近年来，随着城市经济的快速发展，X市汽车保有量高速持续增长，截至2019年底已达到360万辆。汽车给人们的生活带来便利的同时，也导致城市停车供需矛盾突出，不仅影响正常的交通秩序，也造成了停车难、违规停车等一系列问题。城市停车供需矛盾问题已经引起了城市管理者和学者的高度重视，各地政府出台了一系列停车管理政策，相关领域的研究成果也在持续增加。因此，在宏观上对城市停车资源需求总量进行预测、在微观上对城市存量停车资源进行充分利用的相关研究均具有现实意义。

本文采用实证的方法，对X市停车供需现状进行客观分析，梳理停车供需矛盾突出的成因，并从宏、微观两个层面进行停车需求的预测分析。

在宏观层面，对X市未来5年的停车需求总量进行预测。由于城市机动车保有量是预测城市停车需求总量的关键变量，因此本文对城市机动车保有量的影响因素进行了分析，进而选取X市2001-2019年人口总量、GDP、城镇市民人均可支配收入、社会消费品零售总额、固定资产投资和机动车保有量的数据，构建针对机动车保有量的多元回归预测模型。通过模型检验，发现存在多重共线性。在剔除相关性较强的自变量后，得到了机动车保有量关于人口总量、社会消费品零售总额的多元回归预测模型，模型预测精度较高，验证了模型的有效性。预测结果显示，X市2025年机动车保有量将达到582万辆。根据机动车保有量的预测值和弹性需求系数，计算可得X市2025年停车需求总量将达到690万个，其中基本停车泊位需求量为582万个，弹性停车泊位需求量为108万个。

在微观层面，考虑到X市土地、资金等方面的限制，短期内大量建设停车设施不具有现实可行性。因此，本文针对特定区域进行停车需求的短时预测，实现停车诱导，以期提高存量停车资源的利用率。具体选取X市2017年5500万条路内停车数据进行数据预处理，选用具有代表性的小寨商圈停车场作为研究对象，进行车辆停放规律分析，并采用BP模型和基于LSTM的停车需求短时预测模型，预测小寨商圈停车场在不同时间间隔下的有效停车泊位数。结果表明，基于LSTM的停车需求短时模型的预测精度高于BP模型，预测效果更好。

在此基础上，结合在论文相关调研过程中的发现以及对宏、微观层面停车泊位需求量的预测结果，提出城市停车管理的政策建议，希望对城市停车环境的改善有一定的帮助。

In recent years, with the rapid development of urban economy, the number of automobiles in X city has been growing rapidly and continuously, reaching 3.6 million by the end of 2019. While bringing convenience to people's life, cars also lead to prominent contradiction between urban parking supply and demand, which not only affects the normal traffic order, but also causes a series of problems such as difficult parking and illegal parking. The contradiction between supply and demand of urban parking has attracted great attention from city managers and scholars. Local governments have issued a series of parking management policies, and the research results in related fields are also increasing continuously. Therefore, it is of practical significance to predict the total demand of urban parking resources at macro level and make full use of urban parking resources at micro level.

 

In this paper, an empirical method is adopted to objectively analyze the current situation of parking supply and demand in X city, to sort out the causes of the prominent contradiction between parking supply and demand, and to forecast and analyze parking demand from macro and micro levels.

 

At the macro level, the total parking demand of X city in the next 5 years is forecasted. As the urban vehicle ownership is the key variable forecasting of city parking demand total, so in this paper, the influence factors of urban vehicle ownership is analyzed, and then select X city in 2001-2019 population, GDP, urban residents per capita disposable income and total retail sales of social consumer goods, investment in fixed assets, and motor vehicle ownership of data, A multivariate regression prediction model for vehicle ownership was established. Through model test, it is found that there is multicollinearity. After eliminating the independent variables with strong correlation, the multiple regression prediction model of motor vehicle ownership on total population and total retail sales of social consumer goods is obtained. The prediction accuracy of the model is high, which verifies the effectiveness of the model. The forecast results show that in 2025, the number of motor vehicles in X city will reach 5.82 million. According to the predicted value of vehicle ownership and the elastic demand coefficient, the total parking demand in X city will reach 6.9 million in 2025, among which the demand for basic parking berths is 5.82 million and the demand for flexible parking berths is 1.08 million.

 

At the micro level, it is not feasible to construct a large number of parking facilities in the short term considering the limitations of land and capital in X city. Therefore, this paper makes short-term prediction of parking demand in specific areas to realize parking induction, in order to improve the utilization rate of parking resources. Specific to select X city in 2017, 55 million road parking data in data preprocessing, selects the representative small village business circle as the research object, parking lot vehicle parking rules analysis, and USES the BP model and the model of parking demand for short-term prediction based on LSTM, predicting small village business circle in the parking lot under different time intervals of effective number of parking garages. The results show that the prediction accuracy of LSTM based short-term parking demand model is higher than that of BP model, and the prediction effect is better.

 

On this basis, combined with the findings in the related research process of the paper and the prediction results of parking parking demand at macro and micro levels, the paper puts forward policy suggestions for urban parking management, hoping to help improve the urban parking environment to some extent.

[1] Zhu L, Yu FR, Wang Y, et al. Big Data Analytics in Intelligent Transportation Systems: A Survey[J]. IEEE Transactions on Intelligent Transportation Systems, 2019, 20(1), 1-16.
[2] 吴娇蓉,林子旸.分类居住区夜间停车供需诱增关系研究[J].城市交通,2020,18(06):28-36+9.
[3] 刘炳全,度巍.双模式网络停车限制交通需求管理模型与方法研究[J].运筹与管理,2020,29(09):218-223.
[4] 林小围,周晶,卢珂,徐媛.基于合作博弈的停车位分配模型[J].系统管理学报,2019,28(01):62-66+85.
[5] 何胜学,高蕾.停车需求分布的双层规划模型及算法[J].交通运输系统工程与信息,2019,19(01):83-88+96.
[6] Hyeonsup Lim, Grant T. Williams, Dua Abdelqader, et al. Alternative Approach for Forecasting Parking Volumes[J]. Transportation Research Procedia,2017,25.
[7] 杨晓芳,江林成.基于系统总耗时最小的双目标泊位分配模型[J].上海理工大学学报,2021,43(02):179-185.
[8] 刘建明,朱宏伟,彭磊.基于诱导系统与用户博弈的停车实时诱导模型[J].计算机仿真,2019,36(06):166-170.
[9] Jong-Ho Shin, Hong-Bae Jun, Jae-Gon Kim. Dynamic control of intelligent parking guidance using neural network predictive control[J]. Computers & Industrial Engineering,2018,120.
[10] Rizvi S R, Zehra S, Olariu S. ASPIRE: An Agent-Oriented Smart Parking Recommendation System for Smart Cities[J]. IEEE Intelligent Transportation Systems Magazine, 2019, 11(4): 48-61.
[11] Chang I C, Hsu P J, Lin Y S, et al. Design and Implementation of an Internet-of-Things Roadside Parking System Based on Raspberry Pi 3 and Bluetooth Low Energy Mesh Sensor Network[J]. SENSORS AND MATERIALS, 2019, 31(2): 365-376.
[12] 谭冰清,屈挺,徐素秀,等.物联网驱动的停车位共享云平台[J].物流科技,2018,41(06):55-58.
[13] 李伟,梁睿君,宋丹.基于NB-IoT技术和GA-BP神经网络的车位预测系统[J].南京航空航天大学学报,2020,52(03):454-459.
[14] 苏世雄,马新华.基于物联网的智能停车系统的设计与研究[J].实验技术与管理,2020,37(06):68-70.
[15] Adam Millard-Ball, Rachel R. Weinberger, Robert C. Hampshire. Is the curb 80% full or 20% empty? Assessing the impacts of San Francisco’s parking pricing experiment[J]. Transportation Research Part A, 2014, 63, 76-92.
[16] Felix Caicedo, Carola Blazquez, Pablo Miranda. Prediction of parking space availability in real time[J]. Expert Systems With Applications, 2012, 39(8), 7281-7290.
[17] Stephen D. Boyles, Shoupeng Tang, Avinash Unnikrishnan. Parking search equilibrium on a network[J]. Transportation Research Part B, 2015, 81, 390-409.
[18] Andreas Klappenecker, Hyunyoung Lee, Jennifer L. Welch. Finding available parking spaces made easy[J]. Ad Hoc Networks, 2014, 12, 243-249.
[19] Jun Xiao, Yingyan Lou, Joshua Frisby. How likely am I to find parking? – A practical model-based framework for predicting parking availability[J]. Transportation Research Part B,2018,112, 19-39.
[20] 田琼,杨丽,罗婷.基于马尔科夫链的停车寻位模型与仿真[J].交通运输系统工程与信息,2015,15(01):179-184.
[21] 唐克双,郝兆康,衣谢博闻,等.停车场泊位占有率预测方法评价[J].同济大学学报(自然科学版),2017,45(4):533-543.
[22] Eleni I. Vlahogianni, Konstantinos Kepaptsoglou, Vassileios Tsetsos, et al. A Real-Time Parking Prediction System for Smart Cities[J]. Journal of Intelligent Transportation Systems,2016,20(2), 192-20.
[23] 季彦婕,陈晓实,王炜,等.基于小波变换和粒子群小波神经网络组合模型的有效停车泊位短时预测[J].吉林大学学报(工学版),2016,46(2):399-405.
[24] Ji Yanjie, Tang Dounan, Blythe P, et al. Short-term forecasting of available parking space using wavelet neural network model[J]. IET Intelligent Transport Systems, 2015, 9(2): 202-209.
[25] 张军,王书华,杨正瓴.基于粒子群优化的小波神经网络的停车位数量短时预测[J].计算机应用与软件,2015,32(11):66-68.
[26] 段满珍,陈光,张林,等.动态随机有效停车泊位预测方法[J].重庆交通大学学报(自然科学版),2018,37(6):81-86.
[27] 陈海鹏,图晓航,王玉,等.基于小波-ELM神经网络的短期停车泊位预测[J].吉林大学学报:理学版,2017,55(2):388-392.
[28] 韩印,郑喆,赵靖,等.基于灰色—小波神经网络的有效泊位预测[J].交通运输系统工程与信息,2017,17(5):60-67.
[29] 傅晨琳,黄敏,沙志仁.基于EEMD-BP方法的城市轨道交通进站客流短期预测[J].铁道运输与经济,2020,42(03):105-111.
[30] 胡立伟,杨锦青,何越人,孟玲,罗振武.基于改进BP神经网络的城市交通拥塞环境下车辆运行风险识别研究[J].公路交通科技,2019,36(10):105-113.
[31] 户佐安,邹正丰,包天雯.基于BP神经网络的交通信息量预测方法[J].交通运输工程与信息学报,2018,16(04):81-87.
[32] 高广银,丁勇,姜枫,等.基于BP神经网络的停车诱导泊位预测[J].计算机系统应用,2017,26(1),236-239.
[33] 代亮,梅洋,钱超,孟芸,吕金明.基于深度学习的短时交通量预测研究综述[J].计算机科学,2019,46(03):39-47.
[34] 温惠英,张东冉,陆思园.GA-LSTM模型在高速公路交通流预测中的应用[J].哈尔滨工业大学学报,2019,51(9):81-87.
[35] 张阳,杨书敏,辛东嵘.改进小波包与长短时记忆组合模型的短时交通流预测[J].交通运输系统工程与信息,2020,20(02):204-210.
[36] WANG Qing-rong，LI Tong-wei，ZHU Chang-feng. Short - time Prediction for Traffic Flow based on Wavelet De-noising and LSTM Model. Journal of Measurement Science and Instrumentation, 2020, 11: 1-14.
[37] Guo F, Polak J W, Krishnan R. Predictor fusion for short-term traffic forecasting [J].Transportation Research Part C: Emerging Technologies, 2018, 92, 90-100.
[38] 梅杰.城市商业区停车诱导关键技术研究[D].东南大学,2018.
[39] 王丰元,邹旭东,阎岩,李洪民,张洪海.基于用地和交通特征的停车需求预测模型[J].交通运输工程学报,2007(02):84-88.
[40] 温旭丽,孙垚飞,罗婷.基于交通特征的停车需求预测模型构建研究[J].公路交通科技(应用技术版),2014,10(04):297-299.
[41] 杨兆升,陈哓冬.智能化停车诱导系统有效停车泊位数据的预测技术硏究[J].交通运输系统工程与信息,2016,3(4):12-15.
[42] 李博,何鹏举,颜靖艺,等.基于物联网的停车场内停车诱导系统[J].电子设计工程,2019,27(09):26-30.
[43] Yan-jie Ji, Dou-nan Tang, Wei-hong Guo, et al. Forecasting available parking space with largest Lyapunov exponents method[J]. Journal of Central South University, 2014, 21(4), 1624-1632.
[44] 中华人民共和国住房和城乡建设部.住房城乡建设部关于印发城市停车设施规划导则的通知[OL].http://www.mohurd.gov.cn/wjfb/201509/t20150906_224668.html. [2015-09-01].
[45] 刘怡，常玉林，张鹏．交通结构优化下的城市停车需求预测研究［J］．重庆理工大学学报(自然科学)，2019，33(11):41－46.
[46] 邓惠章,李铭.基于机动车保有量的停车需求总量研究[J].交通与港航,2016,3(03):51-54+76.
[47] 郝晖.城市出租汽车保有量影响因素分析[J].中小企业管理与科技(上旬刊),2010(11):169.
[48] 格雷特,李洪成.时间序列预测实践教程[M].清华大学出版社,2012.
[49] 张金梦,刘慧君.遗传算法优化BP神经网络的泊车位数量预测[J].重庆大学学报,2018,41(03):76-81.
[50] LAI J, CHEN B, TAN T, et al. Phone-aware LSTM-RNN for voice conversion[C]//Signal Processing (ICSP),2016 IEEE 13th International Conference on. IEEE,2016: 177-182.
[51] LI D, QIAN J. Text sentiment analysis based on long short-term memory[C]//IEEE International Conference on Computer Communication and the Internet. IEEE, 2016: 471-475.
[52] Hochreiter S, Schmidhuber J. Long Short-Term Memory[J]. Neural Computation, 1997, 9(8), 1735-80.
[53] GRAVES A. Supervised sequence labelling with recurrent neural networks[M]. Springer Berlin Heidelberg, 2012.
[54] 王祥雪,许伦辉.基于深度学习的短时交通流预测研究[J].交通运输系统工程与信息,2018,18(01):81-88.