Application of wavelet and Hilbert transform in the detecting parameter recognition of box milk quality

In order to improve the detecting accuracy of the fast-detection device of the UHT(Ultra Heat Treated) box milk quality, the method of wavelet de-noising is applied to the free damped torsional vibration signals produced by the liquid milk box. And by getting the envelope demodulation using the Hilbert transform, the damping ratio of the system is finally calculated. Under the case that the package of liquid box milk is not destroyed, the change in the quality of the liquid milk can be estimated indirectly with the help of the damping ratio of the system. The theoretical and experimental results indicate that the detecting accuracy of the damping ratio of the system is improved significantly in contrast with the traditional calculation method of the damping ratio. This is due to the application of the additional two filtering process namely, the wavelet de-noising and the Hilbert transform for the envelope. The distributing range of the detecting results of the detection device for box milk quality adopting this calculation method also gets well controlled.