臺灣博碩士論文加值系統

現今資訊和科技的進步將工廠推向了全自動化的時代，工廠中的AGV車扮演著重要的角色，用於運輸和搬運重要的產品和設備，然而確保AGV車的正常運行和安全性需要準確監控其電流與電量情況，也隨著物聯網技術的不斷進步與普及，智能監控系統已被廣泛應用於各個領域，所以做好電量管理和監控是非常重要的一環。因此本研究透過小型電動智能車做為模擬以及配合影像辨識開發的一套電流監測系統，期望未來能將此系統套用在工廠中的AGV車上，以改善傳統昂貴的電流監測儀器設備。
本研究提出了一項研究架構，使用了電流感測元件與單晶片智能車結合樹莓派和影像辨識的一項分析電流研究，透過串聯的方式連接電流感測元件與智能車的供電系統，並透過將智能車設定為前進、後退、停止三個步驟，並即時偵測流訊號，在藉由樹莓派的程式將資料匯入資料庫，並透過Python抓取資料並繪圖後使用電量計算公式進行比對分析，在透過建立類神經模型給予輸入層三種特徵電流狀態，分別為初始滿電狀態、中段電量損耗、末段電量不足的狀態，並透過卷積神經深度學習影像辨識結合系統進行比對分析，達到運行間即時監控電流和電池的耗電變化。

The advancement of information and technology has led to full automation in factories. AGV transport vehicles play a crucial role in handling essential products and equipment. Ensuring their smooth operation and safety requires accurate monitoring of their current conditions. With the growing popularity of IoT technology, intelligent monitoring systems are widely used, making power management and monitoring essential.

This research develops a current monitoring system using a small electric intelligent vehicle and image recognition. The goal is to apply this system to AGV transport vehicles in factories, improving traditional and costly current monitoring equipment.

The framework involves current sensing components, a single-chip intelligent vehicle, Raspberry Pi, and image recognition. Real-time current detection is achieved during vehicle actions (forward, backward, stop). Data is processed, and a neural network model combined with CNN and image recognition analyzes the power consumption for real-time monitoring.

摘要......i
Abstract......ii
誌謝......iii
目錄......iv
表目錄......vi
圖目錄......vii
第一章 緒論......1
1.1 研究背景......1
1.2 研究動機與目的......1
1.3 研究流程......2
1.4 實驗設計......3
1.4.1 實驗設備......3
1.4.2 實驗步驟......3
第二章 文獻探討......4
2.1 卷積神經路(Convolutional neural network, CNN)......4
2.1.1 卷積層(Convolution Layer)......5
2.1.2 池化層(Pooling Layer)......5
2.1.3 全連接層(Fully Connected Layer)......6
2.2 AGG深度學習模型......6
2.3 ACS-712電流感測模組......7
2.4 TT四驅控制馬達智能車......8
2.4.1 智能車控制架構......8
2.5 電池續航公式......10
2.6 CNN負載電流監控......10
第三章 研究方法......11
3.1 研究架構......11
3.2 電流感測器連接供電系統......13
3.3 智能車設定......14
3.4 資料收取......14
3.5 資料集準備......15
3.6 CNN模型訓練......16
3.6.1. 資料集分類處裡......17
3.6.2. 資料集強化......19
3.6.3 參數設定......20
3.6.4 模型訓練......21
3.6.5 模型結果......22
3.7 電池續航分析......24
第四章 研究分析與成果......25
4.1 研究系統設備......25
4.1.1 模型訓練設備......25
4.1.2 模型訓練環境......26
4.2 研究分析......26
4.2.1 智能車運作時間分析......27
4.2.2 模型實驗測試分析......28
4.2.3 模型準確度分析......28
4.3 電流即時監控系統......29
第五章 結論與未來展望......30
5.1 結論......30
5.2 未來展望......30
參考文獻......31
附錄一......33
Extended Abstract......44

[1] S. Li, Y. Wu, C. Zhang, and L. Wang. (2022). A Review of Autonomous Navigation and Path Planning Methods for AGV Tugger Vehicles in Industrial Environments. Journal of Robotics and Automation, 28(3), 315-330.
[2 ]H. Zhang, G. Chen, J. Wang, and M. Liu. (2023). Real-time Electric Current Monitoring System for AGV Tugger Vehicles Based on Image Recognition. Proceedings of the International Conference on Industrial Automation, 189-196.
[3] LeCun Y, Bengio Y, Hinton G. Deep learning. Nature. 2015 May 28;521(7553):436-44.
[4] Allegro MicroSystems, LLC. (2009). ACS712 Datasheet (Rev. 6). https://www.allegromicro.com/-/media/files/datasheets/acs712-datasheet.ashx
[5] T.K. Hareendran. (2017). ACS712 Current Sensor Module Pinout, Specifications, Circuit & Datasheet. CircuitDigest. https://circuitdigest.com/microcontroller-projects/acs712-current-sensor-module
[6] Arduino. (n.d.). ACS712 Current Sensor. https://www.arduino.cc/en/reference/ACS712
[7] 王智民, 高煒. 電動汽車用鋰電池的細微特性研究[J]. 中國汽車工程, 2011, 5(1): 11-15.
[8] Bhangu B S, Singh B, Singh J. Performance Analysis of Lithium-Ion Battery using MATLAB/Simulink[J]. Journal of Electrical and Electronics Engineering Research, 2017, 9(2): 11-18.
[9] Zheng X, Lu J, Huang X. A review on the key issues for lithium-ion battery management in electric vehicles[J]. Journal of Power Sources, 2014, 267: 469-485.
[11] 姚超, 朱峰, 陳威. 基於卷積神經網絡的變壓器內部故障識別[J]. 2018.
[10] 鄭華, 劉中民, 張偉. 基於電流訊號的變頻器內部故障監測研究[J]. 電力自動化設備, 2016.
[12] 鄭華, 劉中民, 張偉. 基於多CNN模型的變頻器內部故障監測[J]. 機械設計與制造, 2017.
[13] 陳彥誠, “設備監測及檢測之異常訊號辨識分類”。碩士論文，長庚大學機械工程學系，2021。
[14] T. Li, H. Zhang, Y. Li, C. Chen, and L. Li, "A Deep Convolutional Neural Network for Fault Diagnosis of Power Systems," IEEE Transactions on Industrial Electronics, vol. 65, no. 11, pp. 8797-8806, Nov. 2018.
[15] L. Huang, Y. Li, J. Sun, and Z. Ding, "Power Equipment Condition Monitoring and Warning System Based on Convolutional Neural Network," Journal of Electrical Engineering & Technology, vol. 14, no. 1, pp. 1-8, Jan. 2019.

[16] J. Zhang, M. Ma, H. Li, and J. Li, "Power Quality Disturbance Recognition Based on Convolutional Neural Network," in Proceedings of the 2019 IEEE International Conference on Power, Energy and Electrical Engineering (PEEE), Apr. 2019, pp. 195-199.
[17] Y. Wang, J. Wu, and Y. Lu, "Fault Diagnosis of Power Systems Based on Convolutional Neural Network with Dropout," in Proceedings of the 2020 IEEE International Conference on Artificial Intelligence and Power Engineering (AIPE), Nov. 2020, pp. 60-65.
[18] X. Chen, L. Wang, and J. Wang, "Convolutional Neural Network for Fault Diagnosis of Power Transformers Based on Differential Evolution Algorithm," in Proceedings of the 2020 IEEE International Conference on Power Electronics, Smart Grid and Renewable Energy (PESGRE), Oct. 2020, pp. 311-315.
[19] https://ithelp.ithome.com.tw/articles/10279717
[20]https://www.cupoy.com/qa/club/ai_tw/0000016D6BA22D97000000016375706F795F72656C656173654B5741535354434C5542/0000017B4407944C0000002E6375706F795F72656C656173655155455354