Design Of Control Terminal Based On Speech Recognition

In recent years,with the development of Internet of Things technology,the traditional human-computer interaction already cannot satisfy people to the intelligence community,the diversification of application domains such as smart home control requirements.At the same time,with the development of speech recognition technology,wireless communication technology and embedded microcontroller processor performance has increased,the use of voice for human-computer interaction has become a popular trend of smart home controller.To this end,the paper put forward the design of a smart home control terminal based on speech recognition,and identify the control terminal can sensor user’s voice commands,and can complete the user instructions by 2.4 G wireless communication mode of wireless transmission and the operation control of the controlled object.First,this paper gives the function definition and design of the controlling terminal,the existing speech recognition algorithm is analyzed,and through the introduction in the endpoint detection threshold-crossing rate instead of the zero crossing rate,avoid the interference of random noise in the matching process into easing short points,on traditional DTW speech recognition algorithm is improved,overcome caused by inaccurate endpoint algorithm test model and the reference model starting point to end point not alignment problem.MATLAB simulation shows that the improved DTW speech recognition algorithm can effectively complete the small speech recognition task of training and recognizing ten Chinese isolated words.Then,this paper describes the hardware and software design and implementation of the control terminal in detail.The hardware of the control terminal consists of speech recognition module,communication module and control module.In terms of hardware design,the speech recognition module,with STM32F4 processor as the core,is composed of STM32 minimum system circuit,voice signal acquisition circuit and GUI circuit to realize the sensing,acquisition,recognition processing and GUI visual display of voice signals.In terms of hardware design,the communication module consists of NRF24l01-based 2.4G wireless communication circuit and STM32F4 full-speed USB circuit,which realize 2.4G wireless transmission and USB communication of user instructions.In terms of hardware design,the control module also takes STM32F4 processor as the core,and controls the motor and other controlled objects through IO,PWM and other peripherals.The software of the control terminal consists of STM32F4 processor program of the speech recognition module and STM32F4 processor program of the control module.Speech recognition module STM32F4 processor program,with the aid of STM32 CubeMx software generated STM32F4 library function implementation,include: the main program of the speech recognition module,the GUI program,the USB communication program,the 2.4G wireless transmission program,and the speech recognition program;the STM32F4 processor program of the control module is also implemented by the STM32F4 library function generated by the STM32 CubeMx software,mainly including: The main program of the module,2.4G wireless receiving program,motor control program,relay control program.Finally,this paper presents various functional tests of the control terminal.The test results show that the control terminal designed in this paper can effectively sense and recognize the speech instructions issued by the user,can complete the wireless transmission of user instructions through 2.4G wireless communication,and can complete the effective control operation of the controlled object.