With the continuous application of modern electronic technology in automotive systems, more and more in-vehicle electrical appliances have been added to the body of electronic vehicles, which has greatly improved the performance of the car, but the switch of the car cab has also come. The more, this puts higher requirements for the operation of the vehicle-mounted electrical appliances in the driver's driving, and also brings unsafe hidden dangers to the driving process. With the improvement of the speech recognition algorithm and the advent of the new generation of Soc-specific speech processing chips, the operation of controlling the automobile electrical appliances by using voice commands can partially replace the functions of the hands with the mouth, thereby reducing the driver's operational burden and improving the driving safety factor.
This article refers to the address: http://
At present, the electronic voice control of the body in China is mainly concentrated on the application of the car navigation system, and the application value of the voice recognition technology in the body electronics is not fully utilized. This paper proposes a design scheme of controlling the car audio operation with the dedicated voice processing chip UniSpeech-SDA80D51 as the core, and realizes the speech recognition and control of non-specific people on the SL1102C1 car audio.
Voice controller hardware
The vehicle voice controller system consists of a directional pickup, a voice recognition module, a control module and an audio module. The function of the system is: collecting the voice command issued by the driver by the pickup, converting the voice physical sound signal output by the pickup into a voice digital signal by using the voice recognition module, and identifying the Chinese character or word corresponding to the voice command, and then generating and The corresponding entry code command transmits the control command to the sound through the system I/O line, and the car audio receives the control command to generate an action, and responds to the driver's voice command, such as fast forward, rewind, or volume adjustment. The system structure and principle block diagram are shown in Figure 1:
The speech recognition module is mainly composed of a UniSpeech-SDA80D51 chip and peripheral circuits. The module mainly realizes the recognition and processing function of the input non-specific human voice signal, and outputs the entry code corresponding to the voice command.
The SDA80D51 is a new dedicated chip from Infineon for speech recognition and speech processing applications. It is manufactured in a 0.18μm semiconductor process using a highly integrated Soc system architecture. It features an 8-bit high-speed enhanced M8051 core (25MIPS) and a 16-bit fixed-point DSP. Core OAK (100MIPS) dual-core architecture, on-chip integration of direct dual access fast SRAM, 2 ADC and 2 DAC (effective 12Bit), a variety of communication interfaces and general-purpose GPIO devices, external expansion of Flash memory, etc. A small number of peripheral circuits can form a complete application system.
The control module is composed of MCU and analog switch circuit. This module mainly completes the logic analysis and processing of the recognition result outputted by the speech recognition module--the word coded signal, and outputs the control signal corresponding to the function through the analog switch circuit to the sound, and controls the sound. Operation. Among them, MCU selects AT89S51 from American ATMEL company. Because the button control panel on the car audio SL1102C1 is a resistive shunt keyboard circuit, the voltage sampling identification mode is used to identify different key values, and the I/O voltage characteristics of the SDA80D51 chip output are integrated. Use the relay to simulate the closing and opening action of the SL1102C1 control panel buttons.
This design is based on the SL1102C1 car audio from Anhui Senli. The SL1102C1 is a car stereo designed for mid-range cars. It features MP3 playback, radio and display time. It is currently used in the JAC Tongyue sedan. The front panel of the SL1102C1 has a total of 15 buttons and a code switch for adjusting the volume. They are switch on/off, sound effect setting, 6 stations, play/pause, random play, repeat play, browse play, select down song (fast forward), select song (rewind), search down/on One song, up lock/next song, mode conversion, station browsing/automatic storage station, band switching, display time/time setting and reset.
The button on the front panel of the audio is the voltage sampling identification mode. The button includes two actions: short press (delay less than 0.2s) and long press (delay greater than 1s). The output voltage of the control module MCU (AT89S51) is TTL level. Direct use of MCU signal to drive the sound button action is easy to cause misidentification, resulting in system misoperation, so this article uses the analog switch circuit, which solves the above problems well. When the AT89S51 receives a voice command recognition result coded signal output by the voice module, it immediately performs logic analysis and outputs a corresponding control signal to drive the corresponding relay pull-in analog button action. The short press and long press functions of the button are realized by software.
System software design
The software of the system includes: a non-specific person speech recognition module and a logic control module.
The non-specific person speech recognition module is based on the HMM model algorithm. The algorithm collects statistics on a large number of voice data, establishes a statistical model speech library for identifying the terms, and then extracts features from the speech to be recognized, matches the model library, and obtains the recognition result by comparing the matching scores, and passes the GPIO port of SDA80D51. The word coded signal corresponding to the recognition result is output. The speech recognition module is mainly composed of signal preprocessing, feature parameter extraction, model matching and Viterbi algorithm.
The signal preprocessing part mainly performs the sampling and analog/digital conversion functions of the input speech signal. The A/D conversion is implemented by the SDA80D51 embedded 12-bit A/D converter, and the sampling frequency is fixed at 8 kHz.
The feature parameter extraction is based on a speech frame, and the feature is extracted by using a frame. First, the speech signal is overlapped and framing, and the previous frame and the latter frame overlap by half (the frame signal overlap is to reflect the correlation between the adjacent two frames of data), the frame length is 25 ms, and the speech feature is extracted once for each frame.
The MFCC parameters belong to the perceptual frequency domain cepstrum parameters, reflecting the characteristics of the short-term amplitude spectrum of the speech signal. The extraction process of p-dimensional MFCC parameters is shown in Figure 1.
Where: m is the frame number and N is the number of sample points in the unit frame.
HMM is a probabilistic statistical model for describing speech signals. The MarKov chain is used to simulate the change of statistical characteristics of speech signals. The HMM model is developed on the basis of Markov chain.
The Viterbi algorithm is a frame synchronization dynamic regularization algorithm. Given a sequence of observations and a model, the Viterbi algorithm gives a sequence of states with the largest probability density P(Q, O | λ).
The main function of the control module is: after the MCU queries the voice entry signal output by the voice module, the table looks for the entry code, and according to the code, the corresponding button is long pressed or short pressed, respectively, and enters the corresponding subroutine processing, long press The subroutine delay is 1s, and the subroutine delay is 0.2s. In the subroutine, the I/O control signal corresponding to the output voice command drives the relay to pull the analog button or the code switch, and resets the I/O port in time. In order to avoid conflicts between voice control and manual control, the voice control module can be fully compatible with manual control, and manual control can be performed at the same time during voice control operation.
Experimental results and conclusions
The prototype experiment of this system is mainly to test the speech recognition rate of non-specific people and the accuracy of analog switch action. Since the voice entry of car audio is 2~4 words, the experimental content of speech recognition rate is 18 for car audio, 12 for 3 words, 12 for 4 words, and 10 for 4 words. 6 people, 4 males and 2 females (Mandarin and dialect), the experimental environment is noise interference environment and relatively quiet environment. The prototype test results are shown in Table 1.
As can be seen from Table 1, the recognition rate of the system is related to the number of words of the voice command, the distance of the microphone receiving, the dialect of the speaker, and the environment. The recognition rate of 4 words is higher than 2 words. The recognition rate can reach more than 90% in a relatively quiet environment, and the recognition rate of male and female voices is close. The recognition rate decreases in a noisy interference environment. In order to improve the recognition rate of the system, the system prototype adopts the Olympus ME52 directional microphone (not reflected in the table), which improves the receiving range of the microphone and increases the overall recognition rate of the system to 95%.
In the system prototype control circuit experiment, the analog switch action reached a high accuracy rate, the test result is more than 98%, as long as the control program is running normally, each relay can perform the closing and opening analog manual switch operation according to the program arrangement.
Realizing the voice control of automotive electrical appliances is the development trend of automotive electrical appliances in the future, and more and more solutions are being continuously proposed and verified. The design proposed in this paper is to use the SDA80D51 chip on the SL1102C1 car audio system to realize the voice recognition and control of the car audio. Due to the high integration of the chip, fewer peripheral modules are required, and the designed hardware circuit is simple, which is convenient for debugging and detection. The prototype obtained by this design has high recognition rate, stable operation and strong scalability, and achieves the expected design goals. The whole design scheme and implementation method are feasible. Since the speech recognition rate varies with the environment and the speaker, although the HMM can obtain a high recognition rate in a low-noise environment, when the test speech or the environment contains different degrees of noise pollution, the speech recognition system Performance will drop quickly. Improving the noise immunity and robustness of the system is one of the keys to the practical application of speech recognition systems.
MLF only advise high quality Car Charger which mainly use for smart phones and other small mobile device, it has many optionals as single Usb Car Charger, dual USB car charger and multi-USB ports car charger , There are several housing shapes for selection, also with ability to make custom housing shape and color.
Diversified services to meet your needs!
Car Charger
Car Charger,Car Phone Adapter,Usb Car Charger,Car Phone Charger
Meile Group Limited , http://www.hkmeile.com