Research On Application Of Instantaneous Frequency Estimation In Wire Rope Ranging And Signal Sampling

In the non-destructive testing method of the wire rope based on the magnetic detection method,in order to facilitate the positioning of the wire rope defects,the main magnetic flux or the magnetic flux leakage signal of the wire rope is often used for equal spacing sampling by using the encoder wheel.During the detection process,the guide wheel connected to the encoder wheel often slips and jams,resulting in the absence of the sampling signal.And the worn guide wheel as well as the low positioning accuracy can be caused by long-term contact sampling.In order to solve the above problems,a kind of system about wire rope ranging and signal sampling without encoder wheel is designed in this dissertation,which is related to the character about the sinusoidal form of the wire rope waveform in the magnetic detection method and the relation between waveform frequency and the speed of the wire rope.Both offline work mode and online work mode are included in this system for different application environment.Three commonly used instantaneous frequency estimation algorithms,such as Wiener-Ville distribution,wavelet spectrum analysis and phase difference method are introduced in this dissertation.The three algorithms are compared and analyzed by using the linear frequency modulated signal for signal processing effect,and the core algorithm of the built system is determined according to the result.Aiming at the influence of WVD distribution on noise and distortion waveform,the Viterbi algorithm and ensemble empirical mode decomposition are used to optimize the peak extraction of WVD and the influence of signal defect waveform.In view of the poor noise immunity of phase difference method,iterative variable weighted least squares method is adopted to reduce noise interference,and the weight function is modified to make it suitable for single operation.In the analysis of the instantaneous frequency method,the traditional full digital multiplier method has also been studied.The error generation of the traditional multiplier method is analyzed and optimized the by the improved dispersion compensation method in this dissertation.As a result,the algorithm is implemented on the DSP system,and the effect of multiplying the arbitrary multiplication factor is achieved.Then the noise of the actual signal is evaluated,and the super servo circuit and the fourth-order Butterworth low-pass filter are analyzed and built to reduce the influence of DC component and noise on the system.According to the above analysis,the offline work and online work structure are designed,and in the related links of online work,the phase smoothing link is designed to further reduce the influence of noise and mutation.After the above research,the online working system with the core phase difference algorithm is realized on the DSP platform to achieve equidistant sampling.The offline working system with the core WVD distribution algorithm is realized in the host computer,and the isochronous sampling points are converted to coordinate points with distance information on the wire rope.Both methods are compared with the defect positioning effect of the traditional encoder wheel to verify the effectiveness of the system.