Road transportation capacity distribution prediction method integrating long-term and short-term fluctuation trends

Abstract

The invention provides a road transportation capacity distribution prediction method integrating long and short term fluctuation trends, which comprises the following steps: performing multi-stage decoupling on original data, respectively extracting transportation capacity spatial-temporal characteristics under different scales, and meanwhile, dynamically updating transportation capacity associated information by using a dynamic warping algorithm; on this basis, transport capacity space-time dependence is captured by using a graph attention mechanism and a diffusion and expansion convolutional network, and a multi-scale fused comprehensive space-time feature representation is generated by using a dynamic weight. The method comprehensively considers the space-time dependence of the transportation capacity under different scales, and facilitates the accurate prediction of the road transportation capacity distribution.

Description

Technical Field

[0001] The present invention relates to the field of highway transport capacity distribution prediction, and in particular to a highway transport capacity distribution prediction method integrating long-term and short-term fluctuation trends. Background Art

[0002] Highway transport capacity distribution is a key indicator to measure the highway transport capacity of a specific geographical area within a specific time frame, covering both freight and passenger transport. Its changes are affected by multiple factors such as regional attributes, geographical environment, climate conditions and human factors. With the rapid development of my country's economy and the acceleration of urbanization, the demand for inter-regional freight transportation is growing, and the demand for highway freight capacity forecasting is becoming increasingly urgent.

[0003] At present, with the optimization of logistics networks and the widespread application of intelligent technologies, how to effectively predict the highway transport capacity in a specific area and accurately perceive the changes in transport capacity has become a core issue to ensure regional logistics efficiency and optimize resource allocation. At present, there is a relative lack of research on the methods of predicting the distribution of highway transport capacity. Existing research mostly focuses on road flow prediction, resulting in a lack of strong technical support for the precise scheduling and management of regional highway transport capacity. Due to the complex transformation characteristics of transport capacity, traditional statistical-based prediction methods are difficult to fully capture the changes in transport capacity within the region, resulting in inefficient resource allocation and excess or insufficient transport capacity in actual transportation management. Therefore, there is an urgent need for a prediction method specifically for highway transport capacity distribution to improve the accuracy and scientificity of logistics and traffic management.

[0004] However, the prediction of regional highway transport capacity faces multiple challenges: First, regional highway transport capacity data presents complex time dependence and spatial heterogeneity, and its spatiotemporal correlation is highly dynamic in different regions and time periods. Second, the distribution of transport capacity is also affected by dynamic factors such as weather and emergencies, and the data shows complex distribution shift phenomena, which increases the difficulty of prediction. How to model and capture the temporal and spatial correlation between event information and individuals, and effectively deal with the distribution shift and spatiotemporal heterogeneity of transport capacity distribution in the prediction problem, is of great significance to improving freight efficiency and resource utilization in the region. Summary of the invention

[0005] The purpose of the present invention is to address the deficiencies of the above-mentioned prior art and provide a highway transport capacity distribution prediction method that integrates long-term and short-term fluctuation trends, which can accurately predict the distribution of auxiliary highway transport capacity.

[0006] To achieve the above object, the present invention adopts the following technical solutions: The present invention provides a method for predicting highway transport capacity distribution based on comprehensive long-term and short-term fluctuation trends, comprising the following steps: S1. Data acquisition and preprocessing, defining highway transport capacity evaluation indicators, and determining regional highway transport core points based on kernel density estimation and centroid analysis; S2. On the basis of determining the coordinates of the regional transportation core points, comprehensively considering the traffic conditions, geographical distance and traffic flow data, and based on the path planning algorithm, construct the initial highway capacity association network for subsequent spatiotemporal feature extraction and network update; S3. Split the historical highway capacity association network into association networks under overlapping time windows by moving steps. The time window size is ; Combining the moving average filter with the wavelet transform to separate the long-term and short-term trend characteristics of transport capacity, decouple the multi-scale trend of regional transport capacity; S4. Use trigonometric functions to encode external events and periodic information. At the same time, for the data in the input time window, use the dynamic regularization algorithm to update the capacity fluctuation correlation matrix. Based on the capacity fluctuation correlation moment, use the diffusion dilation convolution and graph attention mechanism to extract the cross-scale multi-period spatiotemporal characteristics of regional highway transportation capacity. S5. Adaptive weight learning and attention mechanism are used to fuse multi-scale spatiotemporal features and external dynamic influencing factors, and a multi-layer perceptron is used as the prediction head to realize the prediction of highway transport capacity distribution.

[0007] Further, the S1 is specifically: S101, obtaining data: obtaining historical data related to the highway transport capacity of the area to be predicted, including the types and quantities of trucks, regional road network information, truck transport record information, and event information of corresponding time; S102. Define the evaluation index of highway transport capacity: Highway transport capacity is determined by the types, number, carrying capacity and actual utilization rate of trucks in the region; The calculation formula of the highway transport capacity evaluation index is: ; in, The categories of truck types included in each region; For Region Middle Types of trucks; For Region Middle The single-vehicle carrying capacity of the truck; For Region Middle The actual utilization rate of each type of truck; For time; S103. Complete transport capacity data: Determine the update frequency of road transport capacity ; For regional highway transport capacity time series data ; If there are missing data at a time point, the spline interpolation method is used to complete it; For Region At the moment capacity data; S104. Determine the regional transportation core point to represent the regional transportation center: rasterize different data according to a unified spatial resolution and divide them into sub-regions to ensure that regional layout and transportation network data can be analyzed at the same spatial scale; In order to comprehensively consider the impact of adjacent areas on this area, the spatial density of population, economic activities and traffic activities in the sub-region is calculated using kernel density estimation; and the population density, economic activity density and traffic flow density are weighted and superimposed, giving different weights to different attribute spatial densities. The overall density of the sub-regions : ; in, is population density; is the density of economic activities; is the traffic flow density; is the kernel density function corresponding to population density; is the kernel density function corresponding to the density of economic activities; is the kernel density function corresponding to the traffic flow density; The expression of kernel density is: ; in, The sub-region for calculating the kernel density currently; For Region Total number of neutron regions; is the bandwidth parameter to control the smoothness; For the sub-regions; is the Gaussian kernel function; In the integrated density map, the centroid analysis method is used to identify the regions The comprehensive core transportation point, the core point coordinate calculation formula is: in, are the coordinates of the regional transportation core points respectively; are the coordinates of the grid center points of each area in the region; For the The comprehensive density value of each sub-area; is the total number of sub-regions; Further, the S2 is specifically: S201, node information construction: Based on the data obtained in S1, a highway transport capacity association network of regional hubs is constructed; in the association network, Determined by the core transport point coordinates, at each time point, The attribute of is the carrying capacity of the corresponding area; S202, determine the node edge weight: In the network, the edge represents the capacity relationship between regions; Weight Responsive regional nodes and regional nodes The capacity flow between them, the weight of the edge is: ; in, For regional nodes and regional nodes The geographical distance between them is calculated by Dijkstra's algorithm applied on the road network; For regional nodes and regional nodes Between time The transportation time is derived from historical data; For regional nodes and regional nodes Between time Traffic flow, derived from historical data; , , is the weight coefficient; S203, constructing the highway transport capacity association network adjacency matrix: for each time , construct the adjacency matrix : A t = [ a p q t ] ; in, If the regional node and regional nodes There is an edge between them and the distance Less than the threshold ,but ,otherwise ;in, is the edge weight of the adjacency matrix of the highway transport capacity association network; S204. Integrate the historical highway capacity correlation network: Integrate the highway capacity correlation networks of adjacent time slices to form a complete historical highway capacity correlation network ;in, is the node feature set of all time slices; is the set of adjacency matrices for all time slices.

[0008] Further, the S3 is specifically: S301. Preliminary decoupling: For the node capacity data within the time window, a moving average filter with different scale factors is used to remove the long-term and short-term trends of the capacity; the original capacity data is decomposed into long-term trend data and short-term fluctuation data. Node capacity sequence within a time window , long-term trend data is expressed as: ; in, is the original capacity data; An adjustable factor that controls the degree of decoupling for the moving average filter; For time The position index of The short-term trend data is: ; S302. Deep feature decoupling: Through wavelet decomposition, long-term trend data is further decoupled into multiple bands of different time scales to capture the capacity fluctuation characteristics at different time scales; short-term fluctuation data is decomposed into different frequency bands to capture small changes in a short period of time; for long-term trend data, the low-frequency band after decomposition is taken to represent the long-term trend characteristics; for short-term trend data, the high-frequency band after decomposition is taken to represent the short-term trend characteristics.

[0009] Further, the S4 is specifically: S401. For an input time window, identify an external event that affects the transport capacity; if the event occurs, the corresponding code value is 1; if the event does not occur, the code value is 0; Formally: ; in, for whether the practice occurs; In order to capture the changing trend of capacity in daily, weekly and monthly cycles, sine and cosine functions are used to encode the time period features to map the time periodicity into the feature space, ensuring that the model can recognize and utilize these periodic patterns; Assume that the cycle of a day is ,time The position during the day is , then the daily cycle code is: ; ; Assume that the weekly cycle is ,time The position of the week is , then the weekly cycle code is: ; ; Combine the periodic codes into a complete feature vector: F c y c l e ( t ) = [ s i n _ d a y ( t ) , c o s _ d a y ( t ) , s i n _ w e e k ( t ) , c o s _ w e e k ( t ) ] ; Time The external event code and the periodic feature code are combined into a complete feature vector : ; in, For feature stitching.

[0010] S402: For data in the input time window: using a dynamic time normalization algorithm to calculate the similarity between the time series of the input time window nodes , set the similarity threshold , If the node and The sequence similarity between , then update the weight ; If the regional node and regional nodes The sequence similarity between , it remains unchanged; S403, spatial feature extraction: for the capacity association network within a period of time, the graph attention mechanism is used to extract spatial dependency features from the capacity association network; in In the layer graph attention mechanism, the node represents the vector By performing a weighted summation with the features of adjacent nodes and updating the features through nonlinear transformation: ; in, For the The node feature matrix of the layer; is the capacity fluctuation correlation matrix; is the activation function; For the The learned weight matrix of the layer; For regional nodes and regional nodes The attention weight between them is calculated by the features of the node pair: α p q ( l ) = e x p ( L e a k y R e L U ( ο T [ W ( l ) H q ( l ) | | W ( l ) H p ( l ) ] ) ) ∑ p , q ∈ A c e x p ( ο T [ W ( l ) H q ( l ) | | W ( l ) H p ( l ) ] ) ; in, It is the concatenation operation of vectors; is the vector used to calculate attention; S404, Time Feature Extraction: Use dilated convolution to extract time-dependent features in time series; dilated convolution can capture long-term and short-term dependencies in time series by enhancing the hole convolution of the receptive field, and use dilated convolution to operate on the input time series at each layer: ; in, For the The input sequence of the layer; is dilated convolution; is bias; It is the convolution calculation; The calculation formula for the dilated convolution is: C o n v ( l ) = ∑ p o s p o s s − 1 W [ p o s ] ⋅ Z [ t − d ⋅ p o s ] ; in, is the convolution kernel size; is the expansion rate; is the index of the convolution kernel on the input sequence.

[0011] Further, the S5 is specifically: S501, integrating spatiotemporal features: performing feature concatenation on the last layer of temporal features and spatial features encoded in S4: ; in, For the last layer; Dynamically and hierarchically fuse the spatial and temporal features of the original sequence, long-term sequence, and short-term sequence; introduce an adaptive weight learning mechanism to dynamically learn the fusion weights of features of different scales through a neural network and set weight parameters , and Corresponding to the original features , long-term characteristics and short-term characteristics , the multi-scale spatiotemporal features are: ; S502, comprehensive feature fusion: Use the attention mechanism to fuse multi-scale spatiotemporal features with external influencing factor features, specifically: ; in, It is the multi-scale spatiotemporal feature representation after the final fusion; Encoding features for external events and key time nodes; is the fusion feature dimension calculated using the attention mechanism; S503, receiving the fused spatiotemporal feature vector As input, the spatiotemporal feature vector Input into the multi-layer perceptron to generate the final prediction value: ; in, P G ̂ = [ P G t , ̂ P G t + T ̂ , … P G t + u T ̂ ] For space-time capacity network The predicted value of highway transport capacity includes the predicted value of transport capacity and relationship distribution characteristics of each regional node in the graph; To predict forward time steps; S504, model training and updating: The training goal of the model is to minimize the predicted value The actual capacity The model parameters are optimized through training to generate the best prediction values ​​for actual capacity scheduling and management.

[0012] The beneficial effects of the present invention are: combining kernel density estimation and centroid analysis to identify the core points of regional highway transportation, while taking into account the internal correlation of the region, effectively integrating multi-source heterogeneous data information such as population, economic activities and traffic flow, ensuring the accuracy of the representative characteristics of spatial information. At the same time, taking into account information such as traffic conditions and geographical distance, a highway transportation capacity association network is constructed based on the path planning algorithm, laying the foundation for subsequent effective spatiotemporal feature extraction; By decoupling the original data at multiple levels and extracting the spatiotemporal characteristics of capacity at different scales, the dynamic regularization algorithm is used to dynamically update the capacity correlation information. On this basis, the graph attention mechanism and the diffusion expansion convolutional network are used to capture the spatiotemporal dependency of capacity, and the dynamic weights are used to generate a comprehensive spatiotemporal feature representation of multi-scale fusion. This method comprehensively considers the spatiotemporal dependency of capacity at different scales, which is conducive to the accurate prediction of highway transport capacity distribution. By finely encoding external events and periodic information, dynamic external influencing factors are effectively integrated into the model. Multi-scale spatiotemporal information is integrated with external information through the attention mechanism, and the impact of external factors and spatiotemporal information on capacity changes is adaptively learned to improve the model's prediction performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a flowchart of a highway transport capacity distribution prediction method that integrates long-term and short-term fluctuation trends; Figure 2 Schematic diagram of diffusion dilation convolution; Figure 3 A network architecture diagram for highway transport capacity forecast that integrates long-term and short-term trends; Figure 4Regional transportation capacity data of Nanjing, Nantong, Suzhou, Wuxi, Xuzhou and Yancheng from November 2020 to July 2021. DETAILED DESCRIPTION

[0014] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0015] See also Figure 1 , a highway transport capacity distribution prediction method integrating long-term and short-term fluctuation trends, comprising the following steps: S1. Data acquisition and preprocessing, defining highway transport capacity evaluation indicators, and determining regional highway transport core points based on kernel density estimation and centroid analysis; S2. On the basis of determining the coordinates of the regional transportation core points, comprehensively considering the traffic conditions, geographical distance and traffic flow data, and based on the path planning algorithm, construct the initial highway capacity association network for subsequent spatiotemporal feature extraction and network update; S3. Split the historical highway capacity association network into association networks under overlapping time windows by moving steps. The time window size is ; Combining the moving average filter with the wavelet transform to separate the long-term and short-term trend characteristics of transport capacity, decouple the multi-scale trend of regional transport capacity; S4. Use trigonometric functions to encode external events and periodic information. At the same time, for the data in the input time window, use the dynamic regularization algorithm to update the capacity fluctuation correlation matrix; based on the capacity fluctuation correlation moment, use the diffusion dilation convolution and graph attention mechanism to extract the cross-scale multi-period spatiotemporal characteristics of regional highway transportation capacity, such as Figure 2 As shown; S5. Adaptive weight learning and attention mechanism are used to integrate multi-scale spatiotemporal features and external dynamic influencing factors, and multi-layer perceptron is used as the prediction head to realize the prediction of highway transport capacity distribution, such as Figure 3 shown.

[0016] The S1 is specifically: S101, obtaining data: obtaining historical data related to the highway transport capacity of the area to be predicted, including the types and quantities of trucks, regional road network information, truck transport record information, and event information of corresponding time; S102. Define the evaluation index of highway transport capacity: Highway transport capacity is determined by the types, number, carrying capacity and actual utilization rate of trucks in the region; The calculation formula of the highway transport capacity evaluation index is: ; in, The categories of truck types included in each region; For Region Middle Types of trucks; For Region Middle The single-vehicle carrying capacity of the truck; For Region Middle The actual utilization rate of each type of truck; For time; S103. Complete transport capacity data: Determine the update frequency of road transport capacity ; For regional highway transport capacity time series data ; If there are missing data at a time point, the spline interpolation method is used to complete it; For Region At the moment capacity data; S104. Determine the regional transportation core point to represent the regional transportation center: rasterize different data according to a unified spatial resolution and divide them into sub-regions to ensure that regional layout and transportation network data can be analyzed at the same spatial scale; In order to comprehensively consider the impact of adjacent areas on this area, the spatial density of population, economic activities and traffic activities in the sub-region is calculated using kernel density estimation; and the population density, economic activity density and traffic flow density are weighted and superimposed, giving different weights to different attribute spatial densities. The overall density of the sub-regions : ; in, is population density; is the density of economic activities; is the traffic flow density; is the kernel density function corresponding to population density; is the kernel density function corresponding to the density of economic activities; is the kernel density function corresponding to the traffic flow density; The expression of kernel density is: ; in, The sub-region for calculating the kernel density currently; For Region Total number of neutron regions; is the bandwidth parameter to control the smoothness; For the sub-regions; is the Gaussian kernel function; In the integrated density map, the centroid analysis method is used to identify the regions The comprehensive core transportation point, the core point coordinate calculation formula is: in, are the coordinates of the regional transportation core points respectively; are the coordinates of the grid center points of each area in the region; For the The comprehensive density value of each sub-area; is the total number of sub-regions; The S2 is specifically: S201, node information construction: Based on the data obtained in S1, a highway transport capacity association network of regional hubs is constructed; in the association network, Determined by the core transport point coordinates, at each time point, The attribute of is the carrying capacity of the corresponding area; S202, determine the node edge weight: In the network, the edge represents the capacity relationship between regions; Weight Responsive regional nodes and regional nodes The capacity flow between them, the weight of the edge is: ; in, For regional nodes and regional nodes The geographical distance between them is calculated by Dijkstra's algorithm applied on the road network; For regional nodes and regional nodes Between time The transportation time is derived from historical data; For regional nodes and regional nodes Between time Traffic flow, derived from historical data; , , is the weight coefficient; S203, constructing the highway transport capacity association network adjacency matrix: for each time, constructing the adjacency matrix : A t = [ a p q t ] ; in, If the regional node and regional nodes There is an edge between them and the distance Less than the threshold ,but ,otherwise ;in, is the edge weight of the adjacency matrix of the highway transport capacity association network; S204. Integrate the historical highway capacity correlation network: Integrate the highway capacity correlation networks of adjacent time slices to form a complete historical highway capacity correlation network ;in, is the node feature set of all time slices; is the set of adjacency matrices for all time slices.

[0017] The S3 is specifically: S301. Preliminary decoupling: For the node capacity data within the time window, a moving average filter with different scale factors is used to remove the long-term and short-term trends of the capacity; the original capacity data is decomposed into long-term trend data and short-term fluctuation data. Node capacity sequence within a time window , long-term trend data is expressed as: ; in, is the original capacity data; An adjustable factor that controls the degree of decoupling for the moving average filter; For time The position index of The short-term trend data is: ; S302. Deep feature decoupling: Based on the preliminary decoupling, the long-term and short-term trend features are deeply decoupled based on wavelet transform to more finely separate the multi-scale information in the capacity data. Specifically, through wavelet decomposition, the long-term trend data is further decoupled into multiple bands of different time scales to capture the capacity fluctuation characteristics at different time scales; the short-term fluctuation data is decomposed into different frequency bands to capture small changes in a short period of time; for the long-term trend data, the low-frequency band after decomposition is taken to represent the long-term trend characteristics; for the short-term trend data, the high-frequency band after decomposition is taken to represent the short-term trend characteristics.

[0018] The S4 is specifically: S401. For the input time window, identify external events that affect the transport capacity, such as holidays, major events, natural disasters, etc. If the event occurs, the corresponding code value is 1; if the event does not occur, the code value is 0; Formally: ; in, for whether the practice occurs; In order to capture the changing trend of capacity in daily, weekly and monthly cycles, sine and cosine functions are used to encode the time period features to map the time periodicity into the feature space, ensuring that the model can recognize and utilize these periodic patterns; Assume that the cycle of a day is ,time The position during the day is , then the daily cycle code is: ; ; Assume that the weekly cycle is ,time The position of the week is , then the weekly cycle code is: ; ; Combine the periodic codes into a complete feature vector: F c y c l e ( t ) = [ s i n _ d a y ( t ) , c o s _ d a y ( t ) , s i n _ w e e k ( t ) , c o s _ w e e k ( t ) ] ; Time The external event code and the periodic feature code are combined into a complete feature vector : ; in, For feature stitching.

[0019] S402: For data in the input time window: using a dynamic time normalization algorithm to calculate the similarity between the time series of the input time window nodes , set the similarity threshold , if the regional node and regional nodes The sequence similarity between , then update the weight ; If the regional node and regional nodes The sequence similarity between , it remains unchanged; S403, spatial feature extraction: for the capacity association network within a period of time, the graph attention mechanism is used to extract spatial dependency features from the capacity association network; in In the layer graph attention mechanism, the node represents the vector By performing a weighted summation with the features of adjacent nodes and updating the features through nonlinear transformation: ; in, For the The node feature matrix of the layer; is the capacity fluctuation correlation matrix; is the activation function; For the The learned weight matrix of the layer; For regional nodes and regional nodes The attention weight between them is calculated by the features of the node pair: α p q ( l ) = e x p ( L e a k y R e L U ( ο T [ W ( l ) H q ( l ) | | W ( l ) H p ( l ) ] ) ) ∑ p , q ∈ A c e x p ( ο T [ W ( l ) H q ( l ) | | W ( l ) H p ( l ) ] ) ; in, It is the concatenation operation of vectors; is the vector used to calculate attention; S404, Time Feature Extraction: Use dilated convolution to extract time-dependent features in time series; dilated convolution can capture long-term and short-term dependencies in time series by enhancing the hole convolution of the receptive field, and use dilated convolution to operate on the input time series at each layer: ; in, For the The input sequence of the layer; is dilated convolution; is bias; It is the convolution calculation; The calculation formula for the dilated convolution is: C o n v ( l ) = ∑ p o s p o s s − 1 W [ p o s ] ⋅ Z [ t − d ⋅ p o s ] ; Among them, is the convolution kernel size; is the expansion rate; is the index of the convolution kernel on the input sequence.

[0020] The S5 is specifically: S501, integrating spatiotemporal features: performing feature concatenation on the last layer of temporal features and spatial features encoded in S4: ; in, For the last layer; Dynamically and hierarchically fuse the spatial and temporal features of the original sequence, long-term sequence, and short-term sequence; introduce an adaptive weight learning mechanism to dynamically learn the fusion weights of features of different scales through a neural network and set weight parameters , and Corresponding to the original features , long-term characteristics and short-term characteristics , the multi-scale spatiotemporal features are: ; S502, comprehensive feature fusion: Use the attention mechanism to fuse multi-scale spatiotemporal features with external influencing factor features, specifically: ; in, It is the multi-scale spatiotemporal feature representation after the final fusion; Encoding features for external events and key time nodes; is the fusion feature dimension calculated using the attention mechanism; S503, receiving the fused spatiotemporal feature vector As input, the spatiotemporal feature vector Input into the multi-layer perceptron to generate the final prediction value: ; in, P G ̂ = [ P G t , ̂ P G t + T ̂ , … P G t + u T ̂ ] For space-time capacity network The predicted value of highway transport capacity includes the predicted value of transport capacity and relationship distribution characteristics of each regional node in the graph; To predict forward time steps; S504, model training and updating: The training goal of the model is to minimize the predicted value The actual capacity The model parameters are optimized through training to generate the best prediction values ​​for actual capacity scheduling and management.

[0021] A highway transport capacity distribution prediction method that integrates long-term and short-term fluctuation trends: First, the historical capacity data of the area to be tested is obtained, and the data is preprocessed and completed; then, the regional transport core points are identified based on kernel density estimation and centroid analysis, and the initial highway transport capacity association network is constructed based on the path planning algorithm and reachable threshold considering traffic conditions, geographical distance and other data; then, the adjustable moving average filter and wavelet transform are combined to separate the long-term and short-term trend characteristics of the capacity sequence, and the regional node capacity sequence is multi-scale decoupled. At the same time, the external events and key time nodes that affect the capacity are finely encoded, and the capacity association information is dynamically updated using the dynamic regularization algorithm; on this basis, the graph attention mechanism and the diffusion expansion convolutional network are used to mine the spatiotemporal characteristics of the capacity, and the dynamic weights are used to fuse the spatiotemporal characteristics of the long-term and short-term trend capacity and the original capacity to form a comprehensive multi-scale spatiotemporal feature; finally, the multi-scale spatiotemporal features are fused with the external influencing factors using the attention mechanism and input into the multi-layer perceptron for feature mapping to achieve accurate prediction of highway transport capacity distribution.

[0022] Example This paper takes Jiangsu Province as the object of highway transport capacity distribution prediction, and divides it into 13 regions according to the geographical boundaries of prefecture-level cities. According to the highway transport capacity evaluation index formula, the historical highway transport capacity data of each region is calculated, and the transport capacity data of six cities, including Nanjing, Nantong, Suzhou, Wuxi, Xuzhou, and Yancheng, from November 2020 to July 2021 are intercepted, in units of 10,000 tons, with a sampling interval of 2 hours. Figure 4 shown.

[0023] At the same time, the transportation core points of each region are calculated as the regional transportation center, the initial transportation capacity association network is constructed, and the normalized adjacency matrix is ​​generated, as shown in Table 1.

[0024] Table 1 Normalized adjacency matrix of regional transport capacity association network

[0025] The event set of external factors in each region is obtained, and the historical data and event data are input into the model for decoupling, feature encoding and prediction. The prediction step is 5 time steps. The long short-term memory neural network LSTM is used for comparative experiments, and the prediction mean square error MSE and root mean square error RMSE of the two methods are calculated. The results are shown in Table 2. The MSE unit is 10,000 tons squared, and the RMSE unit is 10,000 tons. The comparison results show that compared with the traditional LSTM model, the highway transport capacity distribution prediction method based on the comprehensive long-term and short-term fluctuation trends proposed in this paper has higher accuracy in predicting regional capacity distribution, especially in cities such as Changzhou and Suzhou.

[0026] Table 2 Prediction error table

[0027] The above-mentioned embodiments only express the implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention should be based on the attached claims.

Claims

1. A highway transport capacity distribution prediction method that integrates long-term and short-term fluctuation trends, characterized in that: The following steps are involved: S1. Data acquisition and preprocessing, defining highway transport capacity evaluation indicators, and determining regional highway transport core points based on kernel density estimation and centroid analysis; S2. On the basis of determining the coordinates of the regional transportation core points, comprehensively considering the traffic conditions, geographical distance and traffic flow data, and based on the path planning algorithm, construct the initial highway capacity association network for subsequent spatiotemporal feature extraction and network update; S3. Split the historical highway capacity association network into association networks under overlapping time windows by moving steps. The time window size is ; Combining the moving average filter with the wavelet transform to separate the long-term and short-term trend characteristics of transport capacity, decouple the multi-scale trend of regional transport capacity; S4. Use trigonometric functions to encode external events and periodic information. At the same time, for the data in the input time window, use the dynamic regularization algorithm to update the capacity fluctuation correlation matrix. Based on the capacity fluctuation correlation moment, use the diffusion dilation convolution and graph attention mechanism to extract the cross-scale multi-period spatiotemporal characteristics of regional highway transportation capacity. S5. Adaptive weight learning and attention mechanism are used to fuse multi-scale spatiotemporal features and external dynamic influencing factors, and a multi-layer perceptron is used as the prediction head to realize the prediction of highway transport capacity distribution.

2. The highway transport capacity distribution prediction method based on comprehensive long-term and short-term fluctuation trends according to claim 1 is characterized in that: The S1 is specifically: S101, obtaining data: obtaining historical data related to the highway transport capacity of the area to be predicted, including the types and quantities of trucks, regional road network information, truck transport record information, and event information of corresponding time; S102. Define the evaluation index of highway transport capacity: Highway transport capacity is determined by the types, number, carrying capacity and actual utilization rate of trucks in the region; The calculation formula of the highway transport capacity evaluation index is: ； in, The categories of truck types included in each region; For Region Middle Types of trucks; For Region Middle The single-vehicle carrying capacity of the truck; For Region Middle The actual utilization rate of each type of truck; For time; S103. Complete transport capacity data: Determine the update frequency of road transport capacity ; For regional highway transport capacity time series data ; If there are missing data at a time point, the spline interpolation method is used to complete it; For Region At the moment capacity data; S104. Determine the regional transportation core point to represent the regional transportation center: rasterize different data according to a unified spatial resolution and divide them into sub-regions to ensure that regional layout and transportation network data can be analyzed at the same spatial scale; In order to comprehensively consider the impact of adjacent areas on this area, the spatial density of population, economic activities and traffic activities in the sub-region is calculated using kernel density estimation; and the population density, economic activity density and traffic flow density are weighted and superimposed, giving different weights to different attribute spatial densities. The overall density of the sub-regions : ； in, is population density; is the density of economic activities; is the traffic flow density; is the kernel density function corresponding to population density; is the kernel density function corresponding to the density of economic activities; is the kernel density function corresponding to the traffic flow density; The expression of kernel density is: ； in, The sub-region for calculating the kernel density currently; For Region Total number of neutron regions; is the bandwidth parameter to control the smoothness; For the sub-regions; is the Gaussian kernel function; In the integrated density map, the centroid analysis method is used to identify the regions The comprehensive core transportation point, the core point coordinate calculation formula is: ； ； in, are the coordinates of the regional transportation core points respectively; are the coordinates of the grid center points of each area in the region; For the The comprehensive density value of each sub-area; is the total number of sub-regions.

3. The highway transport capacity distribution prediction method based on comprehensive long-term and short-term fluctuation trends according to claim 2 is characterized in that: The S2 is specifically: S201, node information construction: Based on the data obtained in S1, a highway transport capacity association network of regional hubs is constructed; in the association network, Determined by the core transport point coordinates, at each time point, The attribute of is the carrying capacity of the corresponding area; S202, determine the node edge weight: In the network, the edge represents the capacity relationship between regions; Weight Responsive regional nodes and regional nodes The capacity flow between them, the weight of the edge is: ； in, For regional nodes and regional nodes The geographical distance between them is calculated by Dijkstra's algorithm applied on the road network; For regional nodes and regional nodes Between time The transportation time is derived from historical data; Node area node and regional nodes Between time Traffic flow, derived from historical data; , , is the weight coefficient; S203, constructing the highway transport capacity association network adjacency matrix: for each time, constructing the adjacency matrix : ； in, If the regional node and regional nodes There is an edge between them and the distance Less than the threshold ,but ,otherwise ;in, is the edge weight of the adjacency matrix of the highway transport capacity association network; S204. Integrate the historical highway capacity correlation network: Integrate the highway capacity correlation networks of adjacent time slices to form a complete historical highway capacity correlation network ;in, is the node feature set of all time slices; is the set of adjacency matrices for all time slices.

4. The highway transport capacity distribution prediction method based on comprehensive long-term and short-term fluctuation trends according to claim 3 is characterized in that: The S3 is specifically: S301. Preliminary decoupling: For the node capacity data within the time window, a moving average filter with different scale factors is used to remove the long-term and short-term trends of the capacity; the original capacity data is decomposed into long-term trend data and short-term fluctuation data. Node capacity sequence within a time window , long-term trend data is expressed as: ； in, is the original capacity data; An adjustable factor that controls the degree of decoupling for the moving average filter; For time The position index of The short-term trend data is: ； S302. Deep feature decoupling: Through wavelet decomposition, long-term trend data is further decoupled into multiple bands of different time scales to capture the capacity fluctuation characteristics at different time scales; short-term fluctuation data is decomposed into different frequency bands to capture small changes in a short period of time; for long-term trend data, the low-frequency band after decomposition is taken to represent the long-term trend characteristics; for short-term trend data, the high-frequency band after decomposition is taken to represent the short-term trend characteristics.

5. The highway transport capacity distribution prediction method based on comprehensive long-term and short-term fluctuation trends according to claim 1 is characterized in that: The S4 is specifically: S401. For an input time window, identify an external event that affects the transport capacity; if the event occurs, the corresponding code value is 1; if the event does not occur, the code value is 0; Formally: ； in, for whether the practice occurs; In order to capture the changing trend of capacity in daily, weekly and monthly cycles, sine and cosine functions are used to encode the time period features to map the time periodicity into the feature space, ensuring that the model can recognize and utilize these periodic patterns; Assume that the cycle of a day is ,time The position during the day is , then the daily cycle code is: ； ； Assume that the weekly cycle is ,time The position of the week is , then the weekly cycle code is: ； ； Combine the periodic codes into a complete feature vector: ； Time The external event code and the periodic feature code are combined into a complete feature vector : ； in, For feature stitching; S402: For data in the input time window: using a dynamic time normalization algorithm to calculate the similarity between the time series of the input time window nodes , set the similarity threshold ; If the regional node and regional nodes The sequence similarity between , then update the weight ; If the regional node and regional nodes The sequence similarity between , it remains unchanged; S403, spatial feature extraction: for the capacity association network within a period of time, the graph attention mechanism is used to extract spatial dependency features from the capacity association network; in In the layer graph attention mechanism, the node represents the vector By performing a weighted summation with the features of adjacent nodes and updating the features through nonlinear transformation: ； in, For the The node feature matrix of the layer; is the capacity fluctuation correlation matrix; is the activation function; For the The learned weight matrix of the layer; For regional nodes and regional nodes The attention weight between them is calculated by the features of the node pair: ； in, It is the concatenation operation of vectors; is the vector used to calculate attention; S404, Time Feature Extraction: Use dilated convolution to extract time-dependent features in time series; dilated convolution can capture long-term and short-term dependencies in time series by enhancing the hole convolution of the receptive field, and use dilated convolution to operate on the input time series at each layer: ； in, For the The input sequence of the layer; is dilated convolution; is bias; It is the convolution calculation; The calculation formula for the dilated convolution is: ； in, is the convolution kernel size; is the expansion rate; is the index of the convolution kernel on the input sequence.

6. The highway transport capacity distribution prediction method based on comprehensive long-term and short-term fluctuation trends according to claim 5 is characterized in that: The S5 is specifically: S501, integrating spatiotemporal features: performing feature concatenation on the last layer of temporal features and spatial features encoded in S4: ； in, For the last layer; Dynamically and hierarchically fuse the spatial and temporal features of the original sequence, long-term sequence, and short-term sequence; introduce an adaptive weight learning mechanism to dynamically learn the fusion weights of features of different scales through a neural network and set weight parameters , and Corresponding to the original features , Long-term characteristics and short-term characteristics , the multi-scale spatiotemporal features are: ； S502, comprehensive feature fusion: Use the attention mechanism to fuse multi-scale spatiotemporal features with external influencing factor features, specifically: ； in, It is the multi-scale spatiotemporal feature representation after the final fusion; Encoding features for external events and key time nodes; is the fusion feature dimension calculated using the attention mechanism; S503, receiving the fused spatiotemporal feature vector As input, the spatiotemporal feature vector Input into the multi-layer perceptron to generate the final prediction value: ； in, For space-time capacity network The predicted value of highway transport capacity includes the predicted value of transport capacity and relationship distribution characteristics of each regional node in the graph; To predict forward time steps; S504, model training and updating: The training goal of the model is to minimize the predicted value The actual capacity The model parameters are optimized through training to generate the best prediction values ​​for actual capacity scheduling and management.

Images