter.The results show that the system has low system noise,strong anti-interference ability,high accuracy of harmonic analysis.When the sampling frequency is 1600 Hz,sampling time 0.5 seconds,the maximum amplitude relative error is only 0.63%.