(44.192.112.123) 您好!臺灣時間:2021/03/06 07:43
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

我願授權國圖
: 
twitterline
研究生:陳世瑋
研究生(外文):CHEN, SHIH-WEI
論文名稱:運用AI精確演算法於智慧製造設備系統之訊號傳輸分析
論文名稱(外文):The Signal Transmission Analysis of the Intelligent Manufacturing Equipment Using AI Precise Computing
指導教授:陳建璋陳建璋引用關係
指導教授(外文):CHEN, CHIEN-CHANG
口試委員:陳建璋陸瑞漢邱宗文
口試委員(外文):CHEN, CHIEN-CHANGLU, JUI-HANCHIOU, TZUNG-WERN
口試日期:2020-07-16
學位類別:碩士
校院名稱:國立虎尾科技大學
系所名稱:自動化工程系碩士班
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2020
畢業學年度:108
語文別:中文
論文頁數:61
中文關鍵詞:第五世代通訊智慧製造天線選擇多輸入多輸出
外文關鍵詞:Fifth Generation CommunicationIntelligent ManufacturingAntenna SelectionMulti-Input Multi-Output
相關次數:
  • 被引用被引用:0
  • 點閱點閱:103
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
下一個通訊世代即將到來,將通訊效能提高到比現有傳輸的方式具有更多傳輸資料、更大頻寬、更低的延遲。然而,智慧製造也隨著機台上感測器的數量變多以及訊號傳遞的速度變快,而會經由5G無線發射/接收設備將機台上的感測器所收集即時數據進行大數據分析判斷,達到工業機械預知保養、智慧監控等重要功效;在現行的5G通訊設備應用於智慧製造產線轉型與攜帶型智慧裝置的過程中,將面臨到以下兩個問題是需要解決的。第一個問題是在智慧製造中,當5G高頻天線隨著應用頻段拉高,波長變小,導致能量傳輸隨距離增加而衰減越多,不僅會讓訊號容易受到干擾,更會因為工業廠房中所配置機台的排列方式不同,致使每一個接收訊號的位置皆會因為場地限制而擺放數量、位置不同,導致接收訊號會傳輸不穩定而造成訊號中斷;第二個問題是攜帶型裝置的耗電量問題,因為目前尚未全面施行5G的核心網,大部分還是使用非獨立組網(Non-Standalone , NSA)的模式進行佈線,所以攜帶型裝置中還是會置入4G與5G的天線模組進行收發,導致在多天線模組的操作下,造成攜帶型裝置耗電量大幅增加等問題。有鑑於此,本研究進行智慧製造機台發射/接收訊號與攜帶型裝置多天線選擇的人工智慧演算法開發與驗證。
在智慧製造中,本研究在5.5公尺*3.2公尺的場域中模擬智慧製造設備傳輸無線高頻訊號的情形,藉由分別於四個相距發射基站3公尺的不同實驗機台旁架設高頻接收天線,並即時分析獨立天線的4個特徵參數,開發一套演算法,參數分別為:路徑衰減後的接收能量、誤差向量幅度、訊雜比與通道容量,經綜合判斷後得出每款天線的表現效能。若有接收天線受環境干擾影響較大,導致傳輸量為最低時,所計算出的效能表現值會最低,此時系統會建議開發者調整系統所輸出的編號。為了能夠節省高頻電路開發的時間與製作成本,就能夠接收到穩定且可靠的無線傳輸訊號做資料分析,本實驗使用軟體無線電定義(Software Defined Radio,SDR)機台來建構多輸入多輸出的無線通訊傳輸環境,研究過程僅需通過介面軟體就可以輕鬆且靈活的調整本實驗所需要設定的輸入功率及頻率,無須通過實做複雜的實體電路即可完成多通道的訊號分析。
在攜帶型多天線裝置中,由於未來的天線陣列會增加到現有的1倍以上,傳輸訊號的通道至少會由10款天線所組成的陣列做多輸入多輸出的資料傳輸,依照原有的切換技術做通道切換會造成最大的問題就是接收天線必須全程全功率輸出來偵測通道切換,這會造成耗電量增加以及縮短電池壽命,並且在使用的過程即使傳輸表現較弱的天線,其通道仍會是處於打開的狀態,造成人體不必要的電磁波輻射。因此本研究基於攜帶型裝置實驗在3種模擬使用情境:自由空間、金屬遮蔽、掌上使用,利用3500 MHZz無線射頻收發模組進行接收通道參數的資料收集,經資料處理後套入本研究提出建立在Python環境的快速演算法,綜合參考了於10秒鐘的短時間內使用天線通道參數的16(4天線各4個參數)筆資料中的通道參數,並將收集到的資料進行正規化至0到1間,並依據參數間的天線物理影響給出對應的權重,而對於不同天線所映射出的資料點離散程度同時劃分不同的範圍來做判斷。經實驗結果得知綜合參數評估演算法得出即時的陣列天線中的天線效能表現值,可以得知目前通訊傳輸表現效率最佳的2款天線,藉由使用即時辨識出的接收強度較高的天線,將通道表現較差的天線關閉,達到讓使用者於大多數的時間使用表現最佳的通道傳輸資料。最後,實驗驗證設計將3種不同環境的傳輸情形,確實切換至6種不同的通道組合作驗證,於100組實驗組中3個不同環境的準確率皆高達95%。最後,本研究所設計開發的精準快速演算法達到以下3個特點。特點一:多天線的智慧選擇,可以大幅度的降低手機收發訊號過程所產生的電磁波輻射,也可以減少攜帶型裝置的耗電量。特點二:使用軟體演算法,將可以減少使用射頻硬體開關的使用,達到減少元件成本的功效。特點三:所設計開發的演算法是透過多通道通訊系統進行資料分析,此演算法應用的終端裝置將不局限於攜帶裝置。

The next generation of communication is approaching. It’s going to offer better communication performance, larger data transmission, greater bandwidth, and lower latency than existing transmission methods. However, smart manufacturing also increases with the number of sensors on the machine and the speed of signal transmission. The real-time data collected by the sensors on the machine will be analyzed and judged through 5G wireless transmission/reception equipment to achieve important functions such as predictive maintenance and intelligent monitoring of industrial machinery. Thus, when the current 5G communication equipment is used in the transformation of smart manufacturing production lines and portable smart devices, the following two problems will need to be solved. The first problem is that in smart manufacturing, when the 5G high-frequency antenna is heightened with the application frequency band, the wavelength becomes shorter, resulting in more attenuation of energy transmission with increasing distance. The signal not only will be susceptible to interference, but also due to space limitations, the number and placement of the machines in the industrial plant are different. As a result, the reception of the signal at each position may become unstable and lead to signal interruption; the second problem is the power consumption of portable devices. Because the 5G core network has not yet been fully implemented and most still use the non-standalone network (NSA) mode for wiring, so, 4G and 5G antenna modules will still be placed in the portable devices for transmission and reception. As a result, under the operation of the multi-antenna module, the power consumption of the portable device is greatly increased. In view of this, this study developes and validates artificial intelligence algorithms for transmitting/receiving signals and multi-antenna selection for portable devices.
In smart manufacturing, this study simulates the transmission of wireless high-frequency signals by intelligent manufacturing equipment in a 5.5 m x 3.2 m field by setting up high-frequency receiving antennas next to four different experimental machines 3 meters away from the transmitting base station, and by analyzing the 4 characteristic parameters of the independent antennas in real time, a set of algorithms is developed. The parameters are received energy after path loss, error vector magnitude, signal-to-noise ratio, and channel capacity. The parameters are comprehensively judged to determine the performance of each antenna. If receiving antennas are greatly affected by environmental interference, resulting in the lowest transmission volume, the calculated performance value will be lowest, and the developer will be advised to adjust the number output from the system. To save high-frequency circuit development time, production cost, receive stablity and reliable wireless transmission signals for data analysis, this study uses Software Defined Radio (SDR) machine to construct a multi-input multi-output wireless communication transmission environment. The interface software allows flexibility and ease of adjustment during the study process to adjust the input power and frequency required by this lab. Multi-channel signal analysis can be completed without complicated physical circuits.
In portable multi-antenna devices, since the antenna array in the future will be more than doubled than the existing ones, the signal transmission channel will consist of an array of at least 10 antennas for multi-input multi-output data transmission. The biggest problem here is that the channel switching according to the original switching technology will cause the receiving antenna to operate under full range and full power for the channel switching detection. This will result in increased power consumption and reduced battery life. And in the process, even an antenna with weaker transmission, its channel will still be in an open state, causing unnecessary electromagnetic radiation to humans. So this study is based on the portable device experiment in 3 analog use scenarios: free space, metal shielding, and handheld. The 3500 MHz wireless radio frequency (RF) transceiver module is used to collect the data of the receiving channel parameters. After the data is processed, this study sets up a fast algorithm based on the Python environment and uses the antenna channel parameters of 16 (4 parameters for each of the 4 antennas) data in a short time of 10 seconds as a reference. The collected data is normalized from 0 to 1, and the corresponding weight is given according to the physical influence of the antenna between the parameters, and the dispersion of the data points mapped by different days is divided into different ranges for judgment. According to the experimental results, it is known that the comprehensive parameter evaluation algorithm can obtain the real-time performance value of the antenna in the array antenna. It can be seen that the two antennas with the best communication transmission performance efficiency can be used to turn off the antenna with the poorest channel performance by using the antenna with the real-time recognition of the higher reception strength to allow the user to use the best performing channel to transmit data most of the time. Finally, the experiment verifies that the transmission conditions of 3 different environments are indeed switched to 6 different channel groups for cooperative verification. The accuracy rates of 3 different environments in the 100 experimental groups were as high as 95%. Lastly, this study developes a quick and precise algorithm and achieved the following 3 features. Feature 1: The intelligent selection of multiple antennas can greatly reduce the electromagnetic radiation generated in the process of receiving and sending signals from mobile phones and can also reduce the power consumption of portable devices. Feature 2: The use of software algorithms will reduce the use of RF hardware switches and reduce component costs. Feature 3: The algorithm developed is designed for data analysis through a multichannel communication system. The application of this algorithm will not be limited to portable devices.

摘要.............................i
Abstract.......................iii
誌謝............................vi
目錄...........................vii
表目錄..........................ix
圖目錄...........................x
第一章 緒論.....................1
1.1 前言.....................1
1.2 國內外文獻探討............2
1.3 研究目的.................3
1.4 論文結構.................3
第二章 多通道通訊系統分析........5
2.1 多輸入多輸出系統介紹......5
2.2 軟體無線定義技術原理......5
2.3 高頻物理意義.............6
2.4 深度學習應用於通訊系統分析.7
2.5 半監督式學習.............8
第三章 實驗方法................10
3.1 實驗架構................10
3.2 理論公式................13
3.3 實驗目標................19
3.4 運算後處理設計...........22
3.5 實驗驗證設計.............25
第四章 實驗結果與討論...........29
4.1 實驗結果................29
4.2 實驗結果綜合討...........52
第五章 未來展望.................55
5.1 結論....................55
5.2 未來展望.................55
參考文獻.........................56
Extended Abstract...............59
Abstract........................59

[1].N. Panwar, S. Sharma, A. Kumar Singh, “A survey on 5G: The next generation of mobile communication,” Phys. Commun., vol. 18, pp. 64-84, Mar. 2016.
[2].X. Xiao, Z. Xiaoyi, S. Caiyao, “Method of multi-antenna signal selection based on SNR estimation,” Procedia Computer Science, vol. 107, pp. 646-651, Apr. 2017.
[3].Y. Seki, O. Takyu and Y. Umeda, “Performance evaluation of user selection based on average SNR in base station cooperation multi-user MIMO,” 2010 IEEE Radio and Wireless Symposium (RWS), New Orleans, LA, pp. 685-688, Mar. 2010.
[4].K. Shr, C. Chen, J. Jhang and Y. Huang, “Power-aware space-time-trellis-coded MIMO detector with SNR estimation and state-purging,” IEEE International Symposium on Circuits and Systems (ISCAS), Baltimore, MD, pp. 1-4, May 2017.
[5].A. F. Molisch, M. Z. Win, Y.-S. Choi and J. H. Winters, “Capacity of MIMO systems with antenna selection,” IEEE Trans. Wireless Commun., Helsinki, Finland, vol. 4, no. 4, pp. 1759-1772, Jul. 2005.
[6].Ghayoula, Elies & Bouallegue, A. & Ghayoula, Ridha & Chouinard, Jean-Yves., “Capacity and performance of MIMO systems for wireless communications,” J. Eng. Sci. Technol., pp. 108-111, Jul. 2014.
[7].X. Chen, “Throughput multiplexing efficiency for MIMO antenna characterization,” in IEEE Antennas Wireless Propag. Lett., vol. 12, pp. 1208-1211, Sep. 2013.
[8].J. Fei, R. Guangliang and Z. Zhe, “A new noise variance and post detection SNR estimation method for MIMO OFDM systems,” Proc. 11th IEEE Int. Conf. Commun. Technol. (ICCT), Hangzhou, pp. 179-182, Dec. 2008.
[9].S. Song, Y. Zhu, S. Luo, J. Hou, S. Du, Y. Song, S. Song, Y. Zhu, S. Luo, J. Hou, S. Du, Y. Song, “Millimeter wave based 3D clustered MIMO channel modeling and system simulation for vehicle-to-vehicle communication,” Phys. Commun., vol. 40, Jan. 2020.
[10].W. An, P. Zhang, J. Xu, H. Luo, L. Huang, and S. Zhong, “A novel machine learning aided antenna selection scheme for mimo internet of things,” Sensors. 20, no. 8, Apr. 2020.
[11].J. Wang, Y. Ma, L. Zhang, R. X. Gao, D. Wu, “Deep learning for smart manufacturing: Methods and applications,” J. Manuf. Syst., vol 48, pp. 144-156, Jan. 2018.
[12].S. Schindler, H. Mellein, “Assessing a MIMO Channel White Paper,” Feb. 2011.
[13].J. Li et al., “Dual-band eight-antenna array design for MIMO applications in 5G mobile terminals,” IEEE Access, vol. 7, pp. 71636-71644, Apr. 2019.
[14].洪丞韡,2016,軟體定義無線電設計實例研究,義守大學電機工程學系學位論文。
[15].A. Hussain, B. Por Einarsson and P. Kildal, “MIMO OTA testing of communication system using SDRs in reverberation chamber [measurements corner],” IEEE Antennas Propag. Mag., vol. 57, no. 2, pp. 44-53, Apr. 2015.
[16].M. Bakir, M. Belhachat, J. G. Liu and S. Z. Zhu, “Optimization of guard interval for OFDM performance over fading and AWGN channels using genetic algorithm,” Proceedings of the IEEE 6th Circuits and Systems Symposium on Emerging Technologies: Frontiers of Mobile and Wireless Communication (IEEE Cat. No.04EX710), Shanghai, China, pp. 457-460, May 2004.
[17].J. Joung, “Machine learning-based antenna selection in wireless communications,” IEEE Commun. Lett., vol. 20, no. 11, pp. 2241-2244, Nov. 2016.
[18].Lin, Huifa, Shin, W. Yong, Joung, Jingon., “Support vector machine-based transmit antenna allocation for multiuser communication systems,” Entropy, vol. 21, no.5, May 2019.
[19].Friedman, Jerome et al., “Special invited paper. additive logistic regression: A statistical view of boosting,” The Annals of Statistics, vol. 28, no. 2, pp. 337-374, Apr. 2000.
[20].Y. Lei and H. Yang, “A Gaussian process ensemble modeling method based on boosting algorithm,” Proceedings of the 32nd Chinese Control Conference, Xi'an, pp. 1704-1707, Oct. 2013.
[21].Sebastian Raschka、Vahid Mirjalili,2018,Python 機器學習,劉立民、吳建華譯,新北,博碩文化出版

電子全文 電子全文(網際網路公開日期:20250821)
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
系統版面圖檔 系統版面圖檔