Green vehicle routing method and system for a logistics site

Abstract

The application belongs to the field of intelligent logistics, and specifically discloses a green vehicle path planning method and system for a logistics site, which comprises the following steps: acquiring path planning basic data; based on the basic data, constructing a quantitative model for dynamically calculating carbon emissions based on vehicle driving time, speed and load, setting constraint conditions for path planning based on logistics information, and establishing a multi-objective function including a driving distance target and a carbon emission target obtained based on the quantitative model; using a tent mapping to generate an initial population of a starling optimization algorithm, with each starling individual corresponding to a path planning solution; using an improved starling optimization algorithm with a polynomial mutation disturbance strategy to iteratively solve the multi-objective function under the constraint conditions, and outputting an optimized path planning solution. Through the application, the dynamic and accurate calculation of carbon emissions of vehicles under different load conditions and driving speeds is realized, and a more low-carbon and green vehicle driving path for a logistics site is planned.

Description

Technical Field

[0001] This application belongs to the field of intelligent logistics and low-carbon transportation, and more specifically, relates to a green vehicle route planning method and system for logistics stations. Background Technology

[0002] Green vehicles typically refer to logistics and transportation vehicles powered by clean energy and with low carbon emissions during operation. Route planning for green vehicles is a key technological challenge in intelligent logistics and low-carbon transportation, aiming to minimize energy consumption and carbon emissions during transportation while meeting delivery needs. Existing technologies mainly focus on two main areas: problem models and route search algorithms.

[0003] In terms of problem models, existing technologies have gradually expanded from simply minimizing fuel consumption or carbon emissions to a multi-objective optimization system that integrates economic costs, environmental impact, and social satisfaction, and incorporates the needs of complex scenarios such as dynamic road conditions and time window constraints. Regarding path search algorithms, existing technologies are based on metaheuristic algorithms and have developed genetic algorithms for optimizing parameters, particle swarm optimization algorithms incorporating greedy mutation operators, multi-ant colony optimization models, and hybrid heuristic algorithms.

[0004] However, existing methods still have the following drawbacks: First, carbon emission models are mostly based on driving distance or average fuel consumption for estimation, and have low sensitivity to factors such as load changes and driving speed fluctuations, resulting in insufficient accuracy in carbon emission calculation; second, path search algorithms are prone to getting trapped in local optima in large-scale complex path searches, resulting in low solution efficiency and difficulty in adapting to the dynamic scheduling needs of distributed logistics stations; third, the matching and synergy between multi-objective functions and search algorithms are insufficient, making it difficult to achieve an effective balance between economy and environmental protection. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this application aims to provide a green vehicle route planning method and system for logistics stations, which addresses the problems of insufficient accuracy in carbon emission model calculations, limited global optimization capabilities of route search algorithms, and insufficient synergy between multi-objective functions and search algorithms.

[0006] The first aspect of this application relates to a green vehicle route planning method for logistics stations. The method includes: acquiring basic route planning data, including logistics information of each logistics station and carbon emission parameters of transport vehicles; constructing a quantitative model for dynamically calculating carbon emissions based on vehicle travel time, speed, and load based on the basic data; setting constraints for route planning based on the logistics information; and establishing a multi-objective function including a travel distance target and a carbon emission target obtained based on the quantitative model; generating an initial population for a sparrow optimization algorithm using tent mapping, where each sparrow corresponds to a route planning solution, and the route planning solution includes a service station sequence for each transport vehicle; and employing an improved sparrow optimization algorithm with a polynomial mutation perturbation strategy to iteratively solve the multi-objective function under the constraints, and outputting an optimized route planning solution.

[0007] The method provided in this invention significantly improves the accuracy of carbon emission calculations under different operating conditions by constructing a quantitative model for dynamically calculating carbon emissions based on vehicle travel time, speed, and load, replacing traditional static or load-ignoring estimation methods. Simultaneously, by incorporating both travel distance and dynamic carbon emissions into a multi-objective function, the two objectives of minimizing distance and carbon emissions are optimized collaboratively within the same optimization framework, avoiding local optima or objective conflicts caused by their separation. This effectively solves the problems of insufficient accuracy in carbon emission model calculations and poor synergy between multi-objective functions and search algorithms in existing technologies, achieving comprehensive optimal route planning under the goal of green logistics.

[0008] Furthermore, by utilizing tent mapping to generate the initial population for the sparrow optimization algorithm, the initial solutions are more evenly distributed and more diverse in the solution space, laying a solid foundation for global search. A multinomial mutation perturbation strategy is introduced to improve the algorithm's iterative process, enhancing the ability of individual sparrows to escape local optima and preventing premature convergence. By iteratively solving multi-objective functions using the improved sparrow optimization algorithm, the global optimization capability of the path search algorithm is significantly improved. It can stably output higher-quality path planning solutions under complex constraints, overcoming the technical defect of traditional algorithms that easily get trapped in local optima.

[0009] In some implementations, the construction of a quantitative model for dynamically calculating carbon emissions based on vehicle travel time, speed, and load includes: calculating engine operating fuel consumption, speed-related fuel consumption, and load-related fuel consumption based on the vehicle's current travel time, speed, and load; summing the engine operating fuel consumption, speed-related fuel consumption, and load-related fuel consumption as the vehicle's total fuel consumption; and multiplying the total fuel consumption by the carbon emissions per unit of fuel consumption to construct the quantitative model for dynamically calculating carbon emissions.

[0010] This embodiment decomposes carbon emissions into three independent and quantifiable fuel consumption components: engine operation, speed variation, and load variation, achieving a refined modeling of the vehicle's actual operating conditions. Compared to traditional methods that only consider distance or fixed emission factors, this dynamic quantification model can respond in real time to the combined effects of travel time, speed fluctuations, and load changes on fuel consumption, significantly improving the physical realism and accuracy of carbon emission calculations. It provides accurate objective function inputs for subsequent multi-objective path planning, thereby ensuring the effectiveness and reliability of the green vehicle path optimization results.

[0011] In some implementations, establishing a multi-objective function that includes a travel distance target and a carbon emission target obtained based on the quantification model includes: calculating the total distance traveled by each transport vehicle along the planned route based on the travel distance between logistics stations, as the travel distance target; calculating the total carbon emissions generated by each transport vehicle based on the quantification model, as the carbon emission target; and weighting and summing the travel distance target and the carbon emission target according to their respective weights to form a multi-objective function with the objective of minimizing the weighted sum.

[0012] This embodiment constructs a multi-objective function that incorporates both travel distance and carbon emission optimization objectives, and uses weighting coefficients to weight and synthesize the two, thereby achieving simultaneous optimization of travel costs and environmental impact during route planning. It effectively reduces carbon emissions during transportation while ensuring transportation efficiency, balancing the economic benefits and environmental requirements of logistics operations.

[0013] In some implementations, the expression for the multi-objective function is:

[0014] Among them, the function F ijk ( ) represents a vehicle k After arc ( i , j Arrival at the logistics station j Total fuel consumption t ijk For vehicles k After arc ( i , j Arrival at the logistics station j travel time v ijk For vehicles k After arc ( i , j Arrival at the logistics station j driving speed, w ijk vehicle k After arc (i , j Arrival at the logistics station j Vehicle load capacity, h Carbon emissions per unit of fuel consumed. For logistics stations i to logistics station j driving distance, A value of 1 indicates a logistics station. i and j If there is a transport route, then the vehicle k After arc ( i , j ), A value of 0 indicates a logistics station. i and j There is no transportation route between them. m 1 represents the weight of the target driving distance. m 2 represents the weight of the carbon emission target.

[0015] This embodiment unifies driving distance and dynamic carbon emissions into a differentiable mathematical expression, where the carbon emission term is dynamically calculated based on time, speed, and load over the arc segment. This function is directly compatible with the improved sparrow optimization algorithm, providing an accurate basis for calculating evaluation values ​​for iterative solutions. It achieves synergistic optimization of the two objectives of minimizing distance and carbon emissions, avoiding the objective conflict problem in the traditional weighted method.

[0016] In some implementations, the constraint conditions include at least one of the following: Each logistics station is visited only once by a vehicle; each vehicle must depart from the logistics center, visit several logistics stations, and return to the logistics center; the arrival and departure times of vehicles at each logistics station must satisfy a time logic relationship; the load of a vehicle after visiting a logistics station changes dynamically according to the two-way logistics flow, and the load on any road segment does not exceed its rated load capacity; the arrival time of a vehicle at each logistics station is within the time window of that station; vehicles only travel directly between two logistics stations with a driving route.

[0017] This embodiment constructs a complete feasible solution space for path planning by setting multiple constraints, including coverage access uniqueness, path closure, temporal logical consistency, dynamic load constraints, time window constraints, and connectivity constraints. The dynamic load constraint under bidirectional freight flow ensures that vehicles do not overload on any road segment during delivery and recycling, more closely reflecting actual operational scenarios. These constraints, closely integrated with multi-objective functions and the improved sparrow optimization algorithm, effectively filter out infeasible solutions, guiding the search process to only perform searches on practically meaningful path combinations, thereby improving optimization efficiency and solution quality while ensuring scheduling feasibility.

[0018] In some embodiments, the method further includes: during the iterative solution process, calculating the evaluation value of the path planning solution corresponding to each individual starbird based on the multi-objective function, and penalizing the evaluation value based on the constraints: when violating the unique access constraint of the logistics station or the origin-end point constraint of the vehicle, the evaluation value is reduced by a first preset percentage; when violating the time window constraint or the path existence constraint, the evaluation value is reduced by a second preset percentage; when violating the time logic constraint or the load constraint, the evaluation value is set to a preset minimum value.

[0019] This embodiment employs a tiered penalty mechanism to rigorously address hard constraints (such as unique access and start / endpoint constraints) while moderately adjusting soft constraints (such as time windows), effectively balancing the feasibility of solutions with population diversity. Compared to directly discarding invalid solutions, this method guides the algorithm to gradually approach the feasible region, improving the solution efficiency and convergence stability of path planning under complex constraints.

[0020] In some implementations, the initial population for generating the starbird optimization algorithm using tent mapping includes: determining the population size and problem dimension based on the number of logistics stations; randomly selecting initial values ​​and iteratively generating a chaotic sequence matrix using the tent mapping formula, wherein the length of the chaotic sequence matrix is ​​the product of the population size and the problem dimension; linearly mapping each element value in the chaotic sequence matrix to the solution space range to obtain the position coordinates of each starbird individual in the initial population in each dimension.

[0021] This embodiment utilizes the chaotic properties of the tent mapping to generate an initial population. The chaotic sequence matrix generated through iteration possesses traversal uniformity and randomness, resulting in a more even distribution and higher diversity of sparrow individuals in the solution space. Compared to random initialization, this method effectively avoids population clustering in local regions, providing a broader initial solution set for subsequent iterative searches, thereby enhancing the global optimization capability of the improved sparrow optimization algorithm.

[0022] In some implementations, the improved starbird optimization algorithm employing a multinomial mutation perturbation strategy iteratively solves the multi-objective function under the constraints, including: generating two random numbers in each iteration; and selecting either a foraging and storage strategy or a cache search and recovery strategy to update the starbird position based on the comparison result of the two random numbers; the foraging and storage strategy includes: generating a first random number and a probability threshold corresponding to the current iteration; if the first random number is greater than the probability threshold, the foraging strategy is used to update the starbird position, and a position perturbation based on a multinomial distribution is introduced after the update; otherwise, the storage strategy is used to update the starbird position; the cache search and recovery strategy includes: generating two different position reference points, generating a second random number and a preset probability threshold; if the second random number is greater than the preset probability threshold, the cache recovery strategy is used to update the starbird position based on the position reference points; otherwise, the cache search strategy is used to update the starbird position based on the position reference points.

[0023] This embodiment employs a double random number comparison mechanism to dynamically select between two strategies—foraging and storage or cache search and recovery—in each iteration, achieving an adaptive balance between the algorithm's exploration and development capabilities. Specifically, the foraging strategy introduces a multinomial distribution perturbation, enabling controlled mutations of sparrow individuals near local optima and effectively enhancing their ability to escape local extrema. The cache search and recovery strategy guides position updates through two different reference points, improving population diversity. This multi-strategy collaborative update mechanism, combined with the perturbation term, improves the global optimization performance and convergence stability of the sparrow optimization algorithm in complex, multi-constraint path planning problems.

[0024] In some implementations, the introduction of position perturbation based on multinomial distribution includes: generating a random number for each dimension of the position vector of the starbird to be mutated; calculating the perturbation value of each dimension using an asymmetric multinomial distribution function based on the relationship between the random number and a preset threshold, wherein the distribution index is used to control the magnitude of the perturbation; superimposing the perturbation value onto the original position coordinates of that dimension to obtain the mutated position, and controlling the mutated position to be within the upper and lower bounds of the solution space, thus completing one mutation operation.

[0025] This embodiment achieves refined positional mutation by independently generating random numbers for each dimension and calculating perturbation values ​​based on a multinomial distribution function. The distribution index can flexibly adjust the perturbation intensity, allowing the algorithm to adapt to different exploration needs at different iteration stages. Simultaneously, a boundary control mechanism ensures that the mutated solution remains within the feasible solution space. This perturbation strategy effectively enhances the ability of individual sparrows to escape local optima, increases population diversity, and further improves the algorithm's global optimization performance and solution quality in complex path planning problems.

[0026] The second aspect of this application relates to a green vehicle route planning system for logistics stations. The system includes: a basic data acquisition unit for acquiring basic route planning data, including logistics information for each logistics station and carbon emission parameters of transport vehicles; an objective function construction unit for constructing a quantitative model based on the basic data, dynamically calculating carbon emissions based on vehicle travel time, speed, and load; setting route planning constraints based on the logistics information; and establishing a multi-objective function including a travel distance objective and a carbon emission objective obtained from the quantitative model; an initialization unit for generating an initial population for a sparrow optimization algorithm using tent mapping, where each sparrow corresponds to a route planning solution, and the route planning solution includes a service station sequence for each transport vehicle; and an iterative solution unit for using an improved sparrow optimization algorithm incorporating a polynomial mutation perturbation strategy to iteratively solve the multi-objective function under the constraints, and outputting an optimized route planning solution.

[0027] Thirdly, this application provides an electronic device, comprising: at least one memory for storing a program; and at least one processor for executing the program stored in the memory, wherein when the program stored in the memory is executed, the processor is configured to execute the method described in the first aspect or any possible implementation thereof.

[0028] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when run on a processor, causes the processor to perform the method described in the first aspect or any possible implementation thereof.

[0029] Fifthly, this application provides a computer program product that, when run on a processor, causes the processor to perform the method described in the first aspect or any possible implementation thereof.

[0030] It is understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here.

[0031] Overall, the technical solutions conceived in this application have the following beneficial effects compared with the prior art: (1) By using a quantitative model that dynamically calculates carbon emissions based on vehicle driving time, speed and load, the carbon emissions of vehicles under different load conditions and driving speeds can be dynamically and accurately measured, thus improving the accuracy of carbon emission assessment.

[0032] (2) An improved star sparrow optimization algorithm with tent mapping initialization and polynomial mutation perturbation strategy is adopted. The chaotic sequence generated by tent mapping is used to initialize the population, which effectively improves the uniformity and diversity of the initial solution of the population. By introducing polynomial mutation perturbation in the foraging strategy, the algorithm's ability to escape local optima is enhanced, the global optimization success rate and convergence speed are improved, and it is better adapted to the dynamic scheduling requirements of distributed logistics stations.

[0033] (3) A multi-objective function that comprehensively considers driving distance, load capacity and carbon emissions was established, and intelligent optimization was achieved through the improved star sparrow optimization algorithm. This effectively solved the problem of insufficient synergy between the existing multi-objective function and path search algorithm, effectively reduced the total transportation cost and carbon emissions, and achieved the optimal balance between economic cost and environmental benefits. Attached Figure Description

[0034] Figure 1 This is one of the flowcharts illustrating the green vehicle route planning method for logistics stations provided in this application embodiment.

[0035] Figure 2 This is a flowchart illustrating the search path planning solution provided in the embodiments of this application.

[0036] Figure 3 This is a flowchart illustrating the foraging and storage strategy provided in the embodiments of this application.

[0037] Figure 4 This is a flowchart illustrating the cache search and recovery strategy provided in the embodiments of this application.

[0038] Figure 5 This is a schematic diagram of the structure of a green path planning system provided in an embodiment of this application. Detailed Implementation

[0039] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0040] In this application, the term "and / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A existing alone, A and B existing simultaneously, and B existing alone. In this application, the symbol " / " indicates that the related objects are in an "or" relationship, for example, A / B means A or B.

[0041] In this application, the terms "first" and "second," etc., are used to distinguish different objects, not to describe a specific order of objects. For example, "first response message" and "second response message," etc., are used to distinguish different response messages, not to describe a specific order of response messages.

[0042] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0043] In the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more, for example, multiple processing units means two or more processing units, multiple elements means two or more elements, etc.

[0044] The embodiments of this application are described below with reference to the accompanying drawings. Figure 1 This is one of the flowcharts illustrating the green vehicle route planning method for logistics stations provided in this application embodiment, such as... Figure 1 As shown, the method includes steps 110-140.

[0045] Step 110: Obtain basic route planning data, which includes logistics information for each logistics station and carbon emission parameters for transport vehicles.

[0046] Specifically, logistics information serves as the foundation for route planning, encompassing key information such as the geographical coordinates of logistics centers and various stations, distance matrices between stations, cargo demand at each station (including pickup and delivery volumes), and service time windows. Carbon emission parameters for transport vehicles include performance parameters and environmental constants. Vehicle performance parameters include physical parameters directly related to vehicle energy consumption, such as vehicle weight, rated load capacity, engine displacement, and friction coefficient. Environmental constants involve external environmental factors such as air density, gravitational acceleration, and road rolling resistance coefficient. Basic data can be obtained directly through the logistics management system's database interface, or through manual input or sensor data collection.

[0047] In some implementations, the logistics information for each logistics station includes: (1) Distance matrix parameters: a symmetrical driving distance matrix between the logistics center and all logistics stations, as well as between each logistics station, to clarify the round-trip transportation distance between any two points.

[0048] (2) Two-way flow rate: The flow rate of goods at each logistics station is distinguished by positive and negative values. Positive values ​​indicate the quantity of goods loaded from the logistics station onto the transport vehicle (goods are loaded from the station onto the vehicle), while negative values ​​indicate the quantity of goods unloaded from the transport vehicle onto the logistics station (goods are unloaded from the vehicle onto the station).

[0049] (3) Time constraints: the earliest allowed time and the latest deadline for transport vehicles to arrive at the logistics station.

[0050] Carbon emissions from transport vehicles are influenced by factors such as vehicle type, load, speed, and distance traveled. Assuming... n Indicates the number of logistics stations; n+ 1 indicates returning to the processing center. Since there is only one processing center, it can be set... n + 1 ≡ 0; V This represents the number of transport vehicles in the entire system. Carbon emission parameters are shown in Table 1 below: Table 1 Carbon Emission Parameters for Transport Vehicles

[0051] Step 120: Based on the basic data, construct a quantitative model for calculating carbon emissions dynamically based on vehicle travel time, speed and load, set constraints for route planning based on logistics information, and establish a multi-objective function that includes a travel distance target and a carbon emission target obtained based on the quantitative model.

[0052] Specifically, considering that traditional carbon emission models often estimate emissions based solely on driving distance and ignore the nonlinear effects of vehicle load and speed on fuel consumption, the quantitative model constructed in this embodiment can dynamically calculate carbon emissions based on the vehicle's real-time driving status, thereby more realistically reflecting the actual emissions of vehicles under different road sections and loads, and providing accurate evaluation indicators for green route planning.

[0053] In some implementations, step 120, which involves constructing a quantitative model for dynamically calculating carbon emissions based on vehicle travel time, speed, and load, includes: Step 121: Based on the vehicle's current travel time, speed, and load, calculate the engine operating fuel consumption, speed-related fuel consumption, and load-related fuel consumption respectively. Step 122: The sum of engine operating fuel consumption, speed-related fuel consumption, and load-related fuel consumption is taken as the vehicle's total fuel consumption. Step 123: Multiply the total fuel consumption by the carbon emissions per unit of fuel consumption to construct a quantitative model for dynamically calculating carbon emissions.

[0054] Specifically, this embodiment decomposes the vehicle's total fuel consumption into three physically distinct components. First, there is engine-related fuel consumption. This is mainly related to engine idling, internal friction, and maintaining the engine's basic operating state. Even when the vehicle is stationary, this consumption exists as long as the engine is running, and its magnitude is primarily affected by travel time. Second, there is speed-related fuel consumption. This mainly stems from the work done to overcome air resistance during vehicle movement. According to aerodynamic principles, air resistance is proportional to the square of the speed, while the power required to overcome resistance is proportional to the cube of the speed. Therefore, this consumption is extremely sensitive to changes in speed, reflecting the energy consumption difference between highway driving and urban congestion. Finally, there is load-related fuel consumption. This mainly originates from the work done by the vehicle to overcome rolling resistance and gradient resistance. The greater the load, the greater the tire deformation in contact with the ground, resulting in greater rolling resistance and higher energy required for climbing hills. This consumption directly reflects the reality of "higher consumption with heavier loads, lower consumption with lighter loads" in logistics and distribution.

[0055] As a preferred embodiment, the expression for total fuel consumption is:

[0056] in, t For vehicle travel time, v For vehicle speed, w For vehicle load capacity, For the vehicle's own weight. For driving distance, , , These are coefficients determined based on the performance parameters and environmental constants.

[0057] This refers to the fuel consumption for engine operation. The coefficient is... It integrates performance parameters such as engine friction coefficient, engine speed, engine displacement, and vehicle transmission efficiency to reflect the vehicle's basic energy consumption level per unit time.

[0058] This represents speed-related fuel consumption. The coefficient is... It integrates parameters such as vehicle transmission efficiency, diesel engine efficiency, vehicle front area, aerodynamic drag coefficient, and air density to reflect the vehicle's ability to overcome air resistance.

[0059] This refers to load-related fuel consumption. The coefficient is... It integrates parameters such as vehicle transmission efficiency, diesel engine efficiency, road turning angle, rolling resistance coefficient, and gravitational acceleration to reflect the vehicle's ability to overcome rolling resistance and gradient resistance.

[0060] It should be understood that the above coefficients , , It is not a fixed constant, but is calibrated according to the specific vehicle type (such as light trucks and heavy trucks) and driving environment (such as plains and mountains). In practical applications, it can be calculated and determined through the performance parameter table provided by the vehicle manufacturer and the environmental constant of the driving section.

[0061] In some implementations, the above coefficients , , It can be calculated based on the following formula:

[0062]

[0063]

[0064] Please refer to Table 1 for the physical meaning of each parameter in the formula.

[0065] Finally, the total fuel consumption is multiplied by the carbon emissions per unit of fuel consumption to construct a quantitative model for dynamically calculating carbon emissions. This quantitative model allows for dynamic adjustment of the calculated carbon emissions based on the vehicle's real-time speed and load status, overcoming the shortcomings of traditional static models that cannot reflect actual road conditions and load changes. This provides a precise environmental cost assessment basis for subsequent multi-objective path planning.

[0066] In some embodiments, step 120 establishes a multi-objective function that includes a driving distance target and a carbon emission target obtained based on a quantification model, including: Based on the driving distance between logistics stations, the total distance traveled by each transport vehicle along the planned route is calculated and represented as the driving distance target; Based on a quantitative model, the total carbon emissions generated by each transport vehicle are calculated and represented as a carbon emission target. Based on the weights corresponding to the driving distance target and the carbon emission target, the two are weighted and summed to form a multi-objective function with the goal of minimizing the weighted sum value.

[0067] Specifically, in order to simultaneously consider the economic and environmental benefits of logistics transportation, this embodiment establishes a multi-objective function that includes a travel distance target and a carbon emission target. When calculating the travel distance target, the planned routes of all transport vehicles can be traversed, and the distances between each station visited by each vehicle can be accumulated to obtain the total travel distance of that vehicle. Then, the total travel distances of all vehicles are summed, and the final result is the travel distance target of the entire transportation task.

[0068] When calculating the carbon emission target, a quantitative model is used. This model dynamically calculates the fuel consumption and corresponding carbon emissions for each segment of the journey based on the vehicle's travel time, real-time speed, real-time load, and carbon emissions per unit of fuel consumed. The carbon emissions of all vehicles across all route segments are then aggregated to obtain the overall carbon emission target for the transportation mission.

[0069] Finally, adjustable weighting coefficients are configured for the travel distance target and the carbon emission target respectively. The two targets are multiplied by their respective weights and then added together to obtain the multi-objective function. This function guides the algorithm to automatically find the transportation route scheme that achieves the best balance between distance and emissions in the subsequent solution process.

[0070] Based on the above embodiments, the expression for the multi-objective function is:

[0071] Among them, the function F ijk ( ) represents a vehicle k After arc ( i , j Arrival at the logistics station j Total fuel consumption t ijk For vehicles k After arc ( i , j Arrival at the logistics station j travel time v ijk For vehicles k After arc ( i , j Arrival at the logistics station j driving speed, w ijk vehicle k After arc ( i , j Arrival at the logistics station j Vehicle load capacity, h Carbon emissions per unit of fuel consumed. For logistics stations i to logistics station j driving distance, A value of 1 indicates a logistics station. i and j If there is a transport route, then the vehicle k After arc ( i , j ), A value of 0 indicates a logistics station. i and jThere is no transportation route between them. m 1 represents the weight of the target driving distance. m 2 represents the weight of the carbon emission target.

[0072] Specifically, the summation symbol in the expression iterates through all vehicles k and all logistics stations i and j. By adding the travel distance and carbon emissions, the objective function achieves a unified measure of the total transportation cost.

[0073] In the formula, decision variables Essentially a binary switch variable, it associates discrete path selection behavior with continuous objective function values. This applies if and only if vehicle k is planned to travel from station... i Arrive at the station j hour, The value is 1, which represents the distance cost of that road segment. and carbon emission costs Only those marked with a dot (·) will be included in the overall objective function value; conversely, if the road segment is not selected, When the value is 0, the corresponding cost term is also 0. This design allows the originally complex path combination optimization problem to be transformed into a mathematical programming problem for solution. During the iterative solution process, the algorithm continuously adjusts the decision variables. The optimal solution for minimizing the overall objective function is found by combining different values ​​of the objective function, i.e., changing the vehicle's travel route. This mathematical modeling of the route planning problem not only facilitates efficient solutions using intelligent optimization algorithms but also ensures that the planning results accurately reflect the specific contribution of each route selection to economic costs and carbon emissions, thereby achieving precise quantitative optimization of green logistics routes.

[0074] In some implementations, when constructing a route planning model, a series of constraints need to be introduced to define the boundaries of feasible solutions in order to ensure that the planning results conform to the physical constraints and business rules of actual logistics operations. This method includes the following constraints in a single route planning task: (1) All vehicles depart from the logistics center and return to the logistics center after completing the loading / unloading tasks at the logistics center and various logistics stations.

[0075] (2) The loading / unloading demand of each logistics station is known.

[0076] (3) The loading / unloading tasks at each logistics station are completed by the same vehicle in one go, avoiding repeated trips.

[0077] (4) The sum of the current loading volume of the logistics center and each logistics station and the existing cargo volume of the vehicle shall not exceed the rated load limit of the vehicle.

[0078] (5) The transportation network is a symmetrical graph, that is, the distance between the vehicles and the logistics stations is equal and satisfies the triangle inequality.

[0079] To address the aforementioned constraints, this embodiment sets constraints for the path planning model, including at least one of the following: (1) Each logistics station is accessed only once by one vehicle to avoid resource waste or service conflicts caused by multiple vehicles repeatedly accessing the same station. The expression is:

[0080] (2) Each vehicle must depart from the logistics center, visit several logistics stations, and finally return to the logistics center. This is constrained by the following expression:

[0081] (3) The arrival and departure times of vehicles at each logistics station must satisfy a temporal logical relationship. This mainly refers to the causal logic between the vehicle's travel time between nodes, the service time at each node, and the arrival and departure times. For example, the time a vehicle leaves the current station must be later than the arrival time plus the service time, and the time it arrives at the next station must be later than the departure time from the current station plus the travel time. k The time-based logical relationship for accessing logistics sites is expressed as follows: For sufficiently large positive numbers (4) Vehicles k The load capacity after visiting the logistics station will be determined based on the loading / unloading weight. d j The load may increase or decrease, but not exceed the vehicle's rated load capacity. When the vehicle arrives at a station for delivery, the load decreases; when picking up goods, the load increases. The model needs to track this dynamic change in real time to ensure that the vehicle's load capacity never exceeds the rated load capacity on any road segment. Otherwise, it will be considered overloaded and an infeasible solution. The expression is:

[0082] (5) The start time of service at each logistics station is within a specified time window, that is, the time when vehicles arrive at each logistics station is within the time window of that station. This can be expressed as:

[0083] (6) x ijk = 0 represents a logistics station. i and j There is no transportation route between them. x ijk =1 indicates a logistics station i and j There exists a transportation route, meaning that vehicles only travel directly between two logistics stations with a shared route. This can be represented as:

[0084] Step 130: Use tent mapping to generate the initial population of the starbird optimization algorithm. Each starbird corresponds to a path planning solution, and the path planning solution contains the service station sequence of each transport vehicle.

[0085] Specifically, the Starfinch Optimization Algorithm is a metaheuristic algorithm that simulates the foraging and storage behavior of starfinches. Its solution performance largely depends on the quality of the initial population. The quality of the initial population directly determines the algorithm's convergence speed and global search capability. Traditional random initialization methods tend to lead to uneven distribution of the population in the solution space, making it prone to clustering in certain local regions, thus increasing the risk of the algorithm getting trapped in local optima. To overcome this drawback, this embodiment uses tent mapping to generate the initial population.

[0086] In some implementations, the initial population for generating the starbird optimization algorithm using tent mapping specifically includes: Step 131: Determine the population size and problem dimension based on the number of logistics stations; Step 132: Randomly select initial values ​​and iteratively generate a chaotic sequence matrix using the tent mapping formula. The length of the chaotic sequence matrix is ​​the product of the population size and the problem dimension. Step 133: Linearly map each element value in the chaotic sequence matrix to the solution space range to obtain the position coordinates of each starbird individual in the initial population in each dimension.

[0087] Specifically, if there are many logistics stations, the population size can be reduced; if there are few logistics stations, the population size can be increased. During the initialization phase, a population is formed by randomly generating several initial individuals within the solution space. The population size is set to [value missing]. N The problem dimension is D The search scope for each dimension is [ L j , U j ]( j =1, 2, ..., D ),in L j The lower realm U j The upper boundary; T max This represents the maximum number of iterations.

[0088] Then, the initial population is randomly generated using the tent mapping. The position of each starbird in the population represents a vehicle routing solution, and each solution contains a sequence of service stations for a vehicle.

[0089] The tent mapping is a piecewise linear mathematical function that maintains population diversity through mapping, making it easier for the algorithm to escape local optima while improving global search capabilities. The steps for generating an initial starling population using the tent mapping are as follows: (1) Randomly select initial values x 0∈(0, 1).

[0090] (2) Substituting into the following tent mapping formula, the generated length is... N × D The chaotic sequence matrix.

[0091]

[0092] (3) The chaotic sequence value x ij Linear mapping ∈[0, 1] to the solution space[ L j , U j The formula is: X i, j = L j + x i, j × ( U j L j ) in, X i, j Indicates the first i The first individual j Dimensional position, x i, j The element values ​​of the chaotic sequence matrix generated for the tent mapping.

[0093] The positions of all individuals are obtained through the above mapping, forming the initial population. This uniformly distributed initial population provides a good starting point for subsequent iterative optimization, effectively reducing the probability of the algorithm getting trapped in local optima, thereby significantly improving the global search capability and convergence accuracy of the sparrow optimization algorithm when solving complex path planning problems.

[0094] Based on the above embodiments, in the process of iteratively solving the problem using the improved star-sparrow optimization algorithm, since the solutions generated in the initial stage of the algorithm and during the iteration process may violate the above constraints, this embodiment introduces a hierarchical penalty mechanism to handle infeasible solutions, thereby guiding the algorithm to converge toward the feasible solution space.

[0095] Specifically, in the iterative solution process, the evaluation value of the path planning solution for each individual starbird is first calculated based on a multi-objective function. The formula for calculating the evaluation value is:

[0096] When calculating the evaluation value, constraints must also be considered. If the constraints are violated, the fitness value needs to be reduced; that is, the evaluation value is penalized and adjusted based on the constraints. The specific rules for penalty adjustment are as follows: When the unique access constraint of a logistics station or the origin-destination constraint of a vehicle is violated, the evaluation value is reduced by a first preset percentage. These constraints are fundamental logical constraints of the path structure; violating them indicates a logical error in the path solution, such as duplicate visits or path breaks, thus incurring a moderate penalty. The first preset percentage can be set to 10%, meaning the fitness value is reduced by 10%. It should be understood that the specific value of the first preset percentage can be adjusted based on the algorithm's convergence, for example, set to any value between 5% and 15%.

[0097] When time window constraints or path existence constraints are violated, the evaluation value is reduced by a second preset percentage. These constraints are business rule constraints; although violating them does not affect the physical connectivity of the path, it does affect service quality or actual executability, thus incurring a heavier penalty. In this embodiment, the second preset percentage can be set to 15%, meaning the fitness value is reduced by 15%. Because the second preset percentage is greater than the first preset percentage, the algorithm will prioritize solutions that satisfy the time window and path existence constraints during the search process, thereby reflecting the emphasis on service timeliness.

[0098] When time logic constraints or load constraints are violated, the evaluation value is set to a preset minimum value. These constraints involve driving safety and physical limits, such as overloading or time reversal, and are considered severely infeasible solutions. By setting the evaluation value to a preset minimum value (e.g., 0 or a very small value), the survival probability of such individuals can be greatly reduced during population evolution, causing them to be quickly eliminated, thereby ensuring the safety of the algorithm's search direction.

[0099] This tiered penalty mechanism, by distinguishing the impact of different degrees of constraint violation on solution quality and applying penalties of varying degrees, can effectively balance the algorithm's global search capability and local exploitation capability, guiding the population to quickly converge to the optimal feasible solution region that satisfies both physical constraints and business rules.

[0100] Step 140: An improved star-sparrow optimization algorithm with a polynomial mutation perturbation strategy is adopted to iteratively solve the multi-objective function under constraints and output the optimized path planning solution.

[0101] Specifically, this embodiment introduces a multinomial mutation perturbation strategy based on the standard sparrow optimization algorithm. During algorithm iteration, sparrow individuals update their positions using foraging and storage strategies, and cache search and recovery strategies. To further enhance the algorithm's ability to escape local optima, a multinomial mutation perturbation is introduced after the position update. This perturbation strategy is based on a multinomial distribution function; by adjusting the distribution exponent, the perturbation amplitude can be controlled, enabling both fine-grained local searches and large-scale exploration. In each iteration, the algorithm calculates the fitness value of each individual according to a multi-objective function and applies penalties based on constraints, selecting individuals with better fitness to retain in the next generation. After multiple iterations, the population gradually converges, ultimately outputting the position of the sparrow individual with the best fitness. Decoding this position yields the optimal vehicle path planning scheme. This path planning scheme clearly defines the station sequence and driving route for each vehicle, achieving a comprehensive optimization of total driving distance and carbon emissions while satisfying constraints.

[0102] In some implementations... Figure 2 This is a flowchart illustrating the search path planning solution provided in an embodiment of this application, as shown below. Figure 2 As shown, the iterative solution process specifically includes: Step 141: In each iteration, generate two random numbers. Based on the comparison of the two random numbers, select either the foraging and storage strategy or the cache search and recovery strategy to update the starbird position.

[0103] This embodiment generates two random numbers (for example) s 1 and s 2) To determine which strategy to use for location updates. If s 1 less than s 2. Then, a foraging and storage strategy is implemented, which focuses on developing or utilizing the population's historical best information within the neighborhood of the current solution; if s 1 is greater than or equal to s If the first strategy is to execute a cache search and retrieval strategy, this strategy focuses on using reference points for a wider range of exploration. This mechanism of alternating execution of two strategies allows the algorithm to maintain a dynamic balance between local development and global exploration, avoiding search blind spots caused by a single strategy.

[0104] Step 142, executing the foraging and storage strategy includes: generating a first random number and the probability threshold corresponding to the current iteration; if the first random number is greater than the probability threshold, then the foraging strategy is used to update the position of the starling, and a position perturbation based on a multinomial distribution is introduced after the update; otherwise, the storage strategy is used to update the position of the starling.

[0105] Specifically, Figure 3 This is a flowchart illustrating the foraging and storage strategy provided in an embodiment of this application, such as... Figure 3 As shown, when the algorithm chooses to execute the foraging and storage strategy, it will further generate a first random number (e.g., f ) and a probability threshold (e.g. P a1 Probability threshold P a1 The value is typically designed to decrease linearly with the number of iterations, for example, from 1 to 0. This design causes the algorithm to tend to execute a foraging strategy in the early stages of iteration, performing extensive searches, while in the later stages of iteration it tends to execute a storage strategy, performing fine-grained exploration.

[0106] If the first random number f Greater than the current probability threshold P a1 If the position of the starling is not found in the foraging strategy, then the foraging strategy is executed. In this case, the position update formula for the individual starling combines the average position of the population, the position of a random individual, and the random step size generated by Levy's flight. This design effectively utilizes the overall information of the population and introduces random jumps, enhancing the algorithm's ability to escape local optima. The formula used to update the starling's position using the foraging strategy is as follows:

[0107] in,

[0108] In the above formula, t This represents the current iteration number. It is the first i The star sparrow in the first t The position updated in the next iteration; c It is based on random numbers generated by Levy's flight. A , B , C These are three different starbird individuals randomly selected from the population. X m All the starbirds are in the first t Average position of the next iteration t 1. t 2. t 3. r、r 1 、r 2 、r All 3 are random numbers in the range [0,1]. t 4 is a random number generated from a normal distribution. t 5 is a random number generated by Levy's flight. T max The maximum number of iterations, d = 0.05.

[0109] In some implementations, after updating the starling positions using the foraging strategy, a position perturbation based on a multinomial distribution is introduced. Applying a certain perturbation during the update process of individual search positions can increase population diversity, expand the search range, improve the algorithm's search capability, and thus make it easier to escape local optima.

[0110] The perturbation is introduced using a polynomial mutation operation, which specifically includes: generating a random number for each dimension of the position vector of the star sparrow to be mutated; based on the relationship between the random number and a preset threshold, using a preset distribution index, calculating the perturbation value for each dimension through an asymmetric polynomial distribution function, where the distribution index is used to control the magnitude of the perturbation; superimposing the perturbation value onto the original position coordinates of that dimension to obtain the mutated position, and controlling the mutated position to be within the upper and lower bounds of the solution space, thus completing one mutation operation.

[0111] Specifically, traditional mutation operations often use uniform or Gaussian distributions, making it difficult to flexibly control the perturbation amplitude. This embodiment uses an asymmetric multinomial distribution function to calculate the perturbation value. The formula is as follows:

[0112]

[0113] In the above formula, d j It is a polynomial mutation operator; m ∈[0, 1] is a random number. or It is the distribution index. U j , L j These are the upper and lower limits of the range of values ​​a variable can take.

[0114] For the position vector of the starbird j First, generate a random number. m ,like m If it is less than 0.5, a multinomial distribution calculation formula is used; if m If the value is greater than or equal to 0.5, a different multinomial distribution calculation formula is used. This asymmetric design allows the perturbation value to be positive or negative, and the distribution pattern is controllable.

[0115] Distribution index or It is a key parameter for controlling the degree of disturbance. When the distribution index... or When the value is large, the generated perturbation value is small, and the mutation operation tends to make small adjustments near the original position, which is beneficial for fine-grained local search; when the distribution index is large... or When the value is small, the generated perturbation value is large, and the mutation operation tends to produce larger jumps, which is beneficial for the algorithm to escape local optima. This can be achieved by adjusting the distribution index. or The value of the distribution index in this embodiment allows for a dynamic balance between local development and global exploration based on the algorithm's iteration state or the quality of the solution. For example, in the early stages of iteration, a smaller distribution index can be set to encourage broad exploration; in the later stages of iteration, a larger distribution index can be set for finer development.

[0116] By introducing an improved strategy of polynomial mutation perturbation, the problem of premature convergence of the standard sparrow optimization algorithm in dealing with complex multi-objective path planning problems is effectively solved, and the success rate of the algorithm in finding the global optimum is improved.

[0117] If the first random number f Less than or equal to the probability threshold P a1 If the optimal solution is found, a storage strategy is executed. This strategy primarily utilizes the position information of the best individual in the current population to guide position updates, helping the algorithm to perform a fine-grained search near the optimal solution and improving solution accuracy. This can be expressed by the following formula:

[0118] in, A , B These are two different starbird individuals randomly selected from the population. It is the optimal position for the star-shaped sparrow individual in the current population. l It is based on random numbers generated by Levy's flight. l It is a coefficient that decreases linearly from 1 to 0.

[0119] Step 143, executing the cache search and recovery strategy includes: generating two different location reference points, generating a second random number and a preset probability threshold; if the second random number is greater than the preset probability threshold, then the cache recovery strategy is used to update the starbird position based on the location reference points; otherwise, the cache search strategy is used to update the starbird position based on the location reference points.

[0120] Specifically, Figure 4 This is a flowchart illustrating the cache search and recovery strategy provided in an embodiment of this application, as shown below. Figure 4 As shown, when the algorithm chooses to execute the cache search and retrieval strategy, it first needs to generate two different location reference points. These two reference points are typically calculated based on the current individual position, a random individual position, and a random angle, and are used to simulate the food cache location in the star-bird's memory. If two different reference points are determined... , Help Starbird find the hidden stash.

[0121] The formula for generating the reference point is as follows:

[0122]

[0123] in, ,

[0124] In this formula, It is [0, π The random radians between [A and B], where A and B are two different starbird individuals randomly selected from the population. r 1. r 2. r 3 is a random number within the range [0, 1]. R This is a randomly selected reference point.

[0125] Then, generate a second random number (e.g.) f ) and a preset probability threshold (e.g. P a2 ).like f A random number within [0,1] P a2 = 0.2.

[0126] If the second random number f Greater than the preset probability threshold P a2 If the cache recovery strategy fails, the individual's position is updated by combining the reference point with the current optimal position, simulating the process of the starling retrieving hidden food. The formula used to update the starling's position using the cache recovery strategy is as follows:

[0127] in, t 1~ t All 8 are random numbers between [0, 1].

[0128] If the second random number f Less than or equal to a preset probability threshold P a2 If the current position is not found in the cached search area, a cached search strategy is executed. This strategy determines whether to move by comparing the fitness value of the current individual's position with that of the reference point, simulating the process of the sparrow searching for food in the cached area. This strategy further enriches the algorithm's search patterns and improves its optimization efficiency in complex solution spaces. The formula used to update the sparrow's position using the cached search strategy is as follows:

[0129] In this formula, f ( ) is the formula for calculating the evaluation value.

[0130] Then store the updated starbird positions. Calculate the fitness value for each starbird position and update the starbird position using the following formula:

[0131] Determine if the maximum number of iterations has been reached or if the objective function has converged. If so, output the optimal star position. The corresponding evaluation value (fitness calculation value) is used as the optimal path planning solution; otherwise, the next iteration is performed.

[0132] The green vehicle routing system for logistics stations provided in this application is described below. The green vehicle routing system described below corresponds to the green vehicle routing method described above. The system includes: The basic data acquisition unit is used to acquire basic data for route planning. The basic data includes logistics information of each logistics station and carbon emission parameters of transport vehicles. The objective function construction unit is used to construct a quantitative model for dynamically calculating carbon emissions based on vehicle travel time, speed and load, based on the basic data, set constraints for route planning based on the logistics information, and establish a multi-objective function including a travel distance target and a carbon emission target obtained based on the quantitative model. An initialization unit is used to generate an initial population for the starbird optimization algorithm using tent mapping. Each starbird corresponds to a path planning solution, and the path planning solution contains a service station sequence for each transport vehicle. The iterative solution unit is used to perform iterative solution of the multi-objective function under the constraints by adopting the improved sparrow optimization algorithm with the introduction of a polynomial mutation perturbation strategy, and output the optimized path planning solution.

[0133] In some implementations... Figure 5 This is a schematic diagram of the structure of a green path planning system provided in an embodiment of this application, as shown below. Figure 5 As shown, the system includes the following modules: Logistics Information Management Module: Inputs and stores the distance matrix, two-way cargo volume, time constraints, and carbon emission parameters of transport vehicles that need to reach logistics stations.

[0134] Carbon emission calculation module: Contains a carbon emission calculation model. Input a planned route, and calculate the carbon emissions of taking that route based on vehicle performance and travel distance.

[0135] The Star-Sparrow Algorithm Execution Module contains an improved Star-Sparrow algorithm model and fitness calculation functions. It executes the Star-Sparrow algorithm with the goal of minimizing distance and carbon emissions to obtain the optimal path planning solution.

[0136] Route Output Module: Displays the planned route as a curve on the map, and shows the total carbon emissions and distance of using the route.

[0137] User revision module: Users can modify the planned routes on the map, recalculate and display the total carbon emissions and distance.

[0138] It is understood that the detailed functional implementation of each of the above units can be found in the description of the aforementioned method embodiments, and will not be repeated here.

[0139] It should be understood that the above system is used to execute the methods in the above embodiments. The corresponding program modules in the system are similar in implementation principle and technical effect to those described in the above methods. The working process of the system can be referred to the corresponding process in the above methods, and will not be repeated here.

[0140] Based on the methods in the above embodiments, this application provides an electronic device that may include a processor, a communications interface, a memory, and a communication bus, wherein the processor, communications interface, and memory communicate with each other via the communication bus. The processor may invoke logical instructions stored in the memory to execute the methods in the above embodiments.

[0141] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

[0142] Based on the methods in the above embodiments, this application provides a computer-readable storage medium storing a computer program that, when run on a processor, causes the processor to execute the methods in the above embodiments.

[0143] Based on the methods in the above embodiments, this application provides a computer program product that, when run on a processor, causes the processor to execute the methods in the above embodiments.

[0144] It is understood that the processor in the embodiments of this application can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. A general-purpose processor can be a microprocessor or any conventional processor.

[0145] The method steps in this application embodiment can be implemented in hardware or by a processor executing software instructions. The software instructions can consist of corresponding software modules, which can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, portable hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can reside in an ASIC.

[0146] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0147] It is understood that the various numerical designations used in the embodiments of this application are merely for the convenience of description and are not intended to limit the scope of the embodiments of this application.

[0148] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A green vehicle route planning method for logistics stations, characterized in that, include: Obtain basic data for route planning, including logistics information of each logistics station and carbon emission parameters of transport vehicles; Based on the aforementioned basic data, a quantitative model for dynamically calculating carbon emissions based on vehicle travel time, speed, and load is constructed. Based on the aforementioned logistics information, constraints for route planning are set, and a multi-objective function is established, including a travel distance target and a carbon emission target obtained based on the quantitative model. The initial population of the starbird optimization algorithm is generated using tent mapping. Each starbird corresponds to a path planning solution, which includes the service station sequence of each transport vehicle. An improved star-sparrow optimization algorithm with a polynomial mutation perturbation strategy is adopted to iteratively solve the multi-objective function under the constraints and output the optimized path planning solution.

2. The path planning method according to claim 1, characterized in that, The construction of a quantitative model for calculating carbon emissions based on the dynamic calculation of vehicle travel time, speed, and load includes: Based on the vehicle’s current driving time, speed and load, calculate the engine’s fuel consumption, speed-related fuel consumption and load-related fuel consumption respectively. The total fuel consumption of the vehicle is the sum of the engine operating fuel consumption, speed-related fuel consumption, and load-related fuel consumption. The total fuel consumption is multiplied by the carbon emissions per unit of fuel consumption to construct the quantitative model for dynamically calculating carbon emissions.

3. The path planning method according to claim 2, characterized in that, The establishment of the multi-objective function, which includes a driving distance target and a carbon emission target obtained based on the quantification model, includes: Based on the driving distance between logistics stations, the total distance traveled by each transport vehicle along the planned route is calculated and represented as the driving distance target; Based on the aforementioned quantification model, the total carbon emissions generated by each transport vehicle are calculated and represented as the carbon emission target. Based on the weights corresponding to the driving distance target and the carbon emission target, the two are weighted and summed to form a multi-objective function with the objective of minimizing the weighted sum.

4. The path planning method according to claim 3, characterized in that, The expression for the multi-objective function is: Among them, the function F ijk ( ) represents a vehicle k After arc ( i , j Arrival at the logistics station j Total fuel consumption t ijk For vehicles k After arc ( i , j Arrival at the logistics station j travel time v ijk For vehicles k After arc ( i , j Arrival at the logistics station j driving speed, w ijk For vehicles k After arc ( i , j Arrival at the logistics station j Vehicle load capacity, h Carbon emissions per unit of fuel consumed. For logistics stations i to logistics station j driving distance, A value of 1 indicates a logistics station. i and j If there is a transport route, then the vehicle k After arc ( i , j ), A value of 0 indicates a logistics station. i and j There is no transportation route between them. m 1 represents the weight of the target driving distance. m 2 represents the weight of the carbon emission target.

5. The path planning method according to claim 1, characterized in that, The constraints include at least one of the following: Each logistics station is visited only once by a single vehicle. Each vehicle must depart from the logistics center, visit several logistics stations, and then return to the logistics center. The arrival and departure times of vehicles at each logistics station conform to a time logic relationship; The load of a vehicle after visiting a logistics station changes dynamically according to the two-way flow of goods, and the load on any road segment does not exceed its rated load capacity. The arrival time of vehicles at each logistics station is within the time window of that station. Vehicles only travel directly between two logistics stations with a shared route.

6. The path planning method according to claim 5, characterized in that, The method further includes: During the iterative solution process, the evaluation value of the path planning solution corresponding to each individual starbird is calculated based on the multi-objective function, and the evaluation value is adjusted with penalties based on the constraints: When the unique access constraint of the logistics station or the origin and destination constraint of the vehicle are violated, the evaluation value is reduced by the first preset percentage. When the time window constraint or path existence constraint is violated, the evaluation value is reduced by the second preset percentage; When time logic constraints or load constraints are violated, the evaluation value is set to the preset minimum value.

7. The path planning method according to any one of claims 1 to 6, characterized in that, The initial population generated using the tent mapping optimization algorithm includes: The population size and problem dimensions are determined based on the number of logistics stations; Randomly select initial values ​​and iteratively generate a chaotic sequence matrix using the tent mapping formula. The length of the chaotic sequence matrix is ​​the product of the population size and the problem dimension. By linearly mapping each element value in the chaotic sequence matrix to the solution space range, the position coordinates of each starbird individual in the initial population in each dimension are obtained.

8. The path planning method according to any one of claims 1 to 6, characterized in that, The improved sparrow optimization algorithm, which employs a polynomial mutation perturbation strategy, iteratively solves the multi-objective function under the constraints, including: In each iteration, two random numbers are generated. Based on the comparison result of the two random numbers, either the foraging and storage strategy or the cache search and recovery strategy is selected to update the starbird position. The execution of the foraging and storage strategy includes: generating a first random number and a probability threshold corresponding to the current iteration; if the first random number is greater than the probability threshold, then the foraging strategy is used to update the position of the starling, and a position perturbation based on a multinomial distribution is introduced after the update; otherwise, the storage strategy is used to update the position of the starling. The cache search and recovery strategy includes: generating two different location reference points, generating a second random number and a preset probability threshold; if the second random number is greater than the preset probability threshold, then the cache recovery strategy is used to update the starbird position based on the location reference points; otherwise, the cache search strategy is used to update the starbird position based on the location reference points.

9. The path planning method according to claim 8, characterized in that, The introduction of position perturbations based on multinomial distribution includes: Generate a random number for each dimension of the position vector of the star sparrow to be mutated; Based on the relationship between the random number and the preset threshold, the perturbation value of each dimension is calculated using a preset distribution index and an asymmetric multinomial distribution function. The distribution index is used to control the magnitude of the perturbation. The perturbation value is superimposed onto the original position coordinates of that dimension to obtain the mutated position, and the mutated position is controlled within the upper and lower bounds of the solution space to complete one mutation operation.

10. A green vehicle routing system for a logistics station, characterized in that, include: The basic data acquisition unit is used to acquire basic data for route planning. The basic data includes logistics information of each logistics station and carbon emission parameters of transport vehicles. The objective function construction unit is used to construct a quantitative model for dynamically calculating carbon emissions based on vehicle travel time, speed and load, based on the basic data, set constraints for route planning based on the logistics information, and establish a multi-objective function including a travel distance target and a carbon emission target obtained based on the quantitative model. An initialization unit is used to generate an initial population for the starbird optimization algorithm using tent mapping. Each starbird corresponds to a path planning solution, and the path planning solution contains a service station sequence for each transport vehicle. The iterative solution unit is used to perform iterative solution of the multi-objective function under the constraints by adopting the improved sparrow optimization algorithm with the introduction of a polynomial mutation perturbation strategy, and output the optimized path planning solution.

Images