基於 Esp8266 物聯網平台和深度學習技術的智慧輪胎檢測系統設計與實現

Design and Implementation of Intelligent Tire Detection System Based on Esp8266 IoT Platform and Deep Learning Technology

傳統上，市售的胎壓偵測系統(TPMS)僅能透過測量胎壓、胎溫來判斷輪胎狀態，無法即時檢測如：輪框變形、胎面異常剝離、輪胎附著異物、異常抖動及打滑等的輪胎異常狀況，若駕駛有所疏忽，將造成嚴重的危害。因此，我們研究設計一套智慧輪胎系統，透過量測輪胎運行間的慣性感測資訊，包含：三軸加速度、角速度、絕對方向，再結合 AI 演算法之辨識技術，動態檢測出輪胎異常狀況，並實作設計下述物聯網平台元件 : 1.安裝於輪框上之硬體裝置，可偵測輪胎行進間之異常數值，2.具即時監測與異常提醒、維修保養建議、車友討論平台等功能之手機應用程式，3.基於深度學習神經網路技術辨識各式輪胎異常狀況，4.建立跨平台智慧輪胎網頁介面供職業車隊管理停車場。經實驗測試結果，本系統的 AI 判讀準確率高達 99.17%。

 

The tire pressure monitoring system (TPMS) sold on the market can only judge the tire status by measuring the tire pressure and tire temperature, and cannot detect such things as: wheel frame deformation, abnormal tread peeling, foreign objects attached to the tire, abnormal vibration and skidding, etc. If the tire is abnormal, if the driver is negligent, it will cause serious harm. Therefore, we design a set of smart tire system, through the measurement of inertial sensing information during tire operation, including: three-axis acceleration, angular velocity, absolute direction, combined with the identification technology of AI algorithm, dynamically detect abnormal tire conditions, and Realize the design of the following IoT platform components: 1. The hardware device installed on the wheel frame can detect abnormal values during tire travel. 2. It has real-time monitoring and abnormal reminders, maintenance suggestions, and a discussion platform for riders, etc. Functional mobile phone application, 3. Based on deep learning neural network technology to identify various tire abnormalities, 4. Establish a cross-platform smart tire web interface for professional fleets to manage parking lots. According to the experimental test results, the AI interpretation accuracy rate of this system is as high as 99.17%.