阅读说明：本技术 基于ai智能语音车载氛围灯控制系统及方法 (AI-based intelligent voice vehicle-mounted atmosphere lamp control system and method ) 是由 李磊 于 2021-07-08 设计创作，主要内容包括：本发明公开了基于AI智能语音车载氛围灯控制系统及方法,控制系统包括语音采集模块、深度学习模块、灯控制模块、故障诊断模块,该控制方法包括,步骤S1、采集语音数据并输出语音指令；步骤S2、判断识别语音是否为有效语音；步骤S3、判断灯控制模块是否接收传递的汽车行驶故障信息。本发明对预设语音进行训练得到声学模型,移植该模型到嵌入式平台中,利用声学模型对在线语音的音频特征进行分类判别,具有响应时间短、设别准确率高的特点。(The invention discloses an AI-based intelligent voice vehicle-mounted atmosphere lamp control system and a method, wherein the control system comprises a voice acquisition module, a deep learning module, a lamp control module and a fault diagnosis module, and the control method comprises the following steps of S1, acquiring voice data and outputting a voice instruction; step S2, judging whether the recognized voice is effective voice; and step S3, judging whether the lamp control module receives the transmitted automobile running fault information. The method trains the preset voice to obtain the acoustic model, transplants the acoustic model into the embedded platform, and utilizes the acoustic model to classify and judge the audio frequency characteristics of the online voice, and has the characteristics of short response time and high identification accuracy.)

1. AI-based intelligent voice vehicle-mounted atmosphere lamp control system is characterized by comprising,

the voice acquisition module is used for acquiring voice data and outputting a voice instruction;

the voice processing module is used for processing the voice instruction and identifying whether the voice is effective voice;

the deep learning module is used for constructing an acoustic model and transplanting the trained acoustic model to the voice processing module, and the voice processing module is in communication connection with the lamp control module;

the fault diagnosis module is used for acquiring operation data, extracting diagnosis characteristic data and judging whether a fault occurs or not;

and the lamp control module is used for adjusting the light effect of the atmosphere lamp according to the fault information.

2. The AI-based intelligent voice vehicle-mounted atmosphere lamp control system according to claim 1, further comprising an LED driving module and an LED lamp strip, wherein the lamp control module controls the lighting effect of the LED lamp strip through the LED driving module according to a user voice instruction, and the fault information sent by the fault diagnosis module is transmitted to the lamp control module to adjust the lighting effect of the LED lamp strip.

3. The AI-based intelligent voice vehicle atmosphere lamp control system of claim 1, wherein the voice processing module employs an i.MX6 voice processor having a trained acoustic model embedded therein.

4. The AI-based intelligent voice vehicle-mounted atmosphere lamp control system according to claim 1, further comprising a communication transceiver module, a vehicle control module, and an instrument display module, wherein the communication transceiver module is respectively connected with the vehicle control module, the fault diagnosis module, and the instrument display module through CAN buses, and the communication transceiver module is in communication connection with the lamp control module.

5. An AI-based intelligent voice vehicle-mounted atmosphere lamp control method is characterized by comprising the following steps of,

step S1, collecting voice data and outputting a voice instruction;

step S2, inputting a voice command into a pre-trained acoustic model for training to judge whether the recognized voice is effective voice, if the recognized voice is effective voice, outputting a recognition signal, otherwise, returning to the step S1 to re-collect voice information;

step S3, providing a lamp control module, and judging whether the lamp control module receives the transmitted automobile running fault information;

if yes, the lamp control module adjusts the light effect of the atmosphere lamp according to the fault signal;

otherwise, the lamp control module controls the light effect of the atmosphere lamp according to the identification signal transmitted in the step S2.

6. The AI-based intelligent voice vehicular atmosphere lamp control method of claim 5, wherein in step S2, the specific training process of the acoustic model comprises:

step S21, providing a voice processing module, and building a CNN neural network framework in the voice processing module;

step S22, obtaining a preset user voice instruction, and performing acoustic model training;

step S23, inputting parameters obtained by acoustic model training into the CNN neural network framework constructed in step S21.

7. The AI-based intelligent voice vehicle-mounted ambience lamp control method according to claim 5, wherein the process of the lamp control module receiving the transmitted vehicle driving fault information in the step S3 includes:

step S31, the CAN bus obtains the information output by the fault diagnosis module;

step S32, the communication transceiver module waits for information to be received;

and step S33, after the communication transceiver module receives the information, judging whether the automobile breaks down or runs at an overspeed in the running process, if so, processing the fault information and sending the information to the lamp control module, otherwise, returning to the step S32 to wait for receiving the information.

Technical Field

The invention relates to the technical field of automobiles, in particular to an AI-based intelligent voice vehicle-mounted atmosphere lamp control system and method.

Background

The regulation of traditional on-vehicle atmosphere lamp is realized through button or vehicle navigation interface, and when normally driving, driver's manual operation atmosphere lamp can shift the sight, has traffic accident hidden danger. At present, a voice interaction technology is popularized and used in a vehicle, and a voice recognition technology is applied to vehicle-mounted equipment, so that the use value of a product can be improved, and a user interface is more friendly. Traditional speech recognition algorithms typically employ dynamic time warping and hidden markov model based techniques. However, the DTW algorithm has a problem that it cannot be applied to unspecified persons and a large number of recognized words; the HMM algorithm has the problems of low response speed, low recognition rate and the like. Therefore, a speech recognition algorithm based on deep learning has been proposed. The existing method for realizing voice recognition of vehicle-mounted intelligent products comprises a voice library adopting a special voice recognition chip, a traditional voice recognition algorithm and the like.

Therefore, an AI-based intelligent voice vehicle-mounted atmosphere lamp control system and method are provided to solve the problems.

Disclosure of Invention

The invention aims to solve the problems of low algorithm response speed and low recognition rate in the prior art, and provides a vehicle-mounted atmosphere lamp control system and method based on AI intelligent voice.

In order to achieve the purpose, the invention adopts the following technical scheme:

AI-based intelligent voice vehicle-mounted atmosphere lamp control system, which comprises,

the voice acquisition module is used for acquiring voice data and outputting a voice instruction;

the voice processing module is used for processing the voice instruction and identifying whether the voice is effective voice;

the deep learning module is used for constructing an acoustic model and transplanting the trained acoustic model to the voice processing module, and the voice processing module is in communication connection with the lamp control module;

the fault diagnosis module is used for acquiring operation data, extracting diagnosis characteristic data and judging whether a fault occurs or not;

and the lamp control module is used for adjusting the light effect of the atmosphere lamp according to the fault information.

The control system further comprises an LED driving module and an LED lamp strip, the lamp control module controls the light effect of the LED lamp strip through the LED driving module according to a user voice instruction, and the fault information sent by the fault diagnosis module is transmitted to the lamp control module to adjust the light effect of the LED lamp strip.

The voice processing module adopts an i.MX6 voice processor, and a trained acoustic model is embedded in the i.MX6 voice processor.

The control system further comprises a communication transceiver module, a whole vehicle control module and an instrument display module, wherein the communication transceiver module is respectively connected with the whole vehicle control module, the fault diagnosis module and the instrument display module through a CAN bus, and the communication transceiver module is in communication connection with the lamp control module.

In order to achieve the purpose, the invention also adopts the following technical scheme:

an AI-based intelligent voice vehicle-mounted atmosphere lamp control method comprises the following steps,

step S1, collecting voice data and outputting a voice instruction;

step S2, inputting a voice command into a pre-trained acoustic model for training to judge whether the recognized voice is effective voice, if the recognized voice is effective voice, outputting a recognition signal, otherwise, returning to the step S1 to re-collect voice information;

step S3, providing a lamp control module, and judging whether the lamp control module receives the transmitted automobile running fault information;

if yes, the lamp control module adjusts the light effect of the atmosphere lamp according to the fault signal;

otherwise, the lamp control module controls the light effect of the atmosphere lamp according to the identification signal transmitted in the step S2.

In step S2, the specific training process of the acoustic model includes:

step S21, providing a voice processing module, and building a CNN neural network framework in the voice processing module;

step S22, obtaining a preset user voice instruction, and performing acoustic model training;

step S23, inputting parameters obtained by acoustic model training into the CNN neural network framework constructed in step S21.

In step S3, the process of the lamp control module receiving the transmitted vehicle driving failure information includes:

step S31, the CAN bus obtains the information output by the fault diagnosis module;

step S32, the communication transceiver module waits for information to be received;

and step S33, after the communication transceiver module receives the information, judging whether the automobile breaks down or runs at an overspeed in the running process, if so, processing the fault information and sending the information to the lamp control module, otherwise, returning to the step S32 to wait for receiving the information.

Compared with the prior art, the invention has the beneficial effects that: the method comprises the steps of training preset voice to obtain an acoustic model, transplanting the acoustic model to an embedded platform, classifying and judging audio features of online voice by using the acoustic model, carrying out online test 20 times on each recognition word, and calculating average response time and average accuracy.

Drawings

FIG. 1 is a general framework diagram of an AI-based intelligent voice vehicular atmosphere lamp control system according to the present invention;

fig. 2 is a flowchart of the AI-based intelligent voice vehicle-mounted ambience lamp control method provided by the present invention.

In the figure: the system comprises a voice acquisition module 1, a voice processing module 2, a deep learning module 3, a lamp control module 4, an LED driving module 5, a communication transceiver module 6, a vehicle control module 7, an instrument display module 8 and a fault diagnosis module 9.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments.

Referring to fig. 1-2, the invention relates to an intelligent voice vehicle-mounted atmosphere lamp control system based on an AI, which comprises a voice acquisition module 1, a voice processing module 2, a deep learning module 3, a lamp control module 4, an LED driving module 5, a communication transceiver module 6, a vehicle control module 7, an instrument display module 8 and a fault diagnosis module 9.

The voice acquisition module 1 is used for acquiring voice data and outputting a voice instruction; the voice acquisition module 1 adopts a microphone, the microphone adopts a USB interface, and the microphone has the functions of intelligent noise reduction, noise resistance, interference resistance and the like.

The voice acquisition module 1 is connected with a voice processing module 2, the voice processing module 2 adopts an i.MX6 voice processor, and the processor is widely applied to the fields of industrial automatic control equipment, human-computer interaction equipment, avionic equipment, robots and the like and has good performance; the voice processing module 2 is used for processing voice data and identifying whether the voice is valid voice, if the voice processing module 2 identifies invalid voice, the voice acquisition module 1 performs voice acquisition again, and if the voice processing module 2 identifies valid voice, an identification signal is output;

the deep learning module 3 is connected with the voice processing module 2, the voice processing module 2 is connected with the lamp control module 4, the communication transceiver module 6 is respectively connected with the vehicle control module 7, the fault diagnosis module 9 and the instrument display module 8 through a CAN bus 100, the CAN bus comprises a CAN-H line and a CAN-L line, and the communication transceiver module 6 is connected with the CAN bus through an OBD-II interface. During operation, the fault diagnosis module 9 and the instrument display module 8 transmit CAN bus data to the communication transceiver module 6 through the OBD-II interface, and the communication transceiver module 6 is in communication connection with the lamp control module. The fault diagnosis module 9 is used for acquiring the operation data, extracting the diagnosis characteristic data and judging whether a fault occurs.

The deep learning module 3 is used for constructing an acoustic model and transplanting the trained acoustic model to an i.MX6 voice processor, and the i.MX6 voice processor is communicated with the lamp control module 4; and then the automobile fault code and the automobile driving record are transmitted to the communication transceiver module 6 through the OBD-II interface so as to be communicated with the lamp control module 4.

Specifically, a CNN neural network framework is built in the i.MX6, acoustic model parameters are input into the framework, an acoustic model based on an embedded platform is realized, and the transplantation of the acoustic model is completed.

The lamp control module 4 is connected with an LED driving module 5, the LED driving module 5 is connected with an LED lamp strip, and the LED driving module 5 is used for driving a switch of the LED lamp strip 10. The lamp control module 4 controls the light effect of the LED lamp strip through the LED driving module according to the user instruction information, and when the fault information sent by the fault diagnosis module 9 is transmitted to the lamp control module 4, the light of the LED lamp strip 10 is adjusted correspondingly. The lamp control module 4 adopts S32K144, wherein the S32K144 is a 32-bit ARM processor applied to the automobile industry, and is suitable for general automobiles and high-reliability industrial application.

The LED lamp area is RGB atmosphere lamp, and the LED lamp area has more than one kind of color and luminance, can adjust multiple color like this, improves the practicality.

The control method of the vehicle-mounted atmosphere lamp based on the AI intelligent voice comprises the following steps:

step S1, the voice acquisition module 1 receives the user voice data and outputs a voice instruction;

step S2, inputting a voice command into a pre-trained acoustic model for training to judge whether the recognized voice is effective voice, if the recognized voice is effective voice, outputting a recognition signal, otherwise, returning to the step S1 to re-collect voice information;

in step S2, the specific process of the speech processing module 2 obtaining the pre-trained acoustic model includes:

step S21, building a CNN neural network framework in the voice processing module 2;

step S22, obtaining a preset user voice instruction, and performing acoustic model training;

in step S22, four words "up", "down", "left", and "right" in the user instruction are selected, each recognized word is tested 20 times on line, the average response time and average accuracy are calculated, and the conventional speech recognition method is used to compare with the deep learning method used herein, which shortens the response time by 52.3% on average and improves the recognition accuracy by 7.4% compared with the HMM method. When the recognition word "up" is accurately recognized, the atmosphere lamp presents a multi-color lamp light flow change.

In step S23, parameters obtained by acoustic model training are input into the framework.

Step S3, judging whether the lamp control module receives the transmitted automobile running fault information;

if yes, the lamp control module adjusts the light effect of the atmosphere lamp according to the fault signal;

otherwise, the lamp control module controls the light effect of the atmosphere lamp according to the identification signal transmitted in the step S2.

In the present invention, the process of the lamp control module receiving the transmitted vehicle driving fault information in step S3 includes:

in step S31, the CAN bus 100 obtains the information output by the fault diagnosis module 9,

step S32, the communication transceiver module 6 waits for reception of information;

step S33, after the communication transceiver module 6 receives the information, it determines whether the vehicle is in failure or speeding, if so, the failure information is processed and the information is sent to the lamp control module 4, otherwise, the process returns to step S32 to wait for receiving the information.

The invention designs, develops and provides a system and a control method for realizing voice recognition and controlling an atmosphere lamp based on deep learning, which train preset voice to obtain an acoustic model, transplant the acoustic model into an embedded platform, classify and judge the audio frequency characteristics of online voice by using the acoustic model, use an LED as a light source, carry out key technology research on LED atmosphere lamp light distribution and control systems for automotive interiors according to diversified use requirements of household and commercial automobiles, optimize a light path structure and a multi-mode atmosphere lamp control module of the automobile atmosphere lamp, develop an LED atmosphere lamp based on deep learning and having a multi-level sense organ experience function, and meet the use requirements of intelligentization, energy conservation and compact structure of control of the automobile atmosphere lamp.

The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.