Admittance Circular Series Resonance Tracking Based On BP Neural Network

The ultrasonic processing system includes an ultrasonic power supply, an ultrasonic transducer and a horn. In the whole ultrasonic machining system, the ultrasonic power supply plays a vital role, which determines the amplitude stability of the tool machining, the output efficiency of the ultrasonic transducer and the surface quality of the workpiece. Therefore, a fast and accurate series resonant frequency tracking algorithm is the key to ultrasonic power supply. In this paper, the resonant frequency of the transducer is tracked by the phase difference, and the series resonant frequency is obtained based on the resonant frequency point using the admittance circle characteristic of the transducer. In the process of finding the series resonance frequency, BP neural network is used to predict the dynamic resistance of the transducer. This method can not only quickly obtain the change of the dynamic resistance but also has a relatively high precision.