Fig. 1. Window transducer model

Fig. 2. Impedance characteristic curve of transducer

Fig. 3. Impedance circle at different temperatures （F=0 N）

Fig. 4. Impedance circle at different loads （T=20 ℃）

Fig. 5. Output testing system of the transducer

Fig. 6. Relationship curve between acceleration and frequency at different temperatures （F=0 N）

Fig. 7. Relationship curve between acceleration and frequency at different bias magnetic fields

Fig. 8. Relationship curve between acceleration and bias current at different temperatures （F=0 N）

Fig. 9. Relationship curve between acceleration and bias current under different loads （T=20 ℃）

Fig. 10. BP neural network model

Fig. 11. Flow chart of GA-BP neural networks algorithm

Fig. 12. Prediction results and errors of different models

Fig. 13. Closed-loop control system for high frequency magnetostrictive transducer

Fig. 14. DSP control system based on GA-BP neural network

Fig. 15. Process of bias current change when transducer load is abrupt

Fig. 16. Comparison of peak acceleration of the transducer before and after the closed-loop control system

Fig. 17. Comparison of transducer acceleration waveform before and after control by closed-loop control system