Research On De-noising Technology In Ultrasonic On-line Detecting System

Noise, an important factor which impact on the performance of ultrasonic online testing system, has a direct influence on the signal-to-noise ratio's level of ultrasonic signal acquisition. Noise problem of grain ultrasonic online testing system has been studied in this paper from two aspects, theoretical analysis and engineering practices. In this case, we can solve the noise interference problem in ultrasonic online testing system, and improve the performance of system and signal-to-noise ratio of signal acquisition.Ultrasonic online automatic testing system includes PLC control system, ultrasonic transducer, transceiver circuits, data acquisition card and industrial computer. The noise which influences the ultrasonic online testing system is inseparable with its system composition. In ultrasonic online testing system, the noise which has great effect on the system, also named as electromagnetic compatibility noise is derived from servo motor and variable-frequency motor of the PLC control system. The noise in the ultrasonic online testing system includes the impedance matching noise of transducer and its screened cable, as well as the one produced by ultrasonic defect detector itself.In this paper, the research has been based on the sources of electromagnetic compatibility noise and inherent system noise. The author uses different treatment approaches on denoising separately in view of the respective characteristic. First of all, based on the theoretical analysis and the engineering practices, the author puts forward several ideas on electromagnetic compatibility noise, such as the isolation of power supply, the set of ultrasonic probes, and detecting-sink which is made of insulation material, e.g. bakelite or nylon, and achieves good results in noise reduction. Next, in the view of signal processing, the acquisitied signals have been processed by the traditional filtering method and the wavelet transform respectively for further noise reduction processing. All the studies above lay the foundation for the realization of follow-up ultrasonic signal defect recognition algorithm, and provide conditions for the realization of linear quantitative ultrasonic testing of solid-propellant rocket engine grain. It shows a greater value in the improvement of ultrasonic online system's reliability and stability of solid-propellant rocket engine grain...