基于改进系统动力学模型的公路客运量预测  

Forecast of highway passenger volume based on improved system dynamics model

作　　者：张兵[1] 周丹丹 周珣 张明阳 钟蒙 ZHANG Bing;ZHOU Dan-dan;ZHOU Xun;ZHANG Ming-yang;ZHONG Meng(School of Transportation Engineering,East China Jiaotong University,Nanchang 330013,Jiangxi,China;Jiangxi Comprehensive Transport Development Research Centre,Nanchang 330013,Jiangxi,China;Jiangxi Vocational and Technical College of Transportation,Nanchang 330013,Jiangxi,China)

机构地区：[1]华东交通大学交通运输工程学院,江西南昌330013 [2]江西省综合交通运输发展研究中心,江西南昌330013 [3]江西省交通职业技术学院,江西南昌330013

出　　处：《长安大学学报（自然科学版）》2023年第2期111-119,共9页Journal of Chang’an University(Natural Science Edition)

基　　金：国家自然科学基金项目(52162042,71961006);江西省教育厅科学技术研究项目(GJJ190331);江西省交通运输厅科技项目(2020H0053)。

摘　　要：为了确定公路客运的发展规模和规律,采用改进后的系统动力学模型对公路客运量进行预测。在传统的系统动力学(SD)模型中引入生态学中的种间竞争模型Lotka-Volterra(LV)理论,首先通过定性分析与客运需求相关的经济、社会、交通环境及其他客运方式竞争等方面的影响因素,构建因果回路图对各因素与公路客运量进行因果反馈关系分析,同时引入LV理论判定系统中模糊的动态关系。接着通过交通化改造后的LV判定方程,确定反馈关系因子之间的相互作用系数及系统流图中微分方程,对各影响因素子模型中影响因子与公路客运关系进行量化分析,对传统的系统动力学模型进行改进优化,建立具有时变性的动态LV-SD模型。最后以江西省公路客运量预测为例,根据公路客运量与各影响因素模型中影响因子之间的相关系数,选择地区生产总值、第三产业产值、居民消费水平、公路里程、民航客运量、铁路客运量、人口数量及城市化率等影响因子作为量化指标代入预测模型,用2013~2018年历史数据对模型进行有效性验证。结合城市发展情况,利用Vensim软件分析预测了基于LV-SD模型下江西省2019~2025年经济高、中、低增长3种情景下的公路客运量。结果表明:LV-SD模型预测精度高于SD模型,提高了近7.2%。2021~2025年江西省公路客运量呈现先增后减最后趋于平缓的趋势。In order to determine the development scale and law of highway passenger transportation,the improved system dynamics model was used to predict the highway passenger transportation volume.The Lotka-Volterra(LV)theory of ecology was introduced into the traditional system dynamics(SD)model.Firstly,qualitative analysis was made on the economic,social,transportation environment and other influencing factors related to passenger transport demand,a causal loop diagram was constructed to analyze the causal feedback relationships between various factors and highway passenger capacity,and the fuzzy dynamic relationship in the LV theory was introduced to determine the system.Then,the interaction coefficients between the feedback relationship factors were determined through the LV decision equation after the traffic transformation,and the differential equations in the system flow diagram were used to quantitatively analyze the relationship between the influencing factors and the highway passenger transport in the sub-models of each influencing factor.The system dynamics model was improved and optimized,and a time-varying dynamic LV-SD model was established.The forecast of highway passenger traffic in Jiangxi Province was taken as an example,according to the correlation coefficients between the highway passenger traffic and the influencing factors in the influencing factor model,the regional GDP,the output value of the tertiary industry,the consumption level of residents,the highway mileage,the civil aviation passenger traffic,railway passenger volume,population number and urbanization rate were selected as quantitative indicators to be substituted into the prediction model,and historical data from 2013 to 2018 were analyzed to verify the validity of the model.Combined with the development of the city,the Vensim software was used to analyze and predict the highway passenger traffic volume of Jiangxi Province under the three scenarios of high,medium and low economic growth from 2019 to 2025 based on the LV-SD model.The results

关 键 词：交通工程 需求预测 系统动力学模型 Lotka-Volterra理论 公路客运量 

分 类 号：U491.4[交通运输工程—交通运输规划与管理]