The Implementation of Identification and Prediction of Music Rhythm by Microcontroller

• Main Authors: Yu-Ren Chen, 陳鈺仁
• Other Authors: Yue-Der Lin
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/31247537698787529233

碩士 === 逢甲大學 === 自動控制工程所 === 99 === In the past, most entertainment robots were moved by the database which was already saved. In this research, we developing a module that can analysis and predict real-time rhythm of music. However, this would allow entertainment robots to dance harmoniously according to the music given. We have chosen dsPIC33 chip to achieve the recognition of type and the prediction of beats in the music. Firstly, we use external circuit to filter signal before it enter into the chip due to the huge range of music’s frequency. Secondly, normalized it and do fast fourier transform(FFT), then extracted result of bands based on definition of ANSI’s octave. Thirdly, we obtained the main algorithm which can identified rhythm was through scatter diagram that formed by principle component analysis(PCA) and t-test. Fourthly, we use regression analysis to predict the next beat in order to present the real-time movement. Finally, transmit the result to robots through I2C and UART.