Archive

25 June 2023, Volume 23 Issue 3 Previous Issue    Next Issue

---

For Selected:

Download Citations EndNote Ris BibTeX

Toggle Thumbnails

Select

Strategies for High-quality Development of Air and High-speed Rail Transportation Under Goal of Carbon Peaking and Carbon Neutrality PENG Hong-qin, ZHANG Guo-wu 2023, 23(3): 1-5.  DOI: 10.16097/j.cnki.1009-6744.2023.03.001 Abstract ( )   PDF (3051KB) ( )   Under the background of carbon peaking and carbon neutrality, as a fundamental infrastructure sector of the national economy, transportation plays an important role in supporting the operation and development of the economy and meeting the high-quality travel needs of residents. To achieve carbon peaking and carbon neutrality, it is crucial to effectively utilize both air transportation and high-speed rail transportation by promoting a reasonable market share between them and optimizing the structure of the transportation system. The theme of the forum is "Strategies for the High-quality Development of Air and High-speed Rail Transportation under the Goal of Carbon Peaking and Carbon Neutrality". The forum analyzed the service levels of air transportation and high-speed rail transportation and proposed measures to improve them. It analyzed the challenges of energy conservation and emissions reduction faced by air transportation and proposed suggestion for low- carbon development pathways. It also introduced the low- carbon development history of railway transportation in typical developed countries or regions and analyzed the low- carbon development path and essential tasks of China's railway industry. The forum analyzed the characteristics of transportation mode choice and the impact of travel distance, travel time, and other influencing factors in business travel and personal travel. It investigated the factors, cumulative effects, and potential of different groups of travelers on the modal shift from air to rail. Related Articles | Metrics

Select

Urban Transportation Management from Perspective of General Spatial Equilibrium: Review and Trend XU Shu-xian, LIU Tian-liang, WANG Ting, XIAN Kai, HUANG Hai-jun, MA Shou-feng 2023, 23(3): 6-19.  DOI: 10.16097/j.cnki.1009-6744.2023.03.002 Abstract ( )   PDF (8507KB) ( )   PDF(English version) (660KB) ( 65 )   Urban transportation is the foundation of urban social and economic activities. Coordinated development of urban transportation and land use is of great practical significance to reduce traffic congestion, optimize the urban spatial structure, and realize sustainable urban development from the root of problems. With the development of urban society and the advancement of urbanization, as well as the low-carbon, green, and intelligent development trend of the transportation system, traditional urban transportation strategies purely focusing on supply or demand management cannot meet the needs of rapid urban development and the aspirations of the people to live a better life. Urban transportation management needs to focus on comprehensive governance of employment, housing, and transportation and dynamic equilibrium of supply and demand, to realize the coordinated and integrated development with urban spatial layout and land use. Based on the general spatial equilibrium theory, the literature on urban travel behavior analysis, travel demand management, transportation infrastructure supply, and supply-demand coupling strategies are systematically reviewed in this paper. Besides, the theoretical models, methods, and research problems in this area are also reviewed. It is found that the existing models cannot describe the dynamic process of urban development and the reality of China, and the related studies still focused on the traditional transportation management research problems. In the context of urban renewal, new territorial space planning systems, new transportation technologies and travel patterns, and big data, it is suggested that there is great potential for urban transportation management research from the perspective of general spatial equilibrium. It needs urgently a breakthrough in the corresponding theories and methods. Further research directions are proposed: the first is to analyze the influencing factors of residents' utility in the process of urban development, and put forward household utility decision-making theories and models under the integrated transportation and urban development; the second is to do activity/travel behavior analysis and management based on the data-driven and theory-driven methods; the third is to explore the impact of new technologies and modes of the transportation system on urban spatial structure and traffic characteristics, and study urban transport management issues for the era of digitization and intelligence; and the fourth is to explore theories and methods of urban space reshaping guided by transportation under urban renewal. References | Related Articles | Metrics

Select

Route and Speed Optimization for Green Intermodal Transportation Considering Emission Control Area WU Peng, LI Ze, JI Hai-tao 2023, 23(3): 20-29.  DOI: 10.16097/j.cnki.1009-6744.2023.03.003 Abstract ( )   PDF (5114KB) ( )   This study solves a new green high-seas multimodal transportation route and speed optimization problem considering emission control areas. This study first formulates a multi-objective mixed integer nonlinear programming model under different carbon emission policies and transforms the nonlinear model into an equivalent mixed-integer linear programming model according to the problem characteristics. To effectively solve the models, an improved adaptive genetic algorithm (IAGA) incorporating the characteristics of the problem is proposed, in which a customized multi-layer coding and decoding mechanism and an adaptive genetic evolution operator are proposed. Finally, a case study from the high-sea multimodal transportation system in China is conducted to demonstrate the viability of the proposed model and algorithm and a sensitivity analysis is also done for various time frames and low-sulfur fuel costs. The numerical experimental results show that: 1) Compared with a traditional genetic algorithm and the commercial solver Lingo, the improved adaptive genetic algorithm results in more satisfactory solutions and reduces total multimodal transportation costs by 5.2% and 3.7% . 2) Under the mandatory carbon emissions policy, changes in emission allowances typically do not affect the choice of the route made by an operator, but only affect whether the operator conducts transportation activities. Under the carbon tax policy, the overall cost of intermodal transportation is not significantly affected by the carbon tax price increase. Under the carbon trading policy, multimodal transport options with different emission allowances may be consistent. And 3) the shipping cost can lead to a direct proportion to the price of low- sulfur fuel, and adopting different ship speeds inside and outside the emission control areas can bring clear economic advantages. References | Related Articles | Metrics

Select

Incorporating Travelers' Self-presentation and Environmental Awareness into Heterogeneity Analysis of Travel Mode Choice Behaviors ZHOU Yu-yang, WANG Pei-yu, CHEN Yan-yan 2023, 23(3): 30-38.  DOI: 10.16097/j.cnki.1009-6744.2023.03.004 Abstract ( )   PDF (4306KB) ( )   The effectiveness of "green travel" guidance is influenced by external information and travelers' choice preferences. The heterogeneity of potential attribute categories of travelers needs to be considered. Self- presentation manifests itself as people influencing others' impressions of themselves by controlling the information that relates to them, which reflects the interaction of information and choice preferences. To quantify the influence of travelers' selfpresentation awareness and environmental awareness on mode choice behavior, 1382 valid samples were collected through a questionnaire survey. The latent class model (LCM) is used to classify the travelers into high selfpresentation group (18.23%), medium self-presentation group (20.26%) and low self-presentation group (61.51%). The results of the discrete choice model suggest that travelers pay more attention to travel time and the characteristics of the travel mode itself when making mode choices. The effect of travel expense on the high self-presentation group would be overestimated if mode characteristics are not considered. Travelers from the high self- presentation group have a strong preference for public transit only for short-distance trips, and their tendency for private cars is obvious for trips of 6 to 10 kilometers. The preference value of the low self-presentation group for cycling can offset to some extent the negative utility of excessively long travel times in short to medium distance trips. The construction of a mode choice model that considers travelers' heterogeneity can provide a theoretical basis for the government and related departments to formulate more coordinated and targeted regulation policies and public transportation operation strategies. References | Related Articles | Metrics

Select

Analysis and Improvement Strategy on Freeway Traffic Capacity in Foggy Weather QIN Yan-yan, XIAO Teng-fei, HE Zheng-bing 2023, 23(3): 39-47.  DOI: 10.16097/j.cnki.1009-6744.2023.03.005 Abstract ( )   PDF (4175KB) ( )   This paper studies the freeway traffic capacity in foggy weather. An improvement strategy for freeway traffic capacity in foggy weather is proposed based on vehicle-to-vehicle (V2V) communications. Firstly, a Gipps model calibrated in foggy weather was selected to describe the car-following behavior, and its spacing-speed function was derived to construct the analysis method of freeway capacity. Secondly, the influence of traffic capacity was analyzed from the perspectives of different foggy scenes and speed limit conditions. Under the influence mechanism by foggy weather, we conducted sensitivity analyses on driver reaction time Tn, the maximum brake deceleration bn of the following vehicle, and the estimated maximum brake deceleration bn-1 of the front vehicle by the follower. Finally, considering the influence of the driver reaction time and the braking deceleration on traffic capacity, a car-following control strategy for improving freeway capacity was proposed based on foggy V2V conditions. The results show that speed limit values of 80 km · h-1 and 100 km · h-1 will lead to the maximum traffic capacity under light fog (visibility of 150 meters) and heavy fog (visibility of 60 meters) conditions, respectively. Both conditions of light fog and heavy fog have the minimum traffic capacity when 60 km · h-1 is selected as the speed limit value. Compared with the speed limit of 40 km · h-1, the maximum traffic capacities corresponding to the speed limit of 80 km · h-1 in light fog and 100 km · h-1 in heavy fog increase by 21.83% and 9.68%, respectively. Additionally, the minimum traffic capacities corresponding to the speed limit of 60 km · h- 1 are reduced by 15.88% and 4.61% under light fog and heavy fog conditions, respectively. Traffic capacity improves with the decrease of the driver reaction time and it is positive for capacity improvement when bn-1> bn. The proposed control strategy can effectively improve the freeway capacity in foggy weather, and the control strategy has a significant effect on the improvement of traffic capacity with a confidence level of 95%. The average improvement percentage is 44.22% under different foggy scenes and speed limit conditions. References | Related Articles | Metrics

Select

Influence Mechanism of Rainfall on Traffic State of Urban Road Network WANG Dian-hai, HUANG Li-sha, ZENG Jia-qi, CAI Zheng-yi 2023, 23(3): 48-55.  DOI: 10.16097/j.cnki.1009-6744.2023.03.006 Abstract ( )   PDF (4212KB) ( )   PDF(English version) (1274KB) ( 65 )   It is a common perception that urban roads are prone to experience traffic congestions in rainy days, but is it true that rainfall will inevitably cause traffic congestion? This paper investigates this mechanism from the theoretical view and actual data analysis. It was found that drivers tend to drive conservatively in rainy days, which increases the saturation headway. Based on this hypothesis, a signal-influenced urban road traffic flow model is established, and the changes in average travel speed under different rainfall scenarios are analyzed theoretically. The saturation headway under each rainfall level obtained by using video data statistics verifies the hypothesis that rainfall increases the saturation headway. Simplifying the data of the main road network in Xiaoshan District, Hangzhou, numerical simulations were conducted and the proposed model was verified through VISSIM road network simulation. The results show that the effect of rainfall on the average speed of the road network is negative. When the traffic demand is small, the negative impact of rainfall is almost negligible; as the rainfall level increases, the road network with high traffic demand is more likely to be oversaturated, and the average travel speed of the road network will experience a steep drop and the traffic condition deteriorates rapidly. Taking the maximum demand of the road network as an example, compared with the non-rainfall weather, the average speed of the road network decreases by 8.63% in the light rainfall scenario, 16.51% in the medium rainfall, 23.43% in the heavy rainfall, and 24.94% in the rainstorm. The VISSIM simulation results fit well with the theoretical values of the model, which verifies the effectiveness of the model. Since the traffic demand during the peak hours is much larger than that during the off-peak hours, the rainfall during the peak hours will bring a greater negative impact on the traffic status of the urban road network, which requires more attention from the traffic management. References | Related Articles | Metrics

Select

Optimizing Integrated Layout of Bi-modal Feeder Buses Based on User Equilibrium CHEN Jian, LI Zhao-kang, ZHOU Tao, LIU Ke-liang, SUN Qin-mei 2023, 23(3): 56-65.  DOI: 10.16097/j.cnki.1009-6744.2023.03.007 Abstract ( )   PDF (4506KB) ( )   The feeder bus is an important component of the urban public transit system, serving as the solver of the "first/last mile" problem from/to the rail transit. In our integrated feeder bus system, demand-responsive buses are proposed to serve the peripheral areas where passengers may have long walking time to access fixed-route bus stops. In designing the integrated feeder bus system, a proper partition of the service area of different feeder bus modes was described by an analytical model that balances the travel time by different modes based on user equilibrium, and the genetic algorithm was employed to solve the travel time minimizing problem that finds out the optimal partition scheme among different equilibrium conditions. A case study based on the Chongqingdong Railway Station and its adjacent area was conducted to verify the model and the algorithm, and the results proved that the integration of bi-modal feeder buses can reduce passengers' trip time by 32% and above without deploying more vehicles comparing with the single fixed-route mode. The result also showed that the energy consumption per capita will also rise by 83% with the deployment of a demand-responsive mode. References | Related Articles | Metrics

Select

Lifecycle Optimization Strategy of Single-line Electric Bus Operational Planning Considering Battery Degradation XIE Dong-fan, LUO Yu-chao, ZHOU Guang-jing, YU Ya-peng, WANG Yong-xing, ZHAO Xiao-mei 2023, 23(3): 66-75.  DOI: 10.16097/j.cnki.1009-6744.2023.03.008 Abstract ( )   PDF (4546KB) ( )   The degradation of battery capacity in the lifecycle of battery electric buses will affect long-term operational planning. To solve this problem, this paper develops an optimization model for the whole life period operational planning of battery electric buses considering battery capacity degradation. The goal is minimizing the operating costs, energy consumption costs, and battery degradation costs of public transport vehicles. The model divides the entire service life of the electric bus fleet into multiple periods. Considering the impact of battery degradation costs, the study optimizes the vehicle scheduling and the number of vehicles of each period separately, and obtains the battery replacement caused by battery capacity degradation in each period. A column generation algorithm is designed to solve the model, and a bus line in Beijing is selected for case study. The results show that the charging cost and battery degradation cost are respectively reduced by 6.5% and 17.5% due to considering battery capacity degradation, compared to the method that are not considering battery degradation. The influence of model parameters on cost is also discussed by sensitivity analysis of influencing parameters, which can provide optimization suggestions for operational planning in the whole life period of electric buses. References | Related Articles | Metrics

Select

Dynamic Pricing of Airline Upgrade Service Considering Revenue Management ZHAO Gui-hong, MA Chen-ao, ZHAO Qiao-tong, LI Jian-fu 2023, 23(3): 76-84.  DOI: 10.16097/j.cnki.1009-6744.2023.03.009 Abstract ( )   PDF (4885KB) ( )   In order to achieve a win-win situation for passengers and airlines with upgraded services, a utility model of equilibrium between passengers and airlines is constructed based on the utility equilibrium theory. The Weibull distribution function is used to describe the probability of passengers accepting a specific price. And it is embedded into the equilibrium utility model to obtain the dynamic pricing model of upgrade services. To make the model more realistic, price constraints and seat number constraints are added. And it is proved that the optimal price of upgrade service must be within the price constraint range for profit maximization. Finally, this paper conducts a numerical study on the upgrading of a single class and multiple classes and obtains the optimal price of upgrade services for different classes under the two conditions. The benefits of dynamic pricing strategies for airlines are also calculated. The results show that, compared with the traditional static pricing, if airlines take passenger demand utility into account, it can reduce the vacancy rate of airlines' high-end class seats and improve their revenue, but also enable passengers to enjoy the upgrade service at a lower price. In addition, compared with single-class upgrade services, multi-class upgrade services can bring more benefits to airlines. The example analysis shows that the revenue of airlines selling two-class upgrade services is 39.12% higher than that of selling single-class upgrade services. References | Related Articles | Metrics

Select

Freight Price Conversion and Road/Rail Competition Model Based on Revealed Preference Data LIU Hao, SHEN Jia-qi, ZHANG Rong, WU Hao-tian, ZHANG Zhuo-wei 2023, 23(3): 85-93.  DOI: 10.16097/j.cnki.1009-6744.2023.03.010 Abstract ( )   PDF (4383KB) ( )   To further promote the modal shift from road to rail, rail transport enterprises need to grasp the comparable prices of the same caliber of competitive modes and fully understand the freight mode choice behavior. This paper proposes a method for converting freight prices among various loading and transport modes, and solves the missing data problem on alternative attributes when adopting RP (revealed preference) data for discrete choice modeling. The freight mode choice behavior model was developed by effectively combining multi-source RP data through an improved PPS (probability proportionate to size sampling) method. The results show that the model can successfully predict more than 90% of the observations. Light cargo's VOT (value of time) is higher than its heavy cargo counterparts. The derivation and calculation of price elasticities show that increasing road prices can lead to a more significant increase in rail market share than reducing rail prices. Reducing current rail prices can increase rail transport income. When rail prices fall to the optimal pricing of the income maximization goal, it will not only bring a significant increase in rail share, volume and income, but is also expected to yield some carbon reduction benefits. References | Related Articles | Metrics

Select

Price Analysis of Commercial Real Estate Around Urban Rail Transit Station Based on Spatial Dubin Model DU Peng, HE Si-le 2023, 23(3): 94-100.  DOI: 10.16097/j.cnki.1009-6744.2023.03.011 Abstract ( )   PDF (3603KB) ( )   To investigate the influence of urban rail transit stations on the surrounding commercial real estate prices, this paper considered the spatial dependence between samples, introduced spatial variables to build a Spatial Econometrics Model (SEM) based on the traditional Hedonic Price Model (HPM) and evaluated the training and fitting effects of different models. The combination form of station entrance and exit was introduced in the study to represent accessibility with the distance variable. The relationship between the accessibility of urban rail transit stations and the price of commercial real estate was also analyzed. The fitting training and effect comparison of relevant models show that the fitting effect of the Spatial Dubin Model (SDM) is better than the traditional HPM and other SEMs. Compared with the accuracy of the regression effect of the control group, the mean square error of the basic group considering the station entrance and exit combination form decreased by 21.74%. The introduction of the combination form variable improved the accuracy of the regression result of the SDM. The case analysis in Beijing and Shanghai show that the price of commercial real estate is positively correlated with the tightness of entrance and exit combination form. Compared with the general type, the prices of commercial real estate with the combination form of docking type, attaching type and integrating type were respectively increased by 51.20%, 45.02% and 43.10%. References | Related Articles | Metrics

Select

Two-layer Model to Distinguish Urban Motorized Travel Mode Based on Mobile Phone Signaling Data GUO Yu-dong, YANG Fei, ZHOU Tao, YAO Zhen-xing, ZHANG Chu-liang, WEI Yin-cheng 2023, 23(3): 101-109.  DOI: 10.16097/j.cnki.1009-6744.2023.03.012 Abstract ( )   PDF (4267KB) ( )   PDF(English version) (821KB) ( 64 )   It is always difficult to apply the mobile phone signaling data in the actual urban complex travel environment and also challenging to distinguish the motorized travel mode under dense road networks. This paper proposes a twolayer model considering accurate path fitting and multidimensional spatio-temporal characteristics. At the level of travel path identification, S- G filtering can effectively smooth signaling data fluctuation relative to the actual travel path. The linear interpolation algorithm can fill in the time and space gaps. At the level of travel mode recognition, the key factors are mined, including the similarity of the recognized travel path, travel time similarity, acceleration, and wavelet velocity. The K-nearest neighbor algorithm is used to identify the travel modes (by car or bus). The results show that the proposed method can effectively identify the bus and car in the dense urban road network, and the accuracy rates can reach 88.29% and 82.28% , respectively. Under different travel distances, travel time, congestion conditions, road classes, road types, and path similarity, the proposed method is better than the existing algorithms, such as random forest, in the accuracy rate. The research supports the accurate recognition of travel characteristics based on mobile phone signaling data. It also provides an essential basis for road planning, construction, and public transit network planning. References | Related Articles | Metrics

Select

Variable Speed Limit Control Based on Improved Dueling Double Deep Q Network Under Mixed Traffic Environment HAN Lei, ZHANG Lun, GUO Wei-an 2023, 23(3): 110-122.  DOI: 10.16097/j.cnki.1009-6744.2023.03.013 Abstract ( )   PDF (7199KB) ( )   Existing variable speed limit (VSL) control strategies suffer from poor flexibility, slow response time, and a heavy reliance on the compliance rate and traffic flow prediction models. Additionally, it is difficult to achieve effective control by relying solely on variable message signs (VMS) to post speed limits to drivers in the mixed traffic environment where connected and automated vehicles (CAVs) and human-driven vehicles (HDVs) coexist. To this end, this paper proposes a VSL control strategy based on the improved dueling double deep Q network (IPD3QN) under the mixed traffic flow environment, i.e., IPD3QN-VSL. This strategy integrates the ability of deep reinforcement learning to automatically adapt to complex environments without establishing traffic flow prediction models, and the advantages of controllability of CAVs. Firstly, the prioritized experience replay mechanism is introduced into the dueling double deep Q network (D3QN) framework of deep reinforcement learning to enhance the convergence speed and parameter update efficiency of the network. Meanwhile, a novel adaptive 𝜀-greedy algorithm is proposed to solve the problem of balance between exploration and utilization in D3QN’s learning process. The proposed VSL control strategy aims to minimize the total time spent (TTS) of vehicles on the freeway section. Real-time traffic data and speed limits within the previous control cycle are used as inputs to the IPD3QN algorithm. Then, a reward function is constructed to guide the algorithm to generate the dynamic speed limit value executed in the VSL control area. Finally, the effectiveness of the IPD3QN-VSL control strategy is verified under different conditions and compared to no control, feedback control, and D3QN-VSL control in terms of control performance. Analysis results indicate that the proposed strategy can achieve remarkable control performance at a 30% penetration rate and effectively improve bottleneck traffic efficiency and reduce the spatiotemporal range of traffic congestion in both stable and fluctuating demand scenarios. Compared to the suboptimal D3QN-VSL control, the proposed strategy can achieve improvements of 14.46% and 10.36% on TTS in stable and fluctuating traffic demand scenarios, respectively. References | Related Articles | Metrics

Select

Large Multilane Turbo Roundabout Lane Level Signal Control Method GE Hui-min, ZANG Wen-kai, DONG Lei, ZHOU Li-jun 2023, 23(3): 123-132.  DOI: 10.16097/j.cnki.1009-6744.2023.03.014 Abstract ( )   PDF (4971KB) ( )   The large multilane roundabout normally experiences frequent vehicle lane changes and conflicting traffic flow in different directions at the entrance and exit. This paper proposes a lane level left-turn two-step signal control method combining the unique advantages of the turbo roundabout to control vehicle lane changes. First, through the analysis of the vehicle arrival-departure curves of the first and second stop lines, a signal timing model is established with the minimum average vehicle delay as the target and the signal period, saturation, and green light duration as the constraints. Then, the effect of different control methods, different number of lanes and center island radius on the average vehicle delay is analyzed through modeling. The results show that the lane level left-turn two-step signal control method has a lower delay level compared to the conventional left-turn two-step signal control method, which can reduce the vehicle delay by 6.13 seconds on average. As the number of lanes and traffic volume increase, the proposed method has a lower average delay growth. The larger the radius of the center island, the greater the average vehicle delay, but the increase is marginal. Using the uncontrolled large multilane roundabout at Mengxi Square in Zhenjiang City as an example, VISSIM simulation was performed, and the result shows that the proposed signal control method reduces the average vehicle delay by 8.91% and the stopping rate by 6.85% compared with the traditional left-turn two-step signal control method. This signal control method combining with the turbo roundabout can effectively improve traffic safety and efficiency at large multilane roundabouts. References | Related Articles | Metrics

Select

Gradient Prompting Strategies of Roadside Signs on Curve Segments FU Rui, XU Qing-jin, GE Zhen-zhen, CHEN Tao, WU Fu-wei, YUAN Wei 2023, 23(3): 133-144.  DOI: 10.16097/j.cnki.1009-6744.2023.03.015 Abstract ( )   PDF (6138KB) ( )   Curve segments on roadway are observed with higher crash density than tangent sections due to improper vehicle speed control to navigate through the curve segment. To improve the safety, this paper focuses on the setting of roadside warning signs on curve segments and develops a gradient prompting strategy for these signs. A virtual scene of mountainous curves was established through a high-fidelity driving simulator platform. Three warning signs, including a sharp curve sign, a speed limit sign (SLS), and a variable message sign (Variable message sign, VMS), are determined as the major warning signs. Experiments are conducted on a traditional prompting strategy (sharp curve sign), a second-level gradient prompting strategy (sharp curve sign and SLS), and a third-level gradient prompting strategy (sharp curve sign, SLS, and VMS). Vehicle motion, eye-tracking data, and subjective questionnaires were collected from 33 participants. Parameters such as average speed, longitudinal acceleration, speed compliance rate, single gaze duration, gaze advance distance, reaction time, and subjective evaluation were used as evaluation indicators. One-way analysis of variance (ANOVA) and post-hoc tests were used to compare the effects of different prompting strategies on driver's attention and speed control. The results show that compared with traditional strategy, the gradient prompting strategies can effectively improve driver attention, reduce visual recognition distance and reaction time for warning signs, and help drivers to slow down before entering the curve and maintain a low speed while driving on the curve. This study provides a basis for the continuous setting of roadside safety warning signs on curve segments. References | Related Articles | Metrics

Select

Data-driven Speed Compound Control of High-speed Train HOU Tao, TANG Li, NIU Hong-xia 2023, 23(3): 145-152.  DOI: 10.16097/j.cnki.1009-6744.2023.03.016 Abstract ( )   PDF (3835KB) ( )   Targeting the topic of high-speed train speed control amid external disturbances, a modeling method of a high-dimensional linear model of high-speed trains based on the Koopman operator is proposed, and a composite controller (ESO-K-MPC) combining extended state observer (ESO) and model predictive control (K-MPC) based on Koopman operator is designed. Firstly, the dynamic high-dimensional linear model of a high-speed train with dynamic nonlinear characteristics was developed by approximating the infinite-dimensional linear Koopman operator using the extended dynamic mode decomposition algorithm. Secondly, the model predictive control was implemented, and the extended state observer was built to estimate and correct the system's total disturbance. The ESO-K-MPC-based highspeed train speed control system was created, together with the controller and control algorithm. Finally, combined CRH3 train parameters with the actual line data of Huashan North Station, Xi'an North Station of Zhengzhou-Xi'an High-speed Railway, the designed control methods and algorithms were simulated and studied without disturbance and white noise interference respectively. The simulation indicates that the displacement and velocity forecast accuracy of high-speed train modeling based on Koopman is increased by 83.86 and 87.40%, respectively, when compared to the linear state space model. ESO-K-MPC can accurately estimate and compensate interference during high-speed train operation, the control output curve practically overlaps the expected curve, and high precision tracking of the predicted curve during train operation is achieved. References | Related Articles | Metrics

Select

Travel Mode Prediction Model Based on Ensemble Learning and Resident Attribute Data SU Yue-jiang, WEN Hui-ying, YUAN Min-xian, WU De-xin, QI Wei-wei 2023, 23(3): 153-160.  DOI: 10.16097/j.cnki.1009-6744.2023.03.017 Abstract ( )   PDF (4037KB) ( )   The traditional model of utility theory aims to maximize the utility on a global level. However, the traditional method may disregard some influencing factors of residents' travel mode and the mutual compensation of utilities. The individual preference, difference and compensation of individual travel utility could be expressed effectively by resident travel mode prediction model based on integrated learning. The problem of different utility representation was solved due to different travel characteristics on different groups. First, the travel feature vector of individual residents was constructed by analyzing the factors of personal attributes, travel attributes and environmental attributes. The difference in the perception of utility of different individuals to different influencing factors was considered. The travel mode prediction model based on ensemble learning method was then developed. Taking Guangzhou as an example, the prediction results were evaluated comprehensively by four parameters: accuracy rate, precision rate, recall rate and F1 value. Due to the similarity of some travel mode attributes, the hierarchical LightGBM models were developed and optimized, including all-mode, slow traffic, public transport and individual motorized traffic. The results were compared with other models, including GBDT, Random Forest, and hierarchical LightGBM. The results indicated that the weighted average accuracy of LightGBM model to predict travel mode is respectively 87% and 78%. The highest accuracy of slow traffic mode is 92% in the big class model, and the highest accuracy of walking mode is 89% in the subdivided class model. The model would be suitable for urban level residents' travel mode prediction and play an important role to predict the residents' travel mode. References | Related Articles | Metrics

Select

Risk Tendency Classification of Drivers of Road Transport Vehicles of Hazardous Materials LI Si-xian, XING Zeng-yong, QIAN Da-lin, LI Peng-cheng, YUAN Meng 2023, 23(3): 161-173.  DOI: 10.16097/j.cnki.1009-6744.2023.03.018 Abstract ( )   PDF (7137KB) ( )   In order to strengthen the safety risk source control of road transport vehicles of hazardous materials (RTVHM), this study fully excavates the trajectory big data, monitoring data, and other multi-source heterogeneous traffic data to study the risk tendency classification of RTVHM drivers. Based on the driving behavior patterns and environmental characteristics contained in the widely available GPS trajectory data, this study introduces the concept of time-varying random volatility and extracts five measures of speed volatility to construct an attribute feature set that characterizes driving style. Coupled with the characteristics of behavior inhibition control, cognitive inhibition control, and physiological load, risk tendency classification indexes are established for RTVHM drivers. The weight of each index is calculated based on the CRITIC method, and the four kinds of attributes describing the RTVHM drivers are scored by the VIKOR algorithm. The risk tendency classification model based on the K-medoids clustering algorithm is established. The results showed that using the classification model, RTVHM drivers were divided into four categories. Among them, drivers with aggressive driving styles and weak behavior inhibition control showed greater speed fluctuations and more vehicle control alarms in face of congested roads and bad weather. Drivers with weak cognitive inhibition control had more distracted duration, allocated more attention to distractions, and diverted attention more frequently between distraction objects and road conditions ahead. Fatigue drivers showed more fatigue alarms and overtime driving alarms, and bear greater physiological loads. The research results can provide a theoretical basis for the identification and risk assessment of the main risk tendency types of RTVHM drivers. References | Related Articles | Metrics

Select

Spatial Transplantation for Modeling of Freeway Traffic Crash Risk Based on Dynamic Traffic Flow YANG Yang, HE Kun, WANG Yun-peng, CHEN Yao, YUAN Zhen-zhou 2023, 23(3): 174-186.  DOI: 10.16097/j.cnki.1009-6744.2023.03.019 Abstract ( )   PDF (7649KB) ( )   This research aims at exploring the influence of multi-scale data sets on real-time traffic crash risk modeling towards freeways, and realizing the real-time risk model transplantation for freeways with spatial differences. First, multi-scale data sets were constructed via extracting freeway sections with different detector characteristics: highresolution data sets, small-sample data sets, low-resolution data sets, and data sets with spatial differences under the same scale conditions (both are high precision and large sample size); Furthermore, the influence of various sample sizes on the prediction performance of the traffic crash risk model was quantified by Bayesian Logistic regression, and statistical methods and machine learning methods were introduced to model the high and low resolution data sets respectively. Finally, the real- time traffic crash risk migration model based on the Bayesian updating method was established, and the freeway real-time crash risk prediction model was spatially transplanted, simultaneously its reliability was verified. The results show that: the performance of the model based on Bayesian Logistic regression improves with the increasing sample size; under the condition of high resolution data, the Area Under Curve (AUC) values of the Bayesian Logistic regression model and Random Forest- Support Vector Machine (RF- SVM) model are 0.092 and 0.037 higher than those under the condition of low resolution data, respectively; in the spatial migration with various data resolution, the AUC value of the low-resolution road segment model can be improved from 0.645 to 0.714 by the Bayesian updating method, and in the spatial migration with the same data scale, the AUC value of the updated road segment model can be improved from 0.737 to 0.751 by applying the Bayesian updating method. The conclusions indicate that: the model from the freeway section with a big sample size can boost the model classification accuracy but cannot significantly improve the performance of the prediction model, the results have some fluctuations, while the model from the freeway section with high data resolution can have higher accuracy of classification and prediction of the model; statistical methods have more advantages in model interpretation, and machine learning has better prediction performance under the condition of low resolution data; The Bayesian updating model can improve the accuracy of model spatial transplantation to a certain extent. References | Related Articles | Metrics

Select

Spatial Patterns of Nonlinear Effects of Built Environment on Beijing Subway Ridership HE Peng, LI Wen-xi, LI Yan, XU Qi 2023, 23(3): 187-194.  DOI: 10.16097/j.cnki.1009-6744.2023.03.020 Abstract ( )   PDF (4582KB) ( )   Machine learning is a nonlinear approach to examine the effects of built environment on station-level ridership of urban rail transit. But existing interpretation approach are not able to analyze the spatial patterns of results derived from machine learning models. This paper first utilizes multi-source location-based big data to quantify the indicators of built environment of urban rail transit, and then uses the indictor to analyze the spatial distribution of nonlinear effect of built environment on urban rail transit passenger flow based on the extreme gradient boosting incorporating local regression technique. The case study of Beijing subway shows that the nonlinear effects of built environment factors on egress flow of urban rail transit is significantly different. The top three indictors of built environment, including employment density, public accessibility, and mixed land use, account for 42.51% of total importance of variables. The spatial distributions of indictor of these three built environment factors show significant spatial heterogeneity, indicating the spatial non-stationary relationship between the passenger flow and built environment. The nonlinear spatial patterns indicate that the development of urban rail transit stations should not only adopt differentiated policies and strategies in different areas, but should also determine reasonable lower limit of resource to activate threshold effect of built environment, which would help to increase the ridership of urban rail transit. References | Related Articles | Metrics

Select

An Optimization Method of Train Scheduling for Urban Rapid Rail Transit Based on Flexible Train Composition Mode TAI Guo-xuan, HUANG You-neng, LI Chun-chi, WANG Xi 2023, 23(3): 195-203.  DOI: 10.16097/j.cnki.1009-6744.2023.03.021 Abstract ( )   PDF (4392KB) ( )   The spatial-temporal distribution of passenger demand in urban rapid rail transit is generally unbalanced due to long mileage and periodic operation of mass transit. This paper proposes an optimization method of train scheduling for urban rapid rail transit based on flexible train composition mode. First, the train operation scenario based on the proposed flexible train composition mode is described. Then, an optimization model of train scheduling for urban rapid rail transit based on flexible train composition mode is established to minimize the average passenger waiting time and the car kilometers. The model takes the arrival and departure time of trains at each station and the flexible train composition plan including the selection of train types in each section of the transit line as the decision variables. And a three-phase algorithm is designed to solve the model. The actual data of an urban rapid rail transit line is taken for the case study. The results indicate that compared with the regular timetable and fixed train composition mode, the timetable based on flexible train composition mode can reduce the average passenger waiting time by 14.80% and the car kilometers by 41.09% , which has been proved to be an effective operation strategy to deal with the unbalanced passenger demand of urban rapid rail transit. References | Related Articles | Metrics

Select

An Optimization Model of Whole Chain from Front-end Collection to Terminal Distribution of Cold Chain Less-than-truckload Logistics XIE Ru-he, HE Jia-wen, ZOU Yi-feng, LI Qing-ting, LIU Zi-ling, LIAO Jing 2023, 23(3): 204-213.  DOI: 10.16097/j.cnki.1009-6744.2023.03.022 Abstract ( )   PDF (4857KB) ( )   To address the complex operations and poor connection of the cold chain less- than- truckload logistics, as well as the low circulation rate, a two-stage optimization model was developed. In the first stage, the Minimum Average Dispersion model was constructed to identify the optimal cargo collection points. In the second stage, a path optimization model was established to minimize the overall cost of cold chain transportation from the trunk to the end distribution. With a numerical experiment based on the Z company, the K-means clustering algorithm and the genetic algorithm are used for the simulation and verification of the model's effectiveness. The optimal cargo collection points obtained from the experimental results reduced the average dispersion by 18.8%, making the distribution of suppliers more concentrated than cargo collection points, thereby improving the efficiency of front-end cargo collection. The optimized distribution plan had zero time window penalty cost, which improves customer satisfaction. Among the four transportation scenarios, the total cost of full cold chain transportation was the lowest, with a 69.4% reduction in total cost compared to room temperature transportation. These results indicate that selecting cargo collection points based on the location of front-end suppliers and the amount of cargo, as well as maintaining suitable cold chain conditions throughout the transportation process, can effectively optimize the overall cost of cold chain less-than-truckload logistics for logistics companies. References | Related Articles | Metrics

Select

Optimization Method of Aircraft Regular Check Task Scheduling Based on Combinatorial Optimization HU Xiao-bing, ZHAO Yu-bo, WANG Rui-xin, WU Zhi-dong, ZENG Zhi-hong 2023, 23(3): 214-222.  DOI: 10.16097/j.cnki.1009-6744.2023.03.023 Abstract ( )   PDF (4325KB) ( )   This paper focuses on the Aircraft Regular Check Task Scheduling Problem (ARCTSP). With the goal of minimizing the difference between the actual and expected daily usage of different types of resources, this paper establishes a scheduling optimization model and proposes a heuristic algorithm based on serial scheduling. In the heuristic algorithm, the priority order of task scheduling is determined by establishing a task scheduling candidate set and designing task priority rules to meet complex constraints, and the set greedy strategies are used to schedule tasks and improve search efficiency. The task priority rules and greedy strategies are combined to effectively achieve the optimization goal. In addition, based on the mathematical description of the ARCTSP, an Integer Linear Programming (ILP) model is proposed and solved. The algorithm was tested using real aircraft regular check data from actual production as an experimental object. The experimental results verified the effectiveness of the algorithm. Compared to the current scheduling methods applied to actual production, the proposed algorithm improved the solution quality by more than 64.55% with different human resources allocations. References | Related Articles | Metrics

Select

Containers Allocation and Recycle Transportation Problem with an Improved Branch-and-price Algorithm WEN Bin-bin, HE Shi-wei, CHI Ju-shang, JIN Fu-cai 2023, 23(3): 223-234.  DOI: 10.16097/j.cnki.1009-6744.2023.03.024 1 Abstract ( )   PDF (5747KB) ( )   Considering the drawback of high cost and poor flexibility of equipment transportation under containerization of logistics, a transportation mode of circular allocation and recovery of containers was proposed. In this transportation mode, the container transportation problem belongs to Unpaired Supply-Demand and Split Pickup and Delivery Vehicle Routing Problem (USDSPDVRP). With an accurate problem description, a two-stage model was proposed. And an improved branch pricing algorithm was designed to solve the USDSPDVRP model. The model is decomposed into a main problem of resolving the number of transport schemes used and a sub- problem model of generating new transport schemes based on the results of the main problem by Dantzig-Wolfe decomposition. To solve complex sub-problems efficiently, an appropriate path reduction strategy is designed. Results show that the proposed algorithm is effective. Compared with the result of the Gurobi solver, the average computational time is reduced by 83.9% , and the average gap between the feasible solution and the optimal solution is 2.8% . At the same time, a comparative analysis of cyclic dispatch transportation and direct transportation proves that the cyclic dispatch transportation can reduce the cost and the number of vehicles, and increase the load rating of vehicles. References | Related Articles | Metrics

Select

Optimization of Slot Allocation for Ocean Container Liner with Booking Cancellation Fees YANG Hua-long, WANG Hao, XING Yu-wei 2023, 23(3): 235-242.  DOI: 10.16097/j.cnki.1009-6744.2023.03.025 Abstract ( )   PDF (3939KB) ( )   This paper focuses on the slot allocation for ocean container liners in the case of high probability of customers' online booking cancellation. A new classification method was proposed with two types of customers in the spot market according to whether the booking cancellation fee is charged or not. Two types of forecasting methods of customers' booking cancellation probability in the spot market were designed based on the freight index in different periods. Then, a nonlinear programming model of two-stage slot allocation and pricing in spot market was established with the goal of maximizing the revenue of the shipping companies by combining the probability of the goods not in place. The model was analyzed and verified through the examples of Paclflc China North 1 route of Orient Overseas Container Line (OOCL). The results show that the slot allocation for ocean container liner with booking cancellation fees can increase the revenue of shipping companies by more than 10% compared to callentation without the fees. There is a reasonable value for the long-term contract customer confidence parameter set by the shipping company. The total revenue of the shipping company will decrease with the increase of customers' sensitivity coefficients of freight rate in spot market. There is an approximate linear relationship between the booking cancellation fee and the revenue of the shipping company in spot market. The research conclusion provides a useful reference for shipping companies to make shipping slot allocation and pricing decisions. References | Related Articles | Metrics

Select

Ship Schedule Coordination Considering Horizontal and Vertical Alliance Strategy Under Uncertainty ZHAO Xu, CHEN Jia-qin, HUANG Rui 2023, 23(3): 243-252.  DOI: 10.16097/j.cnki.1009-6744.2023.03.026 Abstract ( )   PDF (4564KB) ( )   In order to explore the relationship between the container ship schedule and the revenue of both the port and the shipping company when a horizontal and vertical alliance is formed between the ports and the shipping company, the optimal coordination of the shipping schedule of a certain shipping company's multiple routes through the port of call is investigated with uncertainty in the ship's time in the port. A mixed integer nonlinear programming model of container ship scheduling coordination under the port horizontal and vertical alliance is constructed with the two objectives of maximizing the revenue of shipping companies and ports, and an improved NSGA- algorithm is designed to solve it. Take KMTC's multiple routes connecting with Shanghai Port, Zhoushan Port, and Nantong Port as an example, numerical simulations are conducted to verify the model and algorithm under different circumstances. The results show that compared with the circumstances without considering the inter-port alliance, the revenue of shipping companies under the port horizontal and vertical alliance increases by 6.2%, and the revenue of the port side increases by 25.9%. Shipping companies, planned port, and cooperative ports all benefit from the alliance. Under different routes or arrival and departure times, the model can still obtain a stable shipping schedule. The research shows that the model and algorithm can provide a reference for shipping companies and ports to make reasonable shipping schedules under different circumstances. References | Related Articles | Metrics

Select

Simulation and Study of Location and Size of One-way Channel Buffers for Large Marine Crude Oil Terminals FENG Xue-jun, WANG Su-yang, XU Bo, ZHANG Yan, HUANG Jian-liang, GU Wei-hua 2023, 23(3): 253-264.  DOI: 10.16097/j.cnki.1009-6744.2023.03.027 Abstract ( )   PDF (6529KB) ( )   The critical bottleneck affecting a terminal's oil throughput capacity is shifting from the limited number of berths to the channel connecting the berths to the anchorage area. This is especially true for the terminals with a oneway channel where large oil tankers can navigate through the channel only at high tides. Buffers can be installed at those terminals to significantly improve the one-way channel's capacity. This paper examines the buffer's location and size choices for a crude oil terminal with a one-way channel. A discrete-event simulator in-house was developed using Python to explicitly and model the buffer's location and size choices. The simulator accounts for a sufficient level of realism, including the tidal time windows for large tankers and the daytime operation constraints. Taking the Rizhao Shihua Oil Terminal of the Rizhao Port as the case study, this paper used the simulator to analyze the effects of various buffer locations and sizes on the port capacity and service level under two priority rules. They are the conventional "first come, first served (FCFS)" rule and a value priority (VP) rule that prioritizes the tankers with higher holding values. The results show that the VP rule can improve the terminal's capacity by up to 2.35% compared to the FCFS rule. The capacity increases as the buffer moves from the anchorage side to the dockyard side of the channel. The resulting improvement is up to 7.56%. Moreover, a one-way channel accompanied with a dockyard-side buffer holding one tanker only can attain 91.54% of the capacity of a two-way channel, manifesting the great potential of buffers. References | Related Articles | Metrics

Select

Robustness Analysis of Seaport-dry Port Container Transport Networks Under Cascading Failure XU Bo-wei, TANG Can-xuan, LI Jun-jun 2023, 23(3): 265-279.  DOI: 10.16097/j.cnki.1009-6744.2023.03.028 Abstract ( )   PDF (8091KB) ( )   When a seaport is attacked and fails, cascading failure effect will be generated in the seaport-dry port container transport network due to the connectivity between ports, which will then spread to the whole network. To this end, this paper conducts a study on the robustness of seaport-dry port container transport networks under cascade failure. This paper designs different state indicators for seaport nodes and dry port nodes, establishes load transfer equations for invalid nodes based on the node failure propagation capability and inter-node gravity model, constructs a cascade failure model and robustness assessment index for seaport-dry port container transport network, and explore the impact of the failure propagation capability threshold of neighbor node, the amount of node redundancy, the proportion of port hopping, the change in the number of cooperative dry ports at seaports and different node attack modes on network robustness, it also explores the change in business volume of its competing ports after the failure of the seaport. The results show that there is an optimal value interval for the node failure propagation capacity threshold and the hopping port ratio in terms of enhancing robustness performance; increasing the node tolerance parameters within a certain range can help reduce the impact of network cascading failures; and the extent of overlap of the dry ports with which the seaports cooperate should not be too high; competing seaports will take a share of the business of the invalid seaport, which is related to the distance between the two ports and the operational capacity of the competing seaport. References | Related Articles | Metrics

Select

Location-routing Problem of Emergency Logistics for Engineering Construction Projects Under Complex Environments ZHANG Jin, ZHU Hong-xing, SHEN Hao, LI Guo-qi 2023, 23(3): 280-289.  DOI: 10.16097/j.cnki.1009-6744.2023.03.029 Abstract ( )   PDF (4501KB) ( )   Emergency material security is the key to post-disaster emergency relief in engineering and construction projects under complex environments. This paper addresses an emergency logistics center location-routing problem, taking into account demand uncertainty, congestion time uncertainty, maximum rescue time requirements, multi-type vehicles, and other factors. This paper adopts triangular fuzzy numbers to inscribe uncertain parameters and constructs a two-stage fuzzy nonlinear location-routing model based on scenarios. A single-objective deterministic model is obtained by integration, linearization and defuzzification, and the Gurobi solver is used to solve it. A super large railroad construction project in the mountainous region of the western plateau of Sichuan Province is taken as an example, and the validity and applicability of the model are verified through model comparison and sensitivity analysis. The results show that as the number of locations increases, storage costs increase by 5.9%, response times for the five scenarios are reduced by an average of 15.2% and the maximum response time is reduced by 7.8%. Compared with the Expected-value model, the model built in this paper is superior in terms of storage cost and emergency response time. The storage cost is linearly related to the demand level, while the maximum response time is affected by both the demand level and the congestion time. The model established in this paper can scientifically select the location of emergency facilities and develop emergency rescue paths to reduce the storage cost and response time of emergency rescue. It can provide decision support for dispatching emergency materials for engineering construction projects in complex and difficult areas. References | Related Articles | Metrics

Select

UAV Cruising for Material Transportation Under Engineering Construction in Complex Mountainous Areas: Modeling and Case Study KANG Liu-jiang, LI Hao, SUN Hui-jun, WU Jian-jun 2023, 23(3): 290-299.  DOI: 10.16097/j.cnki.1009-6744.2023.03.030 Abstract ( )   PDF (4885KB) ( )   This paper focuses on the unmanned aerial vehicle (UAV) cruising and fleet management problems based on the material transportation plan for the YA-LZ railway construction section. An approximate algorithm is proposed to calculate the actual cruising distance of UAV in mountainous areas, considering altitude limits and flight trajectories. Then, a nonlinear UAV cruising model is developed to minimize the base building cost, UAV fixed cost and cruising cost, which determines the UAV base coordinates, cruising coverage and fleet sizes. Moreover, an Adaptive Clustering- Spatial Movement (AC-SM) algorithm is designed to divide the cruising coverage through finite iterations and find the optimal UAV base coordinate as well as the minimum UAV fleet size. This paper applied the model and algorithm to the case study of the YA-LZ section. The results indicate that the proposed approaches reduce approximately 20% of the UAV cruising cost compared with the traditional algorithm. References | Related Articles | Metrics

Select

Spatial and Temporal Variation Characteristics of Urban Traffic Congestion Factors and Source Analysis ZHAO Xue-ting, HU Li-wei 2023, 23(3): 300-310.  DOI: 10.16097/j.cnki.1009-6744.2023.03.031 Abstract ( )   PDF (4850KB) ( )   The traffic congestion factor is an important source of urban traffic congestion effect at the "point-line surface" level. In order to investigate the spatial and temporal characteristics of urban traffic congestion factors and their influence degree, this paper introduces the relevant methods for geological and ecological risk evaluation. Firstly, based on the "source-path-target" model, the multivariate and comprehensive complexity of urban traffic congestion is considered, and the impact index system of urban traffic congestion evaluation is analyzed in depth. to explore the law of periodic changes; improve the Nemero pollution index method to assess the characteristics and risk degree of traffic congestion factors; finally, use person correlation analysis and positive definite matrix factor decomposition (PMF) model to analyze their main sources and assess the uncertainty of the results. The results showed that the magnitude of seven factors, namely, overtaking, traffic rule violation, random lane change, parking, illegal and illegal road occupation, adverse weather conditions, and sudden traffic accidents, varied more significantly, with coefficients of variation exceeding 40%; the urban traffic congestion degree factor Ʃ18Bi was the largest in the morning, afternoon, and evening peaks in January (11.23) and reached the lowest in March ( 8.12). The overall urban traffic congestion shows a time-varying characteristic of evening peak > morning peak > mid-peak, and there are 2 cycles of time variation of 8a (first main cycle) and 4a (second main cycle). The results obtained by the Nemero pollution index method are dominated by moderate and above traffic congestion. The results of the study can provide a scientific basis for urban traffic congestion node management. References | Related Articles | Metrics

Select

Electro-pneumatic Brake for Heavy Haul Locomotive During Automatic Driving at Long Downgrade Segment JIANG Zheng, WANG Rui, DU Hai-bin, YU Jian, CHEN Xue 2023, 23(3): 311-318.  DOI: 10.16097/j.cnki.1009-6744.2023.03.032 Abstract ( )   PDF (3963KB) ( )   The railway transport amount has been promoting with the use of heavy haul freight train, which also brings safety risks. As time went on, such potential safety matters have received wide-spread attention. It is known that the long heavy downgrade railways are the most dangerous part for heavy haul freight train, especially on the continuous long heavy downgrade section. It was found that the key to resolve this problem is to control the speed of the train. This paper focuses on speed control and proposes an electro-pneumatic braker application method, which is based on air brake and with power brake supplement. The paper proposes the method and applies it into the automatic train operation system, which have been used on heavy haul freight HXD2 locomotive. Through the practical tests in QinLing region along the XiKang railway, the running tests performed are more than 40000 kilometers in one year, while the train draw average qualities no less than 3900- ton. Eventually, the method obtains the expected results and the proposed method could give certain guidelines for user. What is more important, this method can effectively ensure the safety and steady for locomotive during automatic driving, especially when the train running in the long heavy downgrade section. References | Related Articles | Metrics