Low-carbon target-oriented synchronous multimodal transport dynamic route planning method and system
By constructing a multimodal transport digital twin network model and a dynamic adjustment mechanism driven by real-time traffic information, the problems of global optimization and dynamic response in multimodal transport route planning are solved, and the reliability and real-time performance of low-carbon goals are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- GUANGZHOU ZHIKA LOGISTICS TECH CO LTD
- Filing Date
- 2026-01-16
- Publication Date
- 2026-05-08
AI Technical Summary
Existing multimodal transport route planning methods struggle to make simultaneous decisions on transport modes, routes, and transshipment nodes within a unified framework, leading to local optima rather than global optima. Furthermore, they cannot dynamically respond to traffic conditions and weather changes, resulting in deviations from carbon emission targets.
A multimodal transport digital twin network model is constructed. An initial low-carbon routing scheme is generated through synchronous integrated optimization model, and dynamic adjustments are made based on real-time traffic information during freight transport. The routing scheme is optimized by combining verification and feedback mechanisms.
It achieves a balance between cost and carbon emissions from a global perspective, enhances the adaptability of the planning scheme to uncertainties in the transportation process, ensures the real-time nature and reliability of low-carbon goals, and continuously optimizes the routing scheme through self-learning.
Smart Images
Figure CN121998536A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of low-carbon emission technology in freight logistics, specifically a synchronous multimodal dynamic routing planning method and system oriented towards low-carbon goals. Background Technology
[0002] In the freight logistics sector, multimodal transport is a key model for improving efficiency and reducing costs. Existing route planning methods typically focus on static optimization for a single objective (such as lowest cost or shortest time), or treat carbon emissions as a fixed constraint rather than a core optimization objective during planning. These methods have significant limitations: First, traditional segmented or sequential planning struggles to simultaneously decide on transport modes, routes, and transshipment nodes within a unified framework, easily leading to local optima rather than global optima, failing to truly achieve synergistic optimization of cost and carbon emissions. Second, most solutions are static plans; once generated, they are executed in a fixed manner, unable to effectively respond to dynamic changes such as traffic conditions and weather during actual transportation processes that can last for days or even weeks. This can lead to significant deviations between actual carbon emissions and planned targets, resulting in the failure to achieve low-carbon goals.
[0003] Therefore, there is an urgent need for an intelligent planning method with dynamic response and synchronous decision-making capabilities to reliably achieve low-carbon transportation goals in complex real-world logistics networks. Summary of the Invention
[0004] The purpose of this application is to provide a synchronous multimodal dynamic routing planning method and system for low-carbon goals, so as to solve the technical problems mentioned in the background art.
[0005] To achieve the above objectives, this application discloses the following technical solutions: In a first aspect, this application discloses a synchronous multimodal dynamic routing planning method for low-carbon objectives, the method comprising: Model construction steps: Based on the topology, node attributes, and directed edge attributes of the multimodal transport physical network, construct a multimodal transport digital twin network model; the node attributes include the transshipment capacity of the transshipment nodes; the directed edge attributes include the mode of transport, distance, time cost, and historical average carbon emission intensity as a benchmark parameter; Input constraints: Receive the origin, destination, and latest delivery time constraints for freight orders; Synchronous optimization steps: Map the starting point and the ending point to the multimodal transport digital twin network model, construct a dual-objective optimization model with the goal of minimizing the total transport cost and the estimated total carbon emissions, and use the latest delivery time constraint and node transshipment capacity as constraints to generate an initial low-carbon routing scheme through synchronous integrated solution. Dynamic adjustment steps: During freight delivery, based on real-time traffic information, update the real-time time cost and real-time carbon emission intensity in the multimodal transport digital twin network model; when the real-time carbon emission intensity of the current path or the delay estimated based on the real-time time cost exceeds a preset threshold, trigger rerouting planning based on the updated multimodal transport digital twin network model, and output a dynamically adjusted low-carbon routing scheme. Verification and feedback steps: For each road segment actually traversed in the dynamically adjusted low-carbon routing scheme, collect its actual transportation mode, actual energy consumption data, and actual travel time; based on the actual energy consumption data, calculate the actual carbon emission intensity of the road segment; fuse the actual carbon emission intensity with the historical average carbon emission intensity corresponding to the road segment in the multimodal transport digital twin network model to obtain the updated historical average carbon emission intensity, and store it in the multimodal transport digital twin network model.
[0006] Optionally, the synchronous integrated solution process includes: A unified decision variable is constructed, which simultaneously encodes the choice of transportation mode, the sequence of path nodes, and the connection relationship at transfer nodes; Based on the decision variables, the bi-objective optimization model is solved, and the final routing scheme obtained is output as the initial low-carbon routing scheme.
[0007] Optionally, solving the bi-objective optimization model and outputting a final routing scheme obtained from the solution includes: The bi-objective optimization model was solved using a decomposition-based multi-objective evolutionary algorithm, yielding a set of Pareto optimal solutions; According to the preset decision rules, a solution is selected from the set of Pareto optimal solutions; The complete transportation mode sequence, path sequence, and transfer node sequence corresponding to the selected solution are output as the initial low-carbon routing scheme.
[0008] Optionally, the optimization objectives of the dual-objective optimization model include a total transportation cost objective function and a predicted total carbon emission objective function. The simultaneous optimization step and the dynamic adjustment step use the same total transportation cost objective function and the predicted total carbon emission objective function for scheme evaluation.
[0009] Optionally, the method for determining whether the real-time carbon emission intensity of the preceding path or the delay estimated based on real-time time cost exceeds a preset threshold includes: Calculate the real-time carbon emission intensity of the planned road section ahead. Compared with the carbon emission intensity prediction used when generating the initial low-carbon routing scheme The proportional relationship between them; if satisfied If it exceeds the preset threshold, then it is determined that it exceeds the threshold. The preset ratio threshold; Based on real-time time costs, the arrival time of goods is re-estimated. and compared with the planned arrival time in the initial low-carbon routing scheme. Compare; if satisfied If it exceeds the preset threshold, then it is determined that it exceeds the threshold. This is a preset time threshold.
[0010] Optionally, the rerouting plan includes: Based on the real-time time cost and real-time carbon emission intensity, a new bi-objective optimization model is constructed for the remaining untransported routes; The actual carbon emissions and time consumption of the current transportation route are used as fixed parameters and input into the reconstructed bi-objective optimization model; Solve the reconstructed bi-objective optimization model to plan new transportation methods, routes, and transfer nodes for the remaining untransported paths.
[0011] Optionally, the process of obtaining the real-time carbon emission intensity includes: Obtain real-time vehicle speed, road congestion index, and vehicle load factor; The real-time vehicle speed, road congestion index, and vehicle load rate are input into the carbon emission model to dynamically calculate the real-time carbon emission intensity.
[0012] Optionally, the carbon emission model performs the following calculation process to obtain the real-time carbon emission intensity: Based on real-time vehicle speed Query or calculate the corresponding baseline emission factor ; The correction factor is determined based on the vehicle load rate. ; According to the formula Real-time carbon emission intensity was calculated .
[0013] Optionally, the updated historical average carbon emission intensity is calculated using the following formula: in, The historical average carbon emission intensity stored in the model before fusion. This represents the actual carbon emission intensity calculated so far. The learning rate is preset, and .
[0014] Secondly, this application discloses a synchronous multimodal dynamic routing planning system for low-carbon objectives, the system comprising: The model building module is configured to: construct a multimodal transport digital twin network model based on the topology, node attributes, and directed edge attributes of the multimodal transport physical network; the node attributes include the transshipment capacity of the transshipment nodes; the directed edge attributes include the mode of transport, distance, time cost, and historical average carbon emission intensity as a benchmark parameter. The constraint input module is configured as follows: receiving the origin, destination, and latest delivery time constraints of freight orders; The synchronous optimization module is configured to: map the starting point and the ending point to the multimodal transport digital twin network model, construct a dual-objective optimization model with the goal of minimizing the total transportation cost and the estimated total carbon emissions, and generate an initial low-carbon routing scheme by synchronously and integratedly solving the constraints of the latest delivery time and the node transshipment capacity. The dynamic adjustment module is configured to: during freight execution, update the real-time time cost and real-time carbon emission intensity in the multimodal transport digital twin network model based on real-time traffic information; when the real-time carbon emission intensity of the current path or the delay estimated based on the real-time time cost exceeds a preset threshold, trigger rerouting planning based on the updated multimodal transport digital twin network model and output a dynamically adjusted low-carbon routing scheme. The verification feedback module is configured to: collect the actual transportation mode, actual energy consumption data, and actual travel time for each road segment actually experienced in the dynamically adjusted low-carbon routing scheme; calculate the actual carbon emission intensity of the road segment based on the actual energy consumption data; fuse the actual carbon emission intensity with the historical average carbon emission intensity corresponding to the road segment in the multimodal transport digital twin network model to obtain the updated historical average carbon emission intensity, and store it in the multimodal transport digital twin network model.
[0015] Beneficial Effects: The synchronous multimodal transport dynamic routing planning method and system proposed in this application, aimed at achieving low-carbon goals, constructs a multimodal transport digital twin network model and organically combines synchronous optimization, dynamic adjustment, and verification feedback. This enables integrated synchronous decision-making on transport modes, routes, and transfer nodes, overcoming the limitations of traditional sequential planning. It can find a more optimal balance between cost and carbon emissions from a global perspective. Through a dynamic adjustment mechanism driven by real-time traffic information, it enhances the adaptability of the planning scheme to uncertainties in the transport process, ensuring that low-carbon routes can be dynamically reconstructed when encountering events such as congestion or weather, thus guaranteeing the real-time nature and reliability of low-carbon goals. Furthermore, through continuous self-learning and optimization capabilities, the multimodal transport digital twin network model becomes increasingly accurate, thereby continuously producing more practical and efficient low-carbon routing schemes. Attached Figure Description
[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0017] Figure 1 A flowchart illustrating the synchronous multimodal dynamic routing planning method for low-carbon objectives provided in this application embodiment; Figure 2 This is a structural block diagram of a synchronous multimodal dynamic routing planning system for low-carbon goals provided in an embodiment of this application. Detailed Implementation
[0018] To facilitate understanding of the technical solutions provided in the embodiments of this application, the background technology involved in the embodiments of this application will be described below.
[0019] In the freight logistics sector, multimodal transport, by combining various modes of transportation such as road, rail, waterway, and air, is considered a key model for improving the efficiency of long-distance transportation and reducing total costs. However, existing multimodal transport route planning technologies still suffer from structural deficiencies when addressing low-carbon goals and complex dynamic environments. These deficiencies are mainly reflected in the following two interrelated aspects: First, at the planning model and optimization objective level, traditional methods typically treat "low carbon" as a rigid constraint or a post-event evaluation indicator, rather than a core optimization objective alongside "economic cost." For example, some methods only plan for the lowest cost under the premise of meeting a preset carbon emission cap. This causes the solution to hover around the constraint boundary, unable to actively explore a better balance between cost and carbon emissions. More importantly, most existing methods adopt "segmented optimization" or "sequential decision-making" strategies, that is, first planning the route, and then assigning transportation modes to each segment of the route, or first determining the sequence of transportation modes, and then planning the specific route. This decoupled decision-making process artificially severs the inherent strong coupling relationship between transportation mode selection, physical route orientation, and transshipment node connections. Due to the huge differences in network topology, cost structure, and carbon emission intensity of different transportation modes, this sequential, segmented decision-making approach is prone to getting trapped in local optima. For example, a route that seems to be the shortest in the route planning stage may be forced to use high-carbon emission road transportation throughout the entire route due to the lack of suitable railway or waterway connection nodes, and ultimately is not optimal from a global perspective.
[0020] Secondly, regarding the dynamism and adaptability of the plans, the vast majority of existing plans are "static plans." These plans are generated based on historical average data or static snapshots of information at the time of planning, and once finalized, they assume the transportation environment remains unchanged and are executed in a fixed manner throughout. However, a single multimodal transport mission often lasts for days or even weeks. During this time, real-time factors such as road network congestion, weather, and port clearance efficiency significantly affect the actual travel time and carbon emission intensity of each route segment. For example, a section of highway might be clear during the planning phase, but a day later it becomes severely congested, resulting in vehicles idling for extended periods and carbon emissions far exceeding the planned expectations, effectively failing to achieve the original low-carbon goals. Existing static plans lack on-the-go dynamic monitoring and response mechanisms, making it impossible to adjust routes and transportation modes in a timely manner when the environment changes. This renders the meticulously crafted static optimizations ineffective in actual implementation, making it difficult to guarantee carbon emission control targets.
[0021] In summary, the fundamental problem with existing technologies lies in their failure to construct an integrated planning framework capable of simultaneously processing multi-objective, multi-dimensional decisions, dynamically responding to real-time information, and continuously evolving. This leads to a dual disconnect between the global optimization level and the actual implementation level of low-carbon multimodal transport. This application proposes a novel solution to address the aforementioned systemic deficiencies.
[0022] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application. Secondly, in this document, the term "comprising" is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements, but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.
[0023] Firstly, this embodiment provides a synchronous multimodal dynamic routing planning method oriented towards low-carbon goals, such as... Figure 1 As shown, the method includes the following steps in sequence: Model construction steps: Based on the topology, node attributes, and directed edge attributes of the multimodal transport physical network, construct a multimodal transport digital twin network model; node attributes include the transshipment capacity of transshipment nodes; directed edge attributes include transport mode, distance, time cost, and historical average carbon emission intensity as a benchmark parameter; Input constraints: Receive the origin, destination, and latest delivery time constraints for freight orders; Synchronous optimization steps: Map the starting point and the destination to the multimodal transport digital twin network model, construct a dual-objective optimization model with the goal of minimizing the total transportation cost and the estimated total carbon emissions, and use the latest delivery time constraint and node transshipment capacity as constraints to generate an initial low-carbon routing scheme through synchronous integrated solution. Dynamic adjustment steps: During freight execution, based on real-time traffic information, update the real-time time cost and real-time carbon emission intensity in the multimodal transport digital twin network model; when the real-time carbon emission intensity of the current path or the delay estimated based on real-time time cost exceeds a preset threshold, trigger rerouting planning based on the updated multimodal transport digital twin network model, and output a dynamically adjusted low-carbon routing scheme. Verification and feedback steps: For each road segment actually traversed in the dynamically adjusted low-carbon routing scheme, collect its actual transportation mode, actual energy consumption data, and actual travel time; based on the actual energy consumption data, calculate the actual carbon emission intensity of the road segment; fuse the actual carbon emission intensity with the historical average carbon emission intensity corresponding to the road segment in the multimodal transport digital twin network model to obtain the updated historical average carbon emission intensity, and store it in the multimodal transport digital twin network model.
[0024] Specifically, in the model construction step, the multimodal transport physical network can be abstracted as a directed graph, where nodes represent logistics hubs (such as ports, railway freight yards, and highway freight stations), and their attributes include geographical location and transshipment capacity. Transshipment capacity describes the equipment types, maximum throughput, and unit transshipment cost that the node supports for switching between different modes of transport. Directed edges represent road segments connecting two nodes, and their attributes include transport mode (such as highway, railway, and waterway), geographical distance, time cost, and historical average carbon emission intensity. Historical average carbon emission intensity is a baseline value calibrated based on the average energy consumption data of vehicles under the historical traffic flow conditions of that road segment. These attribute data of nodes and edges constitute the data foundation of the digital twin network model and can be stored in a database.
[0025] In the constraint input step, the system receives freight orders entered by the user through a graphical user interface or application programming interface. The order must include at least the origin, destination, and latest delivery time. The origin and destination are mapped to corresponding nodes in the multimodal transport digital twin network model.
[0026] In the synchronous optimization step, the system performs initial route planning based on the constructed multimodal transport digital twin network model, using the order's origin and destination as the starting and ending points. This step aims to determine the complete transportation plan in one go.
[0027] During the dynamic adjustment process, as goods are transported according to the initial low-carbon route plan, the system continuously collects real-time traffic information through vehicle IoT terminals and a traffic information service platform. This information is used to update the real-time time cost and real-time carbon emission intensity of relevant directed edges in the model. The system continuously monitors the status of the path ahead and triggers a replanning immediately if it determines that replanning is necessary.
[0028] In the validation and feedback step, after each transport mission is completed, the system uses the actual trajectory and energy consumption data uploaded by the onboard terminal to calibrate the model parameters. This step ensures that the model can continuously evolve in response to changes in actual traffic conditions and fleet emission levels.
[0029] Based on the above, a knowledge foundation for the planning was established through the model building step; a globally collaborative initial decision was achieved through the synchronous optimization step; the dynamic adjustment step endowed the solution with the ability to cope with real-time changes; and finally, the verification and feedback step completed the closed loop of the system's self-learning. These four steps work together to solve the technical problems of the inapplicability of static planning and the lack of coordination in serial decision-making, making the generated low-carbon routing solution not only better globally but also more reliable in dynamic execution.
[0030] To address the challenge of unifying decisions across the three dimensions of transportation mode, route nodes, and connectivity within a single framework, as an optional implementation method in this embodiment, a synchronous and integrated solution process includes: Construct unified decision variables that simultaneously encode the choice of transportation mode, the sequence of path nodes, and the connection relationship at transfer nodes; Based on decision variables, the bi-objective optimization model is solved, and the final routing scheme obtained is output as the initial low-carbon routing scheme.
[0031] Specifically, to achieve integrated decision-making, this embodiment designs a composite encoding method for decision variables. These decision variables can be represented as a sequence, where each element corresponds to a "step" on the path. Each step itself is a tuple containing node and transportation mode information, representing a journey from a node, using a specific transportation mode, to the next node. Here, the starting node and the destination node must be nodes in the multimodal transport digital twin network model that have directed edges connected by that transportation mode. When two adjacent steps use different transportation modes, it means a transportation mode change has occurred at the node, and it is necessary to check whether the node's transshipment capacity supports this change. Through this encoding, the decision variable naturally expresses simultaneously the complete path node sequence, the sequence of transportation modes used in each path segment, and the connection relationships occurring at the nodes. Based on this decision variable, the objective function and constraints in the bi-objective optimization model can both be expressed as functions of this decision variable sequence.
[0032] Based on the above, by constructing the unified decision variables, the transportation mode, route nodes, and transshipment connections, which originally required separate decisions at different stages or in different modules, are integrated into a single decision framework. This allows the optimization algorithm to adjust these three dimensions synchronously and in a correlated manner during the search process, avoiding suboptimal solutions that may result from step-by-step decision-making, and ensuring synchronous integration and global optimization from the implementation mechanism.
[0033] To address the problem of efficiently solving complex bi-objective combinatorial optimization problems and determining a unique execution scheme, as a further optional implementation method in this embodiment, the bi-objective optimization model is solved, and a final routing scheme obtained from the solution is output, including: A decomposition-based multi-objective evolutionary algorithm was used to solve the bi-objective optimization model, yielding a set of Pareto optimal solutions; According to the preset decision rules, a solution is selected from a set of Pareto optimal solutions; The complete transportation mode sequence, route sequence, and transfer node sequence corresponding to the selected solution are output as the initial low-carbon routing scheme.
[0034] Specifically, this embodiment employs a decomposition-based multi-objective evolutionary algorithm for solving the problem. First, the algorithm initializes a population, where each individual corresponds to a sequence of decision variables as defined above. The algorithm assigns a weight vector to each individual, and each weight vector defines a subproblem that scalarizes the two objectives: total transportation cost and estimated total carbon emissions. The algorithm iteratively updates the population through evolutionary operations such as crossover and mutation. In each generation, each subproblem selects a parent from its neighboring subproblems for evolution. After a preset number of iterations, the algorithm terminates, outputting all non-dominated solutions in the current population, forming a Pareto optimal solution set. Subsequently, the system selects a final solution from this solution set according to a preset decision rule. The decision rule can be selecting the solution with the lowest total cost, the solution with the lowest total carbon emissions, or defining a comprehensive utility function and selecting the solution with the highest utility value. Finally, the final solution is decoded into explicit routing instructions and output.
[0035] Based on the above, a decomposition-based multi-objective evolutionary algorithm can effectively explore the complex solution space and obtain a set of Pareto optimal solutions that balance cost and carbon emission objectives. Then, through pre-defined decision rules, the algorithm is automatically selected, ensuring both the uniqueness and executability of the final output solution and incorporating user preferences into the decision-making process. This makes the solution not only satisfy dual-objective optimization but also more closely aligned with actual business needs.
[0036] To ensure consistency in the evaluation criteria between the initial planning and dynamic replanning, as an optional implementation method in this embodiment, the optimization objectives of the dual-objective optimization model include the total transportation cost objective function and the estimated total carbon emission objective function. The same total transportation cost objective function and the estimated total carbon emission objective function are used for scheme evaluation in the simultaneous optimization step and the dynamic adjustment step.
[0037] Specifically, the total transportation cost objective function and the estimated total carbon emissions objective function are the core indicators for evaluating any candidate route. A feasible approach is to calculate the total transportation cost objective function as the sum of the transportation costs of all segments of the route, plus the sum of the transshipment costs at all transfer nodes. Similarly, a feasible approach is to calculate the estimated total carbon emissions objective function as the sum of the products of the distance of all segments of the route and their corresponding carbon emission intensities. Here, in the synchronous optimization step, the historical average carbon emission intensity is used; in the dynamic adjustment step, the real-time carbon emission intensity is used. The specific calculation of these two functions follows conventional methods for path cost and carbon emission accounting in existing technologies. The key point is that, whether in the initial synchronous optimization step or in the rerouting planning triggered by the dynamic adjustment step, the system uses objective functions with the exact same structure to evaluate and compare different candidate routes.
[0038] Based on the above, by explicitly defining and uniformly using these two objective functions, the consistency of the entire method in the optimization direction is ensured. Whether performing global planning before freight transport begins or local replanning for the remaining routes during freight transport, the system strives to minimize the same cost and carbon emission targets. This prevents the replanning scheme from deviating from the original objectives due to inconsistent evaluation criteria.
[0039] To address the problem of how to quantitatively assess the risk of the current path and intelligently trigger replanning, as an optional implementation method in this embodiment, the determination method for whether the real-time carbon emission intensity of the preceding path or the delay estimated based on real-time time cost exceeds a preset threshold includes: Calculate the real-time carbon emission intensity of the planned road section ahead. Compared with the carbon emission intensity prediction used when generating the initial low-carbon routing scheme The proportional relationship between them; if satisfied If it exceeds the preset threshold, then it is determined that it exceeds the threshold. The preset ratio threshold; Based on real-time time costs, the arrival time of goods is re-estimated. and compared with the planned arrival time in the initial low-carbon routing scheme. Compare; if satisfied If it exceeds the preset threshold, then it is determined that it exceeds the threshold. This is a preset time threshold.
[0040] Specifically, the system continuously monitors for risks associated with one or more upcoming critical road sections during execution. Regarding carbon emission risks, the system obtains information about the road sections ahead. Current real-time carbon intensity And read the carbon emission intensity predicted for this road segment when generating the initial scheme. Calculate the percentage exceeding the limit. Preset percentage threshold. It is a constant greater than 0, for example, it can be 0.15. If the real-time intensity exceeds the predicted value... If the carbon emission rate is more than doubled, the route is deemed to pose a high carbon emission risk, requiring replanning. For time risk, the system re-estimates the time for goods to reach their destination or the next critical node based on the real-time time cost of each route segment. Preset time threshold It is a constant in units of time, for example, 4 hours. If the re-estimated arrival time is later than the originally planned arrival time by more than [a certain amount], [then the process continues]. If either condition is met, a serious delay risk is identified, requiring a replanning process to be triggered. The two conditions are related by "OR," and triggering the replanning process occurs if either condition is met.
[0041] Based on the above, by setting quantifiable and explicit threshold judgment rules, the vague description of exceeding a preset threshold in the dynamic adjustment step is transformed into automatically executable calculation logic. This enables the system to objectively and consistently assess abnormal states during transportation, initiating replanning only when a risk truly exists, thus ensuring both the reliability of the plan and the system's decision-making efficiency.
[0042] To address the problem of how to quickly re-optimize remaining freight traffic while taking into account past events, as a further optional implementation method of this embodiment, rerouting planning includes: Based on real-time time cost and real-time carbon emission intensity, a bi-objective optimization model is reconstructed for the remaining untransported routes; The actual carbon emissions and time consumption of the current transportation route are used as fixed parameters and input into the reconstructed bi-objective optimization model; Solve the reconstructed bi-objective optimization model to plan new transportation methods, routes, and transfer nodes for the remaining untransported paths.
[0043] Specifically, when replanning is triggered, the system first determines the current location of the goods and the completed routes. Then, using the current location as the new starting point and the original order's destination as the endpoint, the system reconstructs a bi-objective optimization model based on the multimodal transport digital twin network model with updated real-time parameters. The actual total carbon emissions and actual time consumed by the transported routes are calculated and input as fixed values into the new model. In the new model, the overall objective is the cost and carbon emissions of the remaining routes, while the total time constraint must include the fixed time already consumed. Finally, a synchronous integrated solution method is used to solve this new model, obtaining a new optimal route from the current location to the destination.
[0044] Based on the above, this rerouting planning method ensures the continuity of the planning perspective by incorporating completed transportation processes as sunk costs into the constraints of the new model. It only re-optimizes the remaining, variable paths in the future, fully respecting the actual situation that has already occurred while seeking the optimal subsequent solution based on the latest road network conditions, thus achieving the accuracy and efficiency of dynamic adjustment.
[0045] To support the feasibility of dynamic adjustment, as an optional implementation method in this embodiment, the process of obtaining real-time carbon emission intensity includes: Obtain real-time vehicle speed, road congestion index, and vehicle load factor; Real-time vehicle speed, road congestion index, and vehicle load rate are input into the carbon emission model to dynamically calculate the real-time carbon emission intensity.
[0046] Specifically, real-time vehicle speed and road congestion index can be obtained through a real-time traffic interface connected to commercial map services. Vehicle load factor can be obtained through onboard weight sensors or estimated based on waybill information. These real-time parameters are input into a pre-defined carbon emission model. This model reflects the relationship between the carbon emission rate of a specific type of vehicle and its operating conditions. Internally, the model determines the typical operating mode of the vehicle based on vehicle speed and congestion index, and calls the baseline emission rate for that mode. This baseline emission rate is then adjusted according to the vehicle load factor, as increased load typically leads to increased carbon emissions per unit distance. Finally, the model outputs a dynamic real-time carbon emission intensity value.
[0047] Based on the above, the abstract real-time carbon emission intensity is transformed into a dynamic value calculated from specific measurable parameters through an explicit model. This solves the problem of the source of key input data in the dynamic adjustment step, enabling the system to perceive the actual impact of changes in the road network environment and transportation status on carbon emissions.
[0048] To further clarify the internal calculation logic of the carbon emission model, as a further optional implementation method in this embodiment, the carbon emission model performs the following calculation process to obtain the real-time carbon emission intensity: Based on real-time vehicle speed Query or calculate the corresponding baseline emission factor ; The correction factor is determined based on the vehicle load rate. ; According to the formula Real-time carbon emission intensity was calculated .
[0049] Specifically, the carbon emission model in this embodiment adopts a two-level calculation architecture of baseline emission factor and load correction. First, the baseline emission factor... It is the vehicle speed The function. For a specific vehicle type, a database or fitted curve of carbon emissions per unit distance at different steady-state speeds can be established by consulting the vehicle's emission certification data or based on the vehicle's energy consumption model. When the real-time vehicle speed is input... At that time, the corresponding result can be obtained by looking up a table or by calculation. Secondly, the correction coefficient. Used to characterize the impact of load factor on carbon emissions. One feasible relationship is... ,in It's the load factor. It is a load impact factor greater than 0, which can be obtained through experiments or actual operating data. Finally, the carbon emission intensity under the current real-time conditions is calculated according to the formula.
[0050] Based on the above, the technological transformation path from raw data to core parameters is revealed through a specific carbon emission model calculation process. This model comprehensively considers the two most significant factors affecting vehicle emissions: driving speed and load, enabling the calculated real-time carbon emission intensity to more accurately reflect the vehicle's true emission level.
[0051] To support the closed loop of continuous model learning, as an optional implementation method in this embodiment, the updated historical average carbon emission intensity is calculated using the following formula: in, The historical average carbon emission intensity stored in the model before fusion. This represents the actual carbon emission intensity calculated so far. The learning rate is preset, and .
[0052] Specifically, after each transport mission is completed, for each road segment along the actual route, the system has calculated the actual carbon emission intensity of the transport on that road segment based on the actual average vehicle speed and vehicle load. Simultaneously, the system reads the currently stored historical average carbon emission intensity of this route segment from the multimodal transport digital twin network model. Learning rate It is an adjustable parameter, for example, it can be set to... Then, calculate the updated value according to the exponential smoothing formula described above. After the calculation is complete, use Replace the original in the model This serves as the new historical average carbon emission intensity for this section of road.
[0053] Based on the above, the fusion calculation formula provides a stable and efficient way to continuously refine key parameters in the model using new real-world transportation data. This makes the multimodal transport digital twin network model a dynamically evolving model that follows changes in the real world, thereby continuously improving the performance of routing schemes in real-world environments.
[0054] Secondly, this embodiment provides a synchronous multimodal dynamic routing planning system oriented towards low-carbon goals, applying the synchronous multimodal dynamic routing planning method for low-carbon goals described above, such as... Figure 2 As shown, the system includes: The model building module is configured to: construct a multimodal transport digital twin network model based on the topology, node attributes, and directed edge attributes of the multimodal transport physical network; node attributes include the transshipment capacity of transshipment nodes; directed edge attributes include transport mode, distance, time cost, and historical average carbon emission intensity as a benchmark parameter. The constraint input module is configured as follows: receiving the origin, destination, and latest delivery time constraints of freight orders; The synchronous optimization module is configured to: map the origin and destination to the multimodal transport digital twin network model, construct a dual-objective optimization model with the goal of minimizing the total transport cost and the estimated total carbon emissions, and generate an initial low-carbon route scheme through synchronous integrated solution with constraints on the latest delivery time and node transshipment capacity. The dynamic adjustment module is configured to: during freight execution, update the real-time time cost and real-time carbon emission intensity in the multimodal transport digital twin network model based on real-time traffic information; when the real-time carbon emission intensity of the current path or the delay estimated based on real-time time cost exceeds a preset threshold, trigger rerouting planning based on the updated multimodal transport digital twin network model and output a dynamically adjusted low-carbon routing scheme. The verification feedback module is configured to: collect actual transportation mode, actual energy consumption data, and actual travel time for each road segment actually experienced in the dynamically adjusted low-carbon routing scheme; calculate the actual carbon emission intensity of the road segment based on the actual energy consumption data; and fuse the actual carbon emission intensity with the historical average carbon emission intensity corresponding to the road segment in the multimodal transport digital twin network model to obtain the updated historical average carbon emission intensity, which is then stored in the multimodal transport digital twin network model.
[0055] Specifically, this system can be built using one or more servers, databases, and network communication interfaces. The model building module is responsible for backend data management, connecting to geographic information systems and logistics infrastructure databases to initialize and maintain the digital twin network model. The constraint input module and synchronization optimization module are deployed on the application server, providing frontend interaction and planning calculation services. The dynamic adjustment module is a resident background service process that subscribes to real-time traffic data streams and listens to vehicle network data, responsible for real-time updates, risk monitoring, and triggering replanning. The verification feedback module listens for transportation completion events, pulls trip data from the vehicle network platform, performs fusion calculations, and updates model parameters. All modules communicate and exchange data through well-defined internal interfaces.
[0056] Furthermore, it should be noted that the synchronous multimodal dynamic routing planning system for low-carbon goals in this embodiment corresponds to the aforementioned synchronous multimodal dynamic routing planning method for low-carbon goals. Therefore, the parts of the synchronous multimodal dynamic routing planning system for low-carbon goals that are not described in detail (including but not limited to specific technical means and technical effects) can be referred to the relevant descriptions in the aforementioned synchronous multimodal dynamic routing planning method for low-carbon goals, and will not be repeated here.
[0057] In the embodiments provided in this application, it should be understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code, or any suitable combination thereof. For hardware implementation, the processor may be implemented in one or more of the following: application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, other electronic units designed to implement the functions described herein, or combinations thereof. For software implementation, some or all of the processes of the embodiments may be performed by a computer program instructing the associated hardware. During implementation, the program may be stored in a computer-readable storage medium or transmitted as one or more instructions or code on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein communication media include any medium that facilitates the transmission of a computer program from one place to another. Storage media may be any available medium accessible to a computer. Computer-readable storage media may include, but are not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code having the form of instructions or data structures and accessible to a computer.
[0058] Finally, it should be noted that the above description is only a preferred embodiment of this application and is not intended to limit this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.
Claims
1. A synchronous multimodal dynamic routing planning method for low-carbon objectives, characterized in that, The method includes: Model construction steps: Based on the topology, node attributes, and directed edge attributes of the multimodal transport physical network, construct a multimodal transport digital twin network model; the node attributes include the transshipment capacity of the transshipment nodes; the directed edge attributes include the mode of transport, distance, time cost, and historical average carbon emission intensity as a benchmark parameter; Input constraints: Receive the origin, destination, and latest delivery time constraints for freight orders; Synchronous optimization steps: Map the starting point and the ending point to the multimodal transport digital twin network model, construct a dual-objective optimization model with the goal of minimizing the total transport cost and the estimated total carbon emissions, and use the latest delivery time constraint and node transshipment capacity as constraints to generate an initial low-carbon routing scheme through synchronous integrated solution. Dynamic adjustment steps: During freight delivery, based on real-time traffic information, update the real-time time cost and real-time carbon emission intensity in the multimodal transport digital twin network model; when the real-time carbon emission intensity of the current path or the delay estimated based on the real-time time cost exceeds a preset threshold, trigger rerouting planning based on the updated multimodal transport digital twin network model, and output a dynamically adjusted low-carbon routing scheme. Verification and feedback steps: For each road segment actually traversed in the dynamically adjusted low-carbon routing scheme, collect its actual transportation mode, actual energy consumption data, and actual travel time; based on the actual energy consumption data, calculate the actual carbon emission intensity of the road segment; fuse the actual carbon emission intensity with the historical average carbon emission intensity corresponding to the road segment in the multimodal transport digital twin network model to obtain the updated historical average carbon emission intensity, and store it in the multimodal transport digital twin network model.
2. The synchronous multimodal dynamic routing planning method for low-carbon objectives as described in claim 1, characterized in that, The synchronous integrated solution process includes: A unified decision variable is constructed, which simultaneously encodes the choice of transportation mode, the sequence of path nodes, and the connection relationship at transfer nodes; Based on the decision variables, the bi-objective optimization model is solved, and the final routing scheme obtained is output as the initial low-carbon routing scheme.
3. The synchronous multimodal dynamic routing planning method for low-carbon objectives according to claim 2, characterized in that, Solving the bi-objective optimization model and outputting a final routing scheme obtained from the solution includes: The bi-objective optimization model was solved using a decomposition-based multi-objective evolutionary algorithm, yielding a set of Pareto optimal solutions; According to the preset decision rules, a solution is selected from the set of Pareto optimal solutions; The complete transportation mode sequence, path sequence, and transfer node sequence corresponding to the selected solution are output as the initial low-carbon routing scheme.
4. The synchronous multimodal dynamic routing planning method for low-carbon objectives according to claim 1, characterized in that, The optimization objectives of the dual-objective optimization model include a total transportation cost objective function and a predicted total carbon emission objective function. The simultaneous optimization step and the dynamic adjustment step use the same total transportation cost objective function and the predicted total carbon emission objective function to evaluate the schemes.
5. The synchronous multimodal dynamic routing planning method for low-carbon objectives according to claim 1, characterized in that, The method for determining whether the real-time carbon emission intensity of the preceding path or the delay estimated based on real-time time cost exceeds a preset threshold includes: Calculate the real-time carbon emission intensity of the planned road section ahead. Compared with the carbon emission intensity prediction used when generating the initial low-carbon routing scheme The proportional relationship between them; if satisfied If it exceeds the preset threshold, then it is determined that it exceeds the threshold. The preset ratio threshold; Based on real-time time costs, the arrival time of goods is re-estimated. and compared with the planned arrival time in the initial low-carbon routing scheme. Compare; if satisfied If it exceeds the preset threshold, then it is determined that it exceeds the threshold. This is a preset time threshold.
6. The synchronous multimodal dynamic routing planning method for low-carbon objectives according to claim 5, characterized in that, The rerouting plan includes: Based on the real-time time cost and real-time carbon emission intensity, a new bi-objective optimization model is constructed for the remaining untransported routes; The actual carbon emissions and time consumption of the current transportation route are used as fixed parameters and input into the reconstructed bi-objective optimization model; Solve the reconstructed bi-objective optimization model to plan new transportation methods, routes, and transfer nodes for the remaining untransported paths.
7. The synchronous multimodal dynamic routing planning method for low-carbon objectives according to claim 1, characterized in that, The process of obtaining the real-time carbon emission intensity includes: Obtain real-time vehicle speed, road congestion index, and vehicle load factor; The real-time vehicle speed, road congestion index, and vehicle load rate are input into the carbon emission model to dynamically calculate the real-time carbon emission intensity.
8. The synchronous multimodal dynamic routing planning method for low-carbon objectives according to claim 7, characterized in that, The carbon emission model performs the following calculation process to obtain the real-time carbon emission intensity: Based on real-time vehicle speed Query or calculate the corresponding baseline emission factor ; The correction factor is determined based on the vehicle load rate. ; According to the formula Real-time carbon emission intensity was calculated .
9. The synchronous multimodal dynamic routing planning method for low-carbon objectives according to claim 1, characterized in that, The updated historical average carbon emission intensity is calculated using the following formula: in, The historical average carbon emission intensity stored in the model before fusion. This represents the actual carbon emission intensity calculated so far. The learning rate is preset, and .
10. A synchronous multimodal dynamic routing planning system for low-carbon objectives, characterized in that, The system includes: The model building module is configured to: construct a multimodal transport digital twin network model based on the topology, node attributes, and directed edge attributes of the multimodal transport physical network; the node attributes include the transshipment capacity of the transshipment nodes; the directed edge attributes include the mode of transport, distance, time cost, and historical average carbon emission intensity as a benchmark parameter. The constraint input module is configured as follows: receiving the origin, destination, and latest delivery time constraints of freight orders; The synchronous optimization module is configured to: map the starting point and the ending point to the multimodal transport digital twin network model, construct a dual-objective optimization model with the goal of minimizing the total transportation cost and the estimated total carbon emissions, and generate an initial low-carbon routing scheme by synchronously and integratedly solving the constraints of the latest delivery time and the node transshipment capacity. The dynamic adjustment module is configured to: during freight execution, update the real-time time cost and real-time carbon emission intensity in the multimodal transport digital twin network model based on real-time traffic information; when the real-time carbon emission intensity of the current path or the delay estimated based on the real-time time cost exceeds a preset threshold, trigger rerouting planning based on the updated multimodal transport digital twin network model and output a dynamically adjusted low-carbon routing scheme. The verification feedback module is configured to: collect the actual transportation mode, actual energy consumption data, and actual travel time for each road segment actually experienced in the dynamically adjusted low-carbon routing scheme; calculate the actual carbon emission intensity of the road segment based on the actual energy consumption data; fuse the actual carbon emission intensity with the historical average carbon emission intensity corresponding to the road segment in the multimodal transport digital twin network model to obtain the updated historical average carbon emission intensity, and store it in the multimodal transport digital twin network model.