Improved Social Emotion Optimization Algorithm for Short-Term Traffic Flow Forecasting Based on Back-Propagation Neural Network

doi:10.1007/s12204-019-2055-9

摘要/Abstract

摘要： The back-propagation neural network (BPNN) is a well-known multi-layer feed-forward neural network which is trained by the error reverse propagation algorithm. It is very suitable for the complex of short-term traffic flow forecasting; however, BPNN is easy to fall into local optimum and slow convergence. In order to overcome these deficiencies, a new approach called social emotion optimization algorithm (SEOA) is proposed in this paper to optimize the linked weights and thresholds of BPNN. Each individual in SEOA represents a BPNN. The availability of the proposed forecasting models is proved with the actual traffic flow data of the 2nd Ring Road of Beijing. Experiment of results show that the forecasting accuracy of SEOA is improved obviously as compared with the accuracy of particle swarm optimization back-propagation (PSOBP) and simulated annealing particle swarm optimization back-propagation (SAPSOBP) models. Furthermore, since SEOA does not respond to the negative feedback information, Metropolis rule is proposed to give consideration to both positive and negative feedback information and diversify the adjustment methods. The modified BPNN model, in comparison with social emotion optimization back-propagation (SEOBP) model, is more advantageous to search the global optimal solution. The accuracy of Metropolis rule social emotion optimization back-propagation (MRSEOBP) model is improved about 19.54% as compared with that of SEOBP model in predicting the dramatically changing data.

关键词: urban traffic, short-term traffic flow forecasting, social emotion optimization algorithm (SEOA), back-propagation neural network (BPNN), Metropolis rule

Abstract: The back-propagation neural network (BPNN) is a well-known multi-layer feed-forward neural network which is trained by the error reverse propagation algorithm. It is very suitable for the complex of short-term traffic flow forecasting; however, BPNN is easy to fall into local optimum and slow convergence. In order to overcome these deficiencies, a new approach called social emotion optimization algorithm (SEOA) is proposed in this paper to optimize the linked weights and thresholds of BPNN. Each individual in SEOA represents a BPNN. The availability of the proposed forecasting models is proved with the actual traffic flow data of the 2nd Ring Road of Beijing. Experiment of results show that the forecasting accuracy of SEOA is improved obviously as compared with the accuracy of particle swarm optimization back-propagation (PSOBP) and simulated annealing particle swarm optimization back-propagation (SAPSOBP) models. Furthermore, since SEOA does not respond to the negative feedback information, Metropolis rule is proposed to give consideration to both positive and negative feedback information and diversify the adjustment methods. The modified BPNN model, in comparison with social emotion optimization back-propagation (SEOBP) model, is more advantageous to search the global optimal solution. The accuracy of Metropolis rule social emotion optimization back-propagation (MRSEOBP) model is improved about 19.54% as compared with that of SEOBP model in predicting the dramatically changing data.

Key words: back-propagation neural network (BPNN), Metropolis rule, urban traffic, short-term traffic flow forecasting, social emotion optimization algorithm (SEOA)

中图分类号:

TP 183

ZHANG Jun* (张军), ZHAO Shenwei (赵申卫), WANG Yuanqiang (王远强), ZHU Xinshan (朱新山). Improved Social Emotion Optimization Algorithm for Short-Term Traffic Flow Forecasting Based on Back-Propagation Neural Network[J]. Journal of Shanghai Jiao Tong University (Science), 2019, 24(2): 209-219.

参考文献 22

[1]	GUO J H, HUANG W, WILLIAMS B M. AdaptiveKalman filter approach for stochastic short-term trafficflow rate prediction and uncertainty quantification [J].Transportation Research Part C, 2014, 43: 50-64.
[2]	HOU Y, EDARA P, SUN C. Traffic flow forecasting forurban work zones [J]. IEEE Transactions on IntelligentTransportation Systems, 2015, 16(4): 1761-1770.
[3]	HUANG M L. Intersection traffic flow forecastingbased on V-GSVR with a new hybrid evolutionary algorithm[J]. Neurocomputing, 2015, 147: 343-349.
[4]	HUWB, YAN L P, LIU K Z, et al. A short-term trafficflow forecasting method based on the hybrid PSO-SVR[J]. Neural Processing Letters, 2015, 43: 155-172.
[5]	YU B, SONG X L, GUAN F, et al. k-nearest neighbormodel for multiple-time-step prediction of short-termtraffic condition [J]. Journal of Transportation Engineering,2016, 142(6) 04016018.
[6]	ZHENG W Z, LEE D H, SHI Q X. Short-term freewaytraffic flow prediction: Bayesian combined neuralnetwork approach [J]. Journal of Transportation Engineering,2006, 132(2): 114-121.
[7]	SMITH B L, DEMETSKY M J. Short-term traffic flowprediction models-a comparison of neural network andnonparametric regression approaches [C]//Proceedingsof IEEE International Conference on Systems, Manand Cybernetics. IEEE: [s.n.], 1994: 1706-1709.
[8]	CHAN Y K, DILLON T S. Traffic flow predictionusing orthogonal arrays and Takagi-Sugeno neuralfuzzy models [C]//Proceedings of IEEE InternationalJoint Conference on Neural Networks. Beijing, China:IEEE, 2014: 35-41.
[9]	TONG G, FAN C L, CUI F Y, et al. Fuzzy neuralnetwork model applied in the traffic flow prediction[C]//Proceedings of the 2006 IEEE International Conferenceon Information Acquisition. Weihai, China:IEEE, 2006: 1229-1233.
[10]	PARK B, MESSER C J, URBANIK II T. Short-termfreeway traffic volume forecasting using radial basisfunction neural network [J]. Transportation ResearchRecord:Journal of the Transportation Research Board,1998, 1651: 39-47.
[11]	DIA H. An object-oriented neural network approachto short-term traffic forecasting [J]. European Journalof Operational Research, 2001, 131(2): 253-261.
[12]	ZHANG N, ZHANG Y L, LU H T. Seasonal autoregressiveintegrated moving average and support vectormachine models: Prediction of short term traffic flowon freeways [J]. Transportation Research Record: Journalof the Transportation Research Board, 2011, 2215:85-92.
[13]	HONG W C. Traffic flow forecasting by seasonal SVRwith chaotic simulated annealing algorithm [J]. Neurocomputing,2011, 74(12/13): 2096-2107.
[14]	LI M W, HONG W C, KANG H G. Urban traffic flowforecasting using Gauss-SVR with cat mapping, cloudmodel and PSO hybrid algorithm [J]. Neurocomputing,2013, 99: 230-240.
[15]	CHENG W, FENG P. Network traffic prediction algorithmresearch based on PSO-BP neural network[C]//Proceedings of International Conference on IntelligentSystems Research and Mechatronics Engineering.[s.l.]: Atlantis Press, 2015: 1239-1243.
[16]	KE L J, ZHANG Q F, BATTITI R. MOEA/D-ACO:A multi-objective evolutionary algorithm using decompositionand ant colony [J]. IEEE Transactions on Cybernetics,2013, 43(6): 1845-1859.
[17]	CUI Z H, CAI X J. Using social cognitive optimizationalgorithm to solve nonlinear equations[C]//Proceedings of the 9th IEEE International Conferenceon Cognitive Informatics. Beijing, China:IEEE, 2010: 199-203.
[18]	XU Y C, CUI Z H, ZENG J C. Social emotional optimizationalgorithm for nonlinear constrained optimizationproblems [C]//Proceedings of the 1st InternationalConference on Swarm, Evolutionary, and MemeticComputing. Berlin, Germany: Springer-Verlag, 2011:583-590.
[19]	ZHANG Y Q, ZHANG P L. Machine training and parametersettings with social emotional optimization algorithmfor support vector machine [J]. Pattern RecognitionLetters, 2015, 54: 36-42.
[20]	METROPOLIS N, ROSENBLUTH A W, ROSENBLUTHM N, et al. Equations of states calculationsfor fast computing machines [J]. Journal of ChemicalPhysics, 1953, 21(6): 1087-1091.
[21]	FANG J Y. Hybrid group search optimization algorithmand its application research [D]. Taiyuan,China: Department of Computer Science and Technology,Taiyuan University of Science and Technology,2010 (in Chinese).
[22]	WONG W H, LIANG F M. Dynamic weighting in Monte Carlo and optimization [J]. Proceedings of theNational Academy of Sciences of the United States ofAmerica, 1997, 94(26): 14220-14224.