Logistics transportation multi-target path dynamic planning method for complex road network

Abstract

The invention provides a logistics transportation multi-target path dynamic planning method for a complex road network, and the method comprises the steps: S1, building a weighted graph network model through collecting real-time road network data and transportation mode parameters, obtaining the attribute representation of nodes and edges in the complex road network, extracting a discretized grid of a geographic continuous space according to the node attributes in the weighted graph network model, and carrying out the calculation of the discretized grid; obtaining a set of potential connection points among the transportation modes; s2, obtaining the service time and traffic capacity of the dynamic demand data updating connection point set, and if the service time exceeds a preset time threshold, excluding the connection point to obtain an optimized transfer station list; s3, generating a seamless connection path scheme according to the optimized transfer station list, analyzing resource utilization indexes from the path scheme to adjust dynamic demand matching, and obtaining a final path planning result; and S4, simulating a real-time change scene according to a final path planning result, and backtracking and adjusting connection details if the cost is judged to be high, so as to obtain a stable and efficient logistics transportation scheme.

Description

Technical Field

[0001] This invention relates to the field of intelligent logistics technology, and in particular to a multi-objective path dynamic planning method for logistics transportation in complex road networks. Background Technology

[0002] In the field of modern logistics and transportation, optimizing route planning within complex road networks is particularly crucial. Logistics and transportation are not only about cost control and efficiency improvement, but also a vital link in ensuring supply chain stability and meeting diverse needs. With the deepening of globalization, the complexity of logistics networks is increasing, involving multiple modes of transportation, numerous transit points, and dynamically changing market demands. How to achieve multi-objective optimization in such an environment has become a pressing challenge for the industry.

[0003] Currently, although many route planning methods have been applied to logistics and transportation, these methods often struggle to adapt to the dynamic characteristics of complex road networks involving multiple modes and nodes. Many solutions, when dealing with large-scale networks, tend to overlook the details of connections between different modes of transport, or lack flexibility in the face of real-time changes in road conditions and demand. This limitation leads to low resource utilization, especially in scenarios requiring frequent switching of transport modes, often resulting in time delays or cost overruns.

[0004] From a technical perspective, the core challenge of complex road network route planning lies in transforming continuous geographical space into a computable decision-making framework, and on this basis, achieving seamless connections between different modes of transportation. First, the continuity of geographical space makes route selection possibilities virtually limitless. Without effectively simplifying the computational scope, the optimization process will be excessively time-consuming or even unsolvable. Second, this continuity further complicates the selection of connection points, as the differences in attributes of transfer stations or transit locations (such as service hours and capacity) directly affect the overall efficiency of the route. For example, in a cross-regional transport operation, road transport may need to switch to rail transport at a certain node. However, if the selection of transfer stations is not precise enough, scheduled rail services may be missed, resulting in wasted time or additional costs.

[0005] Therefore, how to rationally simplify path selection in continuous space within complex road networks, while accurately grasping the characteristics and matching needs of transfer points between different modes of transportation, has become a key issue in multi-objective path dynamic programming for logistics transportation. Solving this problem is not only about improving computational efficiency, but also directly related to the smoothness and economy of logistics transportation in actual operations. Summary of the Invention

[0006] To address the technical problems mentioned in the background section, this invention provides a multi-objective path dynamic programming method for logistics transportation in complex road networks, comprising:

[0007] S1. By collecting real-time road network data and transportation mode parameters, a weighted graph network model is constructed to obtain the attribute representations of nodes and edges in the complex road network. Based on the node attributes in the weighted graph network model, a discretized grid of the geographically continuous space is extracted to obtain a set of potential connection points between transportation modes. S2. Dynamic demand data is acquired, and the service time and capacity of the connection point set are updated. If the service time exceeds a preset time threshold, the connection point is excluded, resulting in an optimized list of transit stations. S3. Based on the optimized list of transit stations, a seamless connection route plan is generated. Resource utilization indicators are analyzed from the route plan to adjust dynamic demand matching and obtain the final route planning result. S4. Real-time changing scenarios are simulated for the final route planning result. If costs exceed the budget, the connection details are adjusted retrospectively to obtain a stable and efficient logistics transportation solution.

[0008] Optionally, step S1 includes:

[0009] Step S11: By collecting real-time road network data and transportation mode parameters, a weighted graph network model is constructed to obtain detailed representations of node attributes and edge weights;

[0010] Step S12: Based on the correlation between node attributes and geographic space, the continuous geographic space is divided into discrete grids to generate an initial shortest path sequence and obtain a set of potential connection points between transportation modes.

[0011] Step S13: Perform spatial proximity analysis on the set of connection points to obtain the surrounding road network density and transportation mode coverage of each connection point. If the accessibility intensity of the connection point is lower than the preset intensity threshold, replan the path for the connection point to obtain a new shortest path sequence and determine a better connection point location.

[0012] Step S14: By comparing and analyzing the new connection point locations with the initial sequence, update the edge weights in the weighted graph network model, generate dynamic connection strategies between transportation modes, and determine the final road network transportation efficiency improvement scheme.

[0013] Optionally, step S11 includes:

[0014] Step S111: By collecting real-time road network data and transportation mode parameters, a weighted graph network model is constructed to obtain the specific attributes of nodes and the weight representation of edges. Key connection points in the network structure are analyzed to determine the matching degree between transportation modes and road network data, and the distribution of core nodes is obtained. Step S112: Based on the distribution of core nodes, the correlation strength between each node is obtained. A pre-established adjacency matrix is ​​used to represent the network structure, and the role weight of each node in the transportation mode is determined. Step S113: If the role weight of a core node is lower than a preset role weight threshold, the correlation strength between the node and its surrounding nodes is recalculated through information processing to obtain the adjusted connection relationship. Step S114: Based on the adjusted connection relationship, the weight allocation in the weighted graph network model is updated to determine a new network structure representation. Step S115: Based on the new network structure representation, the adaptability of real-time road network data under different transportation modes is analyzed to obtain the priority ranking of each path. Step S116: Based on the priority ranking of each path, an optimized configuration scheme combining transportation modes and road network data is generated, and the final resource allocation strategy is determined.

[0015] Optionally, step S2 includes:

[0016] Step S21: Obtain dynamic demand-related data from the system, and extract the corresponding service time and access capacity data for each connection node to obtain an initial set of connection nodes;

[0017] Step S22: For the initial set of connecting nodes, analyze the service time data of each node one by one. If the service time exceeds the preset time threshold, remove the node and generate an updated set of nodes.

[0018] Step S23: For the updated node set, evaluate the traffic capacity data of each node. If the traffic capacity is lower than the preset standard, mark it as an inefficient node, remove all inefficient nodes, and generate an optimized node list.

[0019] Step S24: Based on the optimized node list, associate the corresponding transit station information to determine the final set of transit stations;

[0020] Step S25: Using the updated set of transit stations and combined with dynamic demand data, generate a station allocation scheme suitable for the current scenario.

[0021] Optionally, step S3 includes:

[0022] Step S31: Obtain the optimized site distribution information, determine the location of key nodes, and obtain the initial site connection framework;

[0023] Step S32: Generate multiple candidate paths based on the initial site connection framework, evaluate the node connectivity of each candidate path, determine whether it meets the seamless connection condition, and mark it as a candidate path if it does, and obtain a candidate path set.

[0024] Step S33: Extract resource utilization data for each candidate path from the candidate path set, perform multi-dimensional calculations on resource occupancy and transmission efficiency, and determine the comprehensive utilization index for each candidate path.

[0025] Step S34: Sort the candidate paths according to the comprehensive utilization index. If the index of a candidate path is lower than the preset index threshold, the path is removed to obtain a list of selected paths.

[0026] Step S35: Analyze the traffic distribution of each selected path in different time periods through the selected path list, use a dynamic adjustment mechanism to balance the traffic load, and determine the adjusted path allocation scheme.

[0027] Step S36: For the adjusted path allocation scheme, match and verify it with real-time demand data. If the matching degree reaches the predetermined standard, output the final path planning result.

[0028] Step S37: Based on the final route planning result, generate the corresponding station scheduling instruction and transmit it to the relevant system for execution to ensure that the route planning is consistent with the actual operation.

[0029] Optionally, step S4 includes:

[0030] Step S41: Obtain the latest transportation information and determine the current route's operational status.

[0031] Using a pre-defined cost calculation model, the cost consumption of the current path is analyzed to obtain cost analysis results;

[0032] Step S42: If the cost exceeds the preset cost threshold, the overspending judgment mechanism is triggered, the data of the path links that need to be adjusted is obtained, the backtracking adjustment process is started, the connection links are optimized, and the adjusted path scheme is determined.

[0033] Step S43: Based on the adjusted route plan and combined with scenario simulation technology, recalculate the cost changes during transportation, determine whether the standard of a stable plan has been met, and optimize the allocation of transportation resources by combining the logistics management database to obtain the final efficient transportation plan.

[0034] Step S44: For the final efficient transportation plan, update the transportation optimization record and synchronize it to the route planning system.

[0035] The technical solution provided by this invention has the following beneficial effects:

[0036] This invention discloses a multi-objective dynamic path planning method for logistics transportation in complex road networks. Addressing the problems of low path planning efficiency, exceeding service time limits for connecting points, and uneven resource utilization in business scenarios where real-time road network data and dynamic demand are matched, this invention constructs a weighted graph network model to accurately represent the attributes of road network nodes and edges. Combined with discretized grid partitioning and shortest path algorithms, it generates an initial path sequence and a set of potential connecting points. This invention updates the service time and capacity of connecting points based on dynamic demand data, eliminates connecting points that do not meet thresholds, optimizes the list of transit stations, and uses an efficient path search algorithm to generate alternative path solutions with seamless connections. Finally, it ensures cost control and transportation efficiency through resource utilization index adjustments and real-time scenario simulation. If costs exceed the budget, this invention can retrospectively adjust connecting details to form a stable and efficient logistics transportation solution. Its core technical effect lies in improving the dynamic adaptability and resource utilization of path planning, providing intelligent support for logistics transportation in complex road networks. Attached Figure Description

[0037] Figure 1 This is a flowchart of the multi-objective path dynamic planning method for logistics transportation in complex road networks according to the present invention. Detailed Implementation

[0038] The technical solutions of the embodiments of the present invention will be clearly and thoroughly described below with reference to the accompanying drawings. The described embodiments are merely some embodiments of the present invention.

[0039] like Figure 1 As shown, this invention provides a multi-objective path dynamic planning method for logistics transportation in complex road networks, the method comprising:

[0040] S1. By collecting real-time road network data and transportation mode parameters, a weighted graph network model is constructed to obtain the attribute representation of nodes and edges in the complex road network. Based on the node attributes in the weighted graph network model, a discretized grid of the geographically continuous space is extracted to obtain the set of potential connection points between transportation modes.

[0041] Specifically, a weighted graph network model is meticulously constructed based on comprehensive and detailed collection of real-time road network data, along with precise collection of relevant parameters for various transportation modes. This construction process fully considers the diversity and complexity of complex road networks, thereby obtaining attribute representations of each node and edge within the complex network. These attribute representations accurately reflect the actual characteristics and operational patterns of the complex road network.

[0042] Furthermore, based on the node attributes in the constructed weighted graph network model, a discretized grid of the geographic continuous space is extracted. During the extraction process, relevant geospatial analysis principles and methods are strictly followed to ensure that the discretized grid accurately reflects the characteristics and distribution of the geographic space. By extracting the discretized grid, the computable range for path selection is determined, allowing subsequent path calculations to be performed within a reasonable and effective scope.

[0043] Furthermore, after determining the computable range, the classic Dijkstra algorithm was employed to perform calculations on the discretized grid. During the calculation process, the steps and rules of the Dijkstra algorithm were strictly followed to accurately calculate the initial shortest path sequence. Through analysis and research of the initial shortest path sequence, a set of potential connection points between transportation modes was ultimately obtained. This set of potential connection points is of great significance for optimizing transportation schemes and improving transportation efficiency.

[0044] Optionally, this step also includes:

[0045] Step S11: By collecting real-time road network data and transportation mode parameters, a weighted graph network model is constructed to obtain detailed representations of node attributes and edge weights.

[0046] Step S12: Based on the node attributes and the correlation of geographic space, the continuous geographic space is divided into discrete grids to generate an initial shortest path sequence and obtain a set of potential connection points between transportation modes.

[0047] Step S13: Perform spatial proximity analysis on the set of connection points to obtain the surrounding road network density and transportation mode coverage of each connection point. If the accessibility intensity of the connection point is lower than the preset intensity threshold, replan the path for the connection point to obtain a new shortest path sequence and determine a better connection point location.

[0048] Step S14: By comparing and analyzing the new connection point locations with the initial sequence, update the edge weights in the weighted graph network model, generate dynamic connection strategies between transportation modes, and determine the final road network transportation efficiency improvement scheme.

[0049] For example, in urban traffic network optimization scenarios, a weighted graph network model can be constructed by collecting real-time road network data and transportation mode parameters. Assuming a medium-sized city, the road network data includes road length, traffic flow, and traffic light waiting times, while transportation mode parameters cover the speed and coverage of buses, shared bicycles, and walking. Node attributes can represent the location and type of intersections, such as whether it is a bus stop or a transfer hub; the edge weight can be a comprehensive value based on the travel time to the node. For example, if the travel time from a certain road to a node is 5 minutes, the edge weight is set to 5.

[0050] In one possible implementation, when dividing a continuous geographic space into discrete grids, a city can be divided into 1-kilometer by 1-kilometer grid units, with each unit serving as a computing node. By combining geographic information system data, the road density and traffic flow within each grid can be determined, thereby clarifying the computable range for path selection.

[0051] For example, the computational range of a city center grid may be smaller due to the dense road network, while that of a suburban grid may be larger.

[0052] Specifically, when using Dijkstra's algorithm to generate the initial shortest path sequence, assuming a route from point A on the west side of the city to point B on the east side, the algorithm calculates a shortest path traversing 3 grids based on a weighted graph network model. The total path time is 20 minutes. Two potential connection points are identified along the way: a bus stop C and a shared bicycle stop D. The spatial proximity of these two connection points can be evaluated using the road network density and transportation mode coverage within a 500-meter radius.

[0053] For example, point C is surrounded by 3 bus routes and 2 shared bicycle stops, with a high road network density and an accessibility intensity of 8 (out of 10); while point D is surrounded by only 1 bus route, with a low road network density and an accessibility intensity of 4, which is lower than the preset intensity threshold of 6.

[0054] For example, if the accessibility strength of point D is insufficient, the route needs to be replanned. Suppose that by adjusting the route selection, we can detour to another nearby shared bike point E, whose accessibility strength is 7. The recalculated shortest path total time is 22 minutes, an increase of only 2 minutes, but the quality of the connecting points is significantly improved. When updating the weighted graph network model, the weight of the edge at point D is increased to reflect its lower accessibility; the optimized route selection scheme is more inclined to pass through point E.

[0055] In one possible implementation, when generating dynamic connection strategies based on optimized routes, the system can be configured to recommend that users transfer to a shared bicycle at point C to reach point E, and then walk to point B, if the bus delay exceeds 5 minutes. This strategy can dynamically adapt to real-time traffic conditions, improving transportation efficiency by approximately 15%, reducing the uncertainty of users' travel time, and simultaneously alleviating local road network congestion through the rational allocation of transportation resources, thus achieving an overall improvement in road network transportation efficiency. This method not only optimizes route selection but also enhances the system's robustness through the rational adjustment of connection points, providing data support for urban traffic management.

[0056] Optionally, this step also includes:

[0057] Step S111: By collecting real-time road network data and transportation mode parameters, a weighted graph network model is constructed, the specific attributes of nodes and the weight representation of edges are obtained, the key connection points in the network structure are analyzed, the degree of matching between transportation mode and road network data is determined, and the distribution of core nodes is obtained.

[0058] Step S112: Based on the distribution of core nodes, obtain the association strength between each node, use a pre-established adjacency matrix to represent the network structure, and determine the role weight of each node in the transportation mode.

[0059] Step S113: If the role weight of the core node is lower than the preset role weight threshold, the association strength between the node and the surrounding nodes is recalculated through the information processing step to obtain the adjusted connection relationship.

[0060] Step S114: Based on the adjusted connection relationships, update the weight allocation in the weighted graph network model to determine the new network structure representation.

[0061] Step S115: For the new network structure representation, analyze the adaptation of real-time road network data under different transportation modes to obtain the priority ranking of each path.

[0062] Step S116: By prioritizing each path, an optimized configuration scheme combining transportation mode and road network data is generated, and the final resource allocation strategy is determined.

[0063] For example, in the optimization of urban traffic networks, the collection of real-time road network data and the analysis of transportation mode parameters are fundamental to constructing a weighted graph network model. Assuming a medium-sized city's traffic system, the collected data includes road speed, vehicle density, and traffic light cycles, while transportation mode parameters involve bus departure frequency and average walking speed. Specific node attributes can be defined as the geographical location and functional type of an intersection, such as whether it is a transportation hub; edge weights can be represented by numerical values ​​based on travel time. If the travel time from a certain road segment to a node is 6 minutes, the edge weight is set to 6. Using this data, a preliminary model reflecting the characteristics of the road network can be constructed.

[0064] For example, analyzing key connection points in a network structure can identify intersections or road segments that significantly impact traffic flow by analyzing the specific attributes of nodes and the weights of edges. Suppose in a city center area, an intersection connects multiple main roads and has a dense network of bus routes; such an intersection would be of high importance. The degree of matching can be determined by analyzing whether the intersection effectively supports the conversion between multiple modes of transportation, such as whether it offers convenient transfers between public transport and walking, thereby determining the distribution of core nodes.

[0065] For example, after analyzing the distribution of core nodes, an adjacency matrix can be constructed to quantify the strength of the connections between nodes. Assume that the value between two nodes in the matrix represents their degree of direct connection in the road network; a higher value indicates a stronger connection. In this way, the weight of each node's role in the transportation mode can be determined. For instance, a node connecting multiple bus routes might have a weight of 0.8, higher than other ordinary intersections.

[0066] For example, if the influence weight of a core node is found to be lower than the preset influence weight threshold of 0.6, its association strength with surrounding nodes needs to be recalculated. This can be achieved by analyzing the road network density and transportation mode coverage within a 1-kilometer radius of the node through information processing, and adjusting its connectivity accordingly, such as increasing its association with nearby high-weight nodes.

[0067] For example, based on the adjusted connectivity, the weight allocation in a weighted graph network model is updated to form a new network structure representation. Suppose the original weight of an edge was 5, and after adjustment, it decreases to 3 due to the addition of new connection points, reflecting the change in traffic distribution. This update helps to more accurately reflect the network status.

[0068] For example, by analyzing the adaptation of real-time road network data to different modes of transportation for a new network structure representation, the priority ranking of each path can be obtained. Suppose a path combines public transportation and walking, has a shorter travel time, and therefore has a higher priority, while another path, although shorter in distance, suffers from severe congestion and has a lower priority.

[0069] For example, by prioritizing routes, an optimized configuration scheme combining transportation modes and road network data can be generated to determine the final resource allocation strategy. For instance, higher-priority routes might be allocated more public transport resources, such as increased departure frequency, while lower-priority routes would receive less resource allocation. This approach can effectively improve the overall operational efficiency of the road network and optimize the user's travel experience.

[0070] S2: Obtain dynamic demand data, update the service time and throughput of the connection point set, and exclude the connection point if the service time exceeds the preset time threshold to obtain the optimized list of transit stations.

[0071] Optionally, this step also includes:

[0072] Step S21: Obtain dynamic demand-related data from the system, and extract the corresponding service time and traffic capacity data for each connection node to obtain an initial set of connection nodes.

[0073] Step S22: For the initial set of connecting nodes, analyze the service time data of each node one by one. If the service time exceeds the preset time threshold, remove the node and generate an updated set of nodes.

[0074] Step S23: For the updated node set, evaluate the traffic capacity data of each node. If the traffic capacity is lower than the preset standard, mark it as an inefficient node, remove all inefficient nodes, and generate an optimized node list.

[0075] Step S24: Based on the optimized node list, associate the corresponding transit station information to determine the final set of transit stations.

[0076] Step S25: Using the updated set of transit stations and combined with dynamic demand data, generate a station allocation scheme suitable for the current scenario.

[0077] For example, in the scenario of optimizing urban traffic networks, the acquisition of dynamic demand data and the processing of connecting nodes can be analyzed in detail from multiple perspectives. By combining real-time traffic data and station characteristics, the applicability of the solution can be ensured. First, let's address the topic of acquiring dynamic demand-related data from the system.

[0078] Understandably, dynamic demand data includes real-time passenger flow, vehicle dispatch information, and road congestion conditions.

[0079] In one possible implementation, assuming a traffic management system in a medium-sized city, peak passenger flow data and road traffic conditions across the city are updated hourly using sensors and monitoring equipment. This allows for the extraction of data showing a passenger flow of 5,000 people per hour and an average road speed of 20 kilometers per hour in a specific area during the morning peak. This data provides the foundation for subsequent node analysis.

[0080] For example, in the process of extracting service time and traffic capacity data for each connecting node, service time can refer to the average waiting time of passengers at the node, while traffic capacity refers to the maximum daily passenger flow that the node can handle.

[0081] In one possible implementation, assume a bus hub station has an average waiting time of 3 minutes per person and a daily capacity of 20,000 passengers. By comparing this to a preset time threshold, if the service time threshold is 2 minutes, the node is marked for removal, and then all nodes marked for removal are removed. This analysis is applied one by one to the initial node set to ensure that nodes that do not meet the basic requirements are filtered out.

[0082] In one embodiment, during the process of removing nodes to be removed and generating an updated node set, nodes that exceed the service time threshold can be automatically filtered out from the list. Assuming the initial set contains 10 nodes, 3 of which are marked and removed due to service time exceeding the time threshold, the remaining 7 nodes form the updated set. This process ensures the quality of nodes for subsequent analysis.

[0083] In one embodiment, regarding the step of assessing traffic capacity and marking inefficient nodes, assuming that among the updated 7 nodes, 2 nodes have a traffic capacity lower than the preset standard of 15,000 passengers per day (12,000 and 11,000 passengers per day respectively), they are marked as inefficient nodes and removed, ultimately generating an optimized list containing 5 nodes. This screening further improves the overall efficiency of the node set.

[0084] For example, in the process of associating transfer station information and determining the final set of transfer stations, the optimized node list can be combined with geographic information data to clarify the specific location and type of each station. Assuming that out of five nodes, three are bus hubs and two are subway transfer stations, by associating with surrounding facility information, it can be determined whether these stations offer sufficient transfer convenience, ultimately forming the set of transfer stations.

[0085] For example, to generate a station allocation scheme suitable for the current scenario, dynamic demand data can be combined to rationally allocate passenger flow to different stations. Suppose that during the morning rush hour, passenger flow in a certain area is concentrated at two stations. By adjusting vehicle scheduling and route planning, some passenger flow can be guided to stations with greater capacity, thereby balancing the load. This approach can effectively improve the utilization rate of transportation resources, alleviate local congestion, and provide passengers with a smoother travel experience. Through multi-level node selection and data analysis, the entire scheme demonstrates strong adaptability under dynamic demand, providing reliable support for urban traffic management.

[0086] S3. Based on the optimized list of transit stations, generate seamless route plans, analyze resource utilization indicators from the route plans to adjust dynamic demand matching, and obtain the final route planning result.

[0087] Optionally, this step also includes:

[0088] Step S31: Obtain the optimized site distribution information, determine the location of key nodes, and obtain the initial site connection framework.

[0089] Step S32: Based on the initial site connection framework, apply Dijkstra's algorithm to generate multiple candidate paths. Evaluate the node connectivity of each candidate path to determine whether it meets the seamless connection condition. If it does, mark it as a candidate path and obtain a set of candidate paths.

[0090] Step S33: Extract resource utilization data for each candidate path from the candidate path set, perform multi-dimensional calculations on resource occupancy and transmission efficiency, and determine the comprehensive utilization index for each candidate path.

[0091] Step S34: Sort the candidate paths according to the comprehensive utilization index. If the index of a candidate path is lower than the preset index threshold, the path is removed to obtain the selected path list.

[0092] Step S35: Based on the list of selected paths, analyze the traffic distribution of each selected path in different time periods, use a dynamic adjustment mechanism to balance the traffic load, and determine the adjusted path allocation scheme.

[0093] Step S36: For the adjusted path allocation scheme, match and verify it with real-time demand data. If the matching degree reaches the predetermined standard, output the final path planning result.

[0094] Step S37: Based on the final route planning result, generate the corresponding station scheduling instruction and transmit it to the relevant system for execution to ensure that the route planning is consistent with the actual operation.

[0095] Specifically, in the scenario of optimizing urban traffic networks, the distribution of transfer stations and route planning issues can be analyzed in detail from multiple perspectives to ensure the adaptability and practicality of the solution.

[0096] Regarding the step of retrieving optimized site distribution information from the repository, it is understood that site distribution information includes the site's geographical location, type, and basic carrying capacity. In one possible implementation, assuming a traffic management system in a medium-sized city, the repository records detailed data on 50 transfer stations throughout the city, including the average daily passenger flow and surrounding road conditions for each station. Through a hierarchical screening method, stations with an average daily passenger flow exceeding 15,000 are prioritized as key nodes, ultimately identifying 10 core locations to form the initial site connection framework.

[0097] Specifically, regarding the step of generating multiple alternative paths using Dijkstra's algorithm, the core of Dijkstra's algorithm lies in finding the shortest path between nodes, which is suitable for evaluating the connectivity efficiency in transportation networks.

[0098] In one possible implementation, based on the initial station connection framework, the system generates 5 alternative routes from the origin to the destination. The connectivity of each alternative route is evaluated to determine whether it meets the seamless connection conditions, such as whether the transfer time is controlled within 5 minutes. If it meets the conditions, it is marked as a candidate route, and finally a set of 3 candidate routes is obtained.

[0099] In one embodiment, in the step of extracting candidate path resource utilization data and calculating comprehensive utilization indicators, the resource utilization data may include the vehicle occupancy and average travel time of the path.

[0100] Specifically, assuming that among the three candidate routes, the first candidate route has a daily vehicle occupancy of 200 vehicles and an average travel time of 30 minutes, its comprehensive utilization index is 85 points through multi-dimensional calculations, while the preset index threshold is 80 points, so this candidate route is retained. The indices for the other two candidate routes are 78 points and 90 points respectively. Finally, the candidate route with an index of 78 points is eliminated, resulting in the selected route list.

[0101] For example, when analyzing traffic distribution and making dynamic adjustments to a list of selected routes, traffic distribution refers to changes in passenger flow on the routes over different time periods.

[0102] In one possible implementation, assuming that the passenger flow on a certain route reaches 8,000 people per hour during the morning peak hours, exceeding the carrying capacity limit, a dynamic adjustment mechanism is used to guide some of the passenger flow to another route to achieve load balancing and determine the adjusted route allocation scheme.

[0103] For example, in the matching and verification process that combines real-time demand data, the real-time demand data includes current passenger flow and road conditions.

[0104] In one possible implementation, the dynamic adjustment mechanism involves real-time monitoring of passenger flow, occupancy rate, and other data to predict trends. When the load exceeds a threshold, a tiered response is triggered: Level 1 reduces departure intervals; Level 2 activates alternative routes (such as express buses or tidal flow lanes); and Level 3 adjusts route directions. Reinforcement learning algorithms are used to optimize scheduling strategies, diverting 20%-30% of passenger flow to parallel routes. Recommended routes are pushed in real-time through apps, navigation software, and station signage, supplemented by fare discounts to incentivize passengers to choose these diversion routes. Finally, closed-loop monitoring is used to evaluate the load balancing effect, continuously optimizing the adjustment strategy to achieve efficient network operation.

[0105] Specifically, assuming the adjusted plan matches the real-time demand data by 90%, exceeding the predetermined standard of 85%, the final path planning result is output. This process ensures the real-time nature and accuracy of the planning.

[0106] For example, the step of generating site scheduling instructions and transmitting them to relevant systems can be understood as transforming the planning results into executable operation instructions.

[0107] In one possible implementation, the system generates vehicle dispatch instructions for each station based on the final route planning results, such as increasing the departure frequency of a certain route to once every 10 minutes, and transmits these instructions to the traffic dispatch system for execution, ensuring consistency between planning and actual operation. This step improves the accuracy of traffic resource allocation.

[0108] S4 simulates real-time changes in the final route planning results, and if costs exceed the budget, backtracks and adjusts the connection details to obtain a stable and efficient logistics transportation solution.

[0109] Optionally, this step also includes:

[0110] Step S41: Obtain the latest transportation information, determine the current route's operating status, use a preset cost calculation model to analyze the current route's cost consumption, and obtain the cost analysis results.

[0111] Step S42: If the cost exceeds the preset cost threshold, the overspending judgment mechanism is triggered, the data of the path links that need to be adjusted is obtained, the backtracking adjustment process is started, the connection links are optimized, and the adjusted path scheme is determined.

[0112] Step S43: Based on the adjusted route plan and combined with scenario simulation technology, recalculate the cost changes during transportation, and optimize the allocation of transportation resources by combining with the logistics management database to obtain the final efficient transportation plan.

[0113] Step S44: For the final efficient transportation plan, update the transportation optimization record and synchronize it to the route planning system.

[0114] For example, in the field of route planning and optimization for urban traffic networks, the latest transportation information can be obtained from real-time dynamic scenarios to determine the current route operation status. It is understood that real-time dynamic scenarios include information such as traffic flow, road congestion, and emergencies.

[0115] In one embodiment, assuming a public transport dispatch system in a medium-sized city updates traffic data for the city's main roads every 5 minutes using onboard sensors and roadside monitoring equipment, if it is found that the average speed on a certain main road has dropped to 10 kilometers per hour due to temporary construction, the system determines that the route is in a "congested" state and marks it as a route requiring attention.

[0116] For example, when using a pre-set cost calculation model to analyze route cost consumption, the cost calculation model usually takes into account factors such as time cost, vehicle operating cost, and energy consumption.

[0117] In one embodiment, the system analyzes real-time data to determine if a bus route experiences increased travel time from 30 minutes to 50 minutes due to congestion, while vehicle fuel consumption increases by 20%. The system concludes that the cost of this route is 35% higher than normal, exceeding a preset cost threshold by 25%, thus triggering a subsequent adjustment mechanism.

[0118] For example, regarding the topic of cost overrun triggering an overrun judgment mechanism and obtaining data on the path links that need adjustment, the overrun judgment mechanism aims to identify the specific links with abnormal costs.

[0119] In one embodiment, the system uses data backtracking to discover that the cost overrun is mainly due to long waiting times and detours on a certain section of road. Specifically, the 5-kilometer section from station A to station B has an average waiting time of 15 minutes, which is identified as the section that needs optimization.

[0120] Furthermore, in the process of initiating the retrospective adjustment process to optimize the connection and determine the adjusted route plan, the retrospective adjustment process refers to replanning the route to reduce unnecessary waiting and detours.

[0121] In one embodiment, the system analyzes surrounding road data to adjust the original route to a backup route that passes through station C, shortening the waiting time to 5 minutes and ensuring smooth transfer connections, thus forming a new route plan.

[0122] Furthermore, in recalculating cost changes and determining whether a stable solution standard has been reached by combining scenario simulation technology, the scenario simulation technology is used to predict the operational effect of the adjusted path.

[0123] In one embodiment, the system simulates the operation of the new route during peak hours and finds that the one-way time is stable at 35 minutes and the cost consumption drops to 110% of the normal level, which meets the standard of a stable solution.

[0124] Furthermore, in obtaining stable solution evaluation results and optimizing resource allocation in conjunction with the logistics management database, the evaluation results need to be compared with historical data in the database to optimize resources.

[0125] In one embodiment, the system adjusts the vehicle configuration of the route from 10 to 8 vehicles based on the evaluation results, while increasing the departure frequency of the backup route to ensure more balanced resource utilization.

[0126] Furthermore, updating transportation optimization records and synchronizing them with the route planning system is a step aimed at maintaining system data consistency.

[0127] In one embodiment, the system uploads the adjusted route plan and resource allocation data to the central database and synchronizes them to each scheduling terminal to ensure that subsequent operations are carried out according to the latest plan.

[0128] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from the spirit or essential characteristics of this application. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of this application is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within this application. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A multi-objective path dynamic programming method for logistics transportation in complex road networks, characterized in that, The method includes: S1. By collecting real-time road network data and transportation mode parameters, a weighted graph network model is constructed to obtain the attribute representations of nodes and edges in the complex road network. Based on the node attributes in the weighted graph network model, a discretized grid of the geographically continuous space is extracted to obtain a set of potential connection points between transportation modes. S2. Dynamic demand data is acquired, and the service time and capacity of the connection point set are updated. If the service time exceeds a preset time threshold, the connection point is excluded, resulting in an optimized list of transit stations. S3. Based on the optimized list of transit stations, a seamless connection route plan is generated. Resource utilization indicators are analyzed from the route plan to adjust dynamic demand matching and obtain the final route planning result. S4. Real-time changing scenarios are simulated for the final route planning result. If costs exceed the budget, the connection details are adjusted retrospectively to obtain a stable and efficient logistics transportation solution.

2. The method according to claim 1, characterized in that, Step S1 includes: Step S11: By collecting real-time road network data and transportation mode parameters, a weighted graph network model is constructed to obtain detailed representations of node attributes and edge weights; Step S12: Based on the correlation between node attributes and geographic space, the continuous geographic space is divided into discrete grids to generate an initial shortest path sequence and obtain a set of potential connection points between transportation modes. Step S13: Perform spatial proximity analysis on the set of connection points to obtain the surrounding road network density and transportation mode coverage of each connection point. If the accessibility intensity of the connection point is lower than the preset intensity threshold, replan the path for the connection point to obtain a new shortest path sequence and determine a better connection point location. Step S14: By comparing and analyzing the new connection point locations with the initial sequence, update the edge weights in the weighted graph network model, generate dynamic connection strategies between transportation modes, and determine the final road network transportation efficiency improvement scheme.

3. The method according to claim 2, characterized in that, Step S11 includes: Step S111: By collecting real-time road network data and transportation mode parameters, a weighted graph network model is constructed to obtain the specific attributes of nodes and the weight representation of edges. Key connection points in the network structure are analyzed to determine the matching degree between transportation modes and road network data, and the distribution of core nodes is obtained. Step S112: Based on the distribution of core nodes, the correlation strength between each node is obtained. A pre-established adjacency matrix is ​​used to represent the network structure, and the role weight of each node in the transportation mode is determined. Step S113: If the role weight of a core node is lower than a preset role weight threshold, the correlation strength between the node and its surrounding nodes is recalculated through information processing to obtain the adjusted connection relationship. Step S114: Based on the adjusted connection relationship, the weight allocation in the weighted graph network model is updated to determine a new network structure representation. Step S115: Based on the new network structure representation, the adaptability of real-time road network data under different transportation modes is analyzed to obtain the priority ranking of each path. Step S116: Based on the priority ranking of each path, an optimized configuration scheme combining transportation modes and road network data is generated, and the final resource allocation strategy is determined.

4. The method according to claim 1, characterized in that, Step S2 includes: Step S21: Obtain dynamic demand-related data from the system, and extract the corresponding service time and access capacity data for each connection node to obtain an initial set of connection nodes; Step S22: For the initial set of connecting nodes, analyze the service time data of each node one by one. If the service time exceeds the preset time threshold, remove the node and generate an updated set of nodes. Step S23: For the updated node set, evaluate the traffic capacity data of each node. If the traffic capacity is lower than the preset standard, mark it as an inefficient node, remove all inefficient nodes, and generate an optimized node list. Step S24: Based on the optimized node list, associate the corresponding transit station information to determine the final set of transit stations; Step S25: Using the updated set of transit stations and combined with dynamic demand data, generate a station allocation scheme suitable for the current scenario.

5. The method according to claim 1, characterized in that, Step S3 includes: Step S31: Obtain the optimized site distribution information, determine the location of key nodes, and obtain the initial site connection framework; Step S32: Generate multiple candidate paths based on the initial site connection framework, evaluate the node connectivity of each candidate path, determine whether it meets the seamless connection condition, and mark it as a candidate path if it does, and obtain a candidate path set. Step S33: Extract resource utilization data for each candidate path from the candidate path set, perform multi-dimensional calculations on resource occupancy and transmission efficiency, and determine the comprehensive utilization index for each candidate path. Step S34: Sort the candidate paths according to the comprehensive utilization index. If the index of a candidate path is lower than the preset index threshold, the path is removed to obtain a list of selected paths. Step S35: Analyze the traffic distribution of each selected path in different time periods through the selected path list, use a dynamic adjustment mechanism to balance the traffic load, and determine the adjusted path allocation scheme. Step S36: For the adjusted path allocation scheme, match and verify it with real-time demand data. If the matching degree reaches the predetermined standard, output the final path planning result. Step S37: Based on the final route planning result, generate the corresponding station scheduling instruction and transmit it to the relevant system for execution to ensure that the route planning is consistent with the actual operation.

6. The method according to claim 1, characterized in that, Step S4 includes: Step S41: Obtain the latest transportation information and determine the current route's operational status. Using a pre-defined cost calculation model, the cost consumption of the current path is analyzed to obtain cost analysis results; Step S42: If the cost exceeds the preset cost threshold, the overspending judgment mechanism is triggered, the data of the path links that need to be adjusted is obtained, the backtracking adjustment process is started, the connection links are optimized, and the adjusted path scheme is determined. Step S43: Based on the adjusted route plan and combined with scenario simulation technology, recalculate the cost changes during transportation, determine whether the standard of a stable plan has been met, and optimize the allocation of transportation resources by combining the logistics management database to obtain the final efficient transportation plan. Step S44: For the final efficient transportation plan, update the transportation optimization record and synchronize it to the route planning system.

Images