# ind-dist-ai-sec-edge-cloud

**Repository Path**: jtyjtyjty333/ind-dist-ai-sec-edge-cloud

## Basic Information

- **Project Name**: ind-dist-ai-sec-edge-cloud
- **Description**: 将“轻量级ai边缘计算“，“OTA远程升级”，“互联网云端交互”融为一体的嵌入式项目。An embedded project integrating lightweight AI edge computing, OTA remote upgrade, and Internet cloud interaction.
- **Primary Language**: C
- **License**: Not specified
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 2
- **Forks**: 0
- **Created**: 2025-12-04
- **Last Updated**: 2025-12-04

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

### 项目名称:
工业级多节点分布式安全监测与AI边缘-云端协同系统
##
### 项目简介:
多节点分布式系统架构，子节点本地进行多个传感器数据的AI边缘计算（如果超过安全预警则立刻执行保护措施）。而后通过CAN总线将数据发送到主节点，主节点通过MQTT上传至云端（如果检测到无法连接云端，则把数据储存在SD卡中，待成功连上云端后通过HTPP再发送数据）。子节点的tensorflow AI检测模型可以通过OTA更新固件。子节点1使用ESP32-S3＋CAN收发器，基于FreeRTOS，搭载AI边缘检测，拥有温湿度传感器DHT11，振动传感器MPU6050，气体传感器MQ-9，通过本地wifi升级固件。子节点2使用STM32F407VET6+CAN收发器+esp32wifi模块，基于FreeRTOS，搭载AI边缘检测，拥有光电传感器E3F-DS30C4，振动传感器，通过本地wifi升级固件。主节点使用STM32F407ZGT6+ESP8266wifi模块+CAN收发器+OLED显示屏+SD卡，基于RTthread，MQTT协议，HTTP协议。wifi接受OTA固件升级。
##
### 你将学习到：
###### 1.如何训练开源的ai机器学习模型（TensorFlowLite）并部署到嵌入式设备中
###### 2.实现简单的OTA远程升级
###### 3.自己实现搭建CAN通信总线
###### 4.了解多种常见传感器的使用和应用层调用逻辑
###### 5.熟悉目前常见的rtos的使用（freertos,rtthread）
###### 6.了解MQTT,HTPP等互联网通信协议
##
### 文件说明:
### mian
    |—— "Complete code" —— 文件夹里附有完整工程的源码。
    |—— "Main Code Analysis" —— 主要代码文件
        |—— "CAN"
            |—— "ESP32_can"
            |—— "HAL_can"
            |—— "RTthread_can"
        |—— "MQTT"
        |—— "OTA"
            |—— "ESP32_ota"
            |—— "STM32_ota"
        |—— "SD"
        |—— "TensorFlow"
            |—— "ESP32_TF"
            |—— "HAL_TF"
        |—— "WiFi"
            |—— "ESP32_Wifi"
            |—— "HAL_Wifi"
            |—— "RTthread_Wifi"
##
### 相关网址:
##### 巴法云（远端固件传输） https://bemfa.com/
##### TensorFlow（机器学习源码） https://tensorflow.google.cn/?hl=zh-cn
##### ESP-IDF（乐鑫IDE官网） https://docs.espressif.com/projects/esp-idf/zh_CN/stable/esp32/get-started/index.html
##### Platformio（本项目ESP32-S3的编译环境） https://platformio.org/
##### RT-thread（本项目stm32f407vgt6的编译环境） https://www.rt-thread.org/
##### STM32CubeMX - Keil（本项目stm32f407vet6的编译环境） https://www.st.com/en/development-tools/stm32cubemx.html - https://www.keil.com/
##
### 注意事项:
   ##### 1.由于stm32的官方没有开源X-CUBE-AI的源码，因此stm32f407vet6的代码只能在keil里编译
   ##### 2.目前只有在arduino环境才能使用TensorFlow，推断是ESP-IDF的开发环境还没配置好 

##
### 项目实现细节:
本项目使用外接CAN收发器将不同的设备进行挂载CAN总线;两个esp作为stm32的wifi连接设备,通过AT指令实现通信;涉及Freertos和RT-thread两款rtos;".tflite"文件通过python的开源tensorflow库训练,esp32的ai边缘计算调用platformio中的"TensorFlowLite_ESP32"库通过python将".tflite"文件转换为C数组,stm32的ai边缘计算使用STM32CubeMX的X-CUBE-AI组件解析".tflite";stm32f407vgt6的SD,MQTT,AT通过RTThread的插件生成并调用;
##
## !!!详细的视频，图片等，请持续关注后续的更新.....
### github: https://github.com/jtylab/ind-dist-ai-sec-edge-cloud
##
### 1. Project Name:
Industrial-grade Multi-node Distributed Security Monitoring and AI Edge-Cloud Collaboration System
##
### 2. Project Overview:
Adopting a multi-node distributed system architecture, slave nodes perform AI edge computing on data from multiple sensors locally (protective measures are executed immediately if the safety warning threshold is exceeded). Data is then transmitted to the master node via the CAN bus, and the master node uploads it to the cloud through MQTT (if cloud connection failure is detected, data is stored in an SD card and retransmitted via HTTP once the cloud connection is restored). The TensorFlow AI detection models on slave nodes support firmware updates via OTA (Over-the-Air).
Slave Node 1: Based on FreeRTOS, uses ESP32-S3 + CAN transceiver, integrates AI edge detection, and is equipped with DHT11 temperature and humidity sensor, MPU6050 vibration sensor, and MQ-9 gas sensor; firmware upgrade is available via local WiFi.
Slave Node 2: Based on FreeRTOS, uses STM32F407VET6 + CAN transceiver + ESP32 WiFi module, integrates AI edge detection, and is equipped with E3F-DS30C4 photoelectric sensor and vibration sensor; firmware upgrade is available via local WiFi.
Master Node: Based on RT-Thread, uses STM32F407ZGT6 + ESP8266 WiFi module + CAN transceiver + OLED display + SD card, supports MQTT and HTTP protocols, and accepts OTA firmware upgrades via WiFi.
##
### 3. File Description:
### mian
    |—— "Complete code" —— This folder contains the source code of the entire project. 
    |—— "Main Code Analysis" —— ore code files.
        |—— "CAN"
            |—— "ESP32_can"
            |—— "HAL_can"
            |—— "RTthread_can"
        |—— "MQTT"
        |—— "OTA"
            |—— "ESP32_ota"
            |—— "STM32_ota"
        |—— "SD"
        |—— "TensorFlow"
            |—— "ESP32_TF"
            |—— "HAL_TF"
        |—— "WiFi"
            |—— "ESP32_Wifi"
            |—— "HAL_Wifi"
            |—— "RTthread_Wifi"
##
### 4. Related URLs:
##### Bemfa Cloud (Remote Firmware Transmission): https://bemfa.com/
##### TensorFlow (Machine Learning Source Code): https://tensorflow.google.cn/?hl=zh-cn
##### ESP-IDF (Espressif IDE Official Website): https://docs.espressif.com/projects/esp-idf/zh_CN/stable/esp32/get-started/index.html
##### PlatformIO (Compilation Environment for ESP32-S3 in this project): https://platformio.org/
##### RT-Thread (Compilation Environment for STM32F407ZGT6 in this project): https://www.rt-thread.org/
##### STM32CubeMX - Keil (Compilation Environment for STM32F407VET6 in this project): https://www.st.com/en/development-tools/stm32cubemx.html - https://www.keil.com/
##
### 5. Notes:
##### 1.Since STMicroelectronics has not open-sourced the source code of X-CUBE-AI, the code for STM32F407VET6 can only be compiled in Keil.
##### 2.Currently, TensorFlow can only be used in the Arduino environment; it is inferred that the ESP-IDF development environment has not been properly configured.
##
### 6. Project Implementation Details:
External CAN transceivers are used to mount different devices to the CAN bus in this project.
Two ESP modules act as WiFi communication devices for STM32, realizing communication via AT commands.
The project involves two RTOSes: FreeRTOS and RT-Thread.
.tflite files are trained using the open-source TensorFlow library for Python; AI edge computing on ESP32 calls the "TensorFlowLite_ESP32" library in PlatformIO to convert .tflite files into C arrays (via Python).
AI edge computing on STM32 uses the X-CUBE-AI component of STM32CubeMX to parse .tflite files.
SD card, MQTT, and AT command functionalities for STM32F407ZGT6 are generated and invoked via RT-Thread plug-ins.
##
### 7. Additional Note:
!!! For detailed videos, pictures, and other materials, please stay tuned for subsequent updates.....