A method and apparatus for determining a logistics vehicle dispatching scheme

Abstract

This invention provides a method and apparatus for determining a logistics vehicle scheduling scheme. The method includes: acquiring logistics order data, customer point data, and vehicle data; extracting target data from the logistics order data, customer point data, and vehicle data to obtain target data; determining a target multi-vehicle type delivery optimization model based on the target data, a preset total delivery cost objective function, and preset logistics transportation constraints; solving the target multi-vehicle type delivery optimization model to obtain an initial logistics vehicle scheduling scheme; dynamically updating the target multi-vehicle type delivery optimization model based on real-time variable data to obtain a dynamic multi-vehicle type delivery optimization model; and dynamically adjusting the initial logistics vehicle scheduling scheme based on the dynamic multi-vehicle type delivery optimization model to obtain a target logistics vehicle scheduling scheme. This invention can reduce logistics delivery costs and improve logistics delivery efficiency.

Description

Technical Field

[0001] This invention relates to the field of vehicle scheduling technology, and also to a method and apparatus for determining a logistics vehicle scheduling scheme. Background Technology

[0002] With the explosive growth of e-commerce, urban logistics and distribution are becoming a crucial support for the operation of modern cities. However, urban logistics and distribution face multiple challenges: the dynamic and uncertain nature of delivery demand, the efficiency decline caused by traffic congestion, and the conflict between environmental protection requirements and operating costs. Against this backdrop, green logistics, as an important component of sustainable development, is receiving increasing attention. How to achieve a win-win situation for both environmental protection and economic benefits while meeting customer service levels has become a core issue that the industry urgently needs to address.

[0003] Vehicle scheduling, as a core component of logistics operations, directly impacts delivery efficiency, operating costs, and environmental impact. Traditional static scheduling models are no longer adequate for the complex and ever-changing urban traffic environment, failing to achieve the logistics goals of reducing delivery costs and improving delivery efficiency. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method and apparatus for determining a logistics vehicle scheduling scheme, so as to improve delivery efficiency and reduce delivery costs.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0006] A first aspect of the present invention provides a method for determining a logistics vehicle scheduling scheme, comprising:

[0007] Acquire logistics order data, customer location data, and vehicle data;

[0008] Target data is obtained by extracting target data from the logistics order data, the customer point data, and the vehicle data;

[0009] Based on the target data, the preset total delivery cost objective function, and the preset logistics and transportation constraints, a target multi-vehicle delivery optimization model is determined.

[0010] Solving the target multi-vehicle delivery optimization model yields an initial logistics vehicle scheduling scheme;

[0011] The target multi-vehicle delivery optimization model is dynamically updated based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model.

[0012] The initial logistics vehicle scheduling scheme is dynamically adjusted based on the dynamic multi-vehicle delivery optimization model to obtain the target logistics vehicle scheduling scheme.

[0013] Optionally, target data can be extracted from the logistics order data, the customer point data, and the vehicle data to obtain target data, including:

[0014] The logistics order data is filtered according to the preset order status to obtain valid order data;

[0015] Based on the valid order data, target data is extracted from the customer point data and the vehicle data to obtain the target data.

[0016] Optionally, based on the target data, the preset total delivery cost objective function, and the preset logistics transportation constraints, a target multi-vehicle delivery optimization model is determined, including:

[0017] Based on the target data and the preset logistics and transportation constraints, the target constraints are obtained;

[0018] Based on the target constraints, the preset total delivery cost objective function, and the preset model fixed parameters, determine the initial multi-vehicle delivery optimization model;

[0019] The initial multi-vehicle delivery optimization model is dynamically corrected based on preset dynamic correction data to obtain the target multi-vehicle delivery optimization model.

[0020] Optionally, the preset logistics transportation constraints include:

[0021] The actual load of each vehicle shall not exceed its maximum load, as expressed in the following formula: , Where N is the total number of customer points, The required weight for the i-th customer point. Variables are between 0 and 1. The maximum load capacity of the k-th vehicle;

[0022] The actual volume of each vehicle does not exceed its maximum volume, expressed as: , ;in, Let i be the volume required for the i-th customer point. Let the maximum volume of the k-th vehicle be denoted as 'k'.

[0023] The vehicle's arrival time at the customer's location must fall within the customer's time window, expressed as: , ;in, Let the earliest arrival time be the i-th customer's location. Let i be the latest arrival time of the i-th customer point. The time it takes for the vehicle to arrive at the i-th customer point;

[0024] The service time for each customer location is fixed at 20 minutes, expressed as follows: , ;in, The service time for the i-th customer point. Let be the distance between the i-th customer point and the j-th customer point. For vehicle speed, This represents the time it takes for the vehicle to arrive at the (i+1)th customer point.

[0025] The number of each vehicle model used cannot exceed its available quantity, expressed as: , ;in, Let m be the set of vehicles of type m. Let m be the number of available models.

[0026] Optionally, the preset total delivery cost objective function is:

[0027] ;

[0028] Where f is the total delivery cost, and K is the total number of vehicles used. Let the starting cost of the kth vehicle be... The time window penalty cost for the kth vehicle. Let the energy consumption cost of the kth vehicle be... Let be the carbon emission cost of the kth vehicle.

[0029] Optionally, the initial multi-vehicle delivery optimization model is dynamically modified based on preset dynamic correction data to obtain a target multi-vehicle delivery optimization model, including:

[0030] Acquire preset dynamic correction data; the preset dynamic correction data includes vehicle speed correction data, vehicle energy consumption correction data, and vehicle carbon emission correction data.

[0031] The preset model fixed parameters in the initial multi-vehicle delivery optimization model are corrected based on the preset dynamic correction data to obtain the target multi-vehicle delivery optimization model.

[0032] Optionally, the target multi-vehicle delivery optimization model is solved to obtain an initial logistics vehicle scheduling scheme, including:

[0033] Multiple initial vehicle scheduling combinations are obtained; all of the multiple initial vehicle scheduling combinations satisfy the target constraints of the target multi-vehicle delivery optimization model.

[0034] The multiple initial vehicle scheduling combinations are encoded to obtain an initial population containing multiple chromosomes; wherein, one chromosome represents one vehicle scheduling combination, and the population represents the set of all chromosomes.

[0035] Based on the initial population and the preset total delivery cost objective function of the target multi-vehicle delivery optimization model, the fitness value of the chromosome is obtained;

[0036] A new population is determined by selection, crossover, and mutation based on the fitness values ​​of the chromosomes and the initial population.

[0037] Based on the new population and the preset iteration termination condition, an initial logistics vehicle scheduling scheme is obtained.

[0038] Optionally, the target multi-vehicle delivery optimization model is dynamically updated based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model, including:

[0039] Acquire real-time variable data; the real-time variable data includes at least one of real-time order variable data, real-time address variable data, and real-time customer point time window variable data.

[0040] The target constraints of the target multi-vehicle delivery optimization model are dynamically updated based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model.

[0041] A second aspect of the present invention provides an apparatus for determining a logistics vehicle scheduling scheme, comprising:

[0042] The acquisition module is used to acquire logistics order data, customer location data, and vehicle data.

[0043] The processing module is used to extract target data from the logistics order data, customer point data, and vehicle data to obtain target data; determine a target multi-vehicle delivery optimization model based on the target data, a preset total delivery cost objective function, and preset logistics transportation constraints; solve the target multi-vehicle delivery optimization model to obtain an initial logistics vehicle scheduling scheme; dynamically update the target multi-vehicle delivery optimization model based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model; and dynamically adjust the initial logistics vehicle scheduling scheme based on the dynamic multi-vehicle delivery optimization model to obtain a target logistics vehicle scheduling scheme.

[0044] A third aspect of the present invention provides a computing device, comprising: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described in the first aspect.

[0045] The above-described solution of the present invention has at least the following beneficial effects:

[0046] The above-described solution of the present invention acquires logistics order data, customer point data, and vehicle data, and extracts target data from these data to obtain target data. Then, based on the target data, a preset total delivery cost objective function, and preset logistics transportation constraints, a target multi-vehicle delivery optimization model is determined. An initial logistics vehicle scheduling scheme is obtained by solving the target multi-vehicle delivery optimization model. The target multi-vehicle delivery optimization model is then dynamically updated based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model. Finally, the initial logistics vehicle scheduling scheme is dynamically adjusted based on the dynamic multi-vehicle delivery optimization model to obtain the target logistics vehicle scheduling scheme. This invention can reduce logistics delivery costs and improve logistics delivery efficiency. Furthermore, it can dynamically adjust the logistics vehicle scheduling scheme according to unexpected events during the logistics delivery process, exhibiting strong dynamic adaptability and real-time response characteristics. Attached Figure Description

[0047] Figure 1 This is a flowchart illustrating the method for determining a logistics vehicle scheduling scheme in an embodiment of the present invention.

[0048] Figure 2 This is a schematic diagram of the structure of the device for determining the logistics vehicle scheduling scheme in an embodiment of the present invention. Detailed Implementation

[0049] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0050] like Figure 1 As shown, an embodiment of the present invention proposes a method for determining a logistics vehicle scheduling scheme, comprising the following steps:

[0051] Step 101: Obtain logistics order data, customer location data, and vehicle data;

[0052] Step 102: Extract target data from the logistics order data, the customer point data, and the vehicle data to obtain target data;

[0053] Step 103: Determine the target multi-vehicle delivery optimization model based on the target data, the preset total delivery cost objective function, and the preset logistics transportation constraints;

[0054] Step 104: Solve the target multi-vehicle delivery optimization model to obtain the initial logistics vehicle scheduling scheme;

[0055] Step 105: Dynamically update the target multi-vehicle delivery optimization model based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model;

[0056] Step 106: Dynamically adjust the initial logistics vehicle scheduling scheme according to the dynamic multi-vehicle delivery optimization model to obtain the target logistics vehicle scheduling scheme.

[0057] The method for determining a logistics vehicle scheduling scheme according to embodiments of the present invention involves acquiring logistics order data, customer point data, and vehicle data, and extracting target data from the logistics order data, customer point data, and vehicle data to obtain target data. Then, based on the target data, a preset total delivery cost objective function, and preset logistics transportation constraints, a target multi-vehicle delivery optimization model is determined. By solving the target multi-vehicle delivery optimization model, an initial logistics vehicle scheduling scheme is obtained. The target multi-vehicle delivery optimization model is then dynamically updated based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model. The initial logistics vehicle scheduling scheme is then dynamically adjusted based on the dynamic multi-vehicle delivery optimization model to obtain the target logistics vehicle scheduling scheme. This invention can reduce logistics delivery costs and improve logistics delivery efficiency. In addition, it can dynamically adjust the logistics vehicle scheduling scheme according to unexpected events during the logistics delivery process, exhibiting strong dynamic adaptability and real-time response characteristics.

[0058] In an optional embodiment of the present invention, step 101, obtaining logistics order data, customer point data, and vehicle data, may include:

[0059] Step 1011: Obtain logistics order data; the logistics order data includes order status data;

[0060] Specifically, obtaining logistics order data is for the purpose of filtering valid orders based on their status (such as orders in the state of pending delivery) and removing invalid orders (such as orders in the state of delivered or cancelled), so as to avoid invalid orders taking up processing time and improve processing efficiency.

[0061] Step 1012, acquire customer point data; the customer point data includes: customer coordinate data, customer time window data, and customer goods demand data;

[0062] Specifically, customer coordinate data can be the latitude and longitude coordinates of the customer's location. This coordinates can then be used to generate road network data between customers and determine travel distances between them. Customer time window data includes the earliest and latest delivery times, used to fix customer service durations, calculate the arrival time of the next customer node, and provide data support for subsequent constraints and penalty costs in the objective function. Customer cargo demand data includes the weight and volume of goods required in customer orders, providing data support for subsequent constraints and preventing vehicle overloading and exceeding space limits. Additionally, customer location data can include customer IDs or customer identifiers to facilitate counting the total number of customer locations.

[0063] Here, logistics order data corresponds one-to-one with customer point data. Customer point data can be obtained based on logistics order data, making it convenient to filter customer point data based on valid orders later.

[0064] Step 1013: Obtain vehicle data; the vehicle data includes: vehicle parameter data and vehicle model quota data.

[0065] Specifically, vehicle parameter data includes: vehicle type, maximum load capacity, maximum volume, and benchmark speed; vehicle quota data includes: available vehicle quota, starting energy consumption, carbon emission coefficient, and other core parameters.

[0066] In an optional embodiment of the present invention, step 102, extracting target data from the logistics order data, the customer point data, and the vehicle data to obtain target data, may include:

[0067] Step 1021: Filter the logistics order data according to the preset order status to obtain valid order data;

[0068] Specifically, the preset order status is "Pending Delivery," and only logistics orders with "Pending Delivery" will proceed to the next stage of processing. Invalid orders that do not meet the preset order status (such as orders with a status of "Delivered" or "Cancelled") will be deleted from the logistics order data, and only valid orders that meet the preset order status (such as orders with a status of "Pending Delivery") will be retained as valid order data.

[0069] Step 1022: Extract target data from the customer point data and the vehicle data based on the valid order data to obtain target data.

[0070] Specifically, the target customer point data is obtained by filtering the customer point data based on valid order data. In practice, invalid orders that do not meet the preset order status (such as orders with a status of "delivered" or "cancelled") and their corresponding customer point data are deleted from the logistics order data. Only valid orders that meet the preset order status (such as orders with a status of "pending delivery") and their corresponding customer point data are retained as the target customer point data, thereby improving the processing efficiency and effectiveness of the method.

[0071] After obtaining the target customer point data, the latitude and longitude coordinates of each customer point are extracted from the target customer point data to generate a road network distance matrix between any two customer points and between the warehouse and the customer point, determining the driving distance between nodes; the delivery time window for each customer is extracted, i.e., the earliest and latest delivery times, to fix the customer service duration; the required weight and volume of goods for each customer's order are extracted one by one. The data extracted above are used as the target customer data. Vehicle data extraction includes: statistically analyzing core parameters such as maximum load capacity, maximum volume, driving base speed, available vehicle quota, starting energy consumption, and carbon emission coefficient for each vehicle model, which are used as the target vehicle data.

[0072] The target data includes customer target data and vehicle target data, specifically including: total number of customer points, customer demand weight set, customer demand volume set, customer time window parameter set, road network distance matrix, vehicle performance parameter set, vehicle type available quota set, fixed service duration parameters, etc. The obtained target data can be saved in a preset format (such as a table) for easy viewing and management.

[0073] The target data is extracted from customer point data and vehicle data in order to fill in the constraints with values ​​and obtain constraints that conform to the actual application scenario.

[0074] In an optional embodiment of the present invention, step 103, determining the target multi-vehicle delivery optimization model based on the target data, the preset total delivery cost objective function, and the preset logistics transportation constraints, may include:

[0075] Step 1031: Based on the target data and the preset logistics transportation constraints, obtain the target constraints;

[0076] Specifically, the preset logistics transportation constraints include:

[0077] The actual load of each vehicle shall not exceed its maximum load, as expressed in the following formula: , Where N is the total number of customer points, The required weight for the i-th customer point. Variables are between 0 and 1. The maximum load capacity of the k-th vehicle;

[0078] The actual volume of each vehicle does not exceed its maximum volume, expressed as: , ;in, Let i be the volume required for the i-th customer point. Let the maximum volume of the k-th vehicle be denoted as 'k'.

[0079] The vehicle's arrival time at the customer's location must fall within the customer's time window, expressed as: , ;in, Let the earliest arrival time be the i-th customer's location. Let i be the latest arrival time of the i-th customer point. The time it takes for the vehicle to arrive at the i-th customer point;

[0080] The service time for each customer location is fixed at 20 minutes, expressed as follows: , ;in, The service time for the i-th customer point. Let be the distance between the i-th customer point and the j-th customer point. For vehicle speed, This represents the time it takes for the vehicle to arrive at the (i+1)th customer point.

[0081] The number of each vehicle model used cannot exceed its available quantity, expressed as: , ;in, Let m be the set of vehicles of type m. Let m be the number of available models.

[0082] It should be noted that the preset constraints in step 1041 can be adjusted according to the actual situation. For example, in a specific embodiment, the preset constraints may also include: Where F represents the collection of gasoline-powered vehicles. Let i be the distance from the i-th customer point to the city center. Let be the coordinates of the i-th customer point, when If customer point i is located within the green delivery zone, it is considered to be within the green delivery zone; otherwise, it is located outside the green delivery zone. This preset constraint means that fuel-powered vehicles are not allowed to enter the green delivery zone between 8:00 and 16:00.

[0083] Based on the data required in the preset logistics and transportation constraints, the specific values ​​corresponding to the target data are filled into the preset logistics and transportation constraints to obtain the target constraints.

[0084] Step 1032: Determine the initial multi-vehicle delivery optimization model based on the target constraints, the preset total delivery cost objective function, and the preset model fixed parameters;

[0085] Specifically, the preset total delivery cost objective function is:

[0086] ;

[0087] Where f is the total delivery cost, and K is the total number of vehicles used. Let the starting cost of the kth vehicle be... The time window penalty cost for the kth vehicle. Let the energy consumption cost of the kth vehicle be... Let be the carbon emission cost of the kth vehicle.

[0088] Specifically, startup cost is the fixed startup cost of a single vehicle trip, covering vehicle inspection, basic labor costs, vehicle depreciation, and dispatch procedures. New energy vehicles are simpler to maintain and have lower dispatch losses, so their startup cost is lower than that of fuel vehicles. Time window penalty cost is the penalty cost for failure to deliver the vehicle within the customer's specified time window. It is divided into early arrival penalty and late arrival penalty. Here, the early arrival penalty can be set at 15 yuan / minute (the vehicle arrives before the customer's earliest time, resulting in waiting and demurrage costs), and the late arrival penalty can be set at 30 yuan / minute (the vehicle arrives after the customer's latest time, resulting in order breach penalty, with a higher penalty than early arrival).

[0089] It should be noted that the above are just examples, and the startup cost and time window penalty cost can be set according to the specific application scenario or specific implementation.

[0090] Here, the energy consumption costs of gasoline vehicles and new energy vehicles can be obtained using the following formula:

[0091] Fuel consumption of gasoline vehicles: ;

[0092] Energy consumption of new energy vehicles: ;

[0093] The carbon emission costs of gasoline-powered vehicles and new energy vehicles can be obtained using the following formula:

[0094] Carbon emission costs of gasoline vehicles: ;

[0095] Carbon emission costs of new energy vehicles: ;

[0096] in, For fuel consumption of gasoline vehicles, For the energy consumption of new energy vehicles, This represents the actual fuel consumption. FPK represents the fuel consumption per 100 kilometers (L / 100km) for gasoline vehicles. S represents the vehicle's mileage. This represents the vehicle's speed (km / h), and E is the energy load correction coefficient in the preset model's fixed parameters. This represents the actual electricity consumption. EPK represents the energy consumption of a new energy vehicle per 100 kilometers (kWh / 100km). , For oil prices, For electricity price, Indicates the carbon emission cost of gasoline vehicles. Indicates the carbon emission cost of new energy vehicles, This represents the fuel consumption carbon emission conversion coefficient (valued at 2.547 kg / L). This represents the carbon emission conversion coefficient for electricity consumption (with a value of 0.501 kg / kWh).

[0097] The initial multi-vehicle delivery optimization model consists of a preset total delivery cost objective function, objective constraints, and preset fixed parameters, expressed as follows:

[0098] ;

[0099] in, In this embodiment, there are 6 target constraints, but the number of target constraints can be modified depending on the specific implementation. The preset model's fixed parameters are: fixed base speed (the average daily base speed of urban logistics vehicles, which can be set to 35 km / h), E (the fixed load correction coefficient under the rated full load condition of the vehicle, which can be set to 1.2 for fuel vehicles and 1.15 for new energy vehicles), and CE (the fixed carbon emission load correction coefficient under the rated full load condition of the vehicle, which can be set to 1.2 for fuel vehicles and 1.15 for new energy vehicles). All preset model fixed parameters in the initial multi-vehicle delivery optimization model are fixed values.

[0100] Step 1033: Dynamically correct the initial multi-vehicle delivery optimization model according to the preset dynamic correction data to obtain the target multi-vehicle delivery optimization model.

[0101] In an optional embodiment of the present invention, step 1033 includes:

[0102] Step 10331: Obtain preset dynamic correction data; the preset dynamic correction data includes vehicle speed correction data, vehicle energy consumption correction data, and vehicle carbon emission correction data.

[0103] Step 10332: Correct the preset model fixed parameters in the initial multi-vehicle delivery optimization model according to the preset dynamic correction data to obtain the target multi-vehicle delivery optimization model.

[0104] Specifically, depending on the actual situation, vehicles may travel during different times of the day, such as morning rush hour, off-peak hour, or evening rush hour, and vehicle speed will also change dynamically. As the vehicle's load increases, the energy consumption load correction coefficient and carbon emission load correction coefficient will also increase. Therefore, the preset dynamic correction data includes vehicle speed correction data, vehicle energy consumption correction data, and vehicle carbon emission correction data. The vehicle speed correction data is a dynamic vehicle speed that changes over time (real-time vehicle speed, a time-related function that outputs the corresponding road network speed based on different times of the day, such as morning rush hour, off-peak hour, evening rush hour, and nighttime, in km / h). The vehicle energy consumption correction data is an energy consumption load correction function (taking the real-time load w as input and outputting an energy consumption correction coefficient; the greater the load, the higher the coefficient, and the greater the fuel consumption per unit mileage). The vehicle carbon emission correction data is a carbon emission load correction function (taking the real-time load w as input and outputting a carbon emission correction coefficient, tailored to the operating characteristics of heavy-load, high-emission vehicles, used for accurate calculation of low-carbon costs). The steps for dynamically correcting the initial multi-vehicle delivery optimization model include:

[0105] The fixed vehicle speed in the initial multi-vehicle delivery optimization model Vehicle speed in the constraints All replaced with dynamic vehicle speeds that change over time. ;

[0106] Replace the energy consumption load correction coefficient E in the initial multi-vehicle delivery optimization model with the energy consumption load correction function E(w), and replace the carbon emission load correction coefficient CE with the carbon emission load correction function CE(w).

[0107] The revised target multi-vehicle delivery optimization model includes: the total delivery cost objective function, objective constraints, dynamic vehicle speed varying over time, energy consumption load correction function, and carbon emission load correction function. The expression for the target multi-vehicle delivery optimization model is:

[0108] ;

[0109] Specifically, when the delivery operation time t falls within the morning peak period... When the delivery operation time t falls within the daytime off-peak period, When the delivery operation time t falls within the evening peak period, When the delivery operation time t falls within the nighttime off-peak period, ; , , , The average driving speeds for the four time periods are respectively; w represents the vehicle's real-time load. The maximum load capacity of the k-th vehicle; Vehicle carbon emissions are strictly linearly positively correlated with fuel / electricity consumption. The increase in carbon emissions due to load is exactly the same as the increase in energy consumption. Therefore, the carbon emission load correction function and the energy consumption load correction function are completely equivalent.

[0110] It should be noted that when calculating the carbon emission costs of gasoline vehicles and new energy vehicles in the target multi-vehicle delivery optimization model, the "E" in the calculation formulas for fuel consumption per 100 kilometers (FPK) for gasoline vehicles and electricity consumption per 100 kilometers (EPK) for new energy vehicles also needs to be replaced with... .

[0111] Through the above modifications, the target multi-vehicle delivery optimization model has the following advantages: vehicle speed and travel time dynamically change with traffic periods, no longer being fixed constants; energy consumption and carbon emission costs are linked to the real-time vehicle load, conforming to physical operating laws; and it takes into account multiple factors such as economic cost, time efficiency cost, environmental cost, traffic conditions, and vehicle load. This step couples spatiotemporal traffic characteristics with vehicle load characteristics, upgrading the model from an idealized static model to a realistic dynamic model, significantly reducing the deviation between scheduling results and actual operations.

[0112] In an optional embodiment of the present invention, step 104, solving the target multi-vehicle delivery optimization model to obtain an initial logistics vehicle scheduling scheme, may include:

[0113] Step 1041: Obtain multiple initial vehicle scheduling combinations; all of the multiple initial vehicle scheduling combinations satisfy the target constraints of the target multi-vehicle type delivery optimization model.

[0114] Specifically, H initial vehicle scheduling combinations can be obtained. These H initial vehicle scheduling combinations are randomly generated and satisfy all objective constraints. Each initial vehicle scheduling combination represents a combination of vehicle allocation, customer segmentation, and delivery sequence. The scheduling content specifically included in the initial vehicle scheduling combination may include:

[0115] Vehicle selection: Select the corresponding fuel vehicle from the available vehicle quota, not exceeding the maximum number of available vehicles for that vehicle type;

[0116] Customer clustering: All delivery customers are randomly assigned to different delivery vehicles to ensure that the load and volume of a single vehicle do not exceed the limits.

[0117] Delivery route sorting: Generate an initial delivery order for the multiple customer points served by each vehicle;

[0118] Time sequence compliance verification: The initial vehicle scheduling combination strictly meets the time window constraints, service duration constraints, and time-varying travel time rules, and invalid schemes that are overloaded, exceed the time limit, or exceed the quota are eliminated.

[0119] Step 1042: Encode the multiple initial vehicle scheduling combinations to obtain an initial population containing multiple chromosomes; wherein, one chromosome represents one vehicle scheduling combination, and the population represents the set of all chromosomes.

[0120] Specifically, through encoding, complex scheduling information such as vehicle allocation, customer segmentation, and delivery sequence is transformed into integer gene sequences that the algorithm can recognize, providing a data foundation for subsequent iterative optimization.

[0121] Here, all customer nodes are numbered sequentially as 1, 2, 3, ... Each integer gene represents a customer delivery node; an additional number 0 is introduced as a vehicle separator gene to distinguish delivery task sequences of different vehicles.

[0122] A complete chromosome is constructed by splicing together a customer ID gene and a separator gene, with the overall structure: [0, customer sequence 1, 0, customer sequence 2, 0, ..., 0, customer sequence K, 0]. This structure represents the integer sequence between two adjacent 0s, representing the customer delivery order of a single vehicle. It contains K customer sequences, corresponding to the K delivery vehicles deployed in this scheduling. Each customer sequence strictly satisfies the vehicle's load capacity, volume, and time window constraints, ensuring the inherent feasibility of the encoded solution. Each integer sequence uniquely corresponds to one of the following: the number of vehicles activated, vehicle task allocation, customer cluster division, and customer delivery order; that is, it uniquely corresponds to an initial vehicle scheduling combination.

[0123] Encode each initial vehicle scheduling combination one by one: each initial vehicle scheduling combination generates an independent chromosome, ultimately resulting in an initial population containing H chromosomes.

[0124] Step 1043: Obtain the fitness value of the chromosome based on the initial population and the preset total delivery cost objective function of the target multi-vehicle delivery optimization model;

[0125] Specifically, the number of vehicles used, travel routes, delivery sequences, load capacities, and travel periods for each initial vehicle scheduling combination are substituted into the total delivery cost objective function to calculate the startup cost, time penalty cost, dynamic energy consumption cost, and carbon emission cost for each initial vehicle scheduling combination, ultimately yielding the total delivery cost for each scheme. A lower total cost corresponds to a higher fitness value, indicating better economy, low carbon footprint, and timeliness of the initial vehicle scheduling combination. The fitness (total cost) value corresponding to each initial vehicle scheduling combination within the population is calculated.

[0126] Step 1044: Based on the fitness values ​​of the chromosomes and the initial population, perform selection, crossover, and mutation to determine a new population;

[0127] Specifically, based on the chromosome's fitness value and Parent individuals are selected for breeding; individuals with higher fitness values ​​have a greater probability of being selected; among them, It is the probability that the i-th chromosome is selected. It is the fitness value of the j-th chromosome in the initial population. , This represents the fitness value of the i-th chromosome. It is accumulated sequentially starting from the first individual in the initial population. The cumulative probability sequence is obtained; a random number r between 0 and 1 is generated; based on the cumulative probability interval in which r falls, the corresponding individual is selected as the parent.

[0128] Randomly select a crossover point and perform a crossover operation on the parent generation to obtain offspring individuals; for example, exchange gene segments of two parent individuals to generate two offspring individuals; in a specific embodiment, parent individual 1 is [1, 0, 1, 0, 1, 0], parent individual 2 is [0, 1, 0, 1, 0, 1], and the third, fourth and fifth gene segments in parent individual 1 and parent individual 2 are exchanged to obtain offspring individual 1 [1, 0, 1, 1, 0, 1] and offspring individual 2 [0, 1, 0, 0, 1, 0].

[0129] A new population is determined by randomly flipping a gene locus in the offspring individuals according to a preset probability (e.g., 0.01). This involves randomly flipping at least one of the following: vehicle usage quantity, driving route, delivery sequence, load, and driving time period, such as flipping 0 to 1 or 1 to 0. In one specific embodiment, the offspring individuals are [1, 0, 1, 0, 1, 0], and the flipped offspring individuals are [1, 0, 0, 0, 1, 0].

[0130] Step 1045: Based on the new population and the preset iteration termination condition, an initial logistics vehicle scheduling scheme is obtained.

[0131] Specifically, the preset iteration termination condition can be reaching a preset maximum number of iterations. Repeat step 1044 to calculate the fitness value of chromosomes in the new population until the preset number of iterations reaches the preset maximum number of iterations. Then stop the iteration and select the single individual with the lowest total cost and the highest fitness from all generations of the population as the logistics vehicle scheduling scheme.

[0132] In an optional embodiment of the present invention, step 105, dynamically updating the target multi-vehicle delivery optimization model based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model, may include:

[0133] Step 1051: Obtain real-time variable data; the real-time variable data includes at least one of real-time order variable data, real-time address variable data, and real-time customer point time window variable data;

[0134] Specifically, real-time order variable data includes order cancellations or new orders, real-time address variable data includes changes to customer delivery addresses, and real-time customer time window variable data includes adjustments to customer time windows. These real-time variable data represent unexpected events in the logistics and delivery process. To address these unexpected events, the initial logistics vehicle scheduling plan needs to be dynamically adjusted to adapt to changes in the scenario and improve the applicability of the method.

[0135] Step 1052: Dynamically update the target constraints of the target multi-vehicle delivery optimization model based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model.

[0136] Specifically, due to changes in orders, customer addresses, or customer time windows, it is necessary to modify customer location data and / or vehicle data based on these changes (i.e., real-time variable data). Then, the target constraints of the target multi-vehicle delivery optimization model are dynamically updated based on the modified customer location data and / or vehicle data. That is, the corresponding values ​​in the target constraints are modified according to the specific values ​​in the modified customer location data and / or vehicle data. Here, the following steps determine the content to be modified in the customer location data and / or vehicle data, and then the customer location data and / or vehicle data are modified accordingly:

[0137] Order cancellation events directly reduce overall delivery demand, remove delivery tasks from the corresponding customer nodes, reduce vehicle load capacity and delivery mileage. The weight of cancelled orders must be removed from the total delivery demand, and the feasible region of vehicle loading constraints must be updated. The quantitative expression for demand change is as follows: ,in, Let C be the total reduction in delivery demand caused by order cancellations, and let C be the set of customer locations for these cancelled orders. For the first The weight of goods required by each customer who canceled their order. Update the vehicle's remaining load constraints, delete the delivery tasks for the corresponding customer nodes, eliminate invalid routes, and avoid empty runs and resource waste.

[0138] A new order event will generate new delivery demand, expanding the delivery task scale. This requires adding new customer nodes, accumulating cargo weight, and updating the total delivery demand and vehicle loading constraints. The demand increment expression is as follows: ,in, This represents the increase in total delivery demand due to new orders. This is the collection of customer points for this new order. Let J represent the weight of goods required by the customer who placed the j-th new order.

[0139] A change in a customer's delivery address alters the spatial location of nodes, rendering the original road network distance matrix, travel time, and route plans invalid. It is necessary to recalculate the Euclidean distances between nodes and update the spatial road network parameters. The distance update expression is as follows: ,in, This is the updated travel distance from customer point i to customer point j after the address change. For the customer's point i, the changed latitude and longitude coordinates, The latitude and longitude coordinates of customer point j are changed. Based on the updated distance matrix, the mileage, energy consumption, and carbon emission costs of the route are recalculated, and the local delivery route is reconstructed.

[0140] Adjusting the customer's time window will change the timeliness constraint boundaries, rendering the original delivery sequence plan non-compliant. The customer's delivery time interval needs to be updated to form a new dynamic time constraint. The adjusted constraint expression is as follows: ,in, This is the earliest delivery time for the i-th customer after adjustment. This is the latest possible delivery time for the i-th customer after adjustment. This refers to the set of customer locations for this time window adjustment. Using the new time window as the constraint boundary, the delivery sequence is re-verified, node arrival times are optimized, and time penalty costs are avoided.

[0141] To avoid repeated global iterations and reduce dynamic response latency after unexpected events, this embodiment follows the principles of execution locking, optimization and reconstruction, and real-time window updates. It limits the optimization boundary and only performs local rescheduling for incomplete delivery tasks, significantly improving dynamic adjustment efficiency. For customer nodes that have already been delivered and routes that have already been traveled, the task state is irreversible and is locked throughout the process, preventing secondary optimization and ensuring the continuity of the delivery process while avoiding unnecessary repetitive calculations. This set represents a fixed set of nodes whose deliveries have been completed, with all node tasks, paths, and timings permanently retained. Only nodes whose deliveries have not been completed or those affected by unforeseen events undergo path and timing reconstruction to narrow the optimization scope and improve dynamic scheduling response speed. The set of nodes to be optimized is as follows: , Let V be the set of customer nodes to be optimized, and V be the set of all customer points. Based on the current moment of the disturbance, update the effective time window of the nodes to be optimized, remove expired early delivery periods, and align with real-time delivery progress. The dynamic time constraint is as follows: ,in, The time when the emergency occurred. Let the earliest arrival time be the i-th customer's location. Let i be the latest arrival time of the i-th customer point. The time for a vehicle to arrive at the i-th customer point is updated through a time window to ensure that the rescheduling scheme is fully adapted to the real-time delivery progress and to eliminate time sequence logic conflicts.

[0142] The target multi-vehicle delivery optimization model is expressed as:

[0143] ;

[0144] in, The target constraint is dynamically updated, and when the delivery operation time t belongs to the morning peak period, When the delivery operation time t falls within the daytime off-peak period, When the delivery operation time t falls within the evening peak period, When the delivery operation time t falls within the nighttime off-peak period, ; , , , The average driving speeds for the four time periods are respectively; w represents the vehicle's real-time load. The maximum load capacity of the k-th vehicle; Vehicle carbon emissions are strictly linearly positively correlated with fuel / electricity consumption. The increase in carbon emissions due to load is exactly the same as the increase in energy consumption. Therefore, the carbon emission load correction function and the energy consumption load correction function are completely equivalent.

[0145] By dynamically updating the constraints in the target multi-vehicle delivery optimization model using real-time variable data, a dynamic multi-vehicle delivery optimization model is obtained. Here, the constraints of the dynamic multi-vehicle delivery optimization model are... include:

[0146] , ;in, For the set of customer nodes to be optimized, This represents the demand weight for the i-th customer point after dynamic updates. Variables are between 0 and 1. The maximum load capacity of the k-th vehicle;

[0147] , ;in, The updated demand volume for the i-th customer point. Let the maximum volume of the k-th vehicle be denoted as 'k'.

[0148] , ;in, Let the earliest arrival time be the i-th customer's location. Let i be the latest arrival time of the i-th customer point. The time it takes for the vehicle to arrive at the i-th customer point;

[0149] , ;in, The service time for the i-th customer point. This is the updated travel distance from customer point i to customer point j after the address change. The dynamic vehicle speed varies over time. This represents the time it takes for the vehicle to arrive at the (i+1)th customer point.

[0150] , ;in, Let m be the set of vehicles of type m. Let m be the number of available car models.

[0151] Where F represents the collection of gasoline-powered vehicles. This represents the distance from the i-th customer point to the city center after dynamic updates. , For the changed latitude and longitude coordinates of customer point i, when If customer point i is located within the green delivery zone, then it is located outside the green delivery zone.

[0152] In an optional embodiment of the present invention, step 106 involves dynamically adjusting the initial logistics vehicle scheduling scheme according to the dynamic multi-vehicle delivery optimization model to obtain the target logistics vehicle scheduling scheme.

[0153] Specifically, the optimal solution for the dynamic multi-vehicle delivery optimization model is obtained by replacing the initial logistics vehicle scheduling scheme with the optimal solution, i.e., the target logistics vehicle scheduling scheme. The solution can be obtained using the method in step 104, or through the following process:

[0154] Initialize the detection population size and set the number of virtual detectors. Each virtual detector corresponds to a complete customer delivery sequence and vehicle route plan, i.e., a dynamic logistics vehicle scheduling plan, simulating the inspection and optimization process of vehicles traversing customer nodes; the pheromone concentration of the delivery route is initialized, and the initial pheromone values ​​of the paths between all customer nodes are uniformly assigned, using the following formula: ,in, Let pheromone concentration be the concentration along the delivery path from customer point i to customer point j. The number of virtual detectors. This represents the shortest basic path length between customer nodes in the delivery network. The maximum value T is the initial iteration count t.

[0155] Each virtual detector randomly selects a client node as its starting client node and maintains a path memory vector to store the client nodes that the virtual detector traverses sequentially. At each step of path construction, the virtual detector selects the next client node to reach according to a random ratio rule. The random ratio rule is as follows:

[0156]

[0157] in, Let be the probability that the q-th virtual detector travels from client node i to client node j. , This is the updated travel distance from customer point i to customer point j after the address change. Let pheromone intensity be the pheromone intensity from client node i to client node j at time t. This is a collection of client nodes that have not been visited. For pheromones, This is a weighted value for visibility.

[0158] When all virtual detectors reach the last client node, the pheromone concentration on each path must be updated once, using the following formula:

[0159] ;

[0160] Where t is the number of iterations. The number of virtual detectors. The evaporation rate of pheromones, 0 < ≤1, It is usually set to 0.5. Let be the pheromone concentration left by the q-th virtual detector between client node i and client node j.

[0161] ;

[0162] in, Let q be the total path length obtained after the q-th virtual detector has traversed the entire path.

[0163] The process of selecting node paths and updating pheromones is executed repeatedly until the number of iterations reaches the preset maximum number of iterations T. Then, the iteration stops and all dynamic logistics vehicle scheduling schemes are output.

[0164] Input all dynamic logistics vehicle scheduling schemes into the objective function of the target multi-vehicle delivery optimization model, and take the dynamic logistics vehicle scheduling scheme that minimizes the total delivery cost f as the target logistics vehicle scheduling scheme.

[0165] An embodiment of the present invention provides a specific embodiment of a method for determining a logistics vehicle scheduling scheme, including:

[0166] Step 111: Obtain logistics order data, customer location data, and vehicle data;

[0167] Acquire logistics order data and corresponding customer location data, as well as vehicle data, to provide data support for determining subsequent vehicle dispatching plans.

[0168] Step 112: Extract target data from the logistics order data, the customer point data, and the vehicle data to obtain target data;

[0169] The validity of customer point data is filtered based on logistics order data, and the data required for subsequent constraints is extracted from customer point data and vehicle data as target data.

[0170] Step 113: Determine the target multi-vehicle delivery optimization model based on the target data, the preset total delivery cost objective function, and the preset logistics transportation constraints;

[0171] Extract the corresponding inputs from the target data, substitute them one by one into the preset logistics and transportation constraints to obtain the actual target constraints, and combine them with the preset total delivery cost objective function and the preset model fixed parameters to form the target multi-vehicle delivery optimization model.

[0172] Step 114: Solve the target multi-vehicle delivery optimization model to obtain the initial logistics vehicle scheduling scheme;

[0173] The optimal solution is sought for the target multi-vehicle delivery optimization model according to the preset solution method, and the initial logistics vehicle scheduling scheme is obtained.

[0174] Step 115: Dynamically update the target multi-vehicle delivery optimization model based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model;

[0175] The data in the customer point data and vehicle data are modified based on real-time variable data. Based on the modified customer point data and vehicle data, the constraints in the target multi-vehicle delivery optimization model are dynamically updated (the values ​​in the constraints are modified), resulting in a dynamic multi-vehicle delivery optimization model.

[0176] Step 116: Dynamically adjust the initial logistics vehicle scheduling scheme according to the dynamic multi-vehicle delivery optimization model to obtain the target logistics vehicle scheduling scheme.

[0177] The dynamic multi-vehicle delivery optimization model is solved, and the optimal solution is used as the target logistics vehicle scheduling scheme.

[0178] During the logistics and distribution process, the target multi-vehicle delivery optimization model is dynamically modified based on real-time variable data obtained from unexpected events such as orders, addresses, or customer times. The model is then solved to dynamically adjust the initial logistics vehicle scheduling plan and obtain the target logistics vehicle scheduling plan.

[0179] The method for determining logistics vehicle scheduling schemes in this invention breaks through the limitations of traditional single cost modeling. It integrates fixed vehicle dispatch costs, time-delivery penalty costs, dynamic energy consumption costs, and carbon emission environmental protection costs to construct a multi-dimensional cost accounting system. This reduces enterprise operating costs and logistics carbon emissions, aligning with green logistics policy requirements. It covers five major constraints: vehicle physical performance, customer time-delivery requirements, enterprise resource quotas, and on-site operation processes. This avoids problems such as overloading, overtime, resource over-allocation, and unreasonable processes, thereby improving the reliability of logistics vehicle scheduling schemes.

[0180] like Figure 2 As shown, an embodiment of the present invention provides a device 200 for determining a logistics vehicle scheduling scheme, comprising:

[0181] Module 201 is used to acquire logistics order data, customer location data, and vehicle data.

[0182] The processing module 202 is used to extract target data from the logistics order data, customer point data, and vehicle data to obtain target data; determine a target multi-vehicle delivery optimization model based on the target data, a preset total delivery cost objective function, and preset logistics transportation constraints; solve the target multi-vehicle delivery optimization model to obtain an initial logistics vehicle scheduling scheme; dynamically update the target multi-vehicle delivery optimization model based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model; and dynamically adjust the initial logistics vehicle scheduling scheme based on the dynamic multi-vehicle delivery optimization model to obtain a target logistics vehicle scheduling scheme.

[0183] Optionally, target data can be extracted from the logistics order data, the customer point data, and the vehicle data to obtain target data, including:

[0184] The logistics order data is filtered according to the preset order status to obtain valid order data;

[0185] Based on the valid order data, target data is extracted from the customer point data and the vehicle data to obtain the target data.

[0186] Optionally, based on the target data, the preset total delivery cost objective function, and the preset logistics transportation constraints, a target multi-vehicle delivery optimization model is determined, including:

[0187] Based on the target data and the preset logistics and transportation constraints, the target constraints are obtained;

[0188] Based on the target constraints, the preset total delivery cost objective function, and the preset model fixed parameters, determine the initial multi-vehicle delivery optimization model;

[0189] The initial multi-vehicle delivery optimization model is dynamically corrected based on preset dynamic correction data to obtain the target multi-vehicle delivery optimization model.

[0190] Optionally, the preset logistics transportation constraints include:

[0191] The actual load of each vehicle shall not exceed its maximum load, as expressed in the following formula: , Where N is the total number of customer points, The required weight for the i-th customer point. Variables are between 0 and 1. The maximum load capacity of the k-th vehicle;

[0192] The actual volume of each vehicle does not exceed its maximum volume, expressed as: , ;in, Let i be the volume required for the i-th customer point. Let the maximum volume of the k-th vehicle be denoted as 'k'.

[0193] The vehicle's arrival time at the customer's location must fall within the customer's time window, expressed as: , ;in, Let the earliest arrival time be the i-th customer's location. Let i be the latest arrival time of the i-th customer point. The time it takes for the vehicle to arrive at the i-th customer point;

[0194] The service time for each customer location is fixed at 20 minutes, expressed as follows: , ;in, The service time for the i-th customer point. Let be the distance between the i-th customer point and the j-th customer point. For vehicle speed, This represents the time it takes for the vehicle to arrive at the (i+1)th customer point.

[0195] The number of each vehicle model used cannot exceed its available quantity, expressed as: , ;in, Let m be the set of vehicles of type m. Let m be the number of available models.

[0196] Optionally, the preset total delivery cost objective function is:

[0197] ;

[0198] Where f is the total delivery cost, and K is the total number of vehicles used. Let the starting cost of the kth vehicle be... The time window penalty cost for the kth vehicle. Let the energy consumption cost of the kth vehicle be... Let be the carbon emission cost of the kth vehicle.

[0199] Optionally, the initial multi-vehicle delivery optimization model is dynamically modified based on preset dynamic correction data to obtain a target multi-vehicle delivery optimization model, including:

[0200] Acquire preset dynamic correction data; the preset dynamic correction data includes vehicle speed correction data, vehicle energy consumption correction data, and vehicle carbon emission correction data.

[0201] The preset model fixed parameters in the initial multi-vehicle delivery optimization model are corrected based on the preset dynamic correction data to obtain the target multi-vehicle delivery optimization model.

[0202] Optionally, the target multi-vehicle delivery optimization model is solved to obtain an initial logistics vehicle scheduling scheme, including:

[0203] Multiple initial vehicle scheduling combinations are obtained; all of the multiple initial vehicle scheduling combinations satisfy the target constraints of the target multi-vehicle delivery optimization model.

[0204] The multiple initial vehicle scheduling combinations are encoded to obtain an initial population containing multiple chromosomes; wherein, one chromosome represents one vehicle scheduling combination, and the population represents the set of all chromosomes.

[0205] Based on the initial population and the preset total delivery cost objective function of the target multi-vehicle delivery optimization model, the fitness value of the chromosome is obtained;

[0206] A new population is determined by selection, crossover, and mutation based on the fitness values ​​of the chromosomes and the initial population.

[0207] Based on the new population and the preset iteration termination condition, an initial logistics vehicle scheduling scheme is obtained.

[0208] Optionally, the target multi-vehicle delivery optimization model is dynamically updated based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model, including:

[0209] Acquire real-time variable data; the real-time variable data includes at least one of real-time order variable data, real-time address variable data, and real-time customer point time window variable data.

[0210] The target constraints of the target multi-vehicle delivery optimization model are dynamically updated based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model.

[0211] The device for determining a logistics vehicle scheduling scheme according to an embodiment of the present invention acquires logistics order data, customer point data, and vehicle data, and extracts target data from the logistics order data, customer point data, and vehicle data to obtain target data. Then, based on the target data, a preset total delivery cost objective function, and preset logistics transportation constraints, a target multi-vehicle vehicle delivery optimization model is determined. By solving the target multi-vehicle vehicle delivery optimization model, an initial logistics vehicle scheduling scheme is obtained. The target multi-vehicle vehicle delivery optimization model is dynamically updated based on real-time variable data to obtain a dynamic multi-vehicle vehicle delivery optimization model. The initial logistics vehicle scheduling scheme is dynamically adjusted based on the dynamic multi-vehicle vehicle delivery optimization model to obtain the target logistics vehicle scheduling scheme. The present invention can reduce logistics delivery costs and improve logistics delivery efficiency. In addition, it can dynamically adjust the logistics vehicle scheduling scheme according to unexpected events in the logistics delivery process, and has strong dynamic adaptability and real-time response characteristics.

[0212] It should be noted that this device corresponds to the method described above, and all implementations in the method embodiments described above are applicable to the embodiments of this device and can achieve the same technical effect. Further details are omitted in this embodiment.

[0213] This invention also provides a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method as described in any of the above embodiments. All implementations in the above method embodiments are applicable to the embodiments of this device and can achieve the same technical effects. Further details are omitted in this embodiment.

[0214] This invention also provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method as described in any of the above embodiments. All implementations in the above method embodiments are applicable to the embodiments of this device and can achieve the same technical effects. Further details are omitted in this embodiment.

[0215] It should be noted that in the apparatus and method of the present invention, the components or steps can obviously be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent solutions of the present invention. Furthermore, the steps for performing the above series of processes can naturally be performed in the order described and in chronological order, but are not necessarily required to be performed in chronological order. Some steps can be performed in parallel, overlapping, or independently of each other.

[0216] It should be noted that in the above embodiments, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments described above is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0217] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for determining a logistics vehicle dispatching scheme, characterized in that, include: Acquire logistics order data, customer location data, and vehicle data; Target data is obtained by extracting target data from the logistics order data, the customer point data, and the vehicle data; Based on the target data, the preset total delivery cost objective function, and the preset logistics and transportation constraints, a target multi-vehicle delivery optimization model is determined. Solving the target multi-vehicle delivery optimization model yields an initial logistics vehicle scheduling scheme; The target multi-vehicle delivery optimization model is dynamically updated based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model. The initial logistics vehicle scheduling scheme is dynamically adjusted based on the dynamic multi-vehicle delivery optimization model to obtain the target logistics vehicle scheduling scheme.

2. The method for determining a logistics vehicle dispatching scheme according to claim 1, characterized in that, Target data is extracted from the logistics order data, the customer point data, and the vehicle data to obtain target data, including: The logistics order data is filtered according to the preset order status to obtain valid order data; Based on the valid order data, target data is extracted from the customer point data and the vehicle data to obtain the target data.

3. The method for determining a logistics vehicle dispatching scheme according to claim 1, characterized in that, Based on the target data, the preset total delivery cost objective function, and the preset logistics and transportation constraints, a target multi-vehicle delivery optimization model is determined, including: Based on the target data and the preset logistics and transportation constraints, the target constraints are obtained; Based on the target constraints, the preset total delivery cost objective function, and the preset model fixed parameters, determine the initial multi-vehicle delivery optimization model; The initial multi-vehicle delivery optimization model is dynamically corrected based on preset dynamic correction data to obtain the target multi-vehicle delivery optimization model.

4. The method for determining a logistics vehicle dispatching scheme according to claim 3, characterized in that, The preset logistics and transportation constraints include: The actual load of each vehicle shall not exceed its maximum load, as expressed in the following formula: , Where N is the total number of customer points, The required weight for the i-th customer point. Variables are between 0 and 1. The maximum load capacity of the k-th vehicle; The actual volume of each vehicle does not exceed its maximum volume, expressed as: , ;in, Let i be the volume required for the i-th customer point. Let the maximum volume of the k-th vehicle be denoted as 'k'. The vehicle's arrival time at the customer's location must fall within the customer's time window, expressed as: , ;in, Let the earliest arrival time be the i-th customer's location. Let i be the latest arrival time of the i-th customer point. The time it takes for the vehicle to arrive at the i-th customer point; The service time for each customer location is fixed at 20 minutes, expressed as follows: , ;in, The service time for the i-th customer point. Let be the distance between the i-th customer point and the j-th customer point. For vehicle speed, This represents the time it takes for the vehicle to arrive at the (i+1)th customer point. The number of each vehicle model used cannot exceed its available quantity, expressed as: , ;in, Let m be the set of vehicles of type m. Let m be the number of available models.

5. The method for determining a logistics vehicle dispatching scheme according to claim 3, characterized in that, The preset total delivery cost objective function is: ； Where f is the total delivery cost, and K is the total number of vehicles used. Let the starting cost of the kth vehicle be... The time window penalty cost for the kth vehicle. Let the energy consumption cost of the kth vehicle be... Let be the carbon emission cost of the kth vehicle.

6. The method for determining a logistics vehicle dispatching scheme according to claim 3, characterized in that, The initial multi-vehicle delivery optimization model is dynamically corrected based on preset dynamic correction data to obtain the target multi-vehicle delivery optimization model, including: Acquire preset dynamic correction data; the preset dynamic correction data includes vehicle speed correction data, vehicle energy consumption correction data, and vehicle carbon emission correction data; The preset model fixed parameters in the initial multi-vehicle delivery optimization model are corrected based on the preset dynamic correction data to obtain the target multi-vehicle delivery optimization model.

7. The method for determining a logistics vehicle dispatching scheme according to claim 1, characterized in that, Solving the target multi-vehicle delivery optimization model yields an initial logistics vehicle scheduling scheme, including: Multiple initial vehicle scheduling combinations are obtained; all of the multiple initial vehicle scheduling combinations satisfy the target constraints of the target multi-vehicle delivery optimization model. The multiple initial vehicle scheduling combinations are encoded to obtain an initial population containing multiple chromosomes; wherein, one chromosome represents one vehicle scheduling combination, and the population represents the set of all chromosomes. Based on the initial population and the preset total delivery cost objective function of the target multi-vehicle delivery optimization model, the fitness value of the chromosome is obtained; A new population is determined by selection, crossover, and mutation based on the fitness values ​​of the chromosomes and the initial population. Based on the new population and the preset iteration termination condition, an initial logistics vehicle scheduling scheme is obtained.

8. The method for determining a logistics vehicle dispatching scheme according to claim 1, characterized in that, The target multi-vehicle delivery optimization model is dynamically updated based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model, including: Acquire real-time variable data; the real-time variable data includes at least one of real-time order variable data, real-time address variable data, and real-time customer point time window variable data. The target constraints of the target multi-vehicle delivery optimization model are dynamically updated based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model.

9. A device for determining a logistics vehicle dispatching scheme, characterized in that, include: The acquisition module is used to acquire logistics order data, customer location data, and vehicle data. The processing module is used to extract target data from the logistics order data, the customer point data, and the vehicle data to obtain target data; Based on the target data, the preset total delivery cost objective function, and the preset logistics transportation constraints, a target multi-vehicle delivery optimization model is determined; the target multi-vehicle delivery optimization model is solved to obtain an initial logistics vehicle scheduling scheme. The target multi-vehicle delivery optimization model is dynamically updated based on real-time variable data to obtain a dynamic multi-vehicle delivery optimization model; the initial logistics vehicle scheduling scheme is dynamically adjusted based on the dynamic multi-vehicle delivery optimization model to obtain a target logistics vehicle scheduling scheme.

10. A computing device, characterized in that, include: A processor, a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method as described in any one of claims 1 to 8.

Images