Research On Robot Control Technology Based On Speech Recognition

With the development of science and technology, speech recognition technology has got more and more attention. As an important branch in the field of intelligent robots research, the purpose of speech recognition is to make the machines can understand human language to achieve human-machine communication. Therefore, it’s an embodiment of the level of development of current automation to apply speech recognition technology in the field of robot control.The paper was based on speech recognition technology and with Simple robot motion control as the goal, to identify isolated speech signal from specific speaker and focus on research feature extraction algorithm and speech recognition algorithm.Based on basic theoretical knowledge of speech recognition, the collected speech signal was analyzed in time domain and frequency domain, then preliminarily transformed such as normalized, pre-emphasis, plus window and framing, start and end point detection. Mel-Frequency Cepstrum Coefficients (MFCC) was used as the characteristic parameter sequence of speech signals, and MFCC was combined with Wavelet Transform to get more efficient feature extraction WT-MFCC.Comparison of several commonly used speech recognition algorithms, Dynamic Time Warping (DTW) algorithm and neural Network algorithm were used to identify the speech signal. DTW algorithm and BP Neural Network algorithm were deeply researched, and then, MATLAB software was used to write speech recognition program based on DTW and BP Neural Network. What’s more, recognition performance of two algorithms were compared and optimized.The collected speech signal program and the PC interface for speech recognition were written through LabVIEW software. The lower machine DC motor’positive and reverse control circuit was built through Proteus software, the simulation of speech recognition control system in pure software environment was completed through a virtual serial port technology.The collected voice command was identified by upper computer speech recognition system, the MT-UROBOT robot motion control program was written, and the DTD462module was used to achieve wireless communication, so as to realize speech control of MT-UROBOT robot.