基于QPSO__RBF神经网络的网约车需求量预测模型
|
摘要
为了准确地预测乘客对网约车的需求量,指向性地提高部分地区的运力,让乘客更加容易预约到网约车,从而提升乘客的出行体验,通过对网约车需求量的变化规律和影响因素进行灰色关联度分析,选取网约车历史需求量、天气类型和道路拥堵比例作为影响因子,利用量子行为粒子群(QPSO)算法优化径向基(RBF)神经网络的网络权值、中心和基宽来构建QPSO__RBF神经网络预测模型。实际运营数据结果表明,QPSO__RBF神经网络预测模型具有可行性和有效性,其预测精度优于普通RBF神经网络模型的,且无论是改进的RBF神经网络还是普通RBF神经网络,综合考虑历史需求量、天气类型和道路拥堵比例作为影响因子的预测模型均优于只考虑历史需求量的预测模型。
Accurate demand forecasting for online car-hailing services can help companies improve traffic capacity in some areas,so that passengers can call a car easily,which improves their travel experience. Historical demand of online car-hailing services, types of weather condition and proportion of congested roads are chosen as impact factors for gray relational degree analysis,and a QPSO__RBF neural network model which uses the quantum-behaved particle swarm optimization algorithm to optimize the initial parameters of radial basis function neural network,including network weights,basis function centers and basis function widths,is established. It is found that the proposed QPSO__RBF neural network model outperformed the RBF neural network model. Moreimportantly,the model considering historical demand of online car-hailing, types of weather condition and congested roads proportion outperforms the model only considering historical demands of online car-hailing. Thus,feasibility and effectiveness of the proposed model is verified.
Accurate demand forecasting for online car-hailing services can help companies improve traffic capacity in some areas,so that passengers can call a car easily,which improves their travel experience. Historical demand of online car-hailing services, types of weather condition and proportion of congested roads are chosen as impact factors for gray relational degree analysis,and a QPSO__RBF neural network model which uses the quantum-behaved particle swarm optimization algorithm to optimize the initial parameters of radial basis function neural network,including network weights,basis function centers and basis function widths,is established. It is found that the proposed QPSO__RBF neural network model outperformed the RBF neural network model. Moreimportantly,the model considering historical demand of online car-hailing, types of weather condition and congested roads proportion outperforms the model only considering historical demands of online car-hailing. Thus,feasibility and effectiveness of the proposed model is verified.
引文
[1]王静.中国网约车的监管困境及解决[J].行政法学研究,2016(2):49-59.
[2]李霞.城市居民出行生成预测方法研究[D].吉林:吉林大学,2004.
[3]VOORT M V D,DOUNGHERTY M,WATSON S.Combining kohonen maps with ARIMA time series models to forecast traffic flow[J].Transportation Re-search Part C,1996,4(5):307-318.
[4]郝勇,朱海燕.基于客流日均量的地铁客流量的时间序列分析[J].铁道运输与经济,2009,31(10):42-46.
[5]周家中,张殿业.基于空间加权的LS-SVM城市轨道交通车站客流量预测[J].铁道学报,2014,36(1):1-7.
[6]傅贵,韩国强,逯峰,等.基于支持向量机回归的短时交通流预测模型[J].华南理工大学学报(自然科学版),2013,41(9):71-76.
[7]ZHENG W Z,LEE D H.Short-term freeway traffic flow prediction:bayesian combined neural network approach[J].Journal of Transportation Engineering,2006,132:114(2),114-121.
[8]MORETTI F,PIZZUTI S,PANZIERI S,et al.Urban traffic flow forecasting through statistical and neural network bagging ensemble hybrid modeling[J].Neurocomputing,2015,167:3-7.
[9]朱文兴,龙艳萍,贾磊.基于RBF神经网络的交通流量预测算法[J].山东大学学报(工学版),2007,37(4):23-27.
[10]冯明发,卢锦川.粒子群优化RBF神经网络的短时交通流量预测[J].计算机仿真.2010,27(12):323-326.
[11]赵晶.量子行为粒子群优化算法及其应用中的若干问题研究[D].无锡:江南大学,2013.
[12]BOARNET M,CRANE R.The influence of land use on travel behavior:specification and estimation strategies[J].Transportation Research Part A:Policy and Practice,2001,35(9):823-845.
[13]CASTRO-NETO M,JEONG Y S,JEONG M K,et al.Online-SVR for short-term traffic flow prediction under typical and atypical traffic conditions[J].Expert Systems with Applications,2009,36(3):6164-6173.
[14]DUNNE S,GHOSH B.Weather adaptive traffic prediction using neuro-wavelet models[J].IEEE Transactions on Intelligent Transportation Systems.2013,14(1):370-379.
[15]胡洁.BP和RBF神经网络短时交通流预测结果比较[J].长江大学学报(自科版),2007,4(3):74-77.
[16]ZHUJ Z,CAO J X,ZHU Y.Traffic volume forecasting based on radial basis function neural network with the consideration of traffic flows at the adjacent intersections[J].Transportation Research Part C:Emerging Technologies,2014,47(2):139-154.
[17]SUN J,FENG B,XU W B.Particle swarm optimization with particles having quantum behavior[C]//Proceedings of the2004 Congress on Evolutionary Computation.Portland,USA:IEEE Service Center,2004:325-331.
[18]方伟,孙俊,谢振平,等.量子粒子群优化算法的收敛性分析及控及控制参数研究[J].物理学报,2010,59(6):3686-3694.
[2]李霞.城市居民出行生成预测方法研究[D].吉林:吉林大学,2004.
[3]VOORT M V D,DOUNGHERTY M,WATSON S.Combining kohonen maps with ARIMA time series models to forecast traffic flow[J].Transportation Re-search Part C,1996,4(5):307-318.
[4]郝勇,朱海燕.基于客流日均量的地铁客流量的时间序列分析[J].铁道运输与经济,2009,31(10):42-46.
[5]周家中,张殿业.基于空间加权的LS-SVM城市轨道交通车站客流量预测[J].铁道学报,2014,36(1):1-7.
[6]傅贵,韩国强,逯峰,等.基于支持向量机回归的短时交通流预测模型[J].华南理工大学学报(自然科学版),2013,41(9):71-76.
[7]ZHENG W Z,LEE D H.Short-term freeway traffic flow prediction:bayesian combined neural network approach[J].Journal of Transportation Engineering,2006,132:114(2),114-121.
[8]MORETTI F,PIZZUTI S,PANZIERI S,et al.Urban traffic flow forecasting through statistical and neural network bagging ensemble hybrid modeling[J].Neurocomputing,2015,167:3-7.
[9]朱文兴,龙艳萍,贾磊.基于RBF神经网络的交通流量预测算法[J].山东大学学报(工学版),2007,37(4):23-27.
[10]冯明发,卢锦川.粒子群优化RBF神经网络的短时交通流量预测[J].计算机仿真.2010,27(12):323-326.
[11]赵晶.量子行为粒子群优化算法及其应用中的若干问题研究[D].无锡:江南大学,2013.
[12]BOARNET M,CRANE R.The influence of land use on travel behavior:specification and estimation strategies[J].Transportation Research Part A:Policy and Practice,2001,35(9):823-845.
[13]CASTRO-NETO M,JEONG Y S,JEONG M K,et al.Online-SVR for short-term traffic flow prediction under typical and atypical traffic conditions[J].Expert Systems with Applications,2009,36(3):6164-6173.
[14]DUNNE S,GHOSH B.Weather adaptive traffic prediction using neuro-wavelet models[J].IEEE Transactions on Intelligent Transportation Systems.2013,14(1):370-379.
[15]胡洁.BP和RBF神经网络短时交通流预测结果比较[J].长江大学学报(自科版),2007,4(3):74-77.
[16]ZHUJ Z,CAO J X,ZHU Y.Traffic volume forecasting based on radial basis function neural network with the consideration of traffic flows at the adjacent intersections[J].Transportation Research Part C:Emerging Technologies,2014,47(2):139-154.
[17]SUN J,FENG B,XU W B.Particle swarm optimization with particles having quantum behavior[C]//Proceedings of the2004 Congress on Evolutionary Computation.Portland,USA:IEEE Service Center,2004:325-331.
[18]方伟,孙俊,谢振平,等.量子粒子群优化算法的收敛性分析及控及控制参数研究[J].物理学报,2010,59(6):3686-3694.