Nowadays, more and more Electrical appliances are used in cars, and more and more electrical switches that drivers need to manually operate, which not only increases the burden on the driver, but also affects driving safety. This paper takes STM32F103VET6 (hereinafter referred to as STM32) chip as the control core, and adopts high-performance LD3320 speech recognition chip to design a car air conditioning control system based on speech recognition. The system can effectively control the car air conditioner with voice, which reduces the driver's operation burden and ensures the safety during the driving process.
1 system hardware design
Using speech recognition technology, combined with various sensors to detect various parameters such as the temperature inside and outside the body (such as temperature, sunlight intensity, etc.) and the state of the refrigeration compressor, and compare with the set parameters, the microcontroller is processed through arithmetic processing. Judging, the corresponding adjustment and control signals are output. The actuator is adjusted and corrected in real time to achieve full-scale and multi-functional adjustment and control of the air environment inside the cabin. The actuators of the system mainly include temperature damper motor, mode damper motor, circulating damper motor, blower, compressor, defrosting control relay and so on. Figure 1 is a block diagram of the system structure.
1.1 main controller
The main controller is a 32-bit microcontroller based on ARM Cortex-M3 core STM32F103VET6, built-in 64 KB RAM, 512 KBFlash, and a wealth of enhanced I / O ports and peripherals connected to two APB buses, mainly controlling sensor analog signals The acquisition, the transmission and reception of voice signals and the output of control signals for automotive air conditioners.
1.2 Speech Recognition Module
The speech recognition chip uses ICRoute's LD3320 chip. The chip integrates a speech recognition processor and some external circuits, including an A/D converter, a D/A converter, a microphone interface, a sound output interface, and the like. The chip is designed to focus on energy saving and high efficiency. It does not require any external auxiliary chip (such as Flash, RAM, etc.), and can be directly integrated into existing products to realize speech recognition. The list of recognized keyword words can be edited arbitrarily.
Referring to the LD3320 data sheet, the speech recognition control circuit is connected by SPI serial mode using LD3320 and STM32. The speech recognition module control circuit is shown in Figure 2. First, to connect MD to a high level, the chip clock signal CLK is connected to the STM32 clock signal output pin MCO (PA8). The pin MBS is a microphone bias and is connected to an RC circuit to ensure that a floating voltage can be output to the microphone.
1.3 SD card storage module
To implement a voice recognition control system with human-computer interaction function, a large number of MP3 audio files need to be stored. In this system, MP3 audio playback files are stored on the SD card, and voice recognition keywords are also stored on the SD card, so that the keywords to be recognized can be easily changed without changing the program content. The master STM32 reads the MP3 data sequentially from the SD card and feeds it into the LD33 20 chip, so that the sound can be output from the corresponding pin of the chip. The SD card hardware connection is shown in Figure 3. The SD card communicates with the STM32 via SPI. The SD card selection signal CS, the data input signal DI, the data output signal DO, and the clock signal SCLK are respectively connected to the PC11, PD2, PC8, and PC12 pins of the STM32.
FGI SCIENCE AND TECHNOLOGY CO., LTD , https://www.fgi-tech.com