Research on vibration property of Picea wood for piano sound quality neural network and ant colony algorithm

Abstract With its characteristics of wide range and large volume variation, piano has an unshakable position in the Musical Instruments and is widely loved by numerous people. In a upright piano, the basic function of the piano sound board (resonance board) is used to expand the sound emitted by the string. Different wood species would produce different vibration amplitudes and resonance effects. If the other parts of the two pianos are exactly the same, their advantages and weaknesses can only be determined by the resonance board, so the resonance version of the material is particularly important. Based on the vibration theory of piano resonance plate, this study analyzed the vibration characteristics of the resonance plate and used the BP neural network ACO-BP optimized by ant colony algorithm. Using the dynamic elastic modulus, the elastic modulus and shear modulus ratio, acoustic radiation damping coefficient and acoustic impedance of the wood species of Russian Far East Picea koraiensis Nakai, Picea jezoensis, Picea spinlosa and Picea sitchensis Carr. as the prediction inputs, the accuracy of the model for predicting the musical instrument grade reached 95%. It is of great value for improving the acoustic quality of piano by establishing the corresponding acoustic quality evaluation system of national Musical Instruments and deeply investigating the prediction method of the parameters of the acoustic evaluation system to realize the transition from the subjective evaluation to objective evaluation.