Matching control method for mechanical impedance of magnetostrictive precision transducer

A matching control method for mechanical impedance of a magnetostrictive precision transducer, the method comprising steps of:

• Step S1:
• using a compressive stress σ and a driving magnetic field strength H as input neurons, and using a Young's modulus E as an output neuron to develop a three-layer neural network model corresponding to a Young's modulus of a Terfenol-D material;
• Step S2:
• acquiring a plurality of sample data from a predetermined original data set, and forming the acquired plural sample data into a training sample set and a testing sample set, respectively, wherein each said sample data has data feature terms consisting of the compressive stress σ, the driving magnetic field strength H and the Young's modulus E;
• Step S3, introducing the training sample set into the three-layer neural network model and performing training using a predetermined Bayesian regularization training algorithm, and introducing the testing sample set into the three-layer neural network model for testing, thereby obtaining a trained and tested three-layer neural network model;
• Step S4, optimizing connection weights and thresholds among individual layers of the trained and tested three-layer neural network model, thereby obtaining an optimized three-layer neural network model;
• Step S5, according to the training sample set and the testing sample set, training and testing the optimized three-layer neural network model, until the optimized three-layer neural network model is converged into a predetermined error index, thereby obtaining a final three-layer neural network model corresponding to the Young's modulus of the Terfenol-D material;
• Step S6, based on the obtained final three-layer neural network model corresponding to the Young's modulus of the Terfenol-D material, developing a mechanical impedance model of the magnetostrictive precision transducer;
• and
• Step S7, based on the mechanical impedance model of the magnetostrictive precision transducer, developing a matching control network for the mechanical impedance of the magnetostrictive precision transducer, and developing an inverse model that is from the Young's modulus of the Terfenol-D material to a bias magnetic field, thereby obtaining a level of the bias magnetic field required by a target stiffness, and after series compensation between the developed inverse model and the magnetostrictive precision transducer is done, finding a solution to the developed inverse model, so as to obtain the bias magnetic field and input it to the magnetostrictive precision transducer, such that the level of the bias magnetic field can be changed by changing a bias current of a DC bias coil in the magnetostrictive precision transducer and achieving adaptive matching between the mechanical impedance and impedance of a load.