A method and system for predicting inter-regional freight distribution of highways

Abstract

This invention relates to the field of transportation planning technology and discloses a method and system for predicting inter-regional freight distribution on highways. The method includes acquiring ETC freight traffic data for highways within a target area and constructing an intercity freight origin-destination matrix; collecting economic and social indicators and transportation network indicators for each city and extracting a comprehensive quality index using principal component analysis; introducing spatial impedance variables to construct an improved gravity model and calibrating the parameters using the intercity freight origin-destination matrix; predicting economic and social indicators using a grey prediction model and updating the comprehensive quality index based on the prediction results; and substituting the updated comprehensive quality index into the calibrated improved gravity model to obtain the intercity freight origin-destination prediction results. This invention comprehensively characterizes regional attractiveness by constructing a comprehensive quality index and integrates multiple impedance forms with grey prediction, effectively improving the interpretability of the freight distribution model and the accuracy of medium- and long-term predictions.

Description

Technical Field

[0001] This invention relates to the field of transportation planning technology, specifically to a method and system for predicting freight distribution between highway regions. Background Technology

[0002] Highway freight traffic is a core component of regional logistics systems, and its travel patterns have a significant impact on transportation planning and logistics network optimization. Accurately predicting the spatiotemporal distribution of highway freight is a crucial prerequisite for investing in transportation infrastructure, optimizing road network structure, and formulating traffic management policies.

[0003] Currently, gravity models are widely used in traffic flow modeling due to their clear theoretical foundation, simple structure, and ability to effectively characterize spatial interactions and the strength of nodal connections. Researchers are continuously expanding the applications of gravity models, attempting to incorporate socioeconomic indicators into the models, allowing them to act as mass components within the gravitational framework to characterize the attractiveness of cities or regions.

[0004] However, existing highway freight distribution forecasting techniques still have the following shortcomings: First, the quality indicators are too simplistic. Existing techniques typically use single factors such as GDP or population size as urban quality indicators, which fails to comprehensively reflect the city's overall attractiveness in terms of economic activity, consumer demand, and transportation network structure, resulting in insufficient model explanatory power. Second, the selection of impedance variables is limited. In characterizing spatial impedance, traditional studies often use a single impedance function based on distance, rarely constructing and comparing the roles of multiple impedance measures (such as straight-line distance and actual travel time) in freight distribution modeling, failing to fully capture the impact of key factors such as time cost on freight connections. Third, the predictive power is insufficient. Existing techniques are mostly static analyses, lacking a medium- to long-term forecasting framework that dynamically predicts key indicators and effectively combines them with gravity models, making it difficult to support the dynamic planning of regional logistics systems.

[0005] Therefore, how to construct a more comprehensive regional attractiveness index, how to select the optimal form of spatial impedance, and how to achieve dynamic forecasting of freight demand to improve the model's interpretability and forecasting accuracy are key issues that urgently need to be addressed. Summary of the Invention

[0006] To address the aforementioned shortcomings in the existing technology, this invention provides a method and system for predicting freight distribution between highway areas.

[0007] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows:

[0008] In a first aspect, the present invention proposes a method for predicting freight distribution across highway regions, comprising the following steps:

[0009] Acquire ETC freight traffic data for highways within the target area, and construct an intercity freight origin-destination matrix based on the acquired data;

[0010] Economic and social indicators and transportation network indicators of each city were collected, and principal component analysis was used to extract comprehensive quality indicators for each city.

[0011] An improved gravity model was constructed by introducing spatial impedance variables, and the parameters of the improved gravity model were calibrated using the intercity freight origin-destination matrix.

[0012] The grey prediction model is used to predict economic and social indicators, and the comprehensive quality indicators are updated based on the prediction results.

[0013] Substituting the updated comprehensive quality index into the calibrated improved gravity model, we obtain the intercity freight origin-destination prediction results.

[0014] Preferably, before extracting the comprehensive quality indicators for each city using principal component analysis, the following steps are also included:

[0015] Obtain candidate economic and social indicators and transportation network indicators datasets, as well as freight connection strength data represented by the intercity freight origin-destination matrix;

[0016] Calculate the correlation between each candidate indicator and the strength of freight linkage, and perform a significance test;

[0017] Candidate indicators whose significance level does not reach the preset threshold are removed from the dataset, and the indicators that pass the significance test are retained to form the effective indicator set.

[0018] Preferably, a spatial impedance variable is introduced to construct an improved gravity model, and the parameters of the improved gravity model are calibrated using the intercity freight origin-destination matrix, including:

[0019] Construct corresponding improved gravity models by using at least two different spatial impedance variables as inputs;

[0020] The parameters of each constructed improved gravity model are calibrated using the intercity freight origin-destination matrix, and the goodness of fit of each calibrated improved gravity model is calculated.

[0021] The model with the highest goodness of fit was selected as the final calibrated improved gravity model.

[0022] As a preferred embodiment, the improved gravity model is as follows:

[0023] ;

[0024] in, The strength of freight transport links between city i and city j. This represents the comprehensive quality index of city i. Let be the comprehensive quality index of city j. For calibration parameters, This represents the space impedance.

[0025] Preferably, the spatial impedance variable includes at least one of the following: straight-line distance between cities, shortest travel distance between cities, and average travel time between cities.

[0026] Preferably, parameter calibration of each constructed improved gravity model is performed using the intercity freight origin-destination matrix, including:

[0027] A logarithmic transformation of the improved gravity model yields its linear form:

[0028] ;

[0029] Based on the intercity freight origin-destination matrix, comprehensive quality index, and spatial impedance variable, the least squares method is used to perform regression analysis on the linear form of the improved gravity model to obtain the estimated calibration parameters.

[0030] Preferably, using a grey prediction model to predict economic and social indicators includes:

[0031] For each city and each economic and social indicator that needs to be predicted, an independent grey prediction model GM(1,1) is established to make predictions, and the predicted values ​​of economic and social indicators for each city in future years are obtained.

[0032] Preferably, updating the comprehensive quality index based on the prediction results includes:

[0033] By combining the predicted economic and social indicators with transportation network indicators, an indicator dataset for future years is formed.

[0034] By using the loadings and score coefficients determined by principal component analysis, the comprehensive quality indicators of each city in future years are calculated.

[0035] As a preferred option, the economic and social indicators include at least one of the following: regional GDP, total industrial output, foreign trade imports, year-end resident population, and total retail sales of consumer goods.

[0036] Traffic network indicators include at least one of the following: highway mileage and node betweenness centrality calculated based on road network topology.

[0037] Secondly, this invention also proposes a freight distribution prediction system for highways across regions, which applies the method described above, including:

[0038] The data acquisition and processing module is used to acquire ETC freight traffic data of highways within the target area and construct an intercity freight origin-destination matrix based on the acquired data.

[0039] The quality indicator calculation module is used to collect economic and social indicators and transportation network indicators of various cities, and to extract comprehensive quality indicators of each city using principal component analysis.

[0040] The model building and calibration module is used to introduce spatial impedance variables to build an improved gravity model and to calibrate the parameters of the improved gravity model using the intercity freight origin-end point matrix.

[0041] The dynamic prediction module is used to predict economic and social indicators using a grey prediction model, and update the comprehensive quality index based on the prediction results. The updated comprehensive quality index is then substituted into the calibrated improved gravity model to obtain the intercity freight origin-destination prediction results.

[0042] The present invention has the following beneficial effects:

[0043] This invention constructs an intercity freight OD matrix using ETC data; it integrates multi-dimensional indicators through principal component analysis to construct a comprehensive quality index, fully characterizing the region's overall attractiveness; it introduces various spatial impedance forms to construct an improved gravity model and compares them to determine the optimal model structure; and it combines a grey prediction model to establish an integrated analysis framework of "indicator prediction - quality update - freight distribution prediction," achieving medium- and long-term dynamic prediction. This invention solves the technical problems of single quality indicators, limited impedance forms, and insufficient prediction capabilities in existing technologies, providing a reliable technical path for regional logistics network optimization and transportation resource allocation. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of a method for predicting freight distribution between highway areas according to the present invention.

[0045] Figure 2 This is a schematic diagram of the structure of a highway inter-regional freight distribution prediction system according to the present invention. Detailed Implementation

[0046] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0047] like Figure 1As shown, an embodiment of the present invention provides a method for predicting freight distribution across highway regions, comprising the following steps S1 to S5:

[0048] S1. Obtain ETC freight traffic data for highways within the target area, and construct an intercity freight origin-destination matrix based on the obtained data;

[0049] In an optional embodiment of the present invention, step S1 takes the highway network of the target area as the research object and obtains ETC freight traffic data within a specific time period (e.g., September 2023). ETC data records vehicle entry, exit, and passage time information, and has the characteristics of wide coverage and high spatiotemporal accuracy. The raw data is cleaned to remove abnormal and missing records, and the data is aggregated from the toll station scale to the city scale according to the administrative division to which the toll station belongs. According to the vehicle type conversion factor, different types of freight trucks are uniformly converted into standard passenger car equivalents (PCUs), and the bidirectional traffic between cities is merged to construct an undirected intercity freight origin-destination matrix. The undirected intercity freight origin-destination matrix reflects the freight connection strength between cities. Regional freight connections show significant spatial imbalance characteristics, and freight traffic is mainly concentrated between cities with higher economic levels and better transportation conditions.

[0050] Existing methods typically use only a single socioeconomic indicator (such as GDP) as a measure of urban quality, which fails to comprehensively reflect a city's overall attractiveness in terms of economy, industry, consumption, and its position as a transportation network node, resulting in limited explanatory power for the models. Therefore, this embodiment utilizes freight traffic data from the highway ETC gantry system, and through data cleaning, vehicle type conversion, and spatial aggregation, directly constructs an intercity freight origin-destination matrix. This data source has the advantages of wide coverage, high spatiotemporal accuracy, and objectivity, providing high-quality observational input for the model.

[0051] S2. Collect economic and social indicators and transportation network indicators of each city, and use principal component analysis to extract comprehensive quality indicators of each city;

[0052] In an optional embodiment of the present invention, the economic and social indicators collected in step S2 include data such as regional GDP, the proportion of secondary industry, total industrial output, foreign trade imports, year-end resident population, and retail sales of consumer goods, used to characterize the regional economic scale and demand level. Transportation network indicators include highway mileage and node betweenness centrality. Node betweenness centrality is calculated based on a highway network construction topology model. This model uses each city (prefecture) as nodes and the highway connections between cities as edges, forming an undirected network to characterize the intermediary and hub role of cities in the network. Highway mileage characterizes the level of transportation infrastructure.

[0053] In an optional embodiment of the present invention, before step S2 extracts the comprehensive quality indicators of each city using principal component analysis, the method further includes:

[0054] Obtain candidate economic and social indicators and transportation network indicators datasets, as well as freight connection strength data represented by the intercity freight origin-destination matrix;

[0055] Calculate the correlation between each candidate indicator and the strength of freight linkage, and perform a significance test;

[0056] Candidate indicators whose significance level does not reach the preset threshold are removed from the dataset, and the indicators that pass the significance test are retained to form the effective indicator set.

[0057] To improve the explanatory power of the comprehensive quality indicators for freight distribution, this embodiment first tests the correlation between candidate indicators and freight volume before conducting principal component analysis. Using intercity freight volume as the dependent variable, significance analysis is performed on each explanatory variable, eliminating indicators with no significant correlation to freight volume. The test results are shown in Table 1. The significance level of the secondary industry proportion and the nodal centrality indicators is higher than 0.05, failing the significance test, and therefore they are removed from the indicator system. All other variables are statistically significant and are retained.

[0058] Table 1. Results of Correlation Analysis

[0059]

[0060] To eliminate the influence of differences in the units and orders of magnitude of different indicators, this embodiment first performs standardization processing on the original data:

[0061] ;

[0062] in, Let k be the k-th indicator of city i. and These are the mean and standard deviation of the indicator, respectively.

[0063] Then, the standardized data that passed the significance test were subjected to the Kaiser-Meyer-Olkin (KMO) measure and Bartlett's test of sphericity. In this example, the KMO value was 0.8 (>0.6), and the Bartlett's test of sphericity was significant (p<0.001), indicating that the correlation between the variables was strong and suitable for principal component analysis.

[0064] Next, principal component analysis is used to reduce the dimensionality of the multidimensional variables. Let the standardized variable vector be... , Let p be the number of variables, then the p-th principal component can be represented as:

[0065] ;

[0066] in, Principal component loadings, reflecting the first The contribution of each variable to the p-th principal component.

[0067] The selection of principal components is determined based on the magnitude of the eigenvalues ​​and the cumulative variance contribution rate. In this embodiment, the principal component analysis results are shown in Table 2. The eigenvalue of the first principal component is 5.542, and the variance contribution rate reaches 79.168%. Based on the principle that the eigenvalue is greater than 1 or the cumulative variance contribution rate reaches a preset threshold (such as 80%), the first principal component is selected as the comprehensive quality index.

[0068] Table 2 Explanation of Total Variance

[0069]

[0070] The loadings and score coefficients of the first principal component are shown in Table 3.

[0071] Table 3 First Principal Component Loads and Score Coefficients

[0072]

[0073] Calculate the comprehensive quality index for each city based on the loadings and score coefficients of each indicator on the first principal component. :

[0074] ;

[0075] in, The comprehensive score coefficient is determined by both principal component loadings and eigenvalues.

[0076] The above method enables the transformation of multidimensional indicators into a single comprehensive variable while preserving the main characteristics of the original information, thus providing a unified quality metric for subsequent model construction.

[0077] This embodiment breaks through the limitations of single indicators and comprehensively selects multi-dimensional indicators such as regional GDP, industrial output, population, consumption, highway mileage, and network centrality. Through principal component analysis, it performs dimensionality reduction and fusion to extract a "comprehensive quality score" that can comprehensively represent the city's development level and transportation location advantages, which significantly improves the representation ability of quality parameters in the gravity model.

[0078] S3. Introduce spatial impedance variables to construct an improved gravity model, and use the intercity freight origin-destination matrix to calibrate the parameters of the improved gravity model;

[0079] In an optional embodiment of the present invention, step S3 introduces spatial impedance variables to construct an improved gravity model, and uses the intercity freight origin-destination matrix to calibrate the parameters of the improved gravity model, including:

[0080] Construct corresponding improved gravity models by using at least two different spatial impedance variables as inputs;

[0081] The parameters of each constructed improved gravity model are calibrated using the intercity freight origin-destination matrix, and the goodness of fit of each calibrated improved gravity model is calculated.

[0082] The model with the highest goodness of fit was selected as the final calibrated improved gravity model.

[0083] This embodiment focuses on the undirected intensity of intercity connections rather than specific flow directions, and improves the classic gravity model. By constructing undirected "city pair" data and imposing symmetry constraints on the model, the improved gravity model is constructed as follows:

[0084] ;

[0085] in, The strength of freight transport links between city i and city j. This represents the comprehensive quality index of city i. Let be the comprehensive quality index of city j. For calibration parameters, This represents the space impedance.

[0086] This embodiment constructs multiple gravity models with straight-line distance, shortest travel distance, and average travel time as impedances to compare the impact of different spatial impedance variables on the explanatory power of the model. Spatial impedances can be obtained through tools such as map software APIs.

[0087] To facilitate model parameter estimation, this embodiment uses the least squares method to solve the model, and performs a logarithmic transformation on the above equation to obtain a linear form:

[0088] ;

[0089] Based on the intercity freight origin-destination matrix constructed in step S1 ( ) and the comprehensive quality index calculated in step S2 ( , ), and the space impedance variable obtained in step S3 ( The least squares method is used to perform regression analysis on the above linear form to estimate the parameters. , , In this embodiment, the parameter estimation and fitting results for the three gravity models are shown in Table 4.

[0090] Table 4 Comparison of Gravity Model Parameter Estimation and Fitting Results

[0091]

[0092] The three calibrated improved gravity models are as follows:

[0093] Straight-line distance model:

[0094] ;

[0095] Running distance model:

[0096] ;

[0097] Running time impedance model:

[0098] ;

[0099] The optimal model structure was determined by comparing the goodness of fit (e.g., adjusted R²) under different impedance variables. In this embodiment, all three models showed good fit, with the running time model having the highest adjusted R² (0.769) and the largest F-value (151.221), thus it was selected as the optimal model. All core parameters were significant at the 1% level, and the signs were consistent with theoretical expectations: city quality had a positive impact on freight volume (α=1.527), while distance impedance had a negative impact on freight volume (γ=1.189). From the multicollinearity diagnosis, the VIF values ​​of all variables were much less than 10, indicating that the model did not have a serious multicollinearity problem.

[0100] Existing methods for measuring spatial obstacles between cities often rely on single geographical metrics such as straight-line distance, failing to effectively incorporate impedance factors that better reflect real transportation costs, such as actual travel distance and travel time. Furthermore, the lack of comparison and optimization of different impedance forms limits the potential for improving model fitting accuracy. Therefore, this embodiment systematically introduces and compares three spatial impedance variables—"straight-line distance," "shortest travel distance," and "average travel time"—into the gravity model. By constructing and calibrating multiple models, and objectively selecting the optimal impedance form based on adjusting statistical indicators such as R², the model better reflects the time cost sensitivity of freight transport behavior.

[0101] S4. Use the grey prediction model to predict economic and social indicators, and update the comprehensive quality indicators based on the prediction results.

[0102] In an optional embodiment of the present invention, step S4, which uses a grey prediction model to predict economic and social indicators, includes:

[0103] For each city and each economic and social indicator that needs to be predicted, an independent grey prediction model GM(1,1) is established to make predictions, and the predicted values ​​of economic and social indicators for each city in future years are obtained.

[0104] In this embodiment, to achieve medium- and long-term forecasting of regional freight demand, an analytical framework of "indicator forecasting—quality updating—OD forecasting" is constructed. Due to limitations in data availability, city-level economic indicators often exhibit small sample characteristics. Therefore, this invention uses the grey prediction model GM(1,1) to extrapolate and forecast various economic indicators.

[0105] In this embodiment, a grey differential equation is established for each prefecture-level city and each socio-economic characteristic, and the grey prediction model GM(1,1) is constructed through the following process:

[0106] For each economic and social indicator in each city, construct the original data sequence. :

[0107] ;

[0108] in, For the first The observation values ​​at each time point;

[0109] The original data sequence is summed once to generate a cumulative sequence, thereby reducing the randomness of the original sequence:

[0110] ;

[0111] in, For the first accumulator in a sequence The observation value at each moment, For the first The observation values ​​at each time point;

[0112] A grey differential equation is established based on the cumulative sequence, and the development coefficient and grey action quantity are estimated using the least squares method:

[0113] ;

[0114] in, The differential symbol, For time, For development coefficient, This is the gray action quantity;

[0115] Based on the development coefficient and the grey effect, a prediction formula is derived to calculate the future predicted values ​​of economic and social indicators:

[0116] ;

[0117] in, For the first accumulator in a sequence The predicted value at each time point;

[0118] If the average relative error of the prediction is greater than a preset threshold (e.g., 5%), then the residual GM(1,1) model is used for correction.

[0119] ;

[0120] in, For the first data sequence in the original data sequence The predicted value at each time point, For the first accumulator in a sequence The predicted value at each time point.

[0121] The above grey prediction model was used to predict the economic and social indicators (such as GDP and total industrial output) of 14 cities in a certain province, and the predicted values ​​for 2026-2035 were obtained.

[0122] In an optional embodiment of the present invention, step S4, updating the comprehensive quality index based on the prediction results, includes:

[0123] By combining the predicted economic and social indicators with transportation network indicators, an indicator dataset for future years is formed.

[0124] By using the loadings and score coefficients determined by principal component analysis, the comprehensive quality indicators of each city in future years are calculated.

[0125] In this embodiment, when updating the comprehensive quality index, the predicted economic and social indicator data is used in conjunction with transportation network indicators (such as highway mileage and node betweenness centrality, assuming the network structure remains unchanged in the short term or adjusting based on planning data) to construct a future indicator dataset. Then, using the loadings and score coefficients determined during principal component extraction in step S2, the future indicator dataset is calculated to obtain the future comprehensive quality index. .

[0126] S5. Substitute the updated comprehensive quality index into the calibrated improved gravity model to obtain the intercity freight origin-destination prediction results.

[0127] In an optional embodiment of the present invention, step S5 updates the comprehensive quality index for future years predicted in step S4. By substituting the optimal improved gravity model calibrated in step S3 (i.e., the model with running time equal to impedance), the freight transport link strength between cities in future years can be calculated. This enables dynamic prediction of the origin and destination of intercity freight transport.

[0128] Since existing methods are mostly based on static modeling and fitting of historical data, they lack an effective mechanism to incorporate the future development trends of key driving indicators (economic and social indicators) into the prediction framework. This makes them unable to support medium- and long-term, dynamic freight demand forecasting and fails to meet the forward-looking needs of transportation infrastructure planning. Therefore, this embodiment creatively introduces the grey prediction model GM(1,1) to predict the future values ​​of various economic and social indicators, forming a dynamic prediction closed loop of grey prediction + gravity model. The predicted values ​​are used to update the comprehensive quality indicators of each city, and then substituted into the calibrated optimal gravity model to achieve dynamic prediction of the intercity freight OD distribution in future years, forming a complete analytical framework from indicator prediction to result output.

[0129] In summary, this invention starts with ETC data to construct an intercity freight OD matrix; it integrates multi-dimensional indicators through principal component analysis to construct a comprehensive quality index; it introduces various spatial impedance forms to construct and compare an improved gravity model; and it combines a grey prediction model to establish an integrated dynamic prediction framework. This invention solves the problems of single quality indicators, limited impedance forms, and insufficient predictive capabilities in existing technologies, improving the model's interpretability and prediction accuracy, and providing a reliable technical path for regional logistics network optimization and transportation resource allocation.

[0130] like Figure 2 As shown in the figure, an embodiment of the present invention provides a highway inter-regional freight distribution prediction system, comprising:

[0131] The data acquisition and processing module is used to acquire ETC freight traffic data of highways within the target area and construct an intercity freight origin-destination matrix based on the acquired data.

[0132] The quality indicator calculation module is used to collect economic and social indicators and transportation network indicators of various cities, and to extract comprehensive quality indicators of each city using principal component analysis.

[0133] The model building and calibration module is used to introduce spatial impedance variables to build an improved gravity model and to calibrate the parameters of the improved gravity model using the intercity freight origin-end point matrix.

[0134] The dynamic prediction module is used to predict economic and social indicators using a grey prediction model, and update the comprehensive quality index based on the prediction results. The updated comprehensive quality index is then substituted into the calibrated improved gravity model to obtain the intercity freight origin-destination prediction results.

[0135] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0136] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0137] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0138] Specific embodiments have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

[0139] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. A method for predicting freight distribution across highway regions, characterized in that, Includes the following steps: Acquire ETC freight traffic data for highways within the target area, and construct an intercity freight origin-destination matrix based on the acquired data; Economic and social indicators and transportation network indicators of each city were collected, and principal component analysis was used to extract comprehensive quality indicators for each city. An improved gravity model was constructed by introducing spatial impedance variables, and the parameters of the improved gravity model were calibrated using the intercity freight origin-destination matrix. The grey prediction model is used to predict economic and social indicators, and the comprehensive quality indicators are updated based on the prediction results. Substituting the updated comprehensive quality index into the calibrated improved gravity model, we obtain the intercity freight origin-destination prediction results.

2. The method for predicting inter-regional freight distribution on highways according to claim 1, characterized in that, Before using principal component analysis to extract the comprehensive quality indicators for each city, the following steps are also required: Obtain candidate economic and social indicators and transportation network indicators datasets, as well as freight connection strength data represented by the intercity freight origin-destination matrix; Calculate the correlation between each candidate indicator and the strength of freight linkage, and perform a significance test; Candidate indicators whose significance level does not reach the preset threshold are removed from the dataset, and the indicators that pass the significance test are retained to form the effective indicator set.

3. The method for predicting inter-regional freight distribution on highways according to claim 1, characterized in that, An improved gravity model is constructed by introducing spatial impedance variables, and the parameters of the improved gravity model are calibrated using the intercity freight origin-destination matrix, including: Construct corresponding improved gravity models by using at least two different spatial impedance variables as inputs; The parameters of each constructed improved gravity model are calibrated using the intercity freight origin-destination matrix, and the goodness of fit of each calibrated improved gravity model is calculated. The model with the highest goodness of fit was selected as the final calibrated improved gravity model.

4. The method for predicting inter-regional freight distribution on highways according to claim 3, characterized in that, The improved gravity model is as follows: ； in, The strength of freight transport links between city i and city j. This represents the comprehensive quality index of city i. Let be the comprehensive quality index of city j. For calibration parameters, This represents the space impedance.

5. The method for predicting inter-regional freight distribution on highways according to claim 4, characterized in that, The spatial impedance variables include at least one of the following: straight-line distance between cities, shortest travel distance between cities, and average travel time between cities.

6. The method for predicting inter-regional freight distribution on highways according to claim 5, characterized in that, The parameters of each improved gravity model are calibrated using the intercity freight origin-destination matrix, including: A logarithmic transformation of the improved gravity model yields its linear form: ； Based on the intercity freight origin-destination matrix, comprehensive quality index, and spatial impedance variable, the least squares method is used to perform regression analysis on the linear form of the improved gravity model to obtain the estimated calibration parameters.

7. The method for predicting inter-regional freight distribution on highways according to claim 1, characterized in that, Using grey prediction models to forecast economic and social indicators includes: For each city and each economic and social indicator that needs to be predicted, an independent grey prediction model GM(1,1) is established to make predictions, and the predicted values ​​of economic and social indicators for each city in future years are obtained.

8. The method for predicting inter-regional freight distribution on highways according to claim 1, characterized in that, The overall quality indicators are updated based on the forecast results, including: By combining the predicted economic and social indicators with transportation network indicators, an indicator dataset for future years is formed. By using the loadings and score coefficients determined by principal component analysis, the comprehensive quality indicators of each city in future years are calculated.

9. The method for predicting inter-regional freight distribution on highways according to claim 1, characterized in that, Economic and social indicators include at least one of the following: regional GDP, total industrial output, foreign trade imports, year-end resident population, and total retail sales of consumer goods. Traffic network indicators include at least one of the following: highway mileage and node betweenness centrality calculated based on road network topology.

10. A highway inter-regional freight distribution prediction system, using the method described in any one of claims 1 to 9, characterized in that, include: The data acquisition and processing module is used to acquire ETC freight traffic data of highways within the target area and construct an intercity freight origin-destination matrix based on the acquired data. The quality indicator calculation module is used to collect economic and social indicators and transportation network indicators of various cities, and to extract comprehensive quality indicators of each city using principal component analysis. The model building and calibration module is used to introduce spatial impedance variables to build an improved gravity model and to calibrate the parameters of the improved gravity model using the intercity freight origin-end point matrix. The dynamic prediction module is used to predict economic and social indicators using a grey prediction model, and update the comprehensive quality index based on the prediction results. The updated comprehensive quality index is then substituted into the calibrated improved gravity model to obtain the intercity freight origin-destination prediction results.

Images