摘 要：出租车作为一种城市常见的交通方式，在某一时刻面临着上客率低的问题，而在另一时刻又面临着需求过剩的问题。为了提高驾驶员巡航效率，文章基于 Spark 并行分布式平台，提出一种集合经验模态分解的空间注意力机制双向门控循环单元模型（EEMDN-SABiGRU），对乘客热点进行精准预测，与 LSTM、GRU、EMD-LSTM、EMD-GRU、EEMD-LSTM、EEMD-GRU、EMDN-GRU、CNN 和 BP 等模型相比，EEMDN-SABiGRU 模型的 MAPE、MAE、RMSE 和 ME 值至少降低了 58.33%、44.91%、43.19% 和 39.33%，在乘客热点预测上具有良好的效果。
关键词：乘客热点预测；网格映射；空间注意力；GPS 轨迹；EEMD；BiGRU；Spark
DOI:10.19850/j.cnki.2096-4706.2022.22.018
中图分类号：O211.61；TP301.6 文献标识码：A 文章编号：2096-4706（2022）22-0071-07
A Distributed EEMDN-SABiGRU Model on Spark for Passenger Hotspot Prediction
GENG Jian, SHEN Bingqi
(College of Data Science and Information Engineering, Guizhou Minzu University, Guiyang 550025, China)
Abstract: As a common transportation mode in cities, taxis are confronted with the problem of low customer rate at one moment and excess demand at another moment. Based on the Spark parallel distributed platform, in order to improve the cruising efficiency of taxi drivers, this paper proposes a bi-directional gated recurrent unit model of spatial attention mechanism (EEMDN-SABiGRU) of incorporating empirical modal decomposition to predict passenger hotspots accurately. Compared with LSTM, GRU, EMD-LSTM, EMDGRU, EEMD-LSTM, EEMD-GRU, EMDN-RU, CNN and BP and other models, the MAPE, MAE, RMSE and ME values of EEMDNSABiGRU model are decreased by 58.33%,44.91%,43.19% and 39.33% at least. Ithas good effect on the passenger hotspot prediction.
Keywords: passenger hotspot prediction; grid mapping; spatial attention; GPS trajectory; EEMD; BiGRU; Spark
参考文献：
[1] ZHENG Y，CAPRA L，WOLFSON O，etal. Urban computing：concepts，methodologies，and applications [J].ACM Transactions on Intelligent Systems and Technology （TIST），2014， 5（3）：1-55.
[2] ZHENG Y. Urban computing：enabling urban intelligence with big data [J].Frontiers of Computer Science，2017，11（1）：1-3.
[3] BATTY M，AXHAUSEN K W，GIANNOTTI F，et al. Smart cities of the future [J].The European Physical Journal Special Topics，2012，214（1）：481-518.
[4] BI S，XU R，LIU A，et al. Mining Taxi Pick-Up Hotspots Based on Grid Information Entropy Clustering Algorithm [J].Journal of Advanced Transportation，2021，2021：1-11
[5] GONG L，LIU X，WU L，et al. Inferring trip purposes and uncovering travel patterns from taxi trajectory data [J].Cartography and Geographic Information Science，2016，43（2）：103-114.
[6] XIA D，JIANG S，YANG N，et al. Discover in gspatio temporal characteristics of passenger travel with mobile trajectory big data [J].Physica A：Statistical Mechanics and its Applications，2021， 578：126056.
[7] XIA D，ZHANG M，YAN X，et al. A distributed WNDLSTM model on MapReduce for short-term traffic flow prediction [J]. Neural Computing and Applications，2021，33（7）：2393-2410.
[8] JIAMIL M S，AKBAR S. Taxi passenger hotspot prediction using automatic ARIMA model [C]//2017 3rd International Conference on Science in Information Technology （ICSITech）.Bandung： IEEE，2017：23-28.
[9] LI X，PAN G，WU Z，et al. Prediction of urban human mobility using large-scale taxi traces and its applications [J].Frontiers of Computer Science，2012，6（1）：111-121.
[10] HUANG N E，SHEN Z，LONG S R，et al. The empirical mode decomposition and the Hilbert spectrum for nonlinear and nonstationary time series analysis [J].Proceedings of the Royal Society of London. Series A：mathematical，physical and engineering sciences， 1998，454（1971）：903-995.
[11] YAO X，WANG F，ZHANG Y.A Prediction Model of Security Situation Based on EMD-PSO-SVM [C]//Proceedings of the 2015 International Conference on Electrical and Information Technologies for Rail Transportation.Springer，2016：355-363.
[12] NIE Z，SHEN F，XU D，et al. An EMD-SVR model for short-term prediction of ship motion using mirror symmetry and SVR algorithms to eliminate EMD boundary effect [J/OL]. Ocean Engineering，2020，217：107927[2022-05-26].https://www. sciencedirect.com/science/article/abs/pii/S0029801820308854.
[13] LIU X，ZHANG Y，ZHANG Q. Comparison of EEMDARIMA，EEMD-BP and EEMD-SVM algorithms for predicting the hourly urban water consumption [J].Journal of Hydroinformatics， 2022，24（3）：535-558.
[14] YANG X，XUE Q，YANG X，et al. A novel prediction model for the inbound passenger flow of urban rail transit [J].Information Sciences，2021，566：347-363.
[15] LI X，ZHANG Y，DU M，et al. The forecasting of passenger demand under hybrid ridesharing service modes：A combined model based on WT-FCBF-LSTM [J/OL].Sustainable Cities and Society，2020，62：102419[2022-05-26].https://www.sciencedirect.com/ science/article/abs/pii/S2210670720306405.
[16] LOU J，SUN J，ZHU Y，et al. Stp-trellisnets：Spatialtemporal parallel trellisnets for metro station passenger flow prediction [C]//Proceedings of the 29th ACM international conference on information & knowledge management.New York：ACM，2020：1185- 1194.
[17] ZHANG X，ZHANG Q，ZHANG G，et al. A novel hybrid data-driven model for daily land surface temperature forecasting using long short-term memory neural network based on ensemble empirical mode decomposition [J].International journal of environmental research and public health，2018，15（5）：1032.
[18] REZAEI H，FAALJOU H，MANSOURFAR G. Stock price prediction using deep learning and frequency decomposition [J/ OL].Expert Systems with Applications，2021，169：114332[2022- 05-26].https://www.sciencedirect.com/science/article/abs/pii/ S0957417420310228.
作者简介：耿建（1998—），男，汉族，贵州毕节人，硕士研究生在读，研究方向：大数据分析、数据挖掘和机器学习；申冰琪（1997—），女，汉族，河南周口人，硕士研究生在读，研究方向：大数据分析、人工智能和深度学习。