Method and device for planning a route for distribution of finished oil products

Abstract

The application discloses a kind of finished oil distribution path planning method and device, wherein the method includes: with the total transportation cost minimization as objective function, with customer satisfaction, carbon emissions, transportation risk and mileage utilization as additional objective function, establish path planning model;Using density-based hierarchical clustering algorithm, based on the path planning model, according to the vehicle density, the tank truck of different gas stations is combined, to determine the vehicle load path of finished oil between gas stations;Using genetic algorithm, based on path planning model, establish the vehicle coding corresponding to the combined vehicle;According to the standard deviation of the calculated different combined vehicle mileage saving, generate initial population and carry out cross and mutation operation, to determine the combined vehicle load path of finished oil between different oil depots;Generate the path planning scheme of finished oil distribution and output.The application is used to improve the efficiency and accuracy of finished oil distribution path planning.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a method and apparatus for route planning in the delivery of refined oil products. Background Technology

[0002] This section is intended to provide background or context for the embodiments of the invention described herein. The description herein is not an admission that it is prior art simply because it is included in this section.

[0003] Refined oil logistics encompasses multiple functions including transportation, warehousing, loading and unloading, distribution, and information processing. It runs through the petrochemical industry supply chain, connecting refineries, distribution depots, and sales outlets, primarily gas stations. Refined oil distribution is divided into primary and secondary distribution. Primary distribution involves transporting large quantities of refined oil from refineries to sales company depots for storage. Secondary distribution involves delivering small quantities of refined oil from distribution depots to gas stations or end users. As the final stage of the supply chain, secondary distribution's management level significantly impacts the profitability, core competitiveness, and market share of refined oil sales companies. In recent years, the concept of proactive distribution has been promoted, leading sales companies to shift from traditional order-based systems to unified distribution management based on sales forecasts. This makes achieving unified optimization of large-scale distribution plans, unified scheduling of transportation resources, and reducing transportation costs and improving distribution efficiency over large regions the core issues of secondary distribution management.

[0004] Existing technologies for planning secondary distribution routes for refined oil products often focus on optimizing single factors, such as transportation costs, without comprehensively considering multiple objectives such as customer satisfaction, carbon emissions, transportation risks, and mileage utilization, making it difficult to maximize overall benefits. In vehicle allocation decisions, they fail to fully integrate multiple constraints such as distance between gas stations, maximum loading distance, tanker truck carrying capacity, and delivery time windows for accurate allocation. Regarding customer segmentation, they do not adequately consider the distribution characteristics of gas stations relative to oil depots, failing to efficiently allocate gas stations to suitable oil depots, thus impacting delivery efficiency and costs. Traditional algorithms are poorly adapted to large-scale data and complex constraints when dealing with secondary distribution route planning problems for refined oil products, exhibiting slow computation speeds and a tendency to get trapped in local optima, resulting in low optimization efficiency and quality. Summary of the Invention

[0005] This invention provides a route planning method for refined oil product distribution, which improves the efficiency and accuracy of route planning for refined oil product distribution, reduces the transportation costs of secondary distribution of refined oil products, improves distribution efficiency, and enhances overall operational benefits. The method includes:

[0006] A route planning model is established with the goal of minimizing total transportation cost, and additional objective functions of customer satisfaction, carbon emissions, transportation risk, and mileage utilization rate. The route planning model includes parameters related to oil depots, gas stations, tank trucks, and tank trucks.

[0007] A density-based hierarchical clustering algorithm is used to calculate the vehicle density of tanker trucks allocated to different gas stations based on the path planning model. Based on pre-set vehicle merging constraints, tanker trucks from different gas stations are merged according to the vehicle density to obtain merged vehicles. The vehicle merging constraints include that the distance between merged tanker trucks is less than or equal to a pre-set maximum loading distance between tanker trucks. These constraints constrain the tanker trucks' carrying capacity, delivery order, delivery time window, and distance between them. Based on the merged vehicles, the vehicle loading path for refined oil products between gas stations is determined.

[0008] A genetic algorithm, based on a path planning model, is used to establish vehicle codes corresponding to the merged vehicles. The locus number of the vehicle code represents the number of the merged vehicle. The value of the vehicle code represents the assigned oil depot number. An initial population is generated based on the calculated standard deviation of the mileage saved by different merged vehicles. The standard deviation of the mileage saved represents the standard deviation of the mileage saved by assigning different merged vehicles to the target oil depot relative to assigning them to other oil depots. Each individual in the initial population represents a different vehicle code assigned to the initial oil depot. By performing crossover and mutation operations on the initial population, the merging vehicle loading path for refined oil between different oil depots is determined.

[0009] The route planning scheme for refined oil distribution is output, which includes the vehicle loading routes between gas stations and the combined vehicle loading routes between different oil depots.

[0010] This invention also provides a route planning device for refined oil product distribution, which improves the efficiency and accuracy of route planning for refined oil product distribution, reduces the transportation costs of secondary distribution of refined oil products, improves distribution efficiency, and enhances overall operational benefits. The device includes:

[0011] The route planning modeling module is used to establish a route planning model with the goal of minimizing total transportation cost and additional objective functions of customer satisfaction, carbon emissions, transportation risk, and mileage utilization rate. The route planning model includes parameters related to oil depots, gas stations, tank trucks, and tank trucks.

[0012] The vehicle loading route determination module is used to calculate the vehicle density of tanker trucks allocated to different gas stations based on the path planning model using a density-based hierarchical clustering algorithm; based on pre-set vehicle merging constraints, it merges tanker trucks from different gas stations according to the vehicle density to obtain merged vehicles; the vehicle merging constraints include that the distance between the merged tanker trucks is less than or equal to a pre-set maximum loading distance between tanker trucks; the vehicle merging constraints are used to constrain the tanker truck carrying capacity, delivery order, delivery time window, and distance between tanker trucks; and based on the merged vehicles, it determines the vehicle loading route for refined oil between gas stations.

[0013] The merged vehicle loading route determination module uses a genetic algorithm based on a path planning model to establish vehicle codes corresponding to the merged vehicles. The locus number of the vehicle code represents the number of the merged vehicle. The value of the vehicle code represents the assigned oil depot number. An initial population is generated based on the calculated standard deviation of the mileage saved by different merged vehicles. The standard deviation of the mileage saved represents the standard deviation of the mileage saved by assigning different merged vehicles to the target oil depot relative to assigning them to other oil depots. Each individual in the initial population represents a different vehicle code assigned to the initial oil depot. By performing crossover and mutation operations on the initial population, the merged vehicle loading route for refined oil between different oil depots is determined.

[0014] The route planning output module is used to output the vehicle loading routes of refined oil between gas stations and the combined vehicle loading routes of refined oil between different oil depots as route planning schemes for refined oil distribution.

[0015] This invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-mentioned route planning method for refined oil delivery.

[0016] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described route planning method for refined oil product distribution.

[0017] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the aforementioned route planning method for refined oil product distribution.

[0018] In this embodiment of the invention, a route planning model is established with minimizing total transportation cost as the objective function and customer satisfaction, carbon emissions, transportation risk, and mileage utilization rate as additional objective functions. The route planning model includes parameters related to oil depots, gas stations, tank trucks, and the tank trucks themselves. A density-based hierarchical clustering algorithm is used to calculate the vehicle density of tank trucks allocated to different gas stations based on the route planning model. Based on pre-set vehicle merging constraints, tank trucks from different gas stations are merged according to the vehicle density to obtain merged vehicles. The vehicle merging constraints include that the distance between merged tank trucks is less than or equal to a pre-set maximum loading distance between tank trucks. These constraints constrain the tank truck carrying capacity, delivery order, delivery time window, and distance between tank trucks involved in the merging process. Based on the merged vehicles... The process involves determining the vehicle loading routes for refined oil between gas stations; using a genetic algorithm based on a path planning model to establish vehicle codes corresponding to the merged vehicles; the locus number of the vehicle code represents the number of the merged vehicle; the value of the vehicle code represents the assigned oil depot number; generating an initial population based on the calculated standard deviation of the mileage saved by different merged vehicles; the standard deviation of the mileage saved represents the standard deviation of the mileage saved by assigning different merged vehicles to the target oil depot relative to assigning them to other oil depots; each individual in the initial population represents a different vehicle code assigned to the initial oil depot; determining the vehicle loading routes for refined oil between different oil depots through crossover and mutation operations on the initial population; and outputting the vehicle loading routes for refined oil between gas stations and the merged vehicle loading routes for refined oil between different oil depots as a path planning scheme for refined oil delivery.The route planning model constructed in this invention uses minimizing total transportation cost as the objective function, and incorporates customer satisfaction, carbon emissions, transportation risk, and mileage utilization as additional objective functions. It is no longer limited to the traditional single optimization that only considers transportation cost, but can maximize the overall benefits of secondary oil delivery in terms of economy, service, environment, and operational efficiency, overcoming the lack of comprehensive optimization in existing technologies. The density-based hierarchical clustering algorithm is used for vehicle merging, taking into account that the distance between tanker trucks is less than or equal to the farthest loading distance, as well as constraints on carrying capacity, delivery order, and delivery time windows. This helps to accurately combine vehicles according to the actual distribution and demand of gas stations, ensuring that the merged vehicles meet carrying capacity requirements and complete delivery tasks within the specified delivery time window under a reasonable delivery order. This improves the rationality and scientific nature of vehicle allocation, representing a significant improvement compared to existing vehicle allocation methods that do not fully incorporate multiple constraints. The density-based hierarchical clustering algorithm filters out deviation points by calculating vehicle density. The algorithm prioritizes merging vehicles with higher density and shorter distances, using the furthest loading distance as the termination condition. This avoids indiscriminate merging of all stations, effectively reducing unnecessary computational steps, lowering the computational load, and improving efficiency when handling large-scale data. It overcomes the slow computation speed of traditional algorithms when dealing with large-scale delivery planning. Furthermore, the algorithm incorporates standard deviation mileage saving and designs corresponding genetic operators. This standard deviation mileage saving guides the algorithm to find better customer segmentation schemes, avoiding getting trapped in local optima, thus improving the optimization quality and enhancing its adaptability to complex constraints and large-scale delivery route planning problems. It also addresses the weakness of traditional algorithms in getting trapped in local optima when dealing with such problems. Finally, the route planning for refined oil delivery is decomposed into two sub-problems: merging travel distance and vehicle segmentation. These are solved sequentially using a system clustering algorithm and an improved genetic algorithm, achieving optimized planning of large-scale refined oil secondary delivery routes. Under various constraints, this reduces transportation costs, improves delivery efficiency, and enhances overall operational benefits. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention 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 the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0020] Figure 1 This is a flowchart illustrating a route planning method for refined oil product distribution according to an embodiment of the present invention.

[0021] Figure 2This is a specific example diagram of a route planning method for refined oil delivery in an embodiment of the present invention;

[0022] Figure 3 This is a specific example diagram of a route planning method for refined oil delivery in an embodiment of the present invention;

[0023] Figure 4 This is a specific example diagram of a route planning method for refined oil delivery in an embodiment of the present invention;

[0024] Figure 5 This is a specific example diagram of a refined oil delivery process in an embodiment of the present invention;

[0025] Figure 6 This is a schematic diagram of the structure of a route planning device for finished oil delivery according to an embodiment of the present invention;

[0026] Figure 7 This is a specific example diagram of a route planning device for refined oil delivery in an embodiment of the present invention;

[0027] Figure 8 This is a schematic diagram of a computer device used for route planning in the delivery of refined oil products, as described in an embodiment of the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0029] In this document, the term "and / or" merely describes a relationship, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" in this document means any combination of at least two of any one or more elements. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0030] In the description of this specification, the terms "comprising," "including," "having," and "containing" are open-ended terms, meaning that they include but are not limited to. The terms "an embodiment," "a specific embodiment," "some embodiments," and "for example," etc., refer to specific features, structures, or characteristics described in connection with that embodiment or example that are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, or characteristics described can be combined in any suitable manner in one or more embodiments or examples. The order of steps involved in the various embodiments is used to illustrate the implementation of this application, and the order of steps is not limited and can be adjusted appropriately as needed.

[0031] The acquisition, storage, use, and processing of data in this application comply with relevant laws and regulations. The information collected in this application is authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of this data all comply with relevant laws, regulations, and standards, and necessary confidentiality measures have been taken. This does not violate public order and good morals, and corresponding operation entry points are provided for users to choose to authorize or refuse. In addition, this application provides users with corresponding operation entry points to choose to agree to or refuse automated decision-making results. If the user chooses to refuse, the process proceeds to the expert decision-making stage.

[0032] It should be noted that in the embodiments of this application, certain existing solutions in the industry, such as software, components, and models, may be mentioned. For example, some existing software tools, components, algorithm models, or solutions well-known in other technical fields may be cited. These should be considered exemplary, and their purpose is only to illustrate the feasibility of implementing the technical solution of this application. These mentions should be understood as typical examples, and their core purpose is to illustrate and verify the rationality and feasibility of implementing the technical solution proposed in this application. However, this does not mean that the applicant has already used or necessarily used the solution. Such citations do not imply that the applicant has actually adopted these existing solutions, or that it will necessarily adopt these methods in its technical implementation process in the future. In other words, these mentions are only illustrative in nature, helping to understand the connection and transcendence of the innovation points of this application with the prior art, and do not constitute an endorsement or reliance statement on a specific prior art product.

[0033] Refined oil logistics integrates transportation, warehousing, loading and unloading, distribution, and information processing, spanning the entire petrochemical industry supply chain and connecting refineries, distribution depots, and sales outlets, primarily gas stations. Refined oil distribution can be divided into two stages: primary distribution and secondary distribution. Primary distribution refers to the process of transporting refined oil from refineries in large batches to sales company depots for storage. Secondary distribution refers to the process of transporting refined oil from distribution depots in small batches to gas stations or end users. As the final stage of the industry chain, secondary distribution's management level significantly impacts the profitability, core competitiveness, and market share of refined oil sales companies. In recent years, with the gradual promotion of proactive distribution concepts, sales companies have shifted from the traditional order-based model to unified distribution management based on sales forecasts. Therefore, how to achieve unified optimization of large-scale distribution plans, unified scheduling of transportation resources, and comprehensive reduction of transportation costs and improvement of distribution efficiency over large regions has gradually become the core issue of secondary distribution management.

[0034] To address the aforementioned problems, embodiments of the present invention provide a route planning method for refined oil product distribution, which improves the efficiency and accuracy of route planning for refined oil product distribution, reduces transportation costs for secondary distribution of refined oil products, increases distribution efficiency, and enhances overall operational benefits. (See also...) Figure 1 , Figure 1 This is a flowchart illustrating a route planning method for refined oil product distribution according to an embodiment of the present invention. The method may include:

[0035] Step 101: Establish a route planning model with the goal of minimizing total transportation cost, and with additional objective functions of customer satisfaction, carbon emissions, transportation risk, and mileage utilization rate; the route planning model includes parameters related to oil depots, gas stations, tank trucks, and tank trucks.

[0036] Step 102: Using a density-based hierarchical clustering algorithm, and based on the path planning model, calculate the vehicle density of tanker trucks allocated to different gas stations; based on pre-set vehicle merging constraints, merge tanker trucks from different gas stations according to the vehicle density to obtain merged vehicles; the vehicle merging constraints include that the distance between the merged tanker trucks is less than or equal to the pre-set maximum tanker truck loading distance; the vehicle merging constraints are used to constrain the tanker truck carrying capacity, delivery order, delivery time window, and distance between tanker trucks; based on the merged vehicles, determine the vehicle loading path for refined oil between gas stations;

[0037] Step 103: Using a genetic algorithm based on a path planning model, establish vehicle codes corresponding to the merged vehicles; the locus number of the vehicle code represents the number of the merged vehicle; the value of the vehicle code represents the assigned oil depot number; generate an initial population based on the calculated standard deviation of the mileage saved by different merged vehicles; the standard deviation of the mileage saved is used to characterize the standard deviation of the mileage saved by assigning different merged vehicles to the target oil depot relative to assigning them to other oil depots; each individual in the initial population represents a different vehicle code assigned to the initial oil depot; by performing crossover and mutation operations on the initial population, determine the merged vehicle loading path for refined oil between different oil depots;

[0038] Step 104: Output the vehicle loading routes of refined oil between gas stations and the combined vehicle loading routes of refined oil between different oil depots as a route planning scheme for refined oil distribution.

[0039] The route planning model constructed in this invention uses minimizing total transportation cost as the objective function, and incorporates customer satisfaction, carbon emissions, transportation risk, and mileage utilization as additional objective functions. It is no longer limited to the traditional single optimization that only considers transportation cost, but can maximize the overall benefits of secondary oil delivery in terms of economy, service, environment, and operational efficiency, overcoming the lack of comprehensive optimization in existing technologies. The density-based hierarchical clustering algorithm is used for vehicle merging, taking into account that the distance between tanker trucks is less than or equal to the farthest loading distance, as well as constraints on carrying capacity, delivery order, and delivery time windows. This helps to accurately combine vehicles according to the actual distribution and demand of gas stations, ensuring that the merged vehicles meet carrying capacity requirements and complete delivery tasks within the specified delivery time window under a reasonable delivery order. This improves the rationality and scientific nature of vehicle allocation, representing a significant improvement compared to existing vehicle allocation methods that do not fully incorporate multiple constraints. The density-based hierarchical clustering algorithm filters out deviation points by calculating vehicle density. The algorithm prioritizes merging vehicles with higher density and shorter distances, using the furthest loading distance as the termination condition. This avoids indiscriminate merging of all stations, effectively reducing unnecessary computational steps, lowering the computational load, and improving efficiency when handling large-scale data. It overcomes the slow computation speed of traditional algorithms when dealing with large-scale delivery planning. Furthermore, the algorithm incorporates standard deviation mileage saving and designs corresponding genetic operators. This standard deviation mileage saving guides the algorithm to find better customer segmentation schemes, avoiding getting trapped in local optima, thus improving the optimization quality and enhancing its adaptability to complex constraints and large-scale delivery route planning problems. It also addresses the weakness of traditional algorithms in getting trapped in local optima when dealing with such problems. Finally, the route planning for refined oil delivery is decomposed into two sub-problems: merging travel distance and vehicle segmentation. These are solved sequentially using a system clustering algorithm and an improved genetic algorithm, achieving optimized planning of large-scale refined oil secondary delivery routes. Under various constraints, this reduces transportation costs, improves delivery efficiency, and enhances overall operational benefits.

[0040] In specific implementation, the first step is to establish a route planning model with the goal of minimizing total transportation cost, and with additional objective functions such as customer satisfaction, carbon emissions, transportation risk, and mileage utilization rate. The route planning model includes parameters related to oil depots, gas stations, tank trucks, and tank trucks.

[0041] In this embodiment, during the specific implementation process, the present invention first constructs a route planning model to optimize the secondary distribution of refined oil products. The total transportation cost, as a key optimization objective, is calculated rigorously and meticulously. For each tanker truck, the cost from the oil depot to each gas station needs to be calculated in detail. Fuel costs are accurately calculated based on the vehicle's actual mileage, fuel price per unit, and the vehicle's own fuel consumption characteristics; vehicle maintenance costs are determined through a predetermined calculation model, taking into account factors such as vehicle mileage, vehicle model, and maintenance cycle; driver labor costs are precisely calculated based on parameters such as driver working hours and hourly wages. These costs for all tanker trucks are then summed to obtain the total transportation cost.

[0042] Furthermore, to achieve a more comprehensive and efficient delivery planning, this invention introduces customer satisfaction, carbon emissions, transportation risk, and mileage utilization as additional objective functions. Customer satisfaction is calculated by comparing the actual delivery time of the gas station with a preset time window. Different calculation methods are used depending on the delivery time range to accurately measure customer satisfaction with the delivery time. Carbon emissions are calculated based on the fuel load, mileage, and carbon emission characteristics of the tanker truck on different road sections, taking into account these factors to arrive at the carbon emission value. Transportation risk is calculated considering factors such as road distance, average population density within the accident impact area, accident impact range, fuel load, and road resilience, determining the level of transportation risk through precise calculation. Mileage utilization is calculated based on parameters such as the distance from the oil depot to the gas station and the total distance traveled by the vehicle, thereby evaluating the efficiency of vehicle mileage utilization.

[0043] In one embodiment, the objective function is further calculated as follows:

[0044] Calculate the transportation cost of each tanker truck from the oil depot to each gas station, including fuel cost, vehicle maintenance cost, and driver labor cost; add up the transportation costs of all tanker trucks to obtain the total transportation cost.

[0045] In the above embodiments, the route planning model encompasses oil depots, gas stations, tank trucks, and related parameters. Oil depot parameters include the number of oil depots, the number of oil types at each depot, the oil inventory at each depot, and the number of tank trucks at each depot. Gas station parameters include the number of gas stations, the demand for different oil products at each gas station, the delivery execution time at each gas station, the time of capacity cutoff, the time of fuel shortage, and the unloading time. Tank truck parameters include tank truck capacity, vehicle speed, and unit distance transportation cost. These parameters provide fundamental data support for subsequent vehicle allocation, customer segmentation, and route planning, ensuring that the model accurately reflects the actual delivery situation and providing strong support for optimizing delivery.

[0046] For example, consider minimizing total transportation cost as the primary objective function. Total transportation cost encompasses several expenses, including fuel costs for tanker trucks traveling from the oil depot to various gas stations, which depend on the tanker truck's mileage, fuel consumption rate, and fuel price; vehicle maintenance costs, which are related to the tanker truck's mileage, usage frequency, and maintenance cycle; and driver labor costs, calculated based on driver hours and salary standards. When planning delivery routes, these factors must be considered comprehensively. By rationally arranging tanker truck routes and delivery tasks, total transportation cost can be reduced as much as possible. For instance, avoiding long-distance, poorly conditioned, or congested routes can reduce fuel consumption and vehicle wear, thereby lowering costs.

[0047] Customer satisfaction is used as an additional objective function. Customer satisfaction primarily depends on whether the tanker trucks can deliver fuel to the gas station within the expected timeframe. For each gas station, a delivery time window is defined, including the earliest acceptable delivery time (out-of-capacity point) and the latest acceptable delivery time (out-of-fuel point). If the tanker trucks arrive within this time window, customer satisfaction is high; arriving too early or too late will negatively impact customer satisfaction. For example, arriving too early may lead to insufficient storage space at the gas station, while arriving too late may expose the gas station to the risk of fuel shortages, affecting normal operations. Therefore, the route planning model must consider how to optimize the delivery sequence and timing to improve customer satisfaction.

[0048] Carbon emissions are also an important additional objective function. Tanker trucks generate carbon emissions during operation, the amount of which is related to mileage, fuel load, and the tanker's carbon emission factor. By rationally planning routes, reducing unnecessary mileage, and optimizing fuel load distribution, carbon emissions can be reduced. For example, prioritizing short distances and routes with good road conditions, and avoiding empty runs or under-loaded vehicles, helps reduce environmental impact and achieve green delivery.

[0049] Transportation risk, as an additional objective function, considers multiple risk factors. For example, road conditions are a significant factor influencing transportation risk, including road smoothness, gradient, curves, and traffic flow. Complex road conditions can increase the probability of accidents. Simultaneously, population density in the area where the road is located also affects risk; accidents in densely populated areas can result in greater losses. By avoiding high-risk road sections in route planning, such as accident-prone areas or densely populated areas with narrow roads, transportation risk can be reduced.

[0050] Mileage utilization rate, as an additional objective function, aims to improve the efficiency of tanker truck mileage utilization. It calculates the distance from the oil depot to each gas station and the actual mileage traveled by the tanker trucks. By rationally scheduling delivery tasks, it ensures that tanker trucks are as fully loaded as possible and travel reasonable distances during each delivery, avoiding excessive detours or empty runs, thereby improving mileage utilization rate and reducing unit transportation costs.

[0051] The route planning model incorporates numerous relevant parameters. Parameters for oil depots include their location coordinates, oil storage capacity, and the initial number of tanker trucks. Gas station parameters include their location coordinates, oil demand, and delivery time windows (times of inactivity, oil shortages, and unloading). Tanker truck parameters cover vehicle model, carrying capacity, fuel consumption rate, and carbon emission factor. These parameters collectively constitute the foundational data for the route planning model, used to calculate the objective function and constraints, providing a basis for subsequent route planning and optimization. For example, based on the oil depot's storage capacity and the gas station's demand, the delivery workload of the tanker trucks can be determined; based on the tanker trucks' carrying capacity and fuel consumption rate, transportation costs and carbon emissions can be calculated. By comprehensively considering these parameters, a comprehensive and accurate route planning model for secondary refined oil distribution is established to achieve efficient, economical, environmentally friendly, and safe delivery processes.

[0052] In specific implementation, after step 101: establishing a route planning model with minimizing total transportation cost as the objective function and customer satisfaction, carbon emissions, transportation risk, and mileage utilization as additional objective functions, step 102: using a density-based hierarchical clustering algorithm, based on the route planning model, calculate the vehicle density of tanker trucks allocated to different gas stations; based on pre-set vehicle merging constraints, merge tanker trucks from different gas stations according to the vehicle density to obtain merged vehicles; the vehicle merging constraints include that the distance between the merged tanker trucks is less than or equal to the pre-set maximum tanker truck loading distance; the vehicle merging constraints are used to constrain the tanker truck carrying capacity, delivery order, delivery time window, and distance between tanker trucks; based on the merged vehicles, determine the vehicle loading path for refined oil between gas stations.

[0053] In this embodiment, the application of density-based hierarchical clustering algorithm in the refined oil secondary distribution route planning model begins with the vehicle density calculation stage. To obtain the vehicle density of tankers allocated to different gas stations, one tanker is first assigned to each gas station, at which point the distance between vehicles is equivalent to the distance between gas stations. Next, the vehicle density is determined according to specific calculation rules. Specifically, the vehicle density value is obtained by statistically analyzing the number of other vehicles adjacent to each tanker. In this process, a cutoff distance is set to distinguish the relative relationships between vehicles, and the vehicle density is calculated based on this. After the calculation is completed, approximately 10% of the vehicles with relatively low density are designated as deviation points, while the remaining vehicles are prioritized for subsequent merging operations.

[0054] Vehicle merging constraints are crucial throughout the algorithm. One of the most important constraints is the distance limit between the tanker trucks participating in the merging. The distance between tanker trucks undergoing merging must be less than or equal to a pre-defined maximum loading distance between them. This distance limit not only serves as an initial screening factor when selecting vehicles for merging, but also allows the algorithm to stop running when the distance between all mergingable vehicles reaches or exceeds this maximum loading distance. This effectively reduces unnecessary computational steps and lowers the overall computational load of the algorithm.

[0055] In addition to distance constraints, the constraints also cover limitations on the tanker truck's carrying capacity, delivery sequence, and delivery time window. Regarding the tanker truck's carrying capacity, the total demand of all gas stations served by the two merged vehicles must be within the tanker truck's carrying capacity to ensure the tanker trucks can still safely and efficiently transport refined oil after the merger. Regarding the delivery sequence, there must be a reasonable delivery order arrangement such that the transportation distance between any two adjacent stations is always less than or equal to the furthest loading distance, ensuring the continuity and efficiency of the delivery process. Simultaneously, regarding the delivery time window, according to the selected delivery sequence, the tanker trucks must be able to accurately arrive at each gas station within the specified delivery time window to ensure that the normal operation of the gas stations is not affected.

[0056] Based on the vehicle density calculation results and the distance matrix between vehicles, the vehicle merging operation begins. First, two tanker trucks are randomly selected from a large pool of tanker trucks as potential merging candidates. During the selection process, the tanker truck with the highest density value is prioritized and paired with the nearest tanker truck as a potential merging combination. It is crucial to ensure that the distance between the two selected tanker trucks always meets the critical condition of being less than or equal to the pre-set maximum loading distance between tanker trucks.

[0057] Once the candidate tanker trucks for merging are identified, a comprehensive constraint check is required. Careful evaluation is needed to determine if both tanker trucks meet all the aforementioned vehicle merging constraints, including requirements for carrying capacity, delivery sequence, and delivery time windows. Only when all constraints are met will the merging operation be performed on the selected candidate tanker trucks. After merging, the distance between the newly formed vehicle and other vehicles not involved in the merging will be recalculated using the shortest distance method. To ensure smooth operation of the merged vehicle in actual delivery, an enumeration method is used to conduct detailed testing of the merged tanker trucks. This method comprehensively checks all possible delivery sequences and station combinations to determine the practical feasibility of station merging.

[0058] After successfully merging vehicles and obtaining the merged vehicles, the next step is to determine the vehicle distribution routes for refined oil products between gas stations. This process requires comprehensive consideration of multiple factors to ensure the scientific and rational nature of the distribution routes. First, the geographical distribution of gas stations must be fully considered. The most reasonable transportation routes must be planned based on the relative positions of each gas station, reducing unnecessary detours and bypasses to improve transportation efficiency. Simultaneously, the delivery sequence and volume of tanker trucks must be rationally arranged based on the demand of each gas station to ensure that the refined oil product needs of each gas station are met promptly and accurately. Furthermore, the carrying capacity limitations of tanker trucks must be considered to avoid overloading or wasted capacity. Regarding delivery time windows, delivery arrangements must strictly adhere to the time windows stipulated by each gas station to ensure that the normal operation of the gas stations is not affected. Through comprehensive weighing and optimization calculations of these factors, the vehicle distribution routes for refined oil products between gas stations are finally determined, achieving efficient and orderly secondary distribution of refined oil products.

[0059] In one embodiment, Figure 2 This is a specific example diagram of a route planning method for refined oil delivery in an embodiment of the present invention. Based on pre-set vehicle merging constraints, tanker trucks from different gas stations are merged according to the vehicle density to obtain merged vehicles, such as... Figure 2 As shown, it includes:

[0060] Step 201: Based on the vehicle density of different tank trucks and the distance matrix between different tank trucks, randomly select two tank trucks as candidate tank trucks for merging; select the tank truck with the highest density among the candidate tank trucks for merging and merge it with the tank truck with the closest distance; wherein, the distance between the tank trucks being merged is less than or equal to the preset maximum loading distance of the tank trucks.

[0061] Step 202: Determine whether the tank trucks to be merged meet the merging constraints for each vehicle; if they do, then merge the candidate tank trucks.

[0062] Step 203: Use the enumeration method to perform detection among the merged tank trucks to obtain the merged vehicle.

[0063] In the above embodiments, when performing tanker truck merging operations based on pre-set vehicle merging constraints, the first step is to select candidate tanker trucks for merging based on the vehicle density of different tanker trucks and the distance matrix between them. Specifically, two tanker trucks are randomly selected from all available tanker trucks as initial candidate merging vehicles. After determining the initial candidate tanker trucks, their vehicle densities need to be compared. The tanker truck with the highest vehicle density is selected, and then the tanker truck closest to this highest density tanker truck is found among the remaining tanker trucks. Throughout the selection process, it must be ensured that the distance between these two tanker trucks is less than or equal to the pre-set maximum loading distance for tanker trucks. This distance constraint is one of the basic prerequisites for vehicle merging, ensuring that the merged tanker trucks meet the economic and safety requirements of actual operation during delivery.

[0064] Once potential tanker truck combinations for merging are identified, rigorous verification of vehicle merging constraints is required. These constraints encompass multiple aspects, including tanker truck carrying capacity, delivery sequence, and delivery time windows. Regarding tanker truck carrying capacity, the total fuel demand for all gas stations served by the two candidate tankers must be calculated and compared with the tanker trucks' carrying capacity to ensure the total demand remains within that range. For delivery sequence, it must be checked whether a reasonable delivery order exists that ensures the transport distance between any two adjacent gas stations does not exceed the maximum loading distance. Simultaneously, for delivery time windows, information such as gas station inactivity points, fuel shortage points, and unloading times must be considered to determine whether the tankers can arrive at each gas station within the designated delivery time window according to the selected delivery sequence. Only when all these constraints are met can the merging operation on the selected candidate tankers proceed.

[0065] After merging the tanker trucks, an enumeration method was used to comprehensively test the merged vehicles to ensure they could meet actual delivery needs. Enumeration is a systematic and comprehensive testing method that analyzes the merged vehicles in detail by listing all possible scenarios. Specifically, it examines various possible delivery sequences and station combinations among the gas stations involved in the merged vehicles. For each possible combination, its compliance with vehicle merging constraints, including requirements for carrying capacity, delivery order, and delivery time windows, is reassessed. Simultaneously, relevant indicators such as vehicle travel distance and delivery efficiency under different combinations are calculated. Through this comprehensive and meticulous enumeration testing process, merged vehicles that meet the requirements are ultimately obtained, providing a reliable basis for subsequently determining the vehicle distribution routes for refined oil products among gas stations. This process effectively avoids unreasonable or infeasible delivery plans caused by the merging operation, thereby improving the accuracy and effectiveness of the entire refined oil secondary distribution route planning.

[0066] For example:

[0067] Assume a specific region contains several oil depots and numerous gas stations located in different areas. Each oil depot has a certain number of tank trucks, whose carrying capacity, storage layout, and other attributes are known. Meanwhile, the fuel demand, delivery time windows (including periods of no capacity and fuel shortages), and unloading times of each gas station are also known. Based on these realities, a route planning model for refined oil product delivery is established. This model encompasses key elements such as oil depots, gas stations, tank trucks, and the relationships between them.

[0068] Using a density-based hierarchical clustering algorithm, each gas station is first assigned a tanker truck, with the distance between vehicles equal to the distance between gas stations. Next, the vehicle density of each tanker truck is calculated by counting the number of other vehicles adjacent to it. For example, based on the relative positions of the vehicles, which vehicles can be considered adjacent are determined, and then the density value of each vehicle is calculated. The calculated vehicle densities are sorted, and vehicles with relatively low density values ​​are marked as deviation points, while the remaining vehicles are considered candidates for subsequent merging operations.

[0069] Based on the vehicle density and distance matrix, the vehicle merging operation begins. Two tanker trucks are randomly selected from the candidate vehicles, prioritizing the one with higher density. Then, the tanker truck closest to it from the remaining candidate vehicles is chosen as the merging target. During the selection process, it is strictly ensured that the distance between these two tanker trucks is less than or equal to the pre-set maximum loading distance between tanker trucks. This distance constraint not only filters out suitable merging vehicles but also serves as one of the algorithm's termination conditions, avoiding unnecessary computation.

[0070] Once potential vehicles for merging are identified, they are checked to ensure they meet a series of vehicle merging constraints. Regarding the tanker truck's carrying capacity, the total fuel demand of all gas stations served by the two candidate vehicles is calculated to ensure it does not exceed the tanker truck's carrying capacity, guaranteeing the merged vehicle can transport fuel safely and efficiently. For delivery order, the geographical location and demand of gas stations are analyzed to determine if a reasonable delivery sequence exists, such that the transport distance between any two adjacent gas stations is less than or equal to the furthest loading distance, ensuring efficiency and continuity in the delivery process. Simultaneously, based on the gas station's delivery time window information, it is checked whether the tanker truck can arrive at each gas station within the specified time frame according to the selected delivery order, avoiding disruptions to normal gas station operations due to arrivals that are too early or too late.

[0071] If two candidate vehicles satisfy all these constraints, they are merged into a new vehicle. After merging, the distance between the new vehicle and other vehicles not involved in the merger is recalculated using the shortest distance method for subsequent merger decisions. To further verify the feasibility of the merger, an enumeration method is used to test the merged vehicle. Specifically, all possible delivery sequences and station combinations are listed, and it is checked whether the vehicle merger constraints are still satisfied under these combinations. At the same time, feasible vehicle loading routes are recorded to provide data support for subsequently determining the final vehicle loading routes for refined oil between gas stations.

[0072] After completing multiple rounds of vehicle consolidation, a series of consolidated vehicles were obtained. Based on previously recorded vehicle loading route information, combined with the actual needs and geographical distribution of gas stations, the final vehicle loading routes for refined oil between gas stations were determined. For example, considering factors such as the urgency of gas station fuel demand, vehicle carrying capacity, and route rationality, consolidated vehicles were arranged to deliver fuel to each gas station sequentially, ensuring that the fuel needs of each gas station are met promptly and accurately, while optimizing overall delivery efficiency, reducing transportation costs, and improving the operational efficiency of secondary refined oil delivery. In determining the loading routes, actual factors such as road conditions and traffic flow were fully considered to avoid selecting congested or unsuitable routes for tanker trucks, ensuring the smoothness and safety of the delivery process.

[0073] In specific implementation, step 102 involves: using a density-based hierarchical clustering algorithm, based on the path planning model, to calculate the vehicle density of tanker trucks allocated to different gas stations; based on pre-set vehicle merging constraints, merging tanker trucks from different gas stations according to the vehicle density to obtain merged vehicles; the vehicle merging constraints include that the distance between the merged tanker trucks is less than or equal to the pre-set maximum loading distance of the tanker trucks; the vehicle merging constraints are used to constrain the tanker truck carrying capacity, delivery order, delivery time window, and distance between tanker trucks; based on the merged vehicles, determining the vehicle allocation of refined oil between gas stations. After determining the loading path, step 103 is performed: using a genetic algorithm based on a path planning model, vehicle codes corresponding to the merged vehicles are established; the locus number of the vehicle code represents the number of the merged vehicle; the value of the vehicle code represents the assigned oil depot number; an initial population is generated based on the calculated standard deviation of the mileage saved by different merged vehicles; the standard deviation of the mileage saved is used to characterize the standard deviation of the mileage saved by assigning different merged vehicles to the target oil depot relative to assigning them to other oil depots; each individual in the initial population represents a different vehicle code assigned to the initial oil depot; by performing crossover and mutation operations on the initial population, the loading path of the merged vehicles for refined oil between different oil depots is determined.

[0074] In this embodiment, when using a genetic algorithm in the secondary distribution route planning of refined oil products, the first step is to establish vehicle codes for the merged vehicles. The vehicle code is a digital representation of the distribution plan, and its design has clear rules and meanings. The locus number of the vehicle code is set to represent the number of the merged vehicle. This numbering method allows for clear identification and processing of the relevant information of each merged vehicle in subsequent genetic operations. The value of the vehicle code is used to represent the assigned oil depot number. This establishes a relationship between the vehicle and the oil depot, enabling accurate determination of the oil depot origin of each merged vehicle during route planning, laying the foundation for rationally planning the distribution route.

[0075] The standard deviation mileage saving calculation aims to more accurately characterize the standard deviation of the mileage saved by assigning different merged vehicles to the target oil depot compared to assigning them to other oil depots. The calculation process involves a comprehensive consideration of the distance relationships between each oil depot and the gas station, as well as the vehicle allocation situation. Through a specific calculation method, the standard deviation of mileage saving for each merged vehicle under different allocation schemes is obtained. This value reflects the rationality and efficiency of vehicle allocation, providing a crucial basis for generating the initial population. The magnitude of the standard deviation mileage saving directly affects the priority of vehicle allocation; the smaller the value, the more advantageous it is to assign the merged vehicle to the target oil depot in terms of mileage saving compared to other oil depots, thus making it more likely to be prioritized when generating the initial population.

[0076] An initial population is generated based on the calculated standard deviation mileage savings of different merged vehicles. First, after calculating the standard deviation mileage savings for all merged vehicles, priority probabilities for different merged vehicles are calculated based on these values. Priority probabilities include various types such as the priority allocation probability of merged vehicles not assigned to fuel depots, the priority removal probability of each merged vehicle, and the probability of merged vehicles being assigned to different fuel depots. When calculating priority probabilities, standard deviation mileage savings and other relevant factors (such as fuel depot inventory and gas station demand) are comprehensively considered, and specific calculation rules (such as the relevant formulas mentioned in the text) are used to derive the probability value of each merged vehicle under different allocation scenarios.

[0077] Next, merged vehicles are randomly selected according to the calculated priority probabilities. Each selection is based on probability, making the results more random and reasonable. After a merged vehicle is selected, its assigned depot is determined again through random selection from among depots whose remaining inventory exceeds the vehicle's fuel demand. After determining the assigned depot, its remaining inventory is updated promptly to reflect actual inventory changes. This process of randomly selecting merged vehicles and depots is repeated until each merged vehicle is assigned to a different depot. Finally, the vehicle codes formed by the merged vehicles and their corresponding assigned depots are combined to form the initial population. Each individual in the initial population represents a possible vehicle allocation scheme, providing an initial sample set for subsequent genetic operations.

[0078] Crossover and mutation operations are performed on the initial population based on pre-defined crossover, mutation, and correction operators. The crossover operator selects some loci from the initial population for exchange according to specific calculation rules (such as the locus-priority removal probability described in the text), thereby generating new offspring individuals. The mutation operator, based on corresponding calculation methods (such as considering the locus-priority removal probability and redistributing loci), modifies certain loci in individuals, introducing new genetic information and increasing population diversity. After the crossover and mutation operations are completed, the post-operation population is obtained.

[0079] Subsequently, the initial population is replaced with the post-operated population, and the crossover and mutation operations described above are repeated. This iterative process continues until a specific termination condition is met: the maximum value of the objective function (e.g., minimizing total transportation costs) corresponding to the post-operated population is less than a preset value, and the values ​​of additional objective functions (e.g., customer satisfaction, carbon emissions, transportation risk, and mileage utilization) are within their respective preset threshold ranges. When this termination condition is met, the combined vehicle loading routes for refined oil products between different oil depots are determined based on the individual information in the post-operated population. These routes are optimal solutions obtained through multiple iterations of optimization, achieving the optimization objectives of secondary refined oil product distribution route planning while satisfying various constraints, thereby improving distribution efficiency, reducing costs, mitigating risks, and enhancing resource utilization.

[0080] Figure 3 This is a specific example diagram of a route planning method for refined oil delivery according to an embodiment of the present invention. In one embodiment, an initial population is generated based on the calculated mileage savings from the standard deviation of different merged vehicles, such as... Figure 3 As shown, it includes:

[0081] Step 301: Calculate the standard deviation of mileage saved for different merged vehicles;

[0082] Step 302: Calculate the priority probability of different merged vehicles based on the standard deviation of the mileage saved; the priority probability includes the priority allocation probability of merged vehicles that have not been assigned to an oil depot, the priority removal probability of each merged vehicle, and / or the probability of merged vehicles being assigned to different oil depots.

[0083] Step 303: Randomly select merged vehicles according to priority probability; randomly select an oil depot from oil depots whose remaining inventory is greater than the oil demand of the randomly selected merged vehicle, and allocate it to the randomly selected merged vehicle; and update the remaining inventory of the randomly selected oil depot; repeat the above steps of randomly selecting merged vehicles and randomly selecting oil depots until each merged vehicle is allocated to a different oil depot.

[0084] Step 304: Use the merged vehicle codes and the corresponding vehicle codes of the assigned oil depots as the initial population.

[0085] Figure 4 This is a specific example diagram of a route planning method for refined oil product distribution in an embodiment of the present invention. In one embodiment, by performing crossover and mutation operations on the initial population, the vehicle loading route for refined oil products after merging between different oil depots is determined, as shown in Figure 4, including:

[0086] Step 401: Based on the preset crossover operator, mutation operator and correction operator, perform crossover and mutation operations on the initial population to obtain the post-operation population;

[0087] Step 402: Replace the initial population with the post-operation population and repeat the above steps of crossover and mutation until the maximum value of the objective function corresponding to the post-operation population is less than a preset value and the value of the additional objective function is at the corresponding preset threshold.

[0088] Step 403: Based on the post-operation population where the maximum value of the objective function is less than a preset value and the value of the additional objective function is at the corresponding preset threshold, determine the vehicle loading path for the merged refined oil between different oil depots.

[0089] In the above embodiments, to generate the initial population, it is first necessary to calculate the standard deviation mileage savings of different merged vehicles. Calculating the standard deviation mileage savings is a process based on specific algorithms and model parameters, aimed at quantifying the difference in mileage savings when assigning different merged vehicles to the target oil depot compared to assigning them to other oil depots. By comprehensively considering factors such as the distance between each oil depot and the gas station, vehicle carrying capacity, and delivery demand, the standard deviation mileage savings value for each merged vehicle is obtained.

[0090] Based on the calculated standard deviation of mileage savings, the priority probability of different merged vehicles is further calculated. Priority probabilities encompass various types, including the priority allocation probability of merged vehicles not assigned to fuel depots, the priority removal probability of each merged vehicle, and the probability of merged vehicles being assigned to different fuel depots. For the calculation of the priority allocation probability of merged vehicles not assigned to fuel depots, a comprehensive analysis of their standard deviation mileage savings relative to other unassigned vehicles is required, considering factors such as the vehicle's potential adaptability to different fuel depots. A specific calculation rule (such as the relevant formulas and logic given in the text) is used to derive their priority allocation probability value. The priority removal probability of each merged vehicle is calculated based on its current allocation status and its relationship with other vehicles in the delivery plan, combined with information such as standard deviation mileage savings. The calculation of the probability of merged vehicles being assigned to different fuel depots considers various factors, including the inventory status of each fuel depot, the distance between the vehicle and the fuel depot, and the vehicle's impact on overall delivery efficiency under different allocation conditions. Similarly, a specific calculation method is used to derive the corresponding probability value.

[0091] After calculating the priority probability, merged vehicles are randomly selected according to the priority probability. During the random selection process, the calculated priority probability is strictly adhered to as the selection criterion, ensuring that the selection result is random while also reflecting a reasonable tendency in vehicle allocation. Once a merged vehicle is selected, an oil depot is randomly selected again from those with remaining inventory greater than the fuel demand of the randomly selected merged vehicle, and this depot is allocated to the selected merged vehicle. This process requires real-time acquisition and updating of the remaining inventory information of each oil depot to ensure the feasibility and accuracy of the allocation. After the oil depot allocation is completed, the remaining inventory of the randomly selected oil depot is immediately updated, and the allocated fuel demand is deducted from the inventory to reflect dynamic changes in inventory. The above steps of randomly selecting merged vehicles and randomly selecting oil depots are repeated continuously until each merged vehicle is allocated to a different oil depot. Finally, the vehicle codes determined by each merged vehicle and its corresponding allocated oil depot are combined to form the initial population. Each individual in the initial population represents a complete vehicle allocation scheme, and these individuals constitute the basic sample set for subsequent genetic operations.

[0092] In another embodiment, crossover and mutation operations are performed on the initial population based on pre-set crossover, mutation, and correction operators. The crossover operator performs locus exchange operations on individuals in the initial population according to predetermined rules. Specifically, based on the individual's locus information and a pre-set crossover method (such as the crossover point and method determined based on factors like the priority removal probability of loci), some loci are selected from different individuals for exchange, thereby generating new offspring individuals. The mutation operator then modifies certain loci in individuals according to corresponding set rules. During mutation, factors such as the priority removal probability of loci are considered, and the values ​​of loci are reset with a certain probability (such as redistributing oil reservoirs), thereby introducing new genetic information into the population and increasing population diversity. Through crossover and mutation operations, the manipulated population is obtained.

[0093] Subsequently, the initial population is replaced with the post-operated population, and the crossover and mutation operations described above are repeated. During each iteration, the post-operated population is continuously evaluated. It is determined whether the maximum value of the objective function (such as minimizing total transportation costs) corresponding to the post-operated population is less than a preset value, and simultaneously, whether the values ​​of additional objective functions (such as customer satisfaction, carbon emissions, transportation risk, and mileage utilization) are within their corresponding preset threshold ranges. This iterative process continues until the aforementioned termination condition is met. When the termination condition is reached, based on the individual information in the post-operated population at this point, the merged vehicle loading routes for refined oil between different oil depots are determined. These routes are optimal solutions obtained after multiple iterations of optimization. While satisfying various constraints (such as vehicle carrying capacity, delivery time window, and oil depot inventory), they achieve optimization goals in cost, efficiency, and environmental impact for secondary refined oil distribution route planning, providing a scientifically sound route planning scheme for secondary refined oil distribution.

[0094] For example:

[0095] Suppose there are multiple oil depots distributed within a certain area, each with several tanker trucks, and numerous gas stations, each with different geographical locations and fuel demands. We have known the initial inventory of each oil depot, the basic carrying capacity of the tanker trucks, and the fuel demands of the gas stations. First, using a density-based hierarchical clustering algorithm, the gas station demand orders have been merged, resulting in merged vehicles. Preliminary vehicle loading paths between gas stations have also been determined, providing foundational data for subsequent genetic algorithm operations.

[0096] A genetic algorithm is used to encode the merged vehicles. Each merged vehicle is treated as an individual, and its encoding follows specific rules, such as using a numerical sequence. The locus number in the vehicle code is set as the unique identifier for each merged vehicle. The code value represents the assigned fuel depot number, thus establishing the association between the vehicle and the fuel depot.

[0097] Next, the standard deviation of mileage savings for different merged vehicles was calculated. By comprehensively analyzing the factors affecting the degree of mileage savings for each merged vehicle under different oil depot allocation scenarios, the standard deviation of mileage savings for each vehicle was calculated. This indicator can reflect the stability and potential advantages of mileage savings when vehicles are allocated to different oil depots, providing a key basis for the subsequent generation of the initial population.

[0098] Based on the calculated standard deviation of mileage savings, an initial population is generated. First, based on the magnitude of the standard deviation mileage savings and related characteristics, the priority probability of each merged vehicle under different allocation scenarios is calculated, including the priority allocation probability of merged vehicles not assigned to fuel depots, the priority removal probability of each merged vehicle, and the probability of merged vehicles being assigned to different fuel depots. These probabilities comprehensively consider the potential value and adaptability of vehicles in the delivery network.

[0099] Then, merged vehicles are randomly selected according to priority probabilities. During the selection process, probability information is fully utilized to ensure the selection is both random and reasonable. After each vehicle is selected, an oil depot is randomly selected from the set of oil depots whose remaining inventory can meet the vehicle's fuel needs, and the remaining inventory information of the selected oil depot is updated promptly. This process is repeated until every merged vehicle is assigned to a different oil depot. The resulting set of vehicle code combinations constitutes the initial population. Each individual in the initial population represents a possible vehicle allocation scheme, covering different combinations of vehicles and oil depots.

[0100] The initial population is manipulated based on predefined crossover and mutation operators. The crossover operator selects two or more individuals from the initial population according to set rules, and generates new offspring individuals by exchanging coding information at certain loci. For example, crossover operations might be performed on several loci of individuals with a certain probability to simulate information exchange and fusion between different vehicle allocation schemes, in order to generate new schemes with potential advantages.

[0101] The mutation operator randomly alters certain loci within an individual with a low probability. This alteration might involve operations such as reassigning oil depot numbers, aiming to introduce new genetic information into the population, increase its diversity, and prevent the algorithm from prematurely getting trapped in local optima. After the crossover and mutation operations are completed, the resulting population is obtained.

[0102] Subsequently, the initial population is replaced with the post-operated population, and the crossover and mutation operations described above are repeated. This process is continuously iterated, and the post-operated population is evaluated in each iteration. It is determined whether the objective function corresponding to each individual in the population (such as minimizing total transportation costs, maximizing customer satisfaction, etc.) meets the preset optimization criteria, that is, whether the maximum value of the objective function is less than a preset value, and whether the values ​​of additional objective functions (such as carbon emissions, transportation risk, and mileage utilization) are within the corresponding preset threshold ranges.

[0103] When the above termination conditions are met, the merged vehicle loading routes for refined oil products between different oil depots are determined based on the individual information in the population after the operation. These routes are optimal solutions obtained through multiple iterations of optimization. They comprehensively consider various factors and can achieve the optimization goal of secondary distribution route planning for refined oil products while meeting actual distribution constraints (such as tanker truck carrying capacity, gas station delivery time windows, etc.), thereby improving distribution efficiency, reducing costs, and enhancing overall operational benefits. For example, the final route planning scheme may specify which gas stations each tanker truck will pass through from its respective oil depot, as well as detailed information such as delivery time, loading and unloading sequence at each gas station. It also provides comprehensive evaluation indicators such as total cost, total carbon emissions, total transportation risk, and average mileage utilization rate for the entire distribution scheme, providing a scientific and reasonable decision-making basis for actual secondary distribution of refined oil products.

[0104] In specific implementation, step 103 involves using a genetic algorithm based on a path planning model to establish vehicle codes corresponding to the merged vehicles; the locus number of the vehicle code represents the number of the merged vehicle; the value of the vehicle code represents the assigned oil depot number; an initial population is generated based on the calculated standard deviation of the mileage saved by different merged vehicles; the standard deviation of the mileage saved is used to characterize the standard deviation of the mileage saved by assigning different merged vehicles to the target oil depot relative to assigning them to other oil depots; each individual in the initial population represents a different vehicle code assigned to the initial oil depot; after determining the merged vehicle loading path of refined oil between different oil depots by performing crossover and mutation operations on the initial population, step 104 involves outputting the vehicle loading path of refined oil between gas stations and the merged vehicle loading path of refined oil between different oil depots as a route planning scheme for refined oil distribution.

[0105] In one embodiment, the route planning scheme for refined oil product distribution further includes:

[0106] The routes of each tanker truck between gas stations, the delivery time and the amount of refined oil delivered at each gas station, the loading and unloading sequence of each tanker truck, the total cost of the route planning scheme, the total carbon emissions, the total transportation risk, and the average mileage utilization rate.

[0107] In the above embodiments, during the secondary distribution route planning process for refined oil products, the vehicle loading routes between gas stations and the combined vehicle loading routes between different oil depots are taken as core components, together forming the route planning scheme for refined oil product distribution. This output process involves integrating and organizing the results of previous calculations and optimizations to ensure the completeness and usability of the route planning scheme. When outputting the route planning scheme, not only is the driving route information of vehicles between different stations included, but it also covers other important information closely related to the distribution process, thereby providing comprehensive guidance for actual refined oil product distribution operations.

[0108] In one embodiment, the route planning scheme for refined oil product distribution further includes several key elements. For each tanker truck's route between gas stations, the specific path from one gas station to the next is clearly defined, based on a comprehensive consideration of factors such as the gas station's geographical location, road conditions, and delivery efficiency. By rationally planning the routes, the aim is to reduce transportation distance, lower transportation costs, and improve delivery timeliness.

[0109] The delivery time and quantity of refined oil delivered to each gas station are also crucial components of the route planning scheme. Determining the delivery time requires careful consideration of factors such as the gas station's operational needs, periods of limited capacity, and times of fuel shortages. This ensures that tanker trucks arrive at the gas station within the appropriate time window, avoiding disruptions to normal operations or fuel shortages due to deliveries that are too early or too late. Simultaneously, precisely specifying the quantity of refined oil delivered to each gas station satisfies its actual needs while fully utilizing the tanker truck's carrying capacity, preventing wasted capacity or overloading.

[0110] The loading and unloading sequence of each tanker truck is equally important. A well-planned loading sequence can improve loading efficiency and ensure the safe loading and transportation of different types of oil. The unloading sequence, on the other hand, needs to be determined based on factors such as the urgency of the gas station's needs and the availability of unloading facilities, to ensure a smooth unloading process, reduce waiting time, and improve overall delivery efficiency.

[0111] In addition, the route planning scheme includes comprehensive evaluation indicators such as total cost, total carbon emissions, total transportation risk, and average mileage utilization rate. Total cost encompasses all expenses incurred by the tanker trucks from the oil depot to various gas stations and back, including fuel costs, vehicle maintenance costs, and driver labor costs. Accurate calculation of total cost helps assess and compare the economics of delivery schemes. Total carbon emissions are calculated based on factors such as the tanker truck's mileage, fuel load, and carbon emission factors, reflecting the environmental impact of the delivery process and providing data support for pursuing green and environmentally friendly delivery. Total transportation risk comprehensively considers factors such as road conditions, accident impact range, population density, and the special characteristics of oil transportation. Quantifying risk assessment helps in taking corresponding measures to reduce transportation risks. Average mileage utilization rate assesses the efficiency of tanker truck mileage utilization during delivery. Its calculation is based on the relationship between vehicle travel distance and the distance from the oil depot to the gas station. A higher average mileage utilization rate indicates a more efficient and rational use of vehicle resources in the delivery scheme. The inclusion of these comprehensive evaluation indicators makes the route planning scheme more comprehensive and scientific, enabling the evaluation and optimization of the distribution scheme from multiple dimensions, and providing a better and more feasible decision-making basis for the secondary distribution of refined oil products.

[0112] For example, consider a refined oil delivery area comprising multiple oil depots and numerous gas stations. After processing gas station demand orders using a density-based hierarchical clustering algorithm and calculating vehicle allocation and route optimization using a genetic algorithm, the vehicle loading routes between gas stations and the merged vehicle loading routes between different oil depots have been determined. Each oil depot and gas station possesses its own geographical location information, and the tanker trucks have corresponding carrying capacities and other basic attributes. Furthermore, the delivery process must adhere to certain constraints, such as gas station delivery time windows and the maximum loading distance limit for tanker trucks.

[0113] Output of path planning solution:

[0114] 1. Vehicle loading route information

[0115] For the vehicle distribution routes of refined oil products between gas stations, the travel routes of each tanker truck between gas stations should be clearly defined. For example, after a tanker truck departs from a gas station, it will proceed to other gas stations in a planned sequence for delivery. The delivery routes should detail the order in which each tanker truck will arrive at each gas station, as well as the stopping and delivery arrangements at each station, to ensure that the oil products are delivered accurately and promptly to each gas station to meet their operational needs.

[0116] 2. Vehicle loading routes after the merger of oil depots

[0117] For the consolidated vehicle loading routes of refined oil products across different oil depots, the oil depot to which each consolidated vehicle belongs and its delivery sequence between different depots are determined. For example, a consolidated vehicle departing from a specific oil depot may, based on demand and optimization results, return to the original oil depot or proceed to other oil depots to replenish fuel or perform other delivery tasks after completing deliveries to certain gas stations. Clearly defining these vehicle movement routes between oil depots helps in the rational arrangement of oil depot inventory management and vehicle scheduling.

[0118] 3. Details regarding gas station delivery

[0119] Output the delivery schedule for each gas station to ensure that tanker trucks arrive within the designated time window, neither earlier than the depletion time nor later than the fuel shortage time. Simultaneously, determine the amount of refined oil to be delivered to each gas station, and rationally allocate fuel resources based on demand forecasts and actual operating conditions to avoid waste or supply shortages.

[0120] 4. Tanker truck operation sequence

[0121] Detailed planning was implemented for the loading and unloading sequence of each tanker truck. During loading, a reasonable loading order was determined based on the demand for different types of oil and the layout of the tanker truck's storage compartments, ensuring safe loading and preventing mixing of different oil types. During unloading, the unloading sequence at each gas station was arranged according to the urgency of the gas station's needs and the availability of unloading facilities, improving unloading efficiency and reducing waiting time.

[0122] The route planning scheme also includes comprehensive evaluation indicators for the entire delivery process. It calculates and outputs the total cost of the route planning scheme, covering various expenses during vehicle transportation, such as fuel consumption and vehicle maintenance costs. It assesses total carbon emissions, considering factors such as tanker mileage, fuel load, and carbon emission factors, reflecting the environmental impact of delivery activities. It analyzes total transportation risks, taking into account factors such as road conditions, accident impact range, and population density, providing a basis for risk management. It calculates average mileage utilization rate to evaluate the efficiency of vehicle resource utilization, in order to further optimize delivery strategies.

[0123] The route planning scheme, which includes complete information such as vehicle loading routes, delivery times and quantities, operational sequences, and comprehensive evaluation indicators, is output. This scheme can be presented to relevant delivery management and operational personnel in the form of electronic documents, database records, or visual maps. Delivery personnel conduct actual secondary delivery operations of refined oil products based on this scheme. During execution, appropriate adjustments can be made according to actual conditions, but the overall goal is to achieve efficient, safe, and low-cost delivery, ensuring that refined oil products can be smoothly delivered from oil depots to various gas stations to meet market demand, while optimizing the utilization of delivery resources and operational efficiency. For example, managers can use the output scheme to arrange the departure time of tanker trucks, dispatch vehicles between different oil depots, and monitor various indicators during the delivery process, promptly identifying and resolving potential problems to ensure the stable operation of the entire secondary refined oil product delivery system.

[0124] The following is a specific embodiment to illustrate the specific application of the method of the present invention.

[0125] This specific embodiment provides a method for optimizing large-scale secondary distribution of refined oil products. It studies the problem of optimizing secondary distribution routes in the refined oil product distribution process, aiming to minimize the total cost. To achieve this goal, the following two problems need to be solved:

[0126] The first step is vehicle allocation decision-making: based on the distance between gas stations and the preset maximum loading distance, gas station demand orders are merged and allocated to transport vehicles. At the same time, the carrying capacity of tanker trucks and the delivery time window limitations of each gas station need to be considered.

[0127] Secondly, there's the issue of customer segmentation: based on the inventory levels of each oil product at the oil depots and the demand at each gas station, the merged transport vehicles are allocated to each oil depot. Gas stations, after being merged into vehicles, can be considered virtual gas stations, and their allocation to appropriate oil depots needs to be based on their distribution characteristics relative to the oil depots.

[0128] Based on the aforementioned problems in the prior art, this specific embodiment proposes a two-stage optimization scheme for large-scale secondary distribution of refined oil products. The study focuses on the route planning problem for secondary distribution of refined oil products, and is described in detail below:

[0129] I. Problem Description

[0130] The problem scenario can be described as follows: taking several oil depots in the same area, several gas stations distributed in various locations, and three types of oil products as the objects, Figure 5 This is a specific example diagram of a refined oil product distribution process according to an embodiment of the present invention. The refined oil product distribution process is as follows: Figure 5 As shown.

[0131] Given a defined demand at gas stations, strict delivery time constraints, and the possibility of tanker trucks carrying additional cargo, a reasonable transportation plan for delivery vehicles should be developed. This plan aims to meet all demands without violating constraints, thereby minimizing total transportation costs.

[0132] The distribution and transportation of refined oil products is divided into two stages: primary distribution and secondary distribution. Primary logistics refers to the process of transporting refined oil products in large quantities but small batches from refineries to the storage depots of sales companies. The main modes of transportation include waterway, rail, pipeline, and road transport. Secondary logistics refers to the process of transporting refined oil products from transit depots in small quantities but large batches to gas stations or end users. The main mode of transportation is road transport, and this stage accounts for a significant proportion of the overall cost of refined oil product distribution.

[0133] The proactive delivery model of secondary distribution consists of multiple links, including the distribution center collecting and processing the inventory and sales information reported by gas stations, predicting the time when gas stations and oil storage tanks will be out of capacity and stockouts, generating replenishment needs, optimizing the delivery plan, matching transport vehicles, issuing oil pickup and delivery plans, and then the transport fleet organizing and implementing oil distribution.

[0134] This specific embodiment studies the problem of optimizing the secondary distribution route of refined oil products, that is, to formulate a reasonable transportation plan for delivery vehicles, meet all needs without violating constraints, and minimize the total transportation cost.

[0135] Given the tanker capacity, demand at each gas station, transportation distances between each oil depot and each gas station, number of tankers at each oil depot, inventory of various oil products at each oil depot, unloading time at each gas station, time of inability to accommodate oil at each gas station and time of oil shortage at each gas station, and the furthest loading distance.

[0136] The task is to develop a delivery vehicle transportation plan that minimizes total cost while satisfying constraints such as tanker capacity, inventory of each type of oil in each oil depot, total number of vehicles in the oil depot, distance between the furthest loading stations, flow balance, and time window for arrival at the gas station.

[0137] II. Problem description and assumptions are as follows:

[0138] 1) There are multiple oil depots in the area that provide delivery services to gas stations. Each oil depot has varying amounts of different types of oil available for storage each day, depending on its size.

[0139] 2) There are a large number of gas stations in the area, and each gas station has a certain amount of delivery demand for each type of oil every day.

[0140] 3) Each oil depot has several tank trucks for distribution and transportation. Each truck is divided into multiple oil storage compartments with limited capacity by partitions. Oil products in each compartment cannot be mixed.

[0141] 4) The total demand for all types of oil at each gas station at the same time shall not exceed the vehicle's carrying capacity; otherwise, the excess portion will be split into virtual gas stations.

[0142] 5) Gas stations have strict restrictions on delivery time; it cannot be earlier than the time when they stop accepting customers, nor later than the time when they stop accepting customers.

[0143] 6) Tanker trucks are allowed to carry multiple loads and deliver to multiple gas stations within the same trip, provided that the total delivery volume does not exceed the tanker truck's carrying capacity.

[0144] 7) The distance between gas stations where the load is distributed must not exceed the preset upper limit to ensure the economy of logistics and the full load rate and safety of transport vehicles.

[0145] 8) Each gas station can only be served by one vehicle from one oil depot at a time.

[0146] 9) All tanker trucks depart from the oil depot and return to the oil depot after completing the delivery.

[0147] III. The path planning model established in this specific embodiment is shown below:

[0148] 1. The mathematical model is described as follows:

[0149] 1) Parameters

[0150] M: Number of oil depots;

[0151] Q: Number of types of oil depots;

[0152] L: Tanker truck capacity;

[0153] K m The number of tank trucks at oil depot m;

[0154] W mq The inventory of oil product q at oil depot m;

[0155] N: Number of gas stations;

[0156] g iq The demand for fuel at gas stations (i: fuel type q).

[0157] t i Delivery execution time for gas station i;

[0158] ET i The non-capacity period of gas station i;

[0159] LT i The fuel cut-off time at gas station i;

[0160] ST i : Unloading time at gas station i;

[0161] d ij : The transportation distance between gas station i and oil depot j (i, j≤N represents a gas station, otherwise it is an oil depot);

[0162] MD: The furthest distance between gas stations where the gas station is distributed;

[0163] v: Vehicle speed;

[0164] C: Unit distance transportation cost.

[0165] 2) Decision variables

[0166] Let the gas stations be numbered 1, 2, ..., N, and the oil depots be numbered N+1, N+2, ..., N+M. Define the variables as follows.

[0167]

[0168] The resulting mathematical model is as follows:

[0169]

[0170] st

[0171]

[0172]

[0173] or

[0174] Equation (3) is the objective of minimizing total transportation costs.

[0175] Furthermore, without loss of generality, this specific embodiment simplifies some relatively fixed costs, such as the driving cost from the parking lot to the oil depot, the toll costs during the delivery process, and the loss costs of unloading, storing, and transporting oil, etc. Only the driving costs of the most important and commonly used delivery routes are considered.

[0176] Equations (4) to (7) are respectively the tanker capacity constraint, the oil depot inventory constraint, the total number of oil depot vehicles constraint, and the distance constraint between the farthest loading stations.

[0177] Equations (8), (9) and (10) indicate that each station can only be provided with one delivery service by one vehicle from one oil depot.

[0178] Equation (11) indicates that the vehicle departs from the oil depot and eventually returns to the oil depot.

[0179] Equation (12) indicates that vehicles cannot travel from one oil depot to another.

[0180] Equation (13) indicates that after the vehicle arrives at each station, it immediately begins the oil unloading operation without waiting. After a certain service time, it leaves and drives to the next station.

[0181] Equation (14) limits the time for the vehicle to arrive at the gas station to be within a preset time window, which is a path planning problem with a hard time window.

[0182] Equations (15) and (16) are constraints on the values ​​of variables.

[0183] It should be noted that:

[0184] This specific embodiment adds a new additional objective function to the objective function of minimizing the total transportation cost shown in equation (3), thereby transforming the multi-objective problem into a single-objective problem through normalization and weighted summation.

[0185]

[0186] The original objective function, with the first term as: Total transportation cost:

[0187]

[0188] The possible objective functions are as follows:

[0189] Customer satisfaction:

[0190]

[0191] Among them: [EET] i LLT i [ET] represents the maximum tolerance time window for gas station i; i ,LT i [This refers to the satisfaction time window for gas station i.] Remove the time constraint from the constraints: ET i ≤t i ≤LT i ,

[0192] Carbon emissions:

[0193]

[0194] in: Let m be the carbon emissions of vehicle k traveling from gas station i to gas station j. Let L be the amount of oil carried by vehicle k from point i to point j in oil depot m, L be the capacity of the oil tanker, and B be the carbon emission factor of the oil tanker.

[0195] Transportation risks:

[0196]

[0197]

[0198] in: Let d be the road risk value for vehicle k from gas station or oil depot m, traveling from point i to point j. ij Let j be the distance from point i at the gas station or oil depot to point j. Let r be the average population density within the accident's impact zone from point i to point j, and r be the accident's impact zone. Let W be the amount of oil carried by vehicle k in oil depot m from point i to point j. ij Let be the road resilience value from point i to point j.

[0199] Mileage utilization rate:

[0200]

[0201] Where d is the vehicle travel distance, d mi Let m be the distance from the oil depot to the gas station i.

[0202] IV. The solution process of the path planning model in this specific embodiment is as follows:

[0203] 1. Density-based hierarchical clustering

[0204] Used to merge gas station demand orders and allocate transportation vehicles.

[0205] The process is as follows: First, a tanker truck is assigned to each gas station, and the distance between the vehicles is the distance between the gas stations.

[0206] Then, the density of each vehicle is calculated, and the approximately 10% with the lowest density is taken as the deviation point. The remaining vehicles are then merged into new vehicles.

[0207] To address the issue of refined oil product distribution, merging conditions are set to handle model constraints, and vehicle loading routes are generated simultaneously during the merging process.

[0208] First, calculate the vehicle density based on the distance between vehicles:

[0209]

[0210] Where ρ i The density of vehicle i is calculated by the number of other vehicles adjacent to vehicle i. d ij Let d be the distance between vehicle i and vehicle j. c The cutoff distance is set. Approximately 10% of the vehicles with the lowest density are used as the deviation point, and the remaining vehicles participate in the merging.

[0211] Secondly, two vehicles are selected as candidate vehicles for merging based on the vehicle density and distance matrix.

[0212] Prioritize selecting the vehicle with the highest density and merging it with the nearest vehicle.

[0213] The distance between vehicles participating in the merging process is less than the preset maximum load-sharing distance. This constraint serves both as a preliminary screening of vehicles to be merged and as a termination condition for the algorithm, eliminating the need to merge all stations and reducing the computational load.

[0214] Furthermore, all gas stations involved in the merger of the two vehicles must meet three constraints:

[0215] 1) The total demand of all gas stations should be less than or equal to the carrying capacity of the tanker trucks;

[0216] 2) There exists a delivery order such that the transportation distance between any two adjacent stations is less than or equal to the farthest loading distance;

[0217] 3) Following this delivery order, vehicles can arrive at each gas station within its delivery time window. Finally, after each vehicle merge, the distance between the new vehicle and other vehicles is calculated using the shortest distance method.

[0218] An enumeration method is used to check if there is an access order that satisfies the above constraints, to determine the feasibility of site merging, and to record the loading paths for future use.

[0219] A density-based hierarchical clustering algorithm aims to merge and allocate delivery vehicles for gas station demand orders. The process begins by assigning one tanker truck to each gas station, at which point the distance between vehicles is equal to the distance between gas stations. Next, the density of each vehicle is calculated, and the approximately 10% of vehicles with the lowest density are designated as decoupling points. The remaining vehicles are used in subsequent merging operations to form new vehicles. Throughout the process, to accommodate the refined oil delivery problem, a series of merging conditions are set to handle model constraints, and vehicle loading routes are generated simultaneously.

[0220] Vehicle density is calculated based on the distance between vehicles. After calculation, vehicles that deviate from the density are filtered out, and the remaining vehicles are selected for merging. Based on the vehicle density and distance matrix, the vehicle with the highest density and the closest vehicle are selected as merging candidates, and the distance between vehicles participating in the merging must be less than the preset maximum load distribution distance. This distance constraint can filter vehicles and serve as the algorithm termination condition.

[0221] All gas stations involved in the merging of two vehicles must meet three constraints: First, the total demand must not exceed the tanker truck's carrying capacity; second, a reasonable delivery order must exist so that the transportation distance between adjacent stations does not exceed the maximum loading distance; and third, vehicles can arrive at each gas station within their respective delivery time windows according to this delivery order. After each vehicle merge, the shortest distance method is used to calculate the distance between the new vehicle and other vehicles. Finally, an enumeration method is used to check if there is an access order that satisfies the constraints, determine the feasibility of the station merge, and record the loading path for subsequent use.

[0222] The density-based hierarchical clustering in this specific embodiment has the following advantages compared to traditional systematic clustering algorithms:

[0223] Hierarchical clustering, when merging vehicles, does not fully consider the distribution density of the vehicles' surroundings, which may lead to the selection of suboptimal combinations. Density-based hierarchical clustering, by calculating vehicle density, can more accurately identify which vehicles are geographically more closely distributed and more suitable for merging. For example, in real-world delivery scenarios, areas with high vehicle density may indicate concentrated demand at gas stations; prioritizing the merging of these vehicles can improve delivery efficiency and avoid the indiscriminate merging that may occur with hierarchical clustering.

[0224] Excluding the approximately 10% of vehicles with the lowest density from the initial merging scope helps reduce the negative impact of isolated or abnormally distributed vehicles on the overall merging results. Systematic clustering lacks such a screening mechanism, which may introduce unreasonable combinations during the merging process, affecting subsequent delivery route planning.

[0225] In systematic clustering, the matching degree between the carrying capacity of merged vehicles and the total demand of gas stations is not refined enough. Density-based hierarchical clustering emphasizes that the total demand of all gas stations must be less than or equal to the carrying capacity of tanker trucks, ensuring that each merged vehicle can effectively meet the demand in actual delivery, avoiding overloading or wasted capacity.

[0226] Systematic clustering lacks specific strategies for generating delivery sequences and satisfying delivery time windows. Density-based hierarchical clustering explicitly requires a suitable delivery sequence, ensuring that both the distance between adjacent stations and the delivery time window are satisfied. This makes the delivery plan more aligned with actual operational needs, improves the timeliness and reliability of delivery, and reduces the increase in operating costs caused by time window conflicts.

[0227] Systematic clustering may require trying and evaluating more vehicle combinations, resulting in high computational complexity. Density-based hierarchical clustering, using the furthest loading distance constraint as a screening and termination condition, can quickly narrow down the merging range, reduce unnecessary computation, significantly improve algorithm efficiency, and is more suitable for complex scenarios of large-scale refined oil delivery.

[0228] Density-based hierarchical clustering offers greater flexibility in handling diverse gas station distributions and demand patterns. It can dynamically adjust merging strategies based on the actual density distribution, while hierarchical clustering's relatively fixed merging method is less adaptable to diverse delivery scenarios. For example, in areas with uneven gas station distribution or significant demand variations, density-based hierarchical clustering is better positioned to generate more rational delivery routes.

[0229] 2. Standard deviation saves mileage

[0230] Standard deviation mileage is used to replace the simple distance between depots and stations, providing a more comprehensive and accurate description of the distribution characteristics of each gas station relative to the oil depot.

[0231] The formula for calculating the mileage saved by standard deviation is:

[0232]

[0233] Where sgn(x) is the sign function. In the genetic algorithm, the standard deviation mileage is converted into a probability value for use in the calculation of the genetic operator. Consider three typical application scenarios.

[0234] Scenario 1: Calculate the priority allocation probability for gas stations that have not yet been assigned. First, generate subsets V° and Δ° as follows.

[0235]

[0236] This was then standardized to the selection probability value ρ. i , ρ i >0 and ∑ i∈V . ρ i =1.

[0237] If max(△°) = min(△°), then

[0238]

[0239] otherwise,

[0240]

[0241] In equation (20), |△°| represents the number of elements in set △°. If max(△°) > min(△°), then max(ρ) / min(ρ) = |△°| + 1. The exponent 0 ≤ α ≤ ∞ is used to adjust the selection probability. The larger the value of α, the greater the difference in probability; as α → 0, then ρ i →1 / |△°| When i∈V°; α→∞, then ∑ρ i →1 when Other ρ j →0 when

[0242] Scenario 2: Calculate the priority removal probability of all gas stations in a certain partitioning scheme, using the following formula.

[0243]

[0244] Subsequently, it is standardized to a selection probability value using equations (1 9) and (20).

[0245] Scenario 3: Calculate the probability of allocating gas station i to a certain oil depot in a limited number of ways, using the following formula.

[0246] △ i ={δ i =δ im |1<m <M} (22)

[0247] Subsequently, it is standardized to a selection probability value using equations (19) and (20).

[0248] The standard deviation mileage saving in this specific embodiment has the following advantages compared to the expected mileage saving in traditional technologies:

[0249] 1. Expected mileage savings primarily focus on average savings, while standard deviation mileage savings not only incorporates average savings but also considers the fluctuations in the distance relationships between fuel depots and gas stations by calculating the standard deviation. In actual delivery, the distribution of gas stations relative to fuel depots is not uniform or stable, and may exhibit local concentration or dispersion. Standard deviation mileage savings better captures this dispersion. For example, when the distances between some gas stations and different fuel depots vary significantly, standard deviation mileage savings can more accurately reflect the stability and uncertainty of mileage savings when allocating vehicles to different fuel depots, providing more comprehensive information for decision-making.

[0250] 2. In complex distribution networks, such as those with multiple oil depots and widely and unevenly distributed gas stations, standard deviation mileage savings can more precisely characterize the potential changes in each gas station under different allocation schemes. In contrast, expected mileage savings may mask some local special cases, leading to suboptimal allocation schemes in certain situations. Standard deviation mileage savings helps to more accurately identify situations where the average mileage savings seem similar but the actual distribution characteristics differ, thus enabling the development of distribution strategies that better meet actual needs.

[0251] 3. When generating the initial population, the priority probability calculated based on the standard deviation of mileage savings can more effectively reflect the diversity and rationality of vehicle allocation. By considering the standard deviation, vehicles with significant fluctuations in mileage savings under different fuel depot allocations can be assigned a more appropriate selection probability based on their specific circumstances, avoiding over-concentration or unreasonable selection tendencies that might result from simply basing decisions on expected mileage savings. This ensures a higher level of diversity and quality in the initial population, providing a better starting point for subsequent iterative optimization of the genetic algorithm.

[0252] 4. In crossover and mutation operations, the richer information provided by standard deviation mileage saving helps to more accurately determine the operation probability of loci. For example, when calculating the probability of priority removal and redistribution of loci, the introduction of standard deviation allows the algorithm to better weigh the advantages and disadvantages of different vehicles under different allocation conditions, avoiding getting trapped in local optima. By guiding genetic operations more rationally, standard deviation mileage saving can improve the algorithm's search ability and convergence speed in complex delivery environments, find better vehicle allocation schemes faster, and thus achieve more efficient secondary delivery route planning for refined oil products.

[0253] 3. Improved Genetic Algorithm

[0254] Used to allocate the merged vehicles to oil depots in order to construct a complete logistics distribution route.

[0255] 1) Encoding:

[0256] One-dimensional integer encoding is used, with the encoding length being the number of vehicles. The gene locus number in the encoding represents the vehicle's number, and the value represents the assigned oil depot number.

[0257] 2) Initialize the operator:

[0258] Step 1: All stations are unassigned, and the remaining inventory at each oil depot is [missing information]. Calculate the standard deviation of the mileage saved [δ] im ] N×M .

[0259] Step 2: Using equations (18) to (20), calculate the priority probability of unassigned stations by the standard deviation of the mileage saved. Then, randomly select a gas station i according to the probability.

[0260] Step 3: Find the satisfying For the oil depot, the probabilities are calculated using equations (22), (19), and (20). Gas station i is randomly assigned to oil depot m according to these probabilities, and the remaining inventory is updated.

[0261] Step 4: If all sites have been partitioned, output the results; otherwise, go to step 2.

[0262] 3) Crossover operator

[0263] Step 1: Calculate the priority removal probability of each locus in the two individuals using equations (19) to (21).

[0264] Step 2: Randomly select several loci from individual i according to probability. Fill them with alleles from j to generate offspring individual i.

[0265] Step 3: Perform the same operation as in step 2 on individual j to generate offspring individual j.

[0266] Step 4: Apply the correction operator to the offspring individuals.

[0267] 4) Mutation operator

[0268] Step 1: Calculate the probability of preferential removal of the locus using equations (19) to (21).

[0269] Step 2: Randomly select a small number of loci i from individual 1 according to probability, calculate the probability using equations (22), (19) and (20), and allocate gas station i to oil depot m according to probability.

[0270] Step 3: Apply the correction operator to the offspring individuals.

[0271] 5) Correction operator: A greedy algorithm is used to correct all new offspring, transforming newly generated infeasible solutions into feasible solutions, while also gradually improving the convergence speed of the algorithm.

[0272] Step 1: Calculate the remaining inventory of all oil products in all oil depots.

[0273] Step 2: If min(W) L If )>0, the algorithm terminates and outputs the result; otherwise, any ) is selected.

[0274] Step 3: Select all stations {i|g} from the delivery list of oil depot m′. iq′ >0}. Calculate the distance from station i to other oil depots. distance {s im}

[0275] Step 4: According to (s) im In ascending order, assign gas station i to the nearest gas depot m in turn, until... Proceed to step 2.

[0276] 6) Fitness calculation: The fitness of individuals is evaluated using an adjusted objective function. With the goal of minimizing transportation costs, the merged transport vehicles are allocated to each oil depot.

[0277]

[0278] 7) Selection: A sorting selection algorithm is used to calculate the survival probability of individuals, and a roulette wheel algorithm is used to construct the new population. Furthermore, this specific embodiment also employs an elite retention strategy to ensure that the current optimal solution is not lost during the random selection process.

[0279] Of course, it is understood that there may be other variations of the above detailed process, and all such variations should fall within the protection scope of this invention.

[0280] In this embodiment of the invention, a route planning model is established with minimizing total transportation cost as the objective function and customer satisfaction, carbon emissions, transportation risk, and mileage utilization rate as additional objective functions. The route planning model includes parameters related to oil depots, gas stations, tank trucks, and the tank trucks themselves. A density-based hierarchical clustering algorithm is used to calculate the vehicle density of tank trucks allocated to different gas stations based on the route planning model. Based on pre-set vehicle merging constraints, tank trucks from different gas stations are merged according to the vehicle density to obtain merged vehicles. The vehicle merging constraints include that the distance between merged tank trucks is less than or equal to a pre-set maximum loading distance between tank trucks. These constraints constrain the tank truck carrying capacity, delivery order, delivery time window, and distance between tank trucks involved in the merging process. Based on the merged vehicles... The process involves determining the vehicle loading routes for refined oil between gas stations; using a genetic algorithm based on a path planning model to establish vehicle codes corresponding to the merged vehicles; the locus number of the vehicle code represents the number of the merged vehicle; the value of the vehicle code represents the assigned oil depot number; generating an initial population based on the calculated standard deviation of the mileage saved by different merged vehicles; the standard deviation of the mileage saved represents the standard deviation of the mileage saved by assigning different merged vehicles to the target oil depot relative to assigning them to other oil depots; each individual in the initial population represents a different vehicle code assigned to the initial oil depot; determining the vehicle loading routes for refined oil between different oil depots through crossover and mutation operations on the initial population; and outputting the vehicle loading routes for refined oil between gas stations and the merged vehicle loading routes for refined oil between different oil depots as a path planning scheme for refined oil delivery.The route planning model constructed in this invention uses minimizing total transportation cost as the objective function, and incorporates customer satisfaction, carbon emissions, transportation risk, and mileage utilization as additional objective functions. It is no longer limited to the traditional single optimization that only considers transportation cost, but can maximize the overall benefits of secondary oil delivery in terms of economy, service, environment, and operational efficiency, overcoming the lack of comprehensive optimization in existing technologies. The density-based hierarchical clustering algorithm is used for vehicle merging, taking into account that the distance between tanker trucks is less than or equal to the farthest loading distance, as well as constraints on carrying capacity, delivery order, and delivery time windows. This helps to accurately combine vehicles according to the actual distribution and demand of gas stations, ensuring that the merged vehicles meet carrying capacity requirements and complete delivery tasks within the specified delivery time window under a reasonable delivery order. This improves the rationality and scientific nature of vehicle allocation, representing a significant improvement compared to existing vehicle allocation methods that do not fully incorporate multiple constraints. The density-based hierarchical clustering algorithm filters out deviation points by calculating vehicle density. The algorithm prioritizes merging vehicles with higher density and shorter distances, using the furthest loading distance as the termination condition. This avoids indiscriminate merging of all stations, effectively reducing unnecessary computational steps, lowering the computational load, and improving efficiency when handling large-scale data. It overcomes the slow computation speed of traditional algorithms when dealing with large-scale delivery planning. Furthermore, the algorithm incorporates standard deviation mileage saving and designs corresponding genetic operators. This standard deviation mileage saving guides the algorithm to find better customer segmentation schemes, avoiding getting trapped in local optima, thus improving the optimization quality and enhancing its adaptability to complex constraints and large-scale delivery route planning problems. It also addresses the weakness of traditional algorithms in getting trapped in local optima when dealing with such problems. Finally, the route planning for refined oil delivery is decomposed into two sub-problems: merging travel distance and vehicle segmentation. These are solved sequentially using a system clustering algorithm and an improved genetic algorithm, achieving optimized planning of large-scale refined oil secondary delivery routes. Under various constraints, this reduces transportation costs, improves delivery efficiency, and enhances overall operational benefits.

[0281] This invention also provides a route planning device for refined oil product distribution, as described in the following embodiments. Since the principle by which this device solves the problem is similar to the route planning method for refined oil product distribution, the implementation of this device can refer to the implementation of the route planning method for refined oil product distribution; repeated details will not be elaborated further.

[0282] This invention also provides a route planning device for refined oil product distribution, which improves the efficiency and accuracy of route planning for refined oil product distribution, reduces the transportation costs of secondary distribution of refined oil products, improves distribution efficiency, and enhances overall operational benefits. Figure 6 This is a schematic diagram of the structure of a route planning device for finished oil delivery according to an embodiment of the present invention, as shown below. Figure 6 As shown, the device includes:

[0283] The route planning model modeling module 601 is used to establish a route planning model with minimizing total transportation cost as the objective function and customer satisfaction, carbon emissions, transportation risk, and mileage utilization rate as additional objective functions; the route planning model includes oil depots, gas stations, tank trucks, and related parameters of tank trucks;

[0284] The vehicle loading route determination module 602 is used to calculate the vehicle density of tanker trucks allocated to different gas stations based on the path planning model using a density-based hierarchical clustering algorithm; based on preset vehicle merging constraints, it merges tanker trucks from different gas stations according to the vehicle density to obtain merged vehicles; the vehicle merging constraints include that the distance between the merged tanker trucks is less than or equal to a preset maximum loading distance between tanker trucks; the vehicle merging constraints are used to constrain the tanker truck carrying capacity, delivery order, delivery time window, and distance between tanker trucks; and based on the merged vehicles, it determines the vehicle loading route for refined oil between gas stations.

[0285] The merged vehicle loading route determination module 603 is used to establish vehicle codes corresponding to the merged vehicles based on a path planning model using a genetic algorithm. The locus number of the vehicle code represents the number of the merged vehicle. The value of the vehicle code represents the assigned oil depot number. An initial population is generated based on the calculated standard deviation of the mileage saved by different merged vehicles. The standard deviation of the mileage saved is used to characterize the standard deviation of the mileage saved by assigning different merged vehicles to the target oil depot relative to assigning them to other oil depots. Each individual in the initial population represents a different vehicle code assigned to the initial oil depot. By performing crossover and mutation operations on the initial population, the merged vehicle loading route for refined oil between different oil depots is determined.

[0286] The route planning output module 604 is used to output the vehicle loading routes of refined oil between gas stations and the combined vehicle loading routes of refined oil between different oil depots as route planning schemes for refined oil distribution.

[0287] Figure 7 This is a specific example diagram of a route planning device for refined oil delivery according to an embodiment of the present invention. In one embodiment, such as... Figure 7 As shown, it also includes:

[0288] The objective function calculation module 701 is used to calculate the objective function as follows: calculate the transportation cost of each tanker truck from the oil depot to each gas station, including fuel cost, vehicle maintenance cost and driver labor cost; and add up the transportation costs of all tanker trucks to obtain the total transportation cost.

[0289] In one embodiment, based on pre-set vehicle merging constraints, tanker trucks from different gas stations are merged according to the vehicle density to obtain merged vehicles, including:

[0290] Based on the vehicle density of different tank trucks and the distance matrix between different tank trucks, two tank trucks are randomly selected as candidate tank trucks for merging; the tank truck with the highest density among the candidate tank trucks is selected and merged with the tank truck with the closest distance; wherein, the distance between the tank trucks being merged is less than or equal to the preset maximum loading distance of the tank trucks.

[0291] Determine whether the tank trucks to be merged meet the merging constraints for each vehicle; if they do, then merge the candidate tank trucks.

[0292] An enumeration method is used to detect the merged tank trucks to obtain the merged vehicle.

[0293] In one embodiment, an initial population is generated based on the calculated mileage savings from the standard deviation of different merged vehicles, including:

[0294] Calculate the standard deviation of mileage saved for different merged vehicles;

[0295] The priority probability of different merged vehicles is calculated based on the standard deviation of the mileage saved; the priority probability includes the priority allocation probability of merged vehicles that have not been assigned to a fuel depot, the priority removal probability of each merged vehicle, and / or the probability of merged vehicles being assigned to different fuel depots.

[0296] Randomly select merged vehicles according to priority probability; randomly select an oil depot from oil depots whose remaining inventory is greater than the oil demand of the randomly selected merged vehicle, and assign it to the randomly selected merged vehicle; and update the remaining inventory of the randomly selected oil depot; repeat the above steps of randomly selecting merged vehicles and randomly selecting oil depots until each merged vehicle is assigned to a different oil depot.

[0297] The merged vehicles and their corresponding assigned vehicle codes from the oil depots are used as the initial population.

[0298] In one embodiment, the vehicle loading route for refined oil products after merging across different oil depots is determined by performing crossover and mutation operations on the initial population, including:

[0299] Based on pre-defined crossover, mutation, and correction operators, crossover and mutation operations are performed on the initial population to obtain the post-operation population.

[0300] The above-described steps of crossover and mutation are repeated with the post-operation population replacing the initial population until the maximum value of the objective function corresponding to the post-operation population is less than a preset value and the value of the additional objective function is at the corresponding preset threshold.

[0301] Based on the post-operation population where the maximum value of the objective function is less than a preset value and the value of the additional objective function is at the corresponding preset threshold, the vehicle loading path for refined oil is determined after merging between different oil depots.

[0302] In one embodiment, the route planning scheme for refined oil product distribution further includes:

[0303] The routes of each tanker truck between gas stations, the delivery time and the amount of refined oil delivered at each gas station, the loading and unloading sequence of each tanker truck, the total cost of the route planning scheme, the total carbon emissions, the total transportation risk, and the average mileage utilization rate.

[0304] This invention provides an embodiment of a computer device for implementing all or part of the above-described route planning method for refined oil product distribution. The computer device specifically includes the following components:

[0305] The computer device comprises a processor, memory, a communications interface, and a bus; wherein the processor, memory, and communications interface communicate with each other via the bus; the communications interface is used to realize information transmission between related devices; the computer device can be a desktop computer, tablet computer, or mobile terminal, etc., and this embodiment is not limited to these. In this embodiment, the computer device can be implemented with reference to the embodiments of the route planning method for realizing refined oil distribution and the route planning device for realizing refined oil distribution, the contents of which are incorporated herein by reference, and repeated details will not be described again.

[0306] Figure 8 This is a schematic block diagram illustrating the system configuration of the computer device 1000 according to an embodiment of this application. Figure 8 As shown, the computer device 1000 may include a central processing unit 1001 and a memory 1002; the memory 1002 is coupled to the central processing unit 1001. It is worth noting that... Figure 8 This is an example; other types of structures can also be used to supplement or replace this structure to achieve telecommunications functions or other functions.

[0307] In one embodiment, the route planning function for refined oil product delivery can be integrated into the central processing unit 1001. The central processing unit 1001 can be configured to perform the following control:

[0308] A route planning model is established with the goal of minimizing total transportation cost, and additional objective functions of customer satisfaction, carbon emissions, transportation risk, and mileage utilization rate. The route planning model includes parameters related to oil depots, gas stations, tank trucks, and tank trucks.

[0309] A density-based hierarchical clustering algorithm is used to calculate the vehicle density of tanker trucks allocated to different gas stations based on the path planning model. Based on pre-set vehicle merging constraints, tanker trucks from different gas stations are merged according to the vehicle density to obtain merged vehicles. The vehicle merging constraints include that the distance between merged tanker trucks is less than or equal to a pre-set maximum loading distance between tanker trucks. These constraints constrain the tanker trucks' carrying capacity, delivery order, delivery time window, and distance between them. Based on the merged vehicles, the vehicle loading path for refined oil products between gas stations is determined.

[0310] A genetic algorithm, based on a path planning model, is used to establish vehicle codes corresponding to the merged vehicles. The locus number of the vehicle code represents the number of the merged vehicle. The value of the vehicle code represents the assigned oil depot number. An initial population is generated based on the calculated standard deviation of the mileage saved by different merged vehicles. The standard deviation of the mileage saved represents the standard deviation of the mileage saved by assigning different merged vehicles to the target oil depot relative to assigning them to other oil depots. Each individual in the initial population represents a different vehicle code assigned to the initial oil depot. By performing crossover and mutation operations on the initial population, the merging vehicle loading path for refined oil between different oil depots is determined.

[0311] The route planning scheme for refined oil distribution is output, which includes the vehicle loading routes between gas stations and the combined vehicle loading routes between different oil depots.

[0312] In another embodiment, the route planning device for refined oil delivery can be configured separately from the central processing unit 1001. For example, the route planning device for refined oil delivery can be configured as a chip connected to the central processing unit 1001, and the route planning function for refined oil delivery can be realized through the control of the central processing unit.

[0313] like Figure 8 As shown, the computer device 1000 may further include: a communication module 1003, an input unit 1004, an audio processor 1005, a display 1006, and a power supply 1007. It is worth noting that the computer device 1000 does not necessarily need to include... Figure 8 All components shown; in addition, the computer device 1000 may also include Figure 8 For components not shown, please refer to existing technologies.

[0314] like Figure 8 As shown, the central processing unit 1001, sometimes also referred to as a controller or operation control, may include a microprocessor or other processor device and / or logic device. The central processing unit 1001 receives input and controls the operation of various components of the computer device 1000.

[0315] The memory 1002 may be, for example, one or more of a cache, flash memory, hard drive, removable medium, volatile memory, non-volatile memory, or other suitable device. It can store the aforementioned device-related information, and may also store programs for executing that information. The central processing unit 1001 can execute the program stored in the memory 1002 to perform information storage or processing, etc.

[0316] Input unit 1004 provides input to central processing unit 1001. This input unit 1004 may be, for example, a keypad or touch input device. Power supply 1007 provides power to computer device 1000. Display 1006 displays images, text, and other display objects. This display may be, for example, an LCD display, but is not limited to this.

[0317] The memory 1002 can be a solid-state memory, such as a read-only memory (ROM), random access memory (RAM), a SIM card, etc. It can also be a memory that retains information even when power is off, can be selectively erased, and contains more data; examples of this type of memory are sometimes referred to as EPROMs, etc. The memory 1002 can also be some other type of device. The memory 1002 includes a buffer memory 1021 (sometimes referred to as a buffer). The memory 1002 may include an application / function storage unit 1022 for storing application programs and function programs or processes for executing operations of the computer device 1000 via the central processing unit 1001.

[0318] The memory 1002 may also include a data storage unit 1023 for storing data, such as contacts, digital data, pictures, sounds, and / or any other data used by the computer device. The driver storage unit 1024 of the memory 1002 may include various drivers for the computer device for communication functions and / or for performing other functions of the computer device (such as messaging applications, address book applications, etc.).

[0319] The communication module 1003 is a transmitter / receiver that transmits and receives signals via the antenna 1008. The communication module (transmitter / receiver) 1003 is coupled to the central processing unit 1001 to provide input signals and receive output signals, which is the same as in a conventional mobile communication terminal.

[0320] Based on different communication technologies, multiple communication modules 1003 can be configured in the same computer device, such as cellular network modules, Bluetooth modules, and / or wireless LAN modules. The communication module (transmitter / receiver) 1003 is also coupled to a speaker 1009 and a microphone 1010 via an audio processor 1005 to provide audio output via the speaker 1009 and receive audio input from the microphone 1010, thereby realizing typical telecommunications functions. The audio processor 1005 may include any suitable buffer, decoder, amplifier, etc. Furthermore, the audio processor 1005 is also coupled to a central processing unit 1001, enabling on-device recording via the microphone 1010 and on-device playback of stored sound via the speaker 1009.

[0321] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described route planning method for refined oil product distribution.

[0322] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the aforementioned route planning method for refined oil product distribution.

[0323] In this embodiment of the invention, a route planning model is established with minimizing total transportation cost as the objective function and customer satisfaction, carbon emissions, transportation risk, and mileage utilization rate as additional objective functions. The route planning model includes parameters related to oil depots, gas stations, tank trucks, and the tank trucks themselves. A density-based hierarchical clustering algorithm is used to calculate the vehicle density of tank trucks allocated to different gas stations based on the route planning model. Based on pre-set vehicle merging constraints, tank trucks from different gas stations are merged according to the vehicle density to obtain merged vehicles. The vehicle merging constraints include that the distance between merged tank trucks is less than or equal to a pre-set maximum loading distance between tank trucks. These constraints constrain the tank truck carrying capacity, delivery order, delivery time window, and distance between tank trucks involved in the merging process. Based on the merged vehicles... The process involves determining the vehicle loading routes for refined oil between gas stations; using a genetic algorithm based on a path planning model to establish vehicle codes corresponding to the merged vehicles; the locus number of the vehicle code represents the number of the merged vehicle; the value of the vehicle code represents the assigned oil depot number; generating an initial population based on the calculated standard deviation of the mileage saved by different merged vehicles; the standard deviation of the mileage saved represents the standard deviation of the mileage saved by assigning different merged vehicles to the target oil depot relative to assigning them to other oil depots; each individual in the initial population represents a different vehicle code assigned to the initial oil depot; determining the vehicle loading routes for refined oil between different oil depots through crossover and mutation operations on the initial population; and outputting the vehicle loading routes for refined oil between gas stations and the merged vehicle loading routes for refined oil between different oil depots as a path planning scheme for refined oil delivery.The route planning model constructed in this invention uses minimizing total transportation cost as the objective function, and incorporates customer satisfaction, carbon emissions, transportation risk, and mileage utilization as additional objective functions. It is no longer limited to the traditional single optimization that only considers transportation cost, but can maximize the overall benefits of secondary oil delivery in terms of economy, service, environment, and operational efficiency, overcoming the lack of comprehensive optimization in existing technologies. The density-based hierarchical clustering algorithm is used for vehicle merging, taking into account that the distance between tanker trucks is less than or equal to the farthest loading distance, as well as constraints on carrying capacity, delivery order, and delivery time windows. This helps to accurately combine vehicles according to the actual distribution and demand of gas stations, ensuring that the merged vehicles meet carrying capacity requirements and complete delivery tasks within the specified delivery time window under a reasonable delivery order. This improves the rationality and scientific nature of vehicle allocation, representing a significant improvement compared to existing vehicle allocation methods that do not fully incorporate multiple constraints. The density-based hierarchical clustering algorithm filters out deviation points by calculating vehicle density. The algorithm prioritizes merging vehicles with higher density and shorter distances, using the furthest loading distance as the termination condition. This avoids indiscriminate merging of all stations, effectively reducing unnecessary computational steps, lowering the computational load, and improving efficiency when handling large-scale data. It overcomes the slow computation speed of traditional algorithms when dealing with large-scale delivery planning. Furthermore, the algorithm incorporates standard deviation mileage saving and designs corresponding genetic operators. This standard deviation mileage saving guides the algorithm to find better customer segmentation schemes, avoiding getting trapped in local optima, thus improving the optimization quality and enhancing its adaptability to complex constraints and large-scale delivery route planning problems. It also addresses the weakness of traditional algorithms in getting trapped in local optima when dealing with such problems. Finally, the route planning for refined oil delivery is decomposed into two sub-problems: merging travel distance and vehicle segmentation. These are solved sequentially using a system clustering algorithm and an improved genetic algorithm, achieving optimized planning of large-scale refined oil secondary delivery routes. Under various constraints, this reduces transportation costs, improves delivery efficiency, and enhances overall operational benefits.

[0324] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0325] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0326] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0327] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0328] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A route planning method for refined oil product distribution, characterized in that, include: A route planning model is established with the goal of minimizing total transportation cost, and additional objective functions of customer satisfaction, carbon emissions, transportation risk, and mileage utilization rate. The route planning model includes parameters related to oil depots, gas stations, tank trucks, and tank trucks. A density-based hierarchical clustering algorithm is used to calculate the vehicle density of tanker trucks allocated to different gas stations based on the path planning model. Based on pre-set vehicle merging constraints, tanker trucks from different gas stations are merged according to the vehicle density to obtain merged vehicles. The vehicle merging constraints include that the distance between merged tanker trucks is less than or equal to a pre-set maximum loading distance between tanker trucks. These constraints constrain the tanker trucks' carrying capacity, delivery order, delivery time window, and distance between them. Based on the merged vehicles, the vehicle loading path for refined oil products between gas stations is determined. A genetic algorithm, based on a path planning model, is used to establish vehicle codes corresponding to the merged vehicles. The locus number of the vehicle code represents the number of the merged vehicle. The value of the vehicle code represents the assigned oil depot number. An initial population is generated based on the calculated standard deviation of the mileage saved by different merged vehicles. The standard deviation of the mileage saved represents the standard deviation of the mileage saved by assigning different merged vehicles to the target oil depot relative to assigning them to other oil depots. Each individual in the initial population represents a different vehicle code assigned to the initial oil depot. By performing crossover and mutation operations on the initial population, the merging vehicle loading path for refined oil between different oil depots is determined. The route planning scheme for refined oil distribution is output, which includes the vehicle loading routes between gas stations and the combined vehicle loading routes between different oil depots.

2. The method as described in claim 1, characterized in that, It also includes: calculating the objective function as follows: Calculate the transportation cost of each tanker truck from the oil depot to each gas station, including fuel cost, vehicle maintenance cost, and driver labor cost; add up the transportation costs of all tanker trucks to obtain the total transportation cost.

3. The method as described in claim 1, characterized in that, Based on pre-set vehicle merging constraints, tanker trucks from different gas stations are merged according to the vehicle density to obtain merged vehicles, including: Based on the vehicle density of different tank trucks and the distance matrix between different tank trucks, two tank trucks are randomly selected as candidate tank trucks for merging; the tank truck with the highest density among the candidate tank trucks is selected and merged with the tank truck with the closest distance; wherein, the distance between the tank trucks being merged is less than or equal to the preset maximum loading distance of the tank trucks. Determine whether the tank trucks to be merged meet the merging constraints for each vehicle; if they do, then merge the candidate tank trucks. An enumeration method is used to detect the merged tank trucks to obtain the merged vehicle.

4. The method as described in claim 1, characterized in that, Based on the calculated mileage savings from the standard deviation of the merged vehicles, an initial population is generated, including: Calculate the standard deviation of mileage saved for different merged vehicles; The priority probability of different merged vehicles is calculated based on the standard deviation of the mileage saved; the priority probability includes the priority allocation probability of merged vehicles that have not been assigned to a fuel depot, the priority removal probability of each merged vehicle, and / or the probability of merged vehicles being assigned to different fuel depots. Randomly select merged vehicles according to priority probability; randomly select an oil depot from oil depots whose remaining inventory is greater than the oil demand of the randomly selected merged vehicle, and assign it to the randomly selected merged vehicle; and update the remaining inventory of the randomly selected oil depot; repeat the above steps of randomly selecting merged vehicles and randomly selecting oil depots until each merged vehicle is assigned to a different oil depot. The merged vehicles and their corresponding assigned vehicle codes from the oil depots are used as the initial population.

5. The method as described in claim 1, characterized in that, By performing crossover and mutation operations on the initial population, the vehicle loading routes for refined oil products after merging across different oil depots were determined, including: Based on pre-defined crossover, mutation, and correction operators, crossover and mutation operations are performed on the initial population to obtain the post-operation population. The above-described steps of crossover and mutation are repeated with the post-operation population replacing the initial population until the maximum value of the objective function corresponding to the post-operation population is less than a preset value and the value of the additional objective function is at the corresponding preset threshold. Based on the post-operation population where the maximum value of the objective function is less than a preset value and the value of the additional objective function is at the corresponding preset threshold, the vehicle loading path for refined oil is determined after merging between different oil depots.

6. The method as described in claim 1, characterized in that, The route planning scheme for refined oil product distribution also includes: The routes of each tanker truck between gas stations, the delivery time and the amount of refined oil delivered at each gas station, the loading and unloading sequence of each tanker truck, the total cost of the route planning scheme, the total carbon emissions, the total transportation risk, and the average mileage utilization rate.

7. A route planning device for refined oil product distribution, characterized in that, include: The route planning modeling module is used to establish a route planning model with the goal of minimizing total transportation cost and additional objective functions of customer satisfaction, carbon emissions, transportation risk, and mileage utilization rate. The route planning model includes parameters related to oil depots, gas stations, tank trucks, and tank trucks. The vehicle loading route determination module is used to calculate the vehicle density of tanker trucks allocated to different gas stations based on the route planning model using a density-based hierarchical clustering algorithm. Based on pre-set vehicle merging constraints, tanker trucks from different gas stations are merged according to the vehicle density to obtain merged vehicles. The vehicle merging constraints include that the distance between the merged tanker trucks is less than or equal to the pre-set maximum tanker truck loading distance. The vehicle merging constraints are used to constrain the tanker truck carrying capacity, delivery order, delivery time window, and distance between the merged tanker trucks. Based on the merged vehicles, the vehicle loading path for refined oil products between gas stations is determined. The merged vehicle loading route determination module uses a genetic algorithm based on a path planning model to establish vehicle codes corresponding to the merged vehicles. The locus number of the vehicle code represents the number of the merged vehicle. The value of the vehicle code represents the assigned oil depot number. An initial population is generated based on the calculated standard deviation of the mileage saved by different merged vehicles. The standard deviation of the mileage saved represents the standard deviation of the mileage saved by assigning different merged vehicles to the target oil depot relative to assigning them to other oil depots. Each individual in the initial population represents a different vehicle code assigned to the initial oil depot. By performing crossover and mutation operations on the initial population, the merged vehicle loading route for refined oil between different oil depots is determined. The route planning output module is used to output the vehicle loading routes of refined oil between gas stations and the combined vehicle loading routes of refined oil between different oil depots as route planning schemes for refined oil distribution.

8. The apparatus as claimed in claim 7, characterized in that, Also includes: The objective function calculation module is used to calculate the objective function as follows: calculate the transportation cost of each tanker truck from the oil depot to each gas station, including fuel cost, vehicle maintenance cost and driver labor cost; and sum the transportation costs of all tanker trucks to obtain the total transportation cost.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method of any one of claims 1 to 6.

11. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the method of any one of claims 1 to 6.

Images