The Application Research Of Multi-wavelet In Wire Rope Fault Signal Processing

When used in the project, the wire rope is always inevitably frayed, broken off and so on. As the wire rope is used widely, the detection of faults becomes more and more important. Except for adopting the high sensitivity excellent sensor, the methods of signal processing is also very important for the leaking magnetic field is often very weak, which is generated by the small fault.Multi-wavelet is a new wavelet construction theory based on traditional wavelet. It can own properties of symmetry, orthogonality, short support and high-order vanish moments, however, traditional wavelet can't possess all these properties at the same time. The research and applications of multi-wavelet are just in the initial stage, and there is no paper about multi-wavelet application in the wire rope detecting in the world.The thesis introduces the importance of wire rope detection firstly, reviews the history and development of the wire rope nondestructive testing, and then expatiate the magnetic detection principle and the design of wire rope detector based on this principle. The detector is made up of two parts. The function of hardware includes the gathering of signal, A/D conversion, data storage and so on. The function of software includes signal de-noising, the judgment and display of fault location and total number of faults, RS232 serial communication and so on.The thesis puts forward a new threshold function. This threshold function is continuous in the wavelet domain while the value is greater than the threshold value, it also has continuous high differential. Also in the domain, the compression of wavelet coefficients and the distribution of noise signals are decreasing along with wavelet coefficients'increase. Detecting signals are de-noised by using the new threshold function that is proposed in the paper. A mess of simulation results show that the de-noising effect, using new threshold function, is better than others that using traditional soft or hard threshold value.Based on the depth study of multi-wavelet, multi-wavelet is firstly applied in signal processing of wire rope. The detecting data of wire rope is de-noised and compressed by using multi-wavelet analysis and wavelet. The simulation results show that the former is better than latter on de-noising effect, and the compression ratio and mean compression error of reconstruction signal by using wavelet is slightly better than using multi-wavelet. However, the signal-to-noise ratio of reconstruction signal using multi-wavelet is great higher than using wavelet. After de-noising and reconstructing the reconstructed signals, it proves the compression signal using multi-wavelet retains much more information of small singular points.