Freight Logistics Characteristics Analysis Method Based on Highway Network Toll Collection Data

Abstract

The present invention relates to the field of data analysis technology, and in particular to a method for analyzing freight logistics characteristics based on highway network toll data. The method comprises the following steps: obtaining highway section data; performing section identification division processing according to the highway section data to generate divided section data; performing freight section characteristic data analysis on the divided section data to generate freight section characteristic data; establishing a mathematical relationship model of capacity restriction and traffic distribution based on the Logit model and the freight section characteristic data to generate a capacity restriction-traffic distribution model; predicting passenger and freight section flow based on the capacity restriction-traffic distribution model to generate section passenger and freight flow prediction data; identifying the main section of freight logistics for the divided section data according to the section passenger and freight flow prediction data to generate the main section of freight logistics data. The present invention realizes the precise analysis of the logistics characteristics of highway trucks.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to a method for analyzing freight logistics characteristics based on highway network toll data. Background Art

[0002] In recent years, the issue of logistics efficiency has received widespread attention from the society. In order to further improve freight capacity, differentiated charging policies for different sections and types of vehicles are adopted to guide more trucks to travel on highways, reduce the travel costs of trucks, and increase the volume of freight highways. Highway network toll data serves as a key source of information, including information on the departure point, destination, section of the road, travel time, type of goods, etc. of the vehicle. Through the comprehensive analysis of these data, we can understand the key characteristics of freight logistics such as spatiotemporal distribution, flow trend, and congestion. The analysis of freight logistics can optimize transportation routes, reasonably plan distribution plans, reduce transportation time and resource waste, and thus improve overall transportation efficiency by deeply understanding the spatiotemporal distribution, flow trend, and possible bottlenecks and congestion points of freight transportation. However, the traditional analysis method of freight logistics requires the main routes of freight logistics to be monitored manually, which is not objective and detailed enough for truck path selection and flow distribution, and the analysis of the characteristics of highway truck logistics is not comprehensive enough and the effect is poor. Summary of the invention

[0003] Based on this, the present invention provides a method for analyzing freight logistics characteristics based on highway network toll collection data to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a method for analyzing freight logistics characteristics based on highway network toll collection data includes the following steps:

[0005] Step S1: obtaining highway section data; performing section identification division processing according to the highway section data to generate divided section data; performing freight section feature data analysis on the divided section data to generate freight section feature data, wherein the freight section feature data includes section congestion status data, truck section fee data and vehicle travel time data;

[0006] Step S2: Establishing a mathematical relationship model between capacity restriction and traffic distribution based on the Logit model and freight section characteristic data, and generating a capacity restriction-traffic distribution model;

[0007] Step S3: predicting the passenger and freight traffic flow on a road section based on the capacity constraint-traffic allocation model, and generating predicted data on the passenger and freight traffic flow on the road section;

[0008] Step S4: marking the freight logistics main section of the divided section data according to the passenger and freight flow prediction data of the section, and generating the freight logistics main section data.

[0009] The present invention obtains highway section data, and then performs section identification and division processing on these data, so that the highway system can be divided in detail, and each divided section can be independently identified and analyzed, such as the design speed of the highway section and the distance of the highway section, etc., providing a more accurate data basis. The freight section characteristic data analysis is performed on the divided section data, and key freight logistics characteristics are extracted, including the section congestion status, truck section fees and vehicle travel time, etc., which have important information value for freight logistics analysis, such as helping to analyze which sections may encounter congestion in what time period, providing real-time traffic status information for truck route planning and scheduling, understanding the cost of each section, and providing basic data for truck transportation time prediction and route planning. Based on the Logit model and freight section characteristic data, a capacity restriction-traffic distribution mathematical relationship model is established. The Logit model is a probabilistic model commonly used in traffic distribution, which can reflect the probability of different sections being selected. Considering the characteristics of freight logistics, the introduction of capacity restrictions is more in line with actual scenarios. The introduction of capacity constraints makes the model closer to the actual situation. In freight logistics, the traffic capacity of the road section is limited. Considering capacity constraints can more accurately reflect the actual transportation bottlenecks of the freight network and the impact of bottlenecks on traffic distribution. The generated capacity constraint-traffic distribution model is used to predict the traffic flow distribution of freight sections under different conditions, understand the load distribution of each section in the freight network, and guide road network optimization and congestion relief. Based on the capacity constraint-traffic distribution model, passenger and freight section flow prediction is carried out, which helps to predict the flow distribution of freight logistics in different sections and provide data support for logistics planning and section optimization. The predicted flow data can be used to evaluate the load status of the freight network and future congestion trends. According to the passenger and freight flow prediction data of the section, the main sections of freight logistics are marked for the divided section data, which helps to determine which sections are more suitable for freight under the current conditions and helps to determine the main roads for freight.

[0010] Preferably, step S1 comprises the following steps:

[0011] Step S11: Acquire high-speed road section data;

[0012] Step S12: performing road section identification division processing on the expressway section data to generate divided road section data;

[0013] Step S13: using monitoring equipment to collect vehicle data of the divided road sections to generate road section vehicle data;

[0014] Step S14: collecting vehicle travel time data according to the road section vehicle data to generate vehicle travel time data;

[0015] Step S15: calculating the passenger and freight flow of the road section according to the road section vehicle data, and generating the passenger and freight flow data of the road section;

[0016] Step S16: performing a road section congestion state clustering analysis on the vehicle travel time data based on a K-Means clustering algorithm to generate road section congestion state data;

[0017] Step S17: Calculate the truck section fee according to the preset freight vehicle charging requirements and the passenger and freight flow data of the section, and generate the truck section fee data;

[0018] Step S18: Integrate the road section congestion status data, truck section fee data, and vehicle travel time data to generate freight section characteristic data.

[0019] The present invention can obtain comprehensive highway network information, including the geographical location, length, design speed, etc. of the section, by acquiring highway section data. This provides basic data for subsequent section feature analysis and helps understand the structure and scale of the entire highway system. The highway section data is processed by identification division, and the highway is divided into different sections, which is helpful for more detailed analysis and modeling. The divided section data can be used to determine the driving path of trucks, calculate the congestion status of the section, predict passenger and freight flow, etc. The section identification is used to determine the design speed, length and other information of the section. By using monitoring equipment to collect vehicle data on the divided sections, information such as vehicle flow, speed and vehicle type can be obtained in real time, and the real-time traffic status of the section can be monitored, providing a practical transportation data basis for freight logistics feature analysis. By collecting vehicle travel time data based on the section vehicle data, the actual driving time of vehicles on different sections can be obtained, and key indicators such as the driving speed and travel efficiency of the section can be analyzed. Passenger and freight flow calculation using section vehicle data can obtain passenger and freight flow on each section, which is very important for understanding information such as traffic load, hot spots and freight density of different sections. By using the K-Means clustering algorithm to perform congestion status clustering analysis on vehicle travel time data, road sections can be divided into different congestion status categories, peak hours and congestion areas can be identified, and the basis for subsequent congestion feature modeling can be provided. According to the preset freight vehicle charging requirements and the passenger and freight flow data of the road section, the truck section fee calculation can be performed to obtain the fee information required for truck passage on each road section, analyze the freight cost, and evaluate the impact of different road sections on the economic benefits of trucks, thereby providing cost support for the selection of freight logistics routes. The road section congestion status data, truck section fee data, and vehicle travel time data are integrated, including information on multiple aspects such as road section congestion, truck fees, and travel time, to fully understand the characteristics of freight sections.

[0020] Preferably, step S12 comprises the following steps:

[0021] Extract high-speed gantry coordinates from high-speed road section data to obtain high-speed gantry coordinate data;

[0022] Divide the highway section data into sections according to the highway gantry data to generate initial divided section data;

[0023] The initial divided road section data is processed for attribute identification of the divided road sections to generate divided road section data.

[0024] The present invention extracts the coordinates of the high-speed gantry from the high-speed section data, can obtain the accurate position information of the high-speed gantry, establishes the spatial relationship between the gantry and the section, and provides an accurate spatial reference for subsequent data processing and section division. The high-speed gantry data is used to divide the high-speed section data, and the entire highway network is divided into specific section units to form a preliminary section division plan. Each divided section can be regarded as a basic unit in this step, which is convenient for further feature analysis and attribute identification. The attribute identification processing of the divided section is performed on the initial divided section data, which includes the design speed, length and other information of the high-speed section, and important attribute information is given to each section. This information plays a key role in the subsequent freight section feature data analysis. For example, attributes such as design speed and length are basic data for analyzing section congestion, calculating travel time and predicting costs, which are crucial for freight logistics analysis.

[0025] Preferably, step S16 comprises the following steps:

[0026] Step S161: performing cluster center data selection and marking according to vehicle travel time data to obtain cluster center data;

[0027] Step S162: using the Euclidean distance algorithm to calculate the cluster distance of the vehicle travel time data and the cluster center data to generate cluster distance data;

[0028] Step S163: performing iterative updating processing on the cluster center data to obtain updated cluster center data, performing Euclidean distance calculation on the updated cluster center data and the unupdated cluster center data, and generating cluster center deviation data;

[0029] Step S164: Return to step S163 and step S164 to perform clustering iteration. When the cluster center deviation data is less than the preset deviation threshold, stop the clustering iteration; perform road section congestion status clustering analysis on the cluster distance data after the clustering iteration based on the K-Means clustering algorithm to generate road section congestion status data.

[0030] The present invention selects and marks the cluster center data of vehicle travel time data, and can select representative cluster centers from the original data to ensure the accuracy of subsequent clustering, because the selection of appropriate initial cluster centers has a greater impact on the performance of the K-Means algorithm. The cluster distance is calculated for the vehicle travel time data and cluster center data using the Euclidean distance algorithm, and the data samples are divided into different cluster clusters. The calculation of the Euclidean distance can measure the similarity between data points, thereby achieving effective division of the congestion status. The cluster center data is iteratively updated to make the cluster center gradually stabilize, improve the accuracy and stability of clustering, and ensure that the final cluster center can better represent each cluster cluster. The cluster center deviation data calculation and update during the iteration process is to monitor the changes in the cluster center, ensure that the iteration is stopped when the deviation is small, and reduce the calculation cost. The K-Means clustering algorithm performs a clustering analysis on the congestion status of the road section based on the cluster distance data obtained after the clustering iteration. This is helpful in dividing the road section into different congestion states, providing key information on the congestion state for subsequent modeling of the road section utility function, and more accurately considering the congestion situation in the path selection calculation, thereby improving the authenticity and prediction accuracy of the model.

[0031] Preferably, step S161 includes the following steps:

[0032] Perform mean square error and calculation based on preset initial cluster center data and vehicle travel time data to generate mean square error data;

[0033] According to the cluster elbow theory and mean square error data, cluster number parameter analysis is performed to generate cluster number parameters;

[0034] The cluster centers of the vehicle travel time data are randomly selected and marked according to the cluster quantity parameter to obtain the cluster center data.

[0035] The present invention uses the preset initial cluster center data and vehicle travel time data to calculate the mean square error and evaluate the degree of fit of the current cluster center to the data. The mean square error is an indicator to measure the quality of fit, which can quantify the difference between the actual observed value and the predicted value of the cluster center. The generated mean square error data is helpful to understand the current clustering effect and provide a basis for the subsequent cluster number parameter analysis. The cluster elbow theory is used to analyze the mean square error data, which is beneficial to determine the optimal number of clusters. The cluster elbow theory compares the change trend of the mean square error under different cluster numbers to find the number of clusters corresponding to the elbow. This number of clusters is generally considered to be optimal. When selecting a suitable number of clusters, the complexity of the model and the fitting effect are weighed to improve the interpretability of the clustering results. The vehicle travel time data is randomly marked for cluster center selection according to the cluster number parameter to ensure that the selected cluster center has diversity. By randomly selecting the cluster center, different aspects of the data can be better explored, not just limited to the initial cluster center, increasing the flexibility of the algorithm and improving the comprehensive capture of the data structure.

[0036] Preferably, step S2 comprises the following steps:

[0037] Step S21: using the minimum-maximum normalization method to normalize the freight section characteristic data respectively, to generate normalized freight section characteristic data;

[0038] Step S22: calculating the freight section utility data based on the normalized freight section characteristic data to generate the freight section utility data;

[0039] Step S23: A mathematical relationship model of capacity restriction and traffic distribution is established based on the Logit model and the freight section utility data to generate a capacity restriction-traffic distribution model.

[0040] The present invention uses the minimum-maximum normalization method to process the freight section characteristic data, eliminates the dimensional differences between different characteristic data, makes them compare within the same numerical range, ensures that the influence weights of each characteristic on the model are relatively balanced, and improves the stability and convergence of the model. Based on the normalized freight section characteristic data, the freight section utility data is calculated, and a section utility function is established to reflect the utility of trucks in different sections. This utility function is the basis of the Logit model, which takes into account factors such as the section congestion state, section travel time, and section driving cost. The generation of freight section utility data provides a basis for the subsequent capacity restriction-traffic allocation model. The mathematical relationship model of capacity restriction and traffic allocation based on the Logit model and freight section utility data is established, which is beneficial to simulating the path selection process of trucks in the road network, takes into account the capacity restriction of the section, makes the allocation result more in line with the actual situation, and the generated capacity restriction-traffic allocation model provides a theoretical basis for the subsequent passenger and freight section flow prediction.

[0041] Preferably, step S22 includes the following steps:

[0042] Step S221: performing fixed utility data analysis of the freight section according to the normalized freight section characteristic data to generate freight section fixed utility data;

[0043] Step S222: performing random utility data analysis of the freight section according to the preset freight section simulation data to generate random utility data of the freight section;

[0044] Step S223: Calculate the freight section utility data according to the freight section fixed utility data and the freight section random utility data to generate the freight section utility data.

[0045] The present invention performs fixed utility data analysis of freight sections based on normalized freight section feature data to determine the impact of observable factors such as section congestion status on truck route selection. The generated fixed utility data represents observable influencing factors that are directly related to the section characteristics, providing a basis for the interpretability of the model. Performing random utility data analysis of freight sections based on preset freight section simulation data is beneficial to considering the impact of unobservable and difficult-to-quantify factors on route selection, including individual differences of drivers, unexpected events, etc. By introducing random utility, the degree of fit of the model to the actual situation is increased. The freight section utility data is calculated based on the fixed utility data of the freight section and the random utility data of the freight section, and the comprehensive impact of fixed utility and random utility on route selection is comprehensively considered. The generated freight section utility data is the core of the Logit model, reflecting the utility of trucks on different sections and providing input for subsequent traffic distribution models.

[0046] Preferably, step S23 includes the following steps:

[0047] Step S231: Transmitting the freight section utility data to the Logit model to design the associated mapping relationship between capacity restriction and traffic distribution, and generating an initial capacity restriction-traffic distribution model;

[0048] Step S232: using the Halton sequence to select target model parameters for the capacity constraint-traffic assignment model to obtain target model parameters;

[0049] Step S233: Optimize the target model parameters according to the Monte Carlo simulation method to establish a capacity constraint-traffic assignment model.

[0050] The present invention transfers the utility data of the freight section to the Logit model, designs the associated mapping relationship of capacity restriction and traffic distribution, and designs the mapping relationship based on the utility data of the freight section, so that the model can better capture the preferences and behaviors of truck route selection, and provide a basis for subsequent parameter optimization. The Halton sequence is used to select the target model parameters of the capacity restriction-traffic distribution model to improve the efficiency and accuracy of parameter selection. The Halton sequence is a low-discrepancy, uniformly distributed sequence that can be used to uniformly select sample points in the parameter space, thereby better representing the entire parameter space. The target model parameters obtained according to the Halton sequence are optimized using the Monte Carlo simulation method, which is beneficial to improving the fit of the model. The Monte Carlo simulation method uses a large number of random samplings to simulate the extraction of a large number of random samples from the total sample, so that the parameter estimation is more accurate.

[0051] Preferably, step S3 comprises the following steps:

[0052] Step S31: performing real-time updating processing on the road section passenger and freight flow data to generate real-time road section passenger and freight flow data;

[0053] Step S32: predicting the passenger and freight traffic on a road section based on the real-time passenger and freight traffic data and the capacity constraint-traffic allocation model, and generating predicted passenger and freight traffic data on the road section.

[0054] The present invention updates the passenger and freight traffic data of the road section in real time, captures the dynamic changes of the passenger and freight traffic of the road section, and reflects the changes of the actual traffic volume in the traffic network in real time, so that the model is more real-time and accurate, and better predicts the future passenger and freight traffic. By using the passenger and freight traffic data of the road section updated in real time and the capacity restriction-traffic allocation model, the passenger and freight traffic of the road section is predicted, and the estimated value of the passenger and freight traffic of the future road section based on the model is provided, which provides important information support for freight logistics management decision-making. The model prediction results can help optimize the allocation of road network resources, understand the main high-speed routes for freight, and formulate road section cost strategies.

[0055] Preferably, step S4 comprises the following steps:

[0056] Step S41: analyzing and processing the truck saturation of the road section according to the predicted data of passenger and freight flow of the road section, and generating predicted data of the truck saturation of the road section;

[0057] Step S42: Identify the main road sections for freight logistics based on the road section truck saturation prediction data to generate the main road section data for freight logistics.

[0058] The present invention performs a road section truck saturation analysis based on the road section passenger and freight flow prediction data, identifies and analyzes the truck load of the road section. Truck saturation refers to the ratio between the number of trucks passing through a certain road section and the maximum capacity of the road section, which helps to determine which road sections may face congestion, saturation or overload, thereby providing important information for traffic management and planning. The main sections of freight logistics are identified for the divided road section data based on the road section truck saturation prediction data, which helps to determine the key sections in the network, that is, those sections that may become bottlenecks or important nodes for freight logistics. Identifying the main freight sections helps to optimize the flow path of goods, reduce congestion, and improve the overall freight logistics efficiency.

[0059] The beneficial effect of the present application is that the freight logistics characteristics analysis method based on highway network toll data of the present invention automatically monitors the main routes of freight logistics, and considers information such as road congestion, road section tolls, and road section travel time, so that truck path selection and flow distribution are more objective and detailed, and constructs a path selection method based on a multi-logit model, and finally distributes flow to the truck starting point data, calculates the truck driving saturation to identify the main channel of truck logistics, and fully analyzes the characteristics of highway truck logistics. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 A schematic diagram of the steps of a method for analyzing freight logistics characteristics based on highway network toll data according to the present invention;

[0061] Figure 2 for Figure 1 Detailed implementation steps of step S1 in FIG.

[0062] Figure 3 for Figure 2 Detailed implementation steps of step S16;

[0063] Figure 4 for Figure 1 Detailed implementation steps of step S2 in the flowchart;

[0064] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0065] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0066] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0067] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0068] To achieve this, please refer to Figures 1 to 3 The present invention provides a method for analyzing freight logistics characteristics based on highway network toll data, comprising the following steps:

[0069] Step S1: obtaining highway section data; performing section identification division processing according to the highway section data to generate divided section data; performing freight section feature data analysis on the divided section data to generate freight section feature data, wherein the freight section feature data includes section congestion status data, truck section fee data and vehicle travel time data;

[0070] Step S2: Establishing a mathematical relationship model between capacity restriction and traffic distribution based on the Logit model and freight section characteristic data, and generating a capacity restriction-traffic distribution model;

[0071] Step S3: predicting the passenger and freight traffic flow on a road section based on the capacity constraint-traffic allocation model, and generating predicted data on the passenger and freight traffic flow on the road section;

[0072] Step S4: marking the freight logistics main section of the divided section data according to the passenger and freight flow prediction data of the section, and generating the freight logistics main section data.

[0073] The present invention obtains highway section data, and then performs section identification and division processing on these data, so that the highway system can be divided in detail, and each divided section can be independently identified and analyzed, such as the design speed of the highway section and the distance of the highway section, etc., providing a more accurate data basis. The divided section data is analyzed for freight section feature data, and key freight logistics features are extracted, including the congestion status of the section, the cost of the truck section, and the vehicle travel time, etc., which have important information value for freight logistics analysis, such as helping to analyze which sections may encounter congestion in what time period, providing real-time traffic status information for truck route planning and scheduling, understanding the cost of each section, and providing basic data for truck transportation time prediction and route planning. Based on the Logit model and freight section feature data, a capacity restriction-traffic distribution mathematical relationship model is established. The Logit model is a probabilistic model commonly used in traffic distribution, which can reflect the probability of different sections being selected. Considering the characteristics of freight logistics, the introduction of capacity restrictions is more in line with actual scenarios. The introduction of capacity constraints makes the model closer to the actual situation. In freight logistics, the traffic capacity of the road section is limited. Considering capacity constraints can more accurately reflect the actual transportation bottlenecks of the freight network and the impact of bottlenecks on traffic distribution. The generated capacity constraint-traffic distribution model is used to predict the traffic flow distribution of freight sections under different conditions, understand the load distribution of each section in the freight network, and guide road network optimization and congestion relief. Based on the capacity constraint-traffic distribution model, passenger and freight section flow prediction is carried out, which helps to predict the flow distribution of freight logistics in different sections and provide data support for logistics planning and section optimization. The predicted flow data can be used to evaluate the load status of the freight network and future congestion trends. According to the passenger and freight flow prediction data of the section, the main sections of freight logistics are marked for the divided section data, which helps to determine which sections are more suitable for freight under the current conditions and helps to determine the main roads for freight.

[0074] In the embodiment of the present invention, reference Figure 1 The above is a schematic flow chart of the steps of a method for analyzing freight logistics characteristics based on highway network toll data of the present invention. In this embodiment, the method for analyzing freight logistics characteristics based on highway network toll data includes the following steps:

[0075] Step S1: obtaining highway section data; performing section identification division processing according to the highway section data to generate divided section data; performing freight section feature data analysis on the divided section data to generate freight section feature data, wherein the freight section feature data includes section congestion status data, truck section fee data and vehicle travel time data;

[0076] In the embodiment of the present invention, highway section data including the coordinates of the starting and ending points, length, speed limit and other information of the section are obtained through the highway management department or relevant institutions. The highway section is spatially divided using GIS (Geographic Information System) tools to obtain the divided section data by dividing the section into relatively uniform spatial units or dividing it according to the actual road structure. Monitoring equipment is used to collect section vehicle data including vehicle type, speed, density and other information, and the section travel time information of the vehicle is analyzed in real time based on the vehicle data to reflect the driving conditions of the section. The K-Means clustering algorithm is used to perform cluster analysis on the travel time to obtain the congestion status data of sections with different congestion states. The section fee data of trucks is calculated based on the section passenger and freight flow data and the preset freight vehicle charging requirements.

[0077] Step S2: Establishing a mathematical relationship model between capacity restriction and traffic distribution based on the Logit model and freight section characteristic data, and generating a capacity restriction-traffic distribution model;

[0078] In the embodiment of the present invention, the normalized freight section characteristic data is used to perform minimum-maximum normalization processing on the freight section characteristic data according to the basic form of the Logit model to obtain the normalized freight section characteristic data. Calculate the utility data for the freight route segment, where is the fixed utility corresponding to the freight section characteristic data, is the random utility corresponding to the freight section characteristic data, Represented as freight section characteristic data. Based on the freight section utility data, a capacity constraint-traffic assignment model is established. The model can adopt a variant of the Logit model, which contains a capacity constraint term to reflect the constraints of the section's traffic capacity. The capacity constraint can be introduced by introducing the section capacity parameter to ensure that the actual traffic capacity is taken into account in the model to avoid excessive congestion.

[0079] Step S3: predicting the passenger and freight traffic flow on a road section based on the capacity constraint-traffic allocation model, and generating predicted data on the passenger and freight traffic flow on the road section;

[0080] In the embodiment of the present invention, the passenger and freight flow data of the existing road section is updated in real time, and the actual traffic data can be obtained by using traffic monitoring equipment, vehicle-mounted sensors, etc. For example, the passenger and freight flow data of the road section is updated by collecting information such as the number of vehicles, speed, and vehicle type in real time. The passenger and freight flow of the road section is predicted by using the capacity restriction-traffic allocation model in combination with the real-time updated passenger and freight flow data of the road section. The prediction process can adopt iterative solution. According to the current flow data and model parameters, new flow data is obtained through model calculation, and it is continuously iterated until the model converges.

[0081] Step S4: marking the freight logistics main section of the divided section data according to the passenger and freight flow prediction data of the section, and generating the freight logistics main section data.

[0082] In the embodiment of the present invention, the predicted data of passenger and freight flow of each road section is used to calculate the truck saturation of each road section, that is, the ratio of the truck traffic volume to the road section traffic capacity. The calculation formula is: prec 货 is the car saturation data, vol 货 The predicted truck flow for the road section passenger and freight flow prediction data, vol 客 For the predicted passenger car flow of the road section passenger and freight flow prediction data, a saturation threshold is set according to the results of the road section truck saturation analysis, and the road section exceeding the threshold is identified as the main freight logistics section. For example, if the saturation threshold is set to 0.8, when the saturation of a certain road section exceeds 0.8, the road section is identified as the main freight logistics section. According to the identified main freight logistics section, its relevant information is integrated to form the freight logistics main section data, and compared with the actual road network channels for analysis to determine the accuracy of the freight logistics main section data.

[0083] Preferably, step S1 comprises the following steps:

[0084] Step S11: Acquire high-speed road section data;

[0085] Step S12: performing road section identification division processing on the expressway section data to generate divided road section data;

[0086] Step S13: using monitoring equipment to collect vehicle data of the divided road sections to generate road section vehicle data;

[0087] Step S14: collecting vehicle travel time data according to the road section vehicle data to generate vehicle travel time data;

[0088] Step S15: calculating the passenger and freight flow of the road section according to the road section vehicle data, and generating the passenger and freight flow data of the road section;

[0089] Step S16: performing a road section congestion state clustering analysis on the vehicle travel time data based on a K-Means clustering algorithm to generate road section congestion state data;

[0090] Step S17: Calculate the truck section fee according to the preset freight vehicle charging requirements and the passenger and freight flow data of the section, and generate the truck section fee data;

[0091] Step S18: Integrate the road section congestion status data, truck section fee data, and vehicle travel time data to generate freight section characteristic data.

[0092] The present invention can obtain comprehensive highway network information, including the geographical location, length, design speed, etc. of the section, by acquiring highway section data. This provides basic data for subsequent section feature analysis and helps understand the structure and scale of the entire highway system. The highway section data is processed by identification division, and the highway is divided into different sections, which is helpful for more detailed analysis and modeling. The divided section data can be used to determine the driving path of trucks, calculate the congestion status of the section, predict passenger and freight flow, etc. The section identification is used to determine the design speed, length and other information of the section. By using monitoring equipment to collect vehicle data on the divided sections, information such as vehicle flow, speed and vehicle type can be obtained in real time, and the real-time traffic status of the section can be monitored, providing a practical transportation data basis for freight logistics feature analysis. By collecting vehicle travel time data based on the section vehicle data, the actual driving time of vehicles on different sections can be obtained, and key indicators such as the driving speed and travel efficiency of the section can be analyzed. Passenger and freight flow calculation using section vehicle data can obtain passenger and freight flow on each section, which is very important for understanding information such as traffic load, hot spots and freight density of different sections. By using the K-Means clustering algorithm to perform congestion status clustering analysis on vehicle travel time data, road sections can be divided into different congestion status categories, peak hours and congestion areas can be identified, and the basis for subsequent congestion feature modeling can be provided. According to the preset freight vehicle charging requirements and the passenger and freight flow data of the road section, the truck section fee calculation can be performed to obtain the fee information required for truck passage on each road section, analyze the freight cost, and evaluate the impact of different road sections on the economic benefits of trucks, thereby providing cost support for the selection of freight logistics routes. The road section congestion status data, truck section fee data, and vehicle travel time data are integrated, including information on multiple aspects such as road section congestion, truck fees, and travel time, to fully understand the characteristics of freight sections.

[0093] As an example of the present invention, refer to Figure 2 As shown, Figure 1 Detailed implementation steps of step S1 in the flowchart, in this example, step S1 includes:

[0094] Step S11: Acquire high-speed road section data;

[0095] In the embodiment of the present invention, the highway section data is obtained from the highway management agency or the relevant geographic information system (GIS) database, which includes the geographic coordinates, length, speed limit and other information of the section.

[0096] Step S12: performing road section identification division processing on the expressway section data to generate divided road section data;

[0097] In an embodiment of the present invention, a GIS tool is used to divide the expressway section, for example, evenly divide it according to the relative distance of each expressway gantry, and mark the length, speed limit and other information of the divided expressway section data to generate initial divided section data.

[0098] Step S13: using monitoring equipment to collect vehicle data of the divided road sections to generate road section vehicle data;

[0099] In the embodiment of the present invention, monitoring equipment, such as video cameras, sensors, etc., are deployed to cover the divided road sections, and vehicle-related data, including the number of vehicles, vehicle models, speeds, etc., are collected through these devices.

[0100] Step S14: collecting vehicle travel time data according to the road section vehicle data to generate vehicle travel time data;

[0101] In the embodiment of the present invention, based on the vehicle travel time data related to the road section vehicle data, including the vehicle entry and exit timestamps, vehicle model information, etc., for each vehicle, the travel time on the road section is calculated by its entry and exit timestamps. The travel time can be simply obtained by calculating the difference between the timestamps, and the calculation method of the travel time can be adjusted according to the specific situation. For example, if there are multiple monitoring points on the road section, the travel time can be calculated by adding the time difference of the vehicle passing each monitoring point.

[0102] Step S15: calculating the passenger and freight flow of the road section according to the road section vehicle data, and generating the passenger and freight flow data of the road section;

[0103] In the embodiment of the present invention, the information related to the passenger and freight flow of the road section based on the road section vehicle data includes counting the number of passenger cars and trucks passing on the road section, classifying the vehicles passing on the road section according to the vehicle type information or other available data, distinguishing between passenger cars and trucks, and judging according to the size, load and other characteristics of the vehicle, and using the number of passenger cars and trucks obtained by counting as the passenger and freight flow data of the road section. Statistics can be performed according to time units (such as hours, days) to generate corresponding passenger and freight flow time series.

[0104] Step S16: performing a road section congestion state clustering analysis on the vehicle travel time data based on a K-Means clustering algorithm to generate road section congestion state data;

[0105] In an embodiment of the present invention, a K-Means clustering algorithm is used to cluster the travel time data within a road section. K-Means needs to specify the number of clusters for clustering, and the appropriate number of clusters can be selected according to the Elbow Method. Each cluster represents a travel time pattern, and these clustering results can be interpreted as different road section congestion states. For example, a road section in a certain cluster may indicate a shorter travel time, while another cluster may represent a longer travel time, that is, a congested state. According to the results of K-Means clustering, a corresponding congestion status label is assigned to each road section to form road section congestion status data, which can be a discrete congestion level, such as extremely low, relatively low, medium, relatively high, extremely high, or other custom status labels.

[0106] Step S17: Calculate the truck section fee according to the preset freight vehicle charging requirements and the passenger and freight flow data of the section, and generate the truck section fee data;

[0107] In the embodiment of the present invention, the road section fee data for trucks is generated based on the road section passenger and freight flow data and the preset freight vehicle charging standard. l represents the path length (per kilometer), and h represents the set truck type to calculate the truck section fee data.

[0108] Step S18: Integrate the road section congestion status data, truck section fee data, and vehicle travel time data to generate freight section characteristic data.

[0109] In an embodiment of the present invention, the road section congestion status data, truck road section fee data and vehicle travel time data are integrated into a data table or data set. For example, in the data table, each row represents a road section, including information such as the identifier of the road section, the congestion status, the truck road section fee, and the vehicle travel time. The road section identifier is a code that uniquely identifies each road section. The congestion status is a description of the degree of congestion of the road section, which may use classification labels such as "congested" and "unblocked". The truck road section fee is the fee for trucks to pass the road section, which can be numerical data. The vehicle travel time is the time required for a vehicle to pass the road section, which can be minutes or other time units.

[0110] Preferably, step S12 comprises the following steps:

[0111] Extract high-speed gantry coordinates from high-speed road section data to obtain high-speed gantry coordinate data;

[0112] Divide the highway section data into sections according to the highway gantry data to generate initial divided section data;

[0113] The initial divided road section data is processed for attribute identification of the divided road sections to generate divided road section data.

[0114] The present invention extracts the coordinates of the high-speed gantry from the high-speed section data, can obtain the accurate position information of the high-speed gantry, establishes the spatial relationship between the gantry and the section, and provides an accurate spatial reference for subsequent data processing and section division. The high-speed gantry data is used to divide the high-speed section data, and the entire highway network is divided into specific section units to form a preliminary section division plan. Each divided section can be regarded as a basic unit in this step, which is convenient for further feature analysis and attribute identification. The attribute identification processing of the divided section is performed on the initial divided section data, which includes the design speed, length and other information of the high-speed section, and important attribute information is given to each section. This information plays a key role in the subsequent freight section feature data analysis. For example, attributes such as design speed and length are basic data for analyzing section congestion, calculating travel time and predicting costs, which are crucial for freight logistics analysis.

[0115] In an embodiment of the present invention, a geographic information system (GIS) tool or related technology is used to extract the latitude and longitude coordinate information of the high-speed gantry from the high-speed section data. For example, the coordinates of gantry A are (40.123, -74.567); the coordinates of gantry B are (40.456, -74.789). The extracted high-speed gantry coordinate data is used to divide the high-speed section, and the sections between adjacent gantries are divided into different sections. The section between gantry A and gantry B is divided into section 001; the section between gantry B and gantry C is divided into section 002. Attribute identifiers are added to the initially divided section data, including the speed limit information, length and other information of the high-speed section, such as the speed limit of section 001 is 120km / h and the length is 10 kilometers; the speed limit of section 002 is 100km / h and the length is 8 kilometers, and finally the divided section data is generated.

[0116] Preferably, step S16 comprises the following steps:

[0117] Step S161: performing cluster center data selection and marking according to vehicle travel time data to obtain cluster center data;

[0118] Step S162: using the Euclidean distance algorithm to calculate the cluster distance of the vehicle travel time data and the cluster center data to generate cluster distance data;

[0119] Step S163: performing iterative updating processing on the cluster center data to obtain updated cluster center data, performing Euclidean distance calculation on the updated cluster center data and the unupdated cluster center data, and generating cluster center deviation data;

[0120] Step S164: Return to step S163 and step S164 to perform clustering iteration. When the cluster center deviation data is less than the preset deviation threshold, stop the clustering iteration; perform road section congestion status clustering analysis on the cluster distance data after the clustering iteration based on the K-Means clustering algorithm to generate road section congestion status data.

[0121] The present invention selects and marks the cluster center data of vehicle travel time data, and can select representative cluster centers from the original data to ensure the accuracy of subsequent clustering, because the selection of appropriate initial cluster centers has a greater impact on the performance of the K-Means algorithm. The cluster distance is calculated for the vehicle travel time data and cluster center data using the Euclidean distance algorithm, and the data samples are divided into different cluster clusters. The calculation of the Euclidean distance can measure the similarity between data points, thereby achieving effective division of the congestion status. The cluster center data is iteratively updated to make the cluster center gradually stabilize, improve the accuracy and stability of clustering, and ensure that the final cluster center can better represent each cluster cluster. The cluster center deviation data calculation and update during the iteration process is to monitor the changes in the cluster center, ensure that the iteration is stopped when the deviation is small, and reduce the calculation cost. The K-Means clustering algorithm performs a clustering analysis on the congestion status of the road section based on the cluster distance data obtained after the clustering iteration. This is helpful in dividing the road section into different congestion states, providing key information on the congestion state for subsequent modeling of the road section utility function, and more accurately considering the congestion situation in the path selection calculation, thereby improving the authenticity and prediction accuracy of the model.

[0122] As an example of the present invention, refer to Figure 3 As shown, Figure 2 Detailed implementation steps of step S16 are shown in the flowchart. In this example, step S16 includes:

[0123] Step S161: performing cluster center data selection and marking according to vehicle travel time data to obtain cluster center data;

[0124] In an embodiment of the present invention, cluster center data is designed based on vehicle travel time data, the number of cluster centers is determined according to the Elbow Method, and cluster centers are randomly selected according to the number of cluster centers to represent high, medium and low travel time levels, respectively, to generate cluster center data.

[0125] Step S162: using the Euclidean distance algorithm to calculate the cluster distance of the vehicle travel time data and the cluster center data to generate cluster distance data;

[0126] In the embodiment of the present invention, according to the cluster center data μ j(j=1,2,…,k) is used as the cluster center of the vehicle travel time data, and the remaining data in the vehicle travel time data are calculated to the cluster center μ j The Euclidean distance D(x i ,μ j ), and assign the remaining data in the vehicle travel time data to the cluster to which the cluster center closest to it belongs. The cluster distance calculation formula is: D(x i ,μ j )=||x i -μ j || 2 , x i Represented as the i-th vehicle travel time data, μ j Represented as the jth cluster center data.

[0127] Step S163: performing iterative updating processing on the cluster center data to obtain updated cluster center data, performing Euclidean distance calculation on the updated cluster center data and the unupdated cluster center data, and generating cluster center deviation data;

[0128] In the embodiment of the present invention, the clustering rights of all data points in each cluster are recalculated according to the convergence result of each clustering distance. For example, when the Euclidean distance from the cluster center to the vehicle travel time data meets the preset weighted distance threshold, the vehicle travel time data corresponding to the cluster center is weighted, and the cluster center is updated according to the final weighted result, such as x i Represented as the i-th vehicle travel time data, δ i,j It is expressed as the weight of the i-th vehicle travel time data and the j-th cluster center data. The weight is only 0 if it does not meet the weighting condition or 1 if it meets the weighting condition. j It is represented as the updated j-th cluster center data, and finally the updated cluster center data is obtained. The Euclidean distance calculation formula is used to calculate the Euclidean distance between the updated cluster center data and the unupdated cluster center data to obtain the cluster center deviation after the update and before the update.

[0129] Step S164: Return to step S163 and step S164 to perform clustering iteration. When the cluster center deviation data is less than the preset deviation threshold, stop the clustering iteration; perform road section congestion status clustering analysis on the cluster distance data after the clustering iteration based on the K-Means clustering algorithm to generate road section congestion status data.

[0130] In the embodiment of the present invention, step S163 and step S164 are returned to repeat the iteration. When the cluster center deviation data is less than the preset deviation threshold, it means that the deviation between the updated cluster center and the cluster center before the update is extremely small, and the cluster center basically no longer moves, then the iteration is stopped. The K-Means clustering algorithm is used to cluster the final cluster center, and all vehicle travel time data are re-clustered to generate the final clustering result. The clustering result can reflect the congestion status of different road sections. For example, several different congestion levels can be obtained. According to the congestion levels 1, 2, 3, 4, and 5 of the clustering results, the road section congestion status is calibrated and arranged from low to high according to the congestion degree. The congestion status can be divided into very smooth, smooth, light congestion, moderate congestion, and heavy congestion, and the road section congestion status data is generated.

[0131] Preferably, step S161 includes the following steps:

[0132] Perform mean square error and calculation based on preset initial cluster center data and vehicle travel time data to generate mean square error data;

[0133] According to the cluster elbow theory and mean square error data, cluster number parameter analysis is performed to generate cluster number parameters;

[0134] The cluster centers of the vehicle travel time data are randomly selected and marked according to the cluster quantity parameter to obtain the cluster center data.

[0135] The present invention uses the preset initial cluster center data and vehicle travel time data to calculate the mean square error and evaluate the degree of fit of the current cluster center to the data. The mean square error is an indicator to measure the quality of fit, which can quantify the difference between the actual observed value and the predicted value of the cluster center. The generated mean square error data is helpful to understand the current clustering effect and provide a basis for the subsequent cluster number parameter analysis. The cluster elbow theory is used to analyze the mean square error data, which is beneficial to determine the optimal number of clusters. The cluster elbow theory compares the change trend of the mean square error under different cluster numbers to find the number of clusters corresponding to the elbow. This number of clusters is generally considered to be optimal. When selecting a suitable number of clusters, the complexity of the model and the fitting effect are weighed to improve the interpretability of the clustering results. The vehicle travel time data is randomly marked for cluster center selection according to the cluster number parameter to ensure that the selected cluster center has diversity. By randomly selecting the cluster center, different aspects of the data can be better explored, not just limited to the initial cluster center, increasing the flexibility of the algorithm and improving the comprehensive capture of the data structure.

[0136] In the embodiment of the present invention, according to the preset initial cluster center data, for example, the initial three cluster centers with high-speed distances of 100 km and vehicle travel times of 1h, 1.5h and 2h are set, and the mean square error and calculation are performed according to the initial cluster center data and the vehicle travel time data. The square error sum formula can be used for calculation to generate the mean square error data, and the mean square error curve is analyzed. According to the elbow method, that is, the position where the curve begins to stabilize, the number of clusters corresponding to the elbow can be considered as the natural number of clusters of the data, that is, the mean square error no longer decreases significantly after the elbow, indicating that increasing the number of clusters will not significantly improve the clustering effect. Determine the location of the cluster elbow, and obtain the corresponding cluster number parameter, which represents the ideal number of clusters of the data. Use the determined cluster number parameter to execute the K-Means clustering algorithm, but this time the selection of the initial cluster center no longer uses the traditional random initialization, but uses a random selection method based on the cluster number parameter.

[0137] Preferably, step S2 comprises the following steps:

[0138] Step S21: using the minimum-maximum normalization method to normalize the freight section characteristic data respectively, to generate normalized freight section characteristic data;

[0139] Step S22: calculating the freight section utility data based on the normalized freight section characteristic data to generate the freight section utility data;

[0140] Step S23: A mathematical relationship model of capacity restriction and traffic distribution is established based on the Logit model and the freight section utility data to generate a capacity restriction-traffic distribution model.

[0141] The present invention uses the minimum-maximum normalization method to process the freight section characteristic data, eliminates the dimensional differences between different characteristic data, makes them compare within the same numerical range, ensures that the influence weights of each characteristic on the model are relatively balanced, and improves the stability and convergence of the model. Based on the normalized freight section characteristic data, the freight section utility data is calculated, and a section utility function is established to reflect the utility of trucks in different sections. This utility function is the basis of the Logit model, which takes into account factors such as the section congestion state, section travel time, and section driving cost. The generation of freight section utility data provides a basis for the subsequent capacity restriction-traffic allocation model. The mathematical relationship model of capacity restriction and traffic allocation based on the Logit model and freight section utility data is established, which is beneficial to simulating the path selection process of trucks in the road network, takes into account the capacity restriction of the section, makes the allocation result more in line with the actual situation, and the generated capacity restriction-traffic allocation model provides a theoretical basis for the subsequent passenger and freight section flow prediction.

[0142] As an example of the present invention, refer to Figure 4 As shown, Figure 1 Detailed implementation steps of step S2 in the embodiment are shown in the flowchart. In this embodiment, step S2 includes:

[0143] Step S21: using the minimum-maximum normalization method to normalize the freight section characteristic data respectively, to generate normalized freight section characteristic data;

[0144] In the embodiment of the present invention, the minimum-maximum normalization method is used to normalize the freight section feature data respectively. For each feature (such as the congestion status of the section, the truck section fee, the vehicle travel time, etc.), the minimum value (Min) and the maximum value (Max) of its value are calculated, and the normalization formula is used to map the value of each feature to the range of [0,1] to obtain the weird data of each feature respectively.

[0145] Step S22: calculating the freight section utility data based on the normalized freight section characteristic data to generate the freight section utility data;

[0146] In an embodiment of the present invention, a freight section utility function is designed for each feature in the normalized freight section characteristic data (such as section congestion status, truck section fee, vehicle travel time, etc.), wherein the freight section utility function of each feature includes a fixed utility function and a random utility function, the fixed utility function is designed based on the normalized freight section characteristic data, and the fixed utility function is designed based on the parameters designed for the normalized freight section characteristic data, the freight section utility function is calculated to generate freight section utility data, which reflects the efficient use of each section for freight.

[0147] Step S23: A mathematical relationship model of capacity restriction and traffic distribution is established based on the Logit model and the freight section utility data to generate a capacity restriction-traffic distribution model.

[0148] In the embodiment of the present invention, a mathematical relationship model of capacity limitation and traffic distribution is established based on the Logit model and the utility data of the freight section, and a binomial Logit model is used, which is often used for traffic distribution problems. The expression of the binomial Logit model is: P i is the probability that road segment i is selected, V i is the utility of segment i, M is the total segment route, V j is the probability of segment j being selected, and e is a natural constant. The introduction of capacity constraints can be achieved by considering the capacity of the segment in the Logit model, and the new probability expression is: It is expressed as a functional relationship of the capacity constraint-traffic distribution model, where v is the parameter that controls the capacity impact, and c is the section capacity. The above probability expression is combined with the utility data of the freight section to establish a capacity constraint-traffic distribution mathematical relationship model. Specifically, by determining the parameters in the model (kk, etc.), the utility data of the freight section is used as the input of the model to generate the traffic distribution probability of each section. Once the model parameters are determined, the complete capacity constraint-traffic distribution model can be obtained.

[0149] Preferably, step S22 includes the following steps:

[0150] Step S221: performing fixed utility data analysis of the freight section according to the normalized freight section characteristic data to generate freight section fixed utility data;

[0151] Step S222: performing random utility data analysis of the freight section according to the preset freight section simulation data to generate random utility data of the freight section;

[0152] Step S223: Calculate the freight section utility data according to the freight section fixed utility data and the freight section random utility data to generate the freight section utility data.

[0153] The present invention performs fixed utility data analysis of freight sections based on normalized freight section feature data to determine the impact of observable factors such as section congestion status on truck route selection. The generated fixed utility data represents observable influencing factors that are directly related to the section characteristics, providing a basis for the interpretability of the model. Performing random utility data analysis of freight sections based on preset freight section simulation data is beneficial to considering the impact of unobservable and difficult-to-quantify factors on route selection, including individual differences of drivers, unexpected events, etc. By introducing random utility, the degree of fit of the model to the actual situation is increased. The freight section utility data is calculated based on the fixed utility data of the freight section and the random utility data of the freight section, and the comprehensive impact of fixed utility and random utility on route selection is comprehensively considered. The generated freight section utility data is the core of the Logit model, reflecting the utility of trucks on different sections and providing input for subsequent traffic distribution models.

[0154] In an embodiment of the present invention, fixed utility data analysis of the freight section is performed based on the normalized freight section feature data. For example, the fixed utility of the section congestion state is constructed based on the normalized result of the section congestion state calibration. The fixed utility of the section congestion state is expressed as the probability of the vehicle selecting a path under the section congestion state. Random utility data analysis of the freight section is performed based on preset freight section simulation data. The preset freight section simulation data is a random distribution parameter designed based on real data, such as a random distribution parameter designed to fit the section congestion state based on the real section congestion state. The freight section utility data is calculated based on the freight section fixed utility data and the freight section random utility data. The freight section utility data of each feature in the normalized freight section feature data is first calculated, and then the utility data of the freight section utility data of each feature is added to obtain the final freight section utility data of each road.

[0155] Preferably, step S23 includes the following steps:

[0156] Step S231: Transmitting the freight section utility data to the Logit model to design the associated mapping relationship between capacity restriction and traffic distribution, and generating an initial capacity restriction-traffic distribution model;

[0157] Step S232: using the Halton sequence to select target model parameters for the capacity constraint-traffic assignment model to obtain target model parameters;

[0158] Step S233: Optimize the target model parameters according to the Monte Carlo simulation method to establish a capacity constraint-traffic assignment model.

[0159] The present invention transfers the utility data of the freight section to the Logit model, designs the associated mapping relationship of capacity restriction and traffic distribution, and designs the mapping relationship based on the utility data of the freight section, so that the model can better capture the preferences and behaviors of truck route selection, and provide a basis for subsequent parameter optimization. The Halton sequence is used to select the target model parameters of the capacity restriction-traffic distribution model to improve the efficiency and accuracy of parameter selection. The Halton sequence is a low-discrepancy, uniformly distributed sequence that can be used to uniformly select sample points in the parameter space, thereby better representing the entire parameter space. The target model parameters obtained according to the Halton sequence are optimized using the Monte Carlo simulation method, which is beneficial to improving the fit of the model. The Monte Carlo simulation method uses a large number of random samplings to simulate the extraction of a large number of random samples from the total sample, so that the parameter estimation is more accurate.

[0160] In the embodiment of the present invention, the general form of the Logit model is defined as: P i is the probability that road segment i is selected, Vi is the utility of segment i, M is the total segment route, V j is the probability of section j being selected, and e is a natural constant. The calculated utility data of the freight section (e.g., fixed utility data, random utility data, etc.) are transferred to the Logit model. The introduction of capacity restriction can be reflected by adding corresponding constraints in the Logit model. The capacity restriction usually involves the maximum carrying capacity of the section, that is, the traffic flow cannot exceed this value. The utility function in the Logit model is designed to be associated with the actual traffic distribution. Considering factors such as the physical characteristics of the freight section and traffic planning, the probability in the Logit model is closely related to the actual traffic flow, thereby forming an association mapping, and the designed Logit model is combined with the capacity restriction to form an initial capacity restriction-traffic distribution model. A certain number of sequence points are generated using the Halton sequence generation algorithm as candidate points for the parameters. The Halton sequence is usually more uniform than random sampling, which helps to search more comprehensively in the parameter space. Define an evaluation objective function, the input of which is the parameter point generated by the Halton sequence, and the output is the degree of fit of the capacity restriction-traffic distribution model or other evaluation indicators. The parameter points generated by the Halton sequence are used as initial values, and the parameters are adjusted through the Monte Carlo simulation method to minimize the objective function. The optimization goal is to find the parameter values ​​that best suit the actual data and conditions so that the model can better fit the observed traffic flow data. Through the optimization process, the optimal estimated values ​​of the target model parameters are obtained to establish the final capacity constraint-traffic allocation model.

[0161] Preferably, step S3 comprises the following steps:

[0162] Step S31: performing real-time updating processing on the road section passenger and freight flow data to generate real-time road section passenger and freight flow data;

[0163] Step S32: predicting the passenger and freight traffic on a road section based on the real-time passenger and freight traffic data and the capacity constraint-traffic allocation model, and generating predicted passenger and freight traffic data on the road section.

[0164] The present invention updates the passenger and freight traffic data of the road section in real time, captures the dynamic changes of the passenger and freight traffic of the road section, and reflects the changes of the actual traffic volume in the traffic network in real time, so that the model is more real-time and accurate, and better predicts the future passenger and freight traffic. By using the passenger and freight traffic data of the road section updated in real time and the capacity restriction-traffic allocation model, the passenger and freight traffic of the road section is predicted, and the estimated value of the passenger and freight traffic of the future road section based on the model is provided, which provides important information support for freight logistics management decision-making. The model prediction results can help optimize the allocation of road network resources, understand the main high-speed routes for freight, and formulate road section cost strategies.

[0165] In the embodiment of the present invention, the passenger and freight flow data of the road section are updated in real time, and the update frequency is set, for example, once every minute or every 5 minutes, to maintain the real-time nature of the data, so as to update the passenger and freight flow data of the road section in real time and generate real-time passenger and freight flow data of the road section. The passenger and freight flow of the road section is predicted by combining the real-time passenger and freight flow data of the road section using the capacity restriction-traffic allocation model. The passenger and freight flow data updated in real time are embedded in the model to reflect the latest traffic conditions. Considering the timeliness of the model, methods such as time decay can be used to make the latest data more weighted. The prediction results can include the distribution of passenger and freight flow in different time periods, as well as possible congestion conditions and other information, thereby generating the predicted data of passenger and freight flow of the road section.

[0166] Preferably, step S4 comprises the following steps:

[0167] Step S41: analyzing and processing the truck saturation of the road section according to the predicted data of passenger and freight flow of the road section, and generating predicted data of the truck saturation of the road section;

[0168] Step S42: Identify the main road sections for freight logistics based on the road section truck saturation prediction data to generate the main road section data for freight logistics.

[0169] The present invention performs a road section truck saturation analysis based on the road section passenger and freight flow prediction data, identifies and analyzes the truck load of the road section. Truck saturation refers to the ratio between the number of trucks passing through a certain road section and the maximum capacity of the road section, which helps to determine which road sections may face congestion, saturation or overload, thereby providing important information for traffic management and planning. The main sections of freight logistics are identified for the divided road section data based on the road section truck saturation prediction data, which helps to determine the key sections in the network, that is, those sections that may become bottlenecks or important nodes for freight logistics. Identifying the main freight sections helps to optimize the flow path of goods, reduce congestion, and improve the overall freight logistics efficiency.

[0170] In the embodiment of the present invention, the passenger and freight section flow prediction data is used to calculate the truck saturation of each section, and the truck saturation can be estimated by the ratio of the actual truck flow to the total passenger and freight flow. A threshold value of truck saturation is set, such as 80% or 90%, to define the congestion condition of the section, and sections with saturation exceeding the threshold are marked to indicate that these sections are the main routes for trucks, forming the main section data of freight logistics, and comparing and analyzing with the actual road network channels to determine the accuracy of the main section data of freight logistics.

[0171] The beneficial effect of the present application is that the freight logistics characteristics analysis method based on highway network toll data of the present invention automatically monitors the main routes of freight logistics, and considers information such as road congestion, road section tolls, and road section travel time, so that truck path selection and flow distribution are more objective and detailed, and constructs a path selection method based on a multi-logit model, and finally distributes flow to the truck starting point data, calculates the truck driving saturation to identify the main channel of truck logistics, and fully analyzes the characteristics of highway truck logistics.

[0172] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0173] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for analyzing freight logistics characteristics based on highway network toll collection data, characterized in that: The following steps are involved: Step S1: Obtaining highway section data; Performing road section identification division processing according to the expressway section data to generate divided road section data; Perform freight section characteristic data analysis on the divided section data to generate freight section characteristic data, wherein the freight section characteristic data includes section congestion status data, truck section fee data and vehicle travel time data; wherein step S1 includes: Step S11: Acquire high-speed road section data; Step S12: extracting high-speed gantry coordinates from the high-speed road section data to obtain high-speed gantry coordinate data; dividing the high-speed road section data into sections according to the high-speed gantry data to generate initial divided road section data; performing attribute identification processing of the divided road sections on the initial divided road section data to generate divided road section data; Step S13: using monitoring equipment to collect vehicle data of the divided road sections to generate road section vehicle data; Step S14: collecting vehicle travel time data according to the road section vehicle data to generate vehicle travel time data; Step S15: calculating the passenger and freight flow of the road section according to the road section vehicle data, and generating the passenger and freight flow data of the road section; Step S16: Performing a road section congestion state clustering analysis on the vehicle travel time data based on the K-Means clustering algorithm to generate road section congestion state data; wherein step S16 includes: Step S161: performing cluster center data selection and marking according to vehicle travel time data to obtain cluster center data; Step S162: using the Euclidean distance algorithm to calculate the cluster distance of the vehicle travel time data and the cluster center data to generate cluster distance data; Step S163: performing iterative updating processing on the cluster center data to obtain updated cluster center data, performing Euclidean distance calculation on the updated cluster center data and the unupdated cluster center data, and generating cluster center deviation data; Step S164: Return to step S163 and step S164 to perform clustering iteration. When the cluster center deviation data is less than the preset deviation threshold, stop the clustering iteration; perform a road section congestion status clustering analysis on the cluster distance data after the clustering iteration based on the K-Means clustering algorithm to generate road section congestion status data; Step S17: Calculate the truck section fee according to the preset freight vehicle charging requirements and the passenger and freight flow data of the section, and generate the truck section fee data; Step S18: integrating the road section congestion status data, the truck section fee data and the vehicle travel time data to generate freight section characteristic data; Step S2: Establishing a mathematical relationship model between capacity restriction and traffic distribution based on the Logit model and freight section characteristic data, and generating a capacity restriction-traffic distribution model; Step S3: predicting the passenger and freight traffic flow on a road section based on the capacity constraint-traffic allocation model, and generating predicted data on the passenger and freight traffic flow on the road section; Step S4: marking the freight logistics main section of the divided section data according to the passenger and freight flow prediction data of the section, and generating the freight logistics main section data.

2. The freight logistics characteristics analysis method based on highway network toll data according to claim 1 is characterized in that: Step S161 includes the following steps: Perform mean square error and calculation based on preset initial cluster center data and vehicle travel time data to generate mean square error data; According to the cluster elbow theory and mean square error data, cluster number parameter analysis is performed to generate cluster number parameters; The cluster centers of the vehicle travel time data are randomly selected and marked according to the cluster quantity parameter to obtain the cluster center data.

3. The freight logistics characteristics analysis method based on highway network toll data according to claim 1 is characterized in that: Step S2 includes the following steps: Step S21: using the minimum-maximum normalization method to normalize the freight section characteristic data respectively, to generate normalized freight section characteristic data; Step S22: calculating the freight section utility data based on the normalized freight section characteristic data to generate the freight section utility data; Step S23: A mathematical relationship model of capacity restriction and traffic distribution is established based on the Logit model and the freight section utility data to generate a capacity restriction-traffic distribution model.

4. The freight logistics characteristics analysis method based on highway network toll data according to claim 3 is characterized in that: Step S22 includes the following steps: Step S221: performing fixed utility data analysis of the freight section according to the normalized freight section characteristic data to generate freight section fixed utility data; Step S222: performing random utility data analysis of the freight section according to the preset freight section simulation data to generate random utility data of the freight section; Step S223: Calculate the freight section utility data according to the freight section fixed utility data and the freight section random utility data to generate the freight section utility data.

5. The freight logistics characteristics analysis method based on highway network toll collection data according to claim 3 is characterized in that: Step S23 includes the following steps: Step S231: Transmitting the freight section utility data to the Logit model to design the associated mapping relationship between capacity restriction and traffic distribution, and generating an initial capacity restriction-traffic distribution model; Step S232: using the Halton sequence to select target model parameters for the capacity constraint-traffic assignment model to obtain target model parameters; Step S233: Optimize the target model parameters according to the Monte Carlo simulation method to establish a capacity constraint-traffic assignment model.

6. The freight logistics characteristics analysis method based on highway network toll collection data according to claim 1 is characterized in that: Step S3 includes the following steps: Step S31: performing real-time updating processing on the road section passenger and freight flow data to generate real-time road section passenger and freight flow data; Step S32: predicting the passenger and freight traffic on a road section based on the real-time passenger and freight traffic data and the capacity constraint-traffic allocation model, and generating passenger and freight traffic prediction data on the road section.

7. The freight logistics characteristics analysis method based on highway network toll collection data according to claim 1 is characterized in that: Step S4 includes the following steps: Step S41: analyzing and processing the truck saturation of the road section according to the predicted data of passenger and freight flow of the road section, and generating predicted data of the truck saturation of the road section; Step S42: Identify the main road sections for freight logistics based on the road section truck saturation prediction data to generate the main road section data for freight logistics.

Images