Loading...
|
Please use this identifier to cite or link to this item:
https://nccur.lib.nccu.edu.tw/handle/140.119/125044
|
| Title: | 於霧計算架構下以LSTM模型預測空餘路邊停車位
Forecast on-street parking space vacancy with LSTM Model under Fog Computing |
| Authors: | 王彥嵐
Wang, Yan-Lan |
| Contributors: | 張宏慶
Jang, Hung-Chin
王彥嵐
Wang, Yan-Lan |
| Keywords: | iFogSim模擬器
長短期記憶(LSTM)
物聯網
雲端運算
霧計算
開放資料
路邊停車
iFogSim simulator
LSTM(Long Short-Term Memory)
IoT(Internet of Things)
Cloud computing
Fog computing
Open data
On-street parking |
| Date: | 2019 |
| Issue Date: | 2019-08-07 17:07:45 (UTC+8) |
| Abstract: | 在許多大城市中，停車問題一直都是駕駛者最頭痛的事，當需要開車前往市中心或是鬧區附近時，常是遍尋不著停車位，駕駛只能不停地在周遭巡航，找尋可停車的位置。找尋停車位是造成道路交通壅塞的原因之一，駕駛不斷的巡航除了耗費多餘的時間與汽車燃料外，更會造成空氣的汙染。因此，在都市計畫中常會針對停車問題設法進行改善，除了規劃增建停車區域或是提高大眾交通工具的使用率外，更重要的是活用既有的停車位，透過制訂彈性的停車收費原則，提高車輛周轉率；或是讓空餘停車位的閒置時間縮短，增加停車位的使用效率。近年來，在物聯網及人工智慧的蓬勃發展下，透過感測器結合影像辨識技術，可以長時間觀察停車位的使用情況，並將觀測後的數據進行分析及運用，以解決此類問題。本研究依據臺北市松山區的路邊停車情形作為案例，並活用政府公開資料，透過物聯網感測器來進行實驗，我們採用LSTM（Long Short-Term Memory）模型，經由歷史資料的學習及訓練，進一步預測路邊停車位的可用數量；並利用iFogSim模擬器，模擬出停車預測模型分別擺放在雲端運算及霧計算架構下，在能源消耗、網路使用量、整體延遲上的差異。實驗結果顯示，於霧計算架構中的停車預測模型誤差值在MAPE（Mean Absolute Percentage Error）指標下平均可達13.6%，且整體延遲時間較雲端運算下降約9成、網路使用量下降約7成
，也由於大規模佈建霧計算節點的因素，設備數量較多，因此在佈建一定數量的霧計算節點後，整體耗能相對於雲端運算也略為上升。
Drivers driving to urban neighborhoods suffer metropolis parking problems. Cruising down the street in search of parking spaces results in one of the leading causes of traffic congestion. Furthermore, continuous cruising along streets not only wastes time but also leads to excessive energy consumption and air pollution. Therefore, in urban planning, parking issues are crucial to be dealt with. Measures such as the expansion of parking space and increase of public transportation usage rate are taken. On top of that, shorten the idle time for available parking space and efficient parking pricing to increase turnover are also critical methods to solve the problems. With the aid of the rapid growth of IoT (Internet of Things) and AI (Artificial Intelligence), real-time on-street parking usage can be observed with data collected by sensors combined with image recognition technology, and the data is further analyzed to put into practical use. This paper aims at comparing efficiency in forecasting on-street parking vacancy with LSTM (Long Short-Term Memory) model under fog computing and under cloud computing. The data for computing was sourced from open data in Song Shan District of Taipei City, chosen as a case study. Learning from historical data, LSTM model is used to forecast parking availability with iFogSim system adopted to simulate the efficiency of energy consumption, network usage, and total delay when placed under cloud computing and fog computing architecture. MAPE (Mean Absolute Percentage Error) result of LSTM model is 13.6%. Total delay of LSTM under fog computing is about 90% lower than that under cloud computing; network usage, about 70% lower. However, with the increased number of fog nodes, energy consumption rises as well. Therefore, when the nodes are more than a threshold, the energy consumption of LSTM under fog computing is higher than that under cloud computing. |
| Reference: | [1]Sateesh Addepalli,Flavio Bonomi,Rodolfo Milito and Jiang Zhu, "Fog Computing and Its Role in the Internet of Things," ACM New York, NY,USA, pp. 13-16, 17 August 2012.
[2]Enrique Alba,Daniel H. Stolfi and Xin Yao,“Predicting Car Park Occupancy Rates in Smart Cities,”International Conference on Smart Cities, Springer 2017.
[3]Wael Alsafery,Badraddin Alturki,Kamal Jambi and Stephan Reiff-Marganiec, "Smart Car Parking System Solution for the Internet of Things in Smart Cities,"IEEE International Conference on Computer Applications & Information Security (ICCAIS), 4-6 April 2018.
[4]Khaled A. Althelaya, El-Sayed M. El-Alfy, Salahadin Mohammed, "Stock Market Forecast Using Multivariate Analysis with Bidirectional and Stacked (LSTM, GRU)," 2018 21st Saudi Computer Society National Computer Conference (NCC IEEE), 25-26 April 2018.
[5]Yoshua Bengio,Tomas Mikolov and Razvan Pascanu, "On the Difficulty of Training Recurrent Neural Networks," 30th International Conference on Machine Learning, PMLR 28(3):1310-1318, 2013.
[6]Yoshua Bengio, KyungHyun Cho,Junyoung Chung and Caglar Gulcehre, "Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling," arXiv:1412.3555v1[cs.NE], 11 Dec 2014.
[7]Rajkumar Buyya,Amir Vahid Dastjerdi,Soumya K.Ghosh and Harshit Gupta1, "iFogSim: A Toolkit for Modeling and Simulation of Resource Management Techniques in the Internet of Things, Edge and Fog Computing Environments," arXiv:1606.02007 [cs.DC], pp. CLOUDS-TR-2016-2, 7 June 2016.
[8]Rajkumar Buyya and Amir Vahid Dastjerdi, "Fog Computing: Helping the Internet of Things Realize Its Potential," IEEE Communication Society, pp. 112-116, August 2016.
[9]Subhrajit Bhattacharya,Mooi Choo Chuah and Xin Li,“UAV Assisted Smart Parking Solution,”IEEE International Conference on Unmanned Aircraft Systems, pp. 10062-1013, 13-16 Jun 2017.
[10]Jalil Boukhobza,Laurent Lemarchand,Mohammed Islam Naas and Philippe Raipin Parvedy, "An Extension to iFogSim to Enable the Design of Data Placement Strategies," IEEE 2nd International Conference on Fog and Edge Computing (ICFEC), 1-3 May 2018.
[11]Zuzana Bělinovà,Kristýna Hlubučková,Martin Langr, Jirí Růžička and Jan Šilar,"Smart Parking in the Smart City Application," IEEE 2018 Smart City Symposium Prague (SCSP) , pp. 1-5, 24-25 May 2018.
[12]Claudio Badii,Paolo Nesi and Irene Paoli,"Predicting Available Parking Slots on Critical and Regular Services by Exploiting a Range of Open Data,"IEEE Access Volume: 6 ,pp.44059-44071,09 August 2018.
[13]Sekhar Banarjee,Aditya Chaubey,Lasani Hussain, Sumit Kumar and Motahar Reza, "Forecasting Time Series Stock Data using Deep Learning Technique in a Distributed Computing Environment," 2018 International Conference on Computing, Power and Communication Technologies (GUCON IEEE) Galgotias University, Greater Noida, UP, India., 28-29 September 2018.
[14]Anna Corinna Cagliano,Alberto De Marco,Giulio Mangano,Paolo Neirotti and Francesco Scorrano,“Current Trends in Smart City Initiatives: Some Stylised Facts,” Elsevier, pp. 25-36, June 2014.
[15]Yu-Lun Chiang,Cheng-Ying Chou,Chao-Liang Hsieh, Hsiang-Yu Huang,Joe-Air Jiang,Jehn-Yih Juang,Chih-Hong Sun,Jen-Cheng Wang and Tzai-Hung Wen, "Urban Area PM2.5 Prediction with Machine Methods An On-Board Monitoring System," 2018 Twelfth International Conference on Sensing Technology (ICST IEEE), 4-6 Dec 2018.
[16]D. Evans,“The Internet of Things: How the Next Evolution of the Internet is Changing Everything,”Cisco Internet Business Solutions Group (IBSG), pp. 1-11, April 2011.
[17]Jie Gao,Shan Lin,Kin Sun Liu and Xiaobing Wu,"On-Street Parking Guidance with Real-Time Sensing Data for Smart Cities," 2018 15th Annual IEEE International Conference on Sensing, Communication, and Networking (SECON), p. 1-9, 11-13 June 2018.
[18]Sepp Hochreiter and Jurgen Schmidhuber, "Long Short-Term Memory," IEEE 1st International Conference on Neural Networks, 1997.
[19]Soman KP, Poornachandran P. and Vinayakumar R., "Applying Deep Learning Approaches for Network Traffic Prediction," International Conference on Advances in Computing, Communications and Informatics (ICACCI IEEE), p. pp. 2353–2358, 13-16 September 2017.
[20]Colin David Lewis, "Industrial and Business Forecasting Methods: A Practical Guide to Exponential Smoothing and Curve Fitting," Published: London, Butterworch Scientific,1982.
[21]Christopher Leckie,Sutharshan Rajasegarar and Yanxu Zheng,"Parking Availability Prediction for Sensor-Enabled Car Parks in Smart Cities," 2015 IEEE Tenth International Conference on Intelligent Sensors,Sensor Networks and Information Processing(ISSNIP) Singapore,7-9 April 2015.
[22]Christopher Liao,Eran Simhon and David Starobinski," Smart Parking Pricing: A Machine Learning Approach," IEEE Conference on Computer Communications Workshops（INFOCOM WKSHPS）:SDP17: 6th Workshop on Smart Data Pricing, 1-4 May 2017.
[23]T Lin,F Le Mouël and H Rivano,“A Survey of Smart Parking Solutions,”IEEE Transactions on Intelligent, pp. 3229-3253, 12 December 2017.
[24]Wei Ma,Xidong Pi,Sean Qian and Shuguan Yang“A Deep Learning Approach to Real-time Parking Occupancy Prediction in Spatio-Termporal Networks Incorporating Multiple Spatio-Temporal Data Sources,”arXiv:1901. 06758v3 [cs.LG], 11 February 2019.
[25]Sang Nguyen,Zoran Salcic and Xuyun Zhang, "Big Data Processing in Fog– Smart Parking Case Study," IEEE Intl. Conf. on Parallel & Distributed Processing with Applications,Ubiquitous Computing & Communications, Big Data & Cloud Computing,Social Computing & Networking, Sustainable Computing & Communications, 2018.
[26]Gregory Pierce and Donald Shoup, "Getting the Prices Right：An Evaluation of Pricing Parking by Demand in San Francisco," Journal of the American Planning Association, vol.79, no. 1, pp. 67-81, 2013.
[27]Shyam Ravishankar and Nrithya Theetharappan, “Cloud Connected Smart Car Park,” IEEE International conference on I-SMAC, pp. 71-74, 2017.
[28]Donald Shoup, "Cruising for Parking," Transport Policy, pp. 479-486, 24 July 2006.
[29]Arm.com, "Products Overview Cortex-a53," 2019. [Online]. Available: https://developer.arm.com/ip-products/processors/cortex-a/cortex-a53.
[30]Boston Consulting Group and Uber, "台北市交通調查白皮書," 2017. [Online].Available:https://buzzorange.com/ techorange/2017/11/09/uber-survey-about-tp-traffic/.
[31]Colah, "Understanding LSTM Networks," 27 August 2015.[Online].Available:https://colah.github.io/posts/2015-08-Understanding-LSTMs/.
[32]CPUID.com,"Software HWMonitor," 2019. [Online]. Available:https://www.cpuid.com/softwares/hwmonitor.html.
[33]Intel.com, "Intel-core-i7-4770K products," 2019. [Online]. available: https://ark.intel.com/content/www/
tw/zh/ark/products/75123/intel-core-i7-4770k-processor-8m-cache-up-to-3-90-ghz.html.
[34]OpenCV.org, "About OpenCV," 2019. [Online]. Available :https://opencv.org /about/.
[35]ROHM Co., "Nano Energy-Ultra Low LQ Buck Converter For Low Power Applications-BD70522GUL, "2019.[Online]. Available:https://www.rohm.com.tw/ products/power-management/switching-regulators/ integrated-fet/buck-converters-synchronous/ bd70522gul-product.
[36]WWW.Raspberrypi.org, "What is a Raspberry Pi?," 2018. [Online]. Available: https://www.raspberrypi.org/help/ what-%20is-a-raspberry-pi/.
[37]WWW.Raspberrypi.org, "Featured Products," 2018. [Online].Available: https://www.raspberrypi.org/ products/.
[38]WWW.Raspberrypi.org, "Featured Products," 2018. [Online]. Available: https://www.raspberrypi.org/
documentation/usage/camera/.
[39]WWW.Raspberrypi.org, "Documentation Power Supply," 2018.[Online]. Available:https://www.raspberrypi.org/ documentation/hardware/raspberrypi/power/README.md.
[40]臺北市停車管理工程處, "107年度臺北市汽機車停車供需調查," 2018.[Online]. Available:https://pma.gov.taipei/ News_Content.aspx?n=65D85809A5ABC2C9&sms=6AAEA2F8F4E6DFD2 &s=886011F8E3A976AD.
[41]臺北市政府主計處, "土地人口分布," March 2018. [Online]. Available: https://w2.dbas.taipei.gov.tw/statchart/ a2.htm.
[42]臺北市政府資料開放平臺,"路邊停車格位使用情形," 2018 [Online]. Available: https://data.taipei/dataset/detail/ metadata?id=434638ca-8770-42c1-940d-0386a74f6eb9.[Accessed 08~10 2018].
[43]臺北市政府資料開放平臺, "臺北市停車管理工程處「各行政區路邊汽車收費路段資訊(僅供參考)-松山區」," 2018. [Online]. Available: https:// pma.gov.taipei/Content_List.aspx?n=B68496752E5DD5EE.
[44]中華民國交通部, "中華民國交通部停車格位與禁停標線之劃設原則," August 1992. [Online].Available: http://www.motc .gov.tw.
[45]中華民國交通部, "交通部統計查詢網-機動車輛登記數," 2019. [Online].Available:https://stat.motc.gov.tw/mocmo/stmain .jsp?sys=100&funid=b3301. |
| Description: | 碩士
國立政治大學
資訊科學系碩士在職專班
105971022 |
| Source URI: | http://thesis.lib.nccu.edu.tw/record/#G0105971022 |
| Data Type: | thesis |
| DOI: | 10.6814/NCCU201900455 |
| Appears in Collections: | [資訊科學系碩士在職專班] 學位論文
|
Files in This Item:
| File |
Size | Format | |
|---|
| 102201.pdf | 1219Kb | Adobe PDF2 | 173 | View/Open |
|
All items in 政大典藏 are protected by copyright, with all rights reserved.
|
