随着社会经济的发展和城市化的加快推进，交通拥堵问题已经严重影响到了城市的运作效率，着眼于交通信息广泛应用与服务的智能交通是目前有效缓解交通拥堵的重要手段，智能交通可以有效提高既有交通设施的运行效率，而实现高精度的短时交通流预测是智能交通建设的关键部分，本研究通过深入挖掘分析路段拥堵影响因素，在解析交通系统相关要素作用机理的基础上，构建了D-SVM动态干扰预测模型，有效提高了模型的预测精度，模型通过增强时间特征的方式考虑了时间周期变化对路段交通流量的影响，消除了波峰波谷时段流量的差异化特征，可以有效的实现实时预测，为实现快速道路拥堵疏导提供可能性，研究丰富了智能交通理论的建设，既有助于交通管理者制定交通策略提供决策支持，建立拥堵预警机制，又方便出行者提前做好出行规划，提高出行体验。

本文对交通系统的影响因素做了较为全面的分析，首先通过检索FMS管理科学高质量期刊及铁路运输领域高质量科技期刊刊载的城市交通拥堵相关中英文文献，梳理了近十年城市交通拥堵相关影响因素，通过对期刊影响因子、文献引用量、影响因素和期刊等级进行建模分析，测算出高认证度的交通拥堵影响因素作为DEMATEL-ISM因素分析的输入变量，建立五层递阶结构模型，根据模型的层次结构分析了常态化路段拥堵和偶发性路段拥堵成因，揭示了复杂交通系统的拥堵规律。

结合对道路拥堵原因的分析，本文构建了基于系统分析的D-SVM动态干扰预测模型，使用傅里叶变换将交通流量的时域信号转换成频域信号，由此测算出流量数据的周期性特征，此外，为了减少离散程度较大的异常值对模型的干扰，采用指数平滑的方法对异常值进行平稳化处理。研究通过融合道路拥堵干扰，综合考虑了道路流量的变化规律，将影响路段偶发性拥堵的动态干扰因素进行着重处理，并将因周期性变化带来的时间特征差异进行增强处理，消除了波峰波谷时段的预测差异，最后通过差分进化改进支持向量机实现了对道路流量短时预测，通过测算平均绝对误差MAE和均方根误差RMSE证明模型预测精度在总体上得到了明显的提升。

With the development of social economy and the acceleration of urbanization, traffic congestion has seriously affected the operation efficiency of cities. Intelligent transportation focusing on the wide application and service of traffic information is an important means to effectively alleviate traffic congestion. Intelligent transportation can effectively improve the operation efficiency of existing transportation facilities, and the realization of high-precision short-term traffic flow prediction is a key part of intelligent transportation construction, This study deeply excavates and analyzes the influencing factors of road congestion, and constructs a D-SVM dynamic interference prediction model based on the analysis of the action mechanism of relevant factors of the traffic system, which effectively improves the prediction accuracy of the model. The model considers the impact of time cycle changes on road traffic flow by enhancing the time characteristics, eliminates the differential characteristics of flow in peak and trough periods, and can effectively realize real-time prediction, In order to provide the possibility of congestion relief on Expressway, the research enriches the construction of intelligent transportation theory, which not only helps traffic managers formulate traffic strategies, provide decision support, establish congestion early warning mechanism, but also facilitates travelers to make travel planning in advance and improve travel experience.

 

This paper makes a more comprehensive analysis on the influencing factors of the transportation system. Firstly, by searching the Chinese and English literature related to urban traffic congestion published in high-quality journals of FMS management science and high-quality scientific and technological journals in the field of railway transportation, this paper combs the influencing factors related to urban traffic congestion in recent ten years, and makes a modeling analysis on the influencing factors of journals, literature citations, influencing factors and journal grades, The influencing factors of traffic congestion with high certification degree are calculated as the input variables of DEMATEL-ISM factor analysis, and a five-layer hierarchical structure model is established. According to the hierarchical structure of the model, the causes of normal section congestion and occasional section congestion are analyzed, and the congestion law of complex traffic system is revealed.

 

Combined with the analysis of the causes of road congestion, this paper constructs a D-SVM dynamic interference prediction model based on system analysis, and uses Fourier transform to convert the time-domain signal of traffic flow into frequency-domain signal, so as to calculate the periodic characteristics of traffic flow data. In addition, in order to reduce the interference of outliers with large dispersion to the model, the exponential smoothing method is used to stabilize the outliers. By integrating the road congestion interference and comprehensively considering the change law of road flow, the dynamic interference factors affecting the occasional congestion of the road section are emphatically treated, and the time characteristic differences caused by periodic changes are enhanced to eliminate the prediction differences in peak and trough periods. Finally, the short-term prediction of road flow is realized by improving support vector machine through differential evolution, By calculating the mean absolute error MAE and root mean square error RMSE, it is proved that the prediction accuracy of the model has been significantly improved on the whole.

[1] 王启燕. 基于时间序列关联规则挖掘的交通拥堵预测研究 [D]; 北京交通大学, 2020.
[2] 杨兆升, 王媛, 管青. 基于支持向量机方法的短时交通流量预测方法 [J]. 吉林大学学报(工学版), 2006, (06): 881-4.
[3] LV Z, ZHANG S, XIU W. Solving the Security Problem of Intelligent Transportation System With Deep Learning [J]. Ieee Transactions on Intelligent Transportation Systems, 2021, 22(7): 4281-90.
[4] 宗刚, 朱丽南, 魏素豪. 北京市公共交通系统局部拥堵的影响因素研究 [J]. 价格理论与实践, 2016, (06): 88-91.
[5] 赵辉. 北京市交通拥堵特征及其影响因素分析 [D]; 北京交通大学, 2017.
[6] 许彬. 基于分辨矩阵的交通拥堵关键影响因素提取及解析 [J]. 长安大学学报(自然科学版), 2012, 32(05): 91-6+106.
[7] DADASHOVA B, LI X, TURNER S, et al. Multivariate time series analysis of traffic congestion measures in urban areas as they relate to socioeconomic indicators [J]. Socio-Economic Planning Sciences, 2021, 75.
[8] ABDELWAHAB M A, ABDEL-NASSER M, HORI M. Reliable and Rapid Traffic Congestion Detection Approach Based on Deep Residual Learning and Motion Trajectories [J]. Ieee Access, 2020, 8: 182180-92.
[9] 张振龙, 邱煜卿, 蒋灵德, et al. 基于实时路况的交通拥堵时空特征及其影响因素分析——以苏州古城区为例 [J]. 现代城市研究, 2020, (01): 104-12.
[10] BIAN C, YUAN C, KUANG W, et al. Evaluation, Classification, and Influential Factors Analysis of Traffic Congestion in Chinese Cities Using the Online Map Data [J]. Mathematical Problems in Engineering, 2016, 2016.
[11] AFTABUZZAMAN M, CURRIE G, SARVI M. Exploring the underlying dimensions of elements affecting traffic congestion relief impact of transit [J]. Cities, 2011, 28(1): 36-44.
[12] LI S, CHENG Y, JIN P, et al. A Feature-Based Approach to Large-Scale Freeway Congestion Detection Using Full Cellular Activity Data [J]. Ieee Transactions on Intelligent Transportation Systems, 2022, 23(2): 1323-31.
[13] 贾凡. 基于解释结构模型和MICMAC的城市交通拥堵问题分析 [J]. 河北大学学报(自然科学版), 2020, 40(04): 344-50.
[14] 韩春阳, 黄合来, 唐进君, et al. 城市道路交通拥堵状态出行者感知差异性研究 [J]. 交通运输系统工程与信息, 2019, 19(03): 202-7.
[15] 李树彬, 高自友, 吴建军等. 基于事件的交通拥堵模拟与消散策略研究 [J]. 系统仿真学报, 2012, 24(08): 1707-13.
[16] 郑思齐, 徐杨菲, 张晓楠等. “职住平衡指数”的构建与空间差异性研究:以北京市为例 [J]. 清华大学学报(自然科学版), 2015, 55(04): 475-83.
[17] 朱建安, 戴帅. 城市交通限行需求管理措施实践评述 [J]. 交通标准化, 2014, 42(21): 33-9.
[18] 李梦, 黄海军. 考虑共享出行的用户均衡交通分配模型 [J]. 系统工程理论与实践, 2019, 39(07): 1771-80.
[19] 李浩, 丁中俊, 石琴等. 基于平均速度信息的城市交通路径诱导策略 [J]. 合肥工业大学学报(自然科学版), 2018, 41(05): 595-600.
[20] HAMZA-LUP G L, HUA K A, LE M, et al. Dynamic Plan Generation and Real-Time Management Techniques for Traffic Evacuation [J]. Ieee Transactions on Intelligent Transportation Systems, 2008, 9(4): 615-24.
[21] 杨阳. 基于系统动力学模型的交通政策模拟与实证研究 [D]; 南京大学, 2014.
[22] ZHAO L, SONG Y, ZHANG C, et al. T-GCN: A Temporal Graph Convolutional Network for Traffic Prediction [J]. Ieee Transactions on Intelligent Transportation Systems, 2020, 21(9): 3848-58.
[23] SUN S L, ZHANG C S, YU G Q. A Bayesian network approach to traffic flow forecasting [J]. Ieee Transactions on Intelligent Transportation Systems, 2006, 7(1): 124-32.
[24] 王晓全, 邵春福, 尹超英等. 基于ARIMA-GARCH-M模型的短时交通流预测方法 [J]. 北京交通大学学报, 2018, 42(04): 79-84.
[25] GIRAKA O, SELVARAJ V K. Short-term prediction of intersection turning volume using seasonal ARIMA model [J]. Transportation Letters-the International Journal of Transportation Research, 2020, 12(7): 483-90.
[26] 唐毅, 刘卫宁, 孙棣华等. 改进时间序列模型在高速公路短时交通流量预测中的应用 [J]. 计算机应用研究, 2015, 32(01): 146-9.
[27] 冮龙晖. 城市道路交通状态判别及拥挤扩散范围估计方法研究 [D]; 吉林大学, 2007.
[28] XU D-W, WANG Y-D, JIA L-M, et al. Real-time road traffic state prediction based on ARIMA and Kalman filter [J]. Frontiers of Information Technology & Electronic Engineering, 2017, 18(2): 287-302.
[29] 陈茜雯. 多维数据驱动的短时交通流预测模型 [D]; 西安电子科技大学, 2020.
[30] TANGARI G, CHARALAMBIDES M, TUNCER D, et al. Accuracy-Aware Adaptive Traffic Monitoring for Software Dataplanes [J]. Ieee-Acm Transactions on Networking, 2020, 28(3): 986-1001.
[31] LI W, WANG X, ZHANG Y, et al. Traffic flow prediction over muti-sensor data correlation with graph convolution network [J]. Neurocomputing, 2021, 427: 50-63.
[32] 郑义彬, 赖伟伟. 基于支持向量机的高速短时交通流量预测 [J]. 工程与建设, 2020, 34(02): 201-4.
[33] 刘晨. 基于WNN与GA-SVM的交通拥堵情况组合预测研究 [J]. 信息与电脑(理论版), 2021, 33(09): 54-7.
[34] CUI H, YUAN G, LIU N, et al. Convolutional neural network for recognizing highway traffic congestion [J]. Journal of Intelligent Transportation Systems, 2020, 24(3): 279-89.
[35] 曹堉, 王成, 王鑫等. 基于时空节点选择和深度学习的城市道路短时交通流预测 [J]. 计算机应用, 2020, 40(05): 1488-93.
[36] 许伦辉, 游黄阳. 基于特性和影响因素分析的短时交通流预测 [J]. 广西师范大学学报(自然科学版), 2013, 31(01): 1-5.
[37] 李雨倩. 基于天气和时空特征的城市道路交通拥堵态势评估与预测研究 [D]; 长安大学, 2021.
[38] 王传睿. 基于智慧交通的西安市交通拥堵治理研究 [D]; 长安大学, 2017.
[39] 商强. 基于机器学习的交通状态判别与预测方法研究 [D]; 吉林大学, 2017.
[40] SIDDIQUE C, BAN X. State-dependent self-adaptive sampling (SAS) method for vehicle trajectory data [J]. Transportation Research Part C-Emerging Technologies, 2019, 100: 224-37.
[41] WANG L, YAN X, LIU Y, et al. Grid Mapping for Road Network Abstraction and Traffic Congestion Identification Based on Probe Vehicle Data [J]. Journal of Transportation Engineering Part a-Systems, 2021, 147(5).
[42] ZHANG Z, LI M, LIN X, et al. Network-wide traffic flow estimation with insufficient volume detection and crowdsourcing data [J]. Transportation Research Part C-Emerging Technologies, 2020, 121.
[43] DUMBLIAUSKAS V, GRIGONIS V, BARAUSKAS A. APPLICATION OF GOOGLE-BASED DATA FOR TRAVEL TIME ANALYSIS: KAUNAS CITY CASE STUDY [J]. Promet-Traffic & Transportation, 2017, 29(6): 613-21.
[44] 于泉, 丰柱林, 徐红领等. 高速公路交通状态判别模型研究 [J]. 交通运输工程与信息学报, 2015, 13(02): 10-5.
[45] 杜崇. 高速公路实时交通状态判别方法研究 [D]; 北京交通大学, 2017.
[46] 樊超. 基于风险分析的危险货物道路运输优化研究 [D]; 大连海事大学, 2017.
[47] 祝付玲. 城市道路交通拥堵评价指标体系研究 [D]; 东南大学, 2006.
[48] 李明捷, 石荣. 基于集成DEMATEL-ISM的机坪容量影响因素分析 [J]. 数学的实践与认识, 2017, 47(21): 155-64.
[49] 赵晨晨. 低碳背景下电力需求响应效益评估研究 [D]; 华北电力大学, 2015.
[50] 周德群, 章玲. 集成DEMATEL/ISM的复杂系统层次划分研究 [J]. 管理科学学报, 2008, 11(02): 20-6.
[51] FENG X, LING X, ZHENG H, et al. Adaptive Multi-Kernel SVM With Spatial-Temporal Correlation for Short-Term Traffic Flow Prediction [J]. Ieee Transactions on Intelligent Transportation Systems, 2019, 20(6): 2001-13.
[52] 王快妮. 支持向量机鲁棒性模型与算法研究 [D]; 中国农业大学, 2015.
[53] 黄艳国. 城市道路交通拥堵机理及控制方法研究 [D]; 华南理工大学, 2015.
[54] 程铁信, 李奇, 赵颖. 考虑碳排放的城市交通拥堵定价模型及其仿真分析 [J]. 中国管理科学, 2016, 24(S1): 932-7.
[55] 徐曌, 欧国立. 交通拥堵收费的理论依据和政策分析 [J]. 中国工业经济, 2012, (12): 18-30.
[56] 杨浩雄, 张浩, 王晶等. 交通拥堵收费政策效应研究 [J]. 管理世界, 2013, (07): 174-5.
[57] 张彤, 侯幸, 吴昱. 效率、民意与车牌分配方式选择 [J]. 经济学(季刊), 2017, 16(02): 707-28.
[58] 朱永中, 宗刚. 时间价值偏好下北京交通拥堵收费设计研究 [J]. 中国软科学, 2014, (09): 67-77.
[59] 邵晓双, 谭德庆. 燃油价格波动、羊群效应与私家车厂商市场策略 [J]. 管理评论, 2018, 30(10): 119-27.
[60] 王雪清, 刘勇. 大气污染与环境规制对企业库存的影响机制研究——基于广东省工业企业数据 [J]. 管理评论, 2021, 33(12): 60-70.
[61] 李妍峰, 高自友, 李军. 基于实时交通信息的城市动态网络车辆路径优化问题 [J]. 系统工程理论与实践, 2013, 33(07): 1813-9.
[62] 王卉彤, 刘传明, 赵浚竹. 交通拥堵与雾霾污染:基于职住平衡的新视角 [J]. 财贸经济, 2018, 39(01): 147-60.
[63] 姬杨蓓蓓, 成枫, 张小宁. 城市中心区域毗邻私有停车场的共享停车位优化设置研究 [J]. 系统工程理论与实践, 2020, 40(11): 2934-45.
[64] 姬杨蓓蓓, 陈欣萌, 成枫. 考虑排放成本的共享电动汽车和共享停车位最优组合定价研究 [J]. 管理工程学报, 2022, 36(01): 134-45.
[65] 侯幸, 彭时平, 马烨. 北京上牌摇号与上海车牌拍卖政策下消费者成本比较 [J]. 中国软科学, 2013, (11): 58-65.
[66] 汪锋, 李善军. 机动车限牌的社会福利影响和最优配额研究 [J]. 管理科学学报, 2016, 19(12): 1-13.
[67] 佟琼, 王稼琼, 王静. 北京市道路交通外部成本衡量及内部化研究 [J]. 管理世界, 2014, (03): 1-9+40.
[68] 张敖木翰, 高自友. 突发事故下交通拥堵控制策略设计 [J]. 系统工程理论与实践, 2013, 33(05): 1307-17.
[69] 宗刚, 曾庆华, 魏素豪. 基于时间价值的交通出行方式选择行为研究 [J]. 管理工程学报, 2020, 34(03): 142-50.
[70] 梁若冰, 席鹏辉. 轨道交通对空气污染的异质性影响——基于RDID方法的经验研究 [J]. 中国工业经济, 2016, (03): 83-98.
[71] 杨浩雄, 李金丹, 张浩等. 基于系统动力学的城市交通拥堵治理问题研究 [J]. 系统工程理论与实践, 2014, 34(08): 2135-43.
[72] 张雪玫. 基于DEMATEL-ISM模型的ISO 9001认证有效性的障碍因素研究 [D]; 山东大学, 2021.
[73] 杨传明, 李晓峰. 多粒度复杂产品供应链低碳设计 [J]. 计算机集成制造系统, 2016, 22(07): 1717-25.
[74] 陆化普. 城市交通供给策略与交通需求管理对策研究 [J]. 城市交通, 2012, 10(03): 1-6.
[75] 张碧瑜. 浅谈呼和浩特市交通拥堵原因及缓解对策 [J]. 内蒙古公路与运输, 2018, (03): 53-4.
[76] 陈希韬. 道路交叉口的交通流量分析及其信号配时方法研究 [D]; 南京航空航天大学, 2016.
[77] 李歆. 多智能体交通拥堵自组织控制策略研究 [D]; 长沙理工大学, 2014.
[78] 张文佳, 王梅梅. 交通拥堵治理的空间与基础设施政策综述 [J]. 人文地理, 2021, 36(02): 20-6.
[79] 宁华晶. 城市道路施工期临时交通组织影响分析 [J]. 湖北理工学院学报, 2018, 34(05): 35-8.
[80] 王笑. 暴雨天气下城市快速路交通流特性研究 [D]; 北京交通大学, 2015.
[81] 龚彭钰, 邬群勇. 基于BF-SVR-GRU的短时交通流预测方法 [J]. 贵州大学学报(自然科学版), 2022, 39(02): 111-8.
[82] HAN L, ZHENG K, ZHAO L, et al. Content-Aware Traffic Data Completion in ITS Based on Generative Adversarial Nets [J]. Ieee Transactions on Vehicular Technology, 2020, 69(10): 11950-62.
[83] DUAN Z, YANG Y, ZHANG K, et al. Improved Deep Hybrid Networks for Urban Traffic Flow Prediction Using Trajectory Data [J]. Ieee Access, 2018, 6: 31820-7.
[84] MAENPAA H, LOBOV A, LASTRA J L M. Travel mode estimation for multi-modal journey planner [J]. Transportation Research Part C-Emerging Technologies, 2017, 82: 273-89.