郑凯;林培杰;赖云锋;程树英;陈志聪;吴丽君;
ZHENG Kai;LIN Peijie;LAI Yunfeng;CHENG Shuying;CHEN Zhicong;WU Lijun;College of Physics and Information Engineering, Fuzhou University;

• 1:福州大学物理与信息工程学院

摘要(Abstract)：

提出了一种基于长短期记忆神经网络的智慧路灯控制方法，实现了能够根据路面能见度情况进行自适应调光的智慧路灯。选择PM2.5、PM10、湿度、累积风速四种气象因子作为输入，采用长短期记忆神经网络实现对路面能见度的建模，并使用Adam算法优化模型。智慧路灯根据建模所得能见度与照度信息，在高能见度时，自动采用普通亮度与高色温照明模式，有效节约能源；在低能见度时输出更高的亮度与更低的色温，增强路灯透雾能力，保证路面照度符合需求。通过实验分析，该模型的预测值与真实能见度之间正则化均方根误差为0.13194、平均绝对误差为0.69785 km以及决定系数为0.85725,优于所选的对比模型。相比于传统路灯控制方式本文方法在保障车辆行驶安全的同时有更好的节能效果。
A smart street lamp control approach based on Long Short-Term Memory Network(LSTM) is proposed, which realizes a smart street lamp capable of adaptive dimming according to the visibility of the road. Four meteorological factors including PM2.5, PM10, humidity and cumulated wind speed are selected as input to train the LSTM for modeling the road visibility, and the Adam algorithm is used to optimize the model. According to the visibility and illuminance, the smart street lamp control approach adopts normal brightness and high color temperature lighting modes in high visibility condition to effectively saving energy, and applies higher brightness output and lower color temperature in low visibility condition to enhance the fog penetration ability and ensure the road illumination meets the requirement. Experimental analysis demonstrates that the NRMSE, MAE and R^{(2 )}obtained by the proposed model are 0.13194, 0.69785 km and 0.85725, respectively, which is superior to other compared approaches. Additional, the studied method can save 19% energy comparing with the conventional timing control method.

关键词(KeyWords)： 智慧路灯;长短期记忆神经网络;自适应控制;节能;能见度建模
smart street lamp;LSTM;adaptive control;energy saving;visibility modeling

Abstract:

Keywords:

基金项目(Foundation): 福建省自然科学基金“基于多源数据特征分析的光伏阵列故障诊断方法研究”(批准号：2018J01774);; 福州市科技计划项目“基于物联网的智慧道路资源综合管理平台的研发及应用”(批准号：2021-P-059)

作者(Authors): 郑凯;林培杰;赖云锋;程树英;陈志聪;吴丽君;
ZHENG Kai;LIN Peijie;LAI Yunfeng;CHENG Shuying;CHEN Zhicong;WU Lijun;College of Physics and Information Engineering, Fuzhou University;

参考文献(References)：

• [1]ECHELON.Monitored Outdoor Lighting:Market,Challenges,Solutions,and Next Steps[R].2007．
• [2]刘义平，张明明，张畅，等．智慧路灯建设的实践与思考[J]．照明工程学报，2017,28(5):103-105．
• [3]Northeast Group.Global LED and Smart Street Lighting:Market Forecast 2015-2025[R].2015．
• [4]田芳，杨紫琼，郝帅．人本视角下城市开敞空间夜景照明评价指标体系研究[J]．照明工程学报，2021,32(3):46-50．
• [5]张宏征，吴雨婷，冯子龙，等．智慧照明应用在智慧城市中的作用[J]．照明工程学报，2020,31(5):125-130．
• [6]叶炜，吕伟，洪宽，等．基于NB-IoT技术的道路照明智能控制系统[J]．照明工程学报，2017,28(5):20-23．
• [7]Sanchez-sutil F,Cano-ortega A.Smart regulation and efficiency energy system for street lighting with LoRa LPWAN[J].Sustainable Cities and Society,2021,70(3):102912．
• [8]崔刚刚，徐方卉，周小丽．模拟雾环境下图像特征与主观评价的关系[J]．照明工程学报，2021,32(3):152-158．
• [9]Jin H,Jin S,Chen L,et al.Research on the lighting performance of LED street lights with different color temperatures[J].IEEE Photonics Journal,2015,7(6):1-9．
• [10]Dong L,Zhao E,Chen Y,et al.Impact of LED color temperatures on perception luminance in the interior zone of a tunnel considering fog transmittance[J].Advances in Civil Engineering,2020,2020(55):1-13．
• [11]Todorovic'B M,Samard?ija D.Road lighting energy-saving system based on wireless sensor network[J].Energy Efficiency,2016,10(1):1-9．
• [12]Daely P T,Reda H T,Satrya G B,et al.Design of smart LED streetlight system for smart city with web-based management system[J].IEEE Sensors Journal,2017,17(18):6100-6110．
• [13]李云霞，徐跃通．石家庄市大气颗粒物与能见度相关性研究[J]．鲁东大学学报(自然科学版)，2015,31(1):92-96．
• [14]Duddu V R,Pulugurtha S S,Mane A S,et al.Backpropagation neural network model to predict visibility at a road link-level[J].Transportation Research Interdisciplinary Perspectives,2020,8:100250．
• [15]王震洲，聂亚宁，于平平．基于神经网络的多城市协同能见度预测研究[J]．电子测量与仪器学报，2019,33(11):73-78．
• [16]Hochreiter S,Schmidhuber J.Long short-term memory[J].Neural Computation,1997,9(8):1735-1780．
• [17]Sadeque Z A,Bui F M.A deep learning approach to predict weather data using Cascaded LSTM Network[C]//Proceedings of the 2020 IEEE Canadian Conference on Electrical and Computer Engineering(CCECE).IEEE,2020．
• [18]Li Shuo,Liang Xuan,Zou Tao,et al.Assessing Beijing’s PM2.5 pollution:severity,weather impact,APEC and winter heating[J].Proceedings of the Royal Society Mathematical,physical and engineering sciences,2015,471(2182):1-20．
• [19]刘健，刘银坤．基于Pearson系数对气象因子与污染物的相关性研究[J]．华北科技学院学报，2019,16(4):93-97．
• [20]姜江，张国平，高金兵．北京大气能见度的主要影响因子[J]．应用气象学报，2018,29(2):188-199．
• [21]城市道路照明设计标准:CJJ 45-2015[S]．北京:中国建筑工业出版社，2015．
• [22]Pal N R,Pal S,Das J,et al.SOFM-MLP:a hybrid neural network for atmospheric temperature prediction[J].IEEETransactions on Geoscience and Remote Sensing,2004,41(12):2783-2791．
• [23]Parvez I,Sarwat A,Debnath A,et al.Multi-layer perceptron based photovoltaic forecasting for rooftop pv applications in smart grid[C]//Proceedings of the 2020Southeast Con.IEEE,2020．
• [24]朱丹，翟丹华，吴志鹏，等．基于Xgboost算法的短时强降水预报方法[J]．气象科技，2021,49(3):406-408．
• [25]黄骞，郑颖尔，邓钰桥．基于XGBoost节假日路网流量预测研究[J]．公路，2018,63(12):234-238．
• [26]郑晅，魏倩，李雪，等．基于DIALux的公路隧道洞口减光防眩措施评价[J]．照明工程学报，2019,30(4):75-81．
• [27]徐博林，李玮晟，陈神飞，等．雾环境下道路照明“白墙效应”的模拟实验[J]．照明工程学报，2018,29(4):106-109．
• [28]李璇，金尚忠，王乐，等．中间视觉条件下不同色温光源对道路照明的影响[J]．光电子激光，2011,22(7):997-999．
• [29]李璇．LED路灯路面照度测量及中间视觉的研究[D]．北京:中国计量学院，2012.