Design And Implementation Of A Control System For Voice Car Based On

With the development of information and communication technology, intelligent voice control application is used more and more widely in people’s daily life. This makes people need higher requirement of intelligent product performance, real-time, processing speed. Nowadays, the speech recognition technology is mature, but mostly was implemented by computer software, the real-time deficiencies of defects of software realization of speech recognition can’t meet the demand for real-time speech recognition is the industrial field. Aiming at improving the problem of real-time speech recognition, this paper considers mainly from two aspects:one is to improve the parts in the processing of speech recognition, improve the processing speed of a link to increase the recognition speed of the whole system; Second, by using the combined technology of hardware or soft hard, using hardware’s own good real-time, system stability, fast processing speed advantage to increase the speed of the speech recognition.The paper designed an autonomous driving control system of voice car according to the above considerations. The system used the voice acquisition chip WM8731of DE2to acquire voice, and extracted MFCC voice features after preprocessing these voice data collected, carrying out FFT transform, feeding with Mel filter banks and processing DCT operation, where FFT transform is a dual-core processor which used a CORDIC algorithm based on pipeline architecture to carry on the spectral analysis of voice signals more stably and rapidly. This system would produce recognition results after comparing the data commands between the test library and the template library for these voice parameters extracted with the HMM model constructed by HTK toolkits. This paper finally achieved voice control driving of the car through realizing the communication between DE2and microcontroller, and completing the command transmission through the drive circuits controlled by the microcontroller.The system on the DE2development platform for downloading test, The results show it is completed in real-time response platform to voice command, and the system has a simple hardware structure, real-time, high reliability, can be widely used in speech recognition control.