A split center site selection method considering fuel consumption of transport vehicles

Abstract

The application discloses a kind of methods for site selection of distribution center considering fuel consumption of transport vehicle, by collecting the geographic location information of express station, the demand of shared express box of each express station, the express station information such as the amount to be recycled, shared express box recovery rate, shared express box loss rate in transit, customer shared express box loss rate, shared express box loss rate of express station, vehicle operation information such as vehicle maximum capacity, vehicle basic rent, vehicle driving speed, while adding vehicle fuel consumption information for different situations, including the road condition fuel consumption of shared express box collected and transported from each express station to distribution center by transport vehicle, distribution center internal transportation fuel consumption, fuel consumption when loading and unloading shared express box and other factors, the application solves the site selection model of distribution center considering fuel consumption of transport vehicle by improved whale optimization algorithm, to obtain the site selection method of distribution center with low fuel consumption index of transport vehicle and low total logistics cost.

Description

Technical Field

[0001] This invention relates to the field of forward and reverse logistics distribution management technology for shared express boxes, and in particular to a method for selecting the location of a distribution center that takes into account the fuel consumption of transport vehicles. Background Technology

[0002] Against the backdrop of booming e-commerce, the volume of express delivery business continues to climb, placing higher demands on the efficiency, cost control, and environmental performance of express delivery systems. Traditional express delivery models typically rely on a large number of transport vehicles for the collection, distribution, and delivery of goods. To address these challenges, the express delivery industry has begun exploring the application of shared express boxes, aiming to achieve resource conservation and environmental protection by reducing the use of packaging materials and reusing express boxes.

[0003] However, simply relying on shared parcel boxes is insufficient to comprehensively solve the problem of resource waste in express delivery. Factors such as the driving distance, load capacity, and route planning of transport vehicles directly affect fuel consumption, and these factors largely depend on the location and layout of express distribution centers. Reasonable location selection can not only shorten transportation distances and reduce empty loads and redundant transportation, thereby lowering fuel consumption, but also improve delivery efficiency and enhance customer satisfaction.

[0004] In the current express delivery system, the location of express distribution centers is often based on experience or simple cost-benefit analysis, rarely considering the crucial factor of vehicle fuel consumption. This practice has led to some distribution centers being poorly located, with transport vehicles frequently traveling to remote or inaccessible areas, resulting in high fuel consumption, low efficiency, and increased operating costs.

[0005] To address this issue, this invention proposes a distribution center location method that considers vehicle fuel consumption. This method aims to optimize the express delivery network, reduce vehicle fuel consumption, and improve overall delivery efficiency through scientific location decisions. Summary of the Invention

[0006] To overcome or at least partially solve the above problems, the present invention provides a method for selecting a distribution center location that takes into account the fuel consumption of transport vehicles, thereby reducing costs.

[0007] The technical solution adopted in this invention is: a method for selecting a distribution center considering the fuel consumption of transport vehicles, comprising the following steps:

[0008] Step S1: Obtain information on express delivery stations, shared express box recycling, vehicle operation, and vehicle fuel consumption under different conditions;

[0009] Step S2: Establish a total transportation cost evaluation mechanism that takes into account vehicle fuel consumption, and add a penalty mechanism to this evaluation mechanism to improve efficiency;

[0010] S3: A diversion center location model that takes into account transportation vehicle fuel consumption is designed by combining factors such as express station information, shared express box recycling information, vehicle operation information, and vehicle fuel consumption information under different conditions.

[0011] S4: Based on the established diversion center location model, an improved whale optimization algorithm is designed to optimize the location method. Combined with the total transportation cost evaluation mechanism that considers vehicle fuel consumption, the optimal total cost diversion center location result is obtained.

[0012] Furthermore, the express station location information of this invention includes the express station's geographical location information, the demand for shared express boxes at each express station, and the amount to be recycled. The express station location information set is V. i = (v1, v2, v3, v4), where v1 represents longitude, v2 represents latitude, v3 represents the demand for shared express boxes, and v4 represents the amount of shared express boxes to be recycled.

[0013] Furthermore, the shared express box recycling information of this invention includes: shared express box recycling rate, shared express box loss rate during transportation, customer shared express box loss rate, and express station shared express box loss rate.

[0014] Furthermore, the vehicle operation information of this invention includes the maximum capacity of the transport vehicle, the basic rental fee for the vehicle, and the vehicle's driving speed. The vehicle fuel consumption information for different situations includes road condition fuel consumption when the transport vehicle collects and transports shared express boxes from various express delivery stations to the distribution center, fuel consumption during internal transportation within the distribution center, and fuel consumption during loading and unloading of shared express boxes.

[0015] Furthermore, this invention establishes a total transportation cost evaluation mechanism that considers vehicle fuel consumption. The vehicle fuel consumption evaluation index is S, and the evaluation mechanism set is as follows:

[0016] B = (B1, B2, B3)

[0017] Where B1 is the description of the shared parcel box, scored according to the actual situation; B2 is the unpaid shipping fee; and B3 is the penalty situation for the shared parcel box, scored according to the actual situation. Let the comprehensive score set of each factor in the factor set B = (B1, B2, B3) be B = (B′1, B′2, B′3), B′ n The maximum estimated transportation cost is defined by the shared parcel box recycling company itself, and the total transportation cost evaluation mechanism is b. h ,

[0018] b h =B′1+B′2+(B′ n -B′3)

[0019]

[0020] Where F iD represents the fuel consumption rate from the i-th express delivery station to the distribution center. i G represents the length of the transportation path from the i-th express delivery station to the distribution center. fuel H represents the unit price of fuel. i Q represents the road condition coefficient from the i-th express delivery station to the distribution center. i Let Q represent the loading and unloading cost coefficient of the i-th express station. maintenance L represents the unit loading and unloading cost. i L represents the internal transportation coefficient of the vehicle at the i-th express station within the distribution center. maintenance This represents the internal transportation cost within the distribution center.

[0021] Furthermore, after each transportation operation, the present invention evaluates the diversion center according to an evaluation mechanism to select the route with the lowest total fuel consumption cost, thereby ensuring the stability of fuel consumption for transport vehicles.

[0022] Furthermore, the vehicle fuel consumption evaluation index of this invention is determined by vehicle performance, driving behavior, road conditions, and load conditions, and is used to assist the vehicle fuel consumption evaluation mechanism.

[0023] S=γ×VPFCI+δ×DBFCI+μ×RCFCI+ρ×LCFCI

[0024] Where S is the total score, γ, δ, μ, and ρ represent the weighting coefficients of each indicator, and their sum is 1, i.e., γ + δ + μ + ρ = 1. VPFCI represents the vehicle performance fuel consumption index, and its formula is:

[0025]

[0026] FCR represents the fuel consumption rate per 100 kilometers, and its formula is: FCR represents the baseline fuel consumption per 100 kilometers, and EEI represents the engine efficiency index, which reflects the engine's operating condition. It is the ratio of the engine's actual power to its rated power, and its formula is: ω1 and ω2 are weighting coefficients, set to default values ​​of ω1 = 0.6 and ω2 = 0.4.

[0027] DBFCI stands for Driving Behavior Fuel Consumption Index, and its formula is:

[0028] DBFCI=ω3×(e HAT )+ω4×(e HDT )+ω5×SDR,

[0029] ω3, ω4, and ω5 are weighting coefficients, set to default values: ω3 = 0.4, ω4 = 0.5, ω5 = 0.1. HAT represents the number of rapid accelerations (the number of times the vehicle accelerates rapidly per unit time), HDT represents the number of rapid decelerations (the number of times the vehicle decelerates rapidly per unit time), and SDR represents the smooth driving rate, reflecting the smoothness of vehicle speed changes during driving.

[0030] Wherein, RCFCI represents the fuel consumption index under road conditions, and its formula is:

[0031] RCFCI=ω6×(1-CI)+ω7×AS

[0032] ω6 and ω7 are weighting coefficients, set to default values: ω6 = 0.4 and ω7 = 0.6. Cl represents the degree of congestion, an indicator used to reflect the degree of road congestion. AS represents the average vehicle speed, which represents the average speed of a vehicle during its journey.

[0033] Wherein, LCFCI represents the fuel consumption index under load conditions, and its formula is:

[0034]

[0035] T M Indicates the actual load. Indicates the full load capacity, C M This indicates fuel consumption.

[0036] S is a vehicle fuel consumption evaluation index, which is determined by vehicle performance, driving behavior, road conditions, and load conditions, and is used to assist in the vehicle fuel consumption evaluation mechanism.

[0037] S=γ×VPFCI+δ×DBFCI+μ×RCFCI+ρ×LCFCI

[0038] Where S is the total score, γ, δ, μ, and ρ represent the weighting coefficients of each indicator, and their sum is 1, i.e., γ + δ + μ + ρ = 1. VPFCI represents the vehicle performance fuel consumption index, and its formula is:

[0039]

[0040] FCR represents the fuel consumption rate per 100 kilometers, and its formula is: FCR' represents the baseline fuel consumption per 100 kilometers, and EEI represents the engine efficiency index, which reflects the engine's operating condition. It is the ratio of the engine's actual power to its rated power, and its formula is: ω1 and ω2 are weighting coefficients, set to default values ​​of ω1 = 0.6 and ω2 = 0.4.

[0041] DBFCI stands for Driving Behavior Fuel Consumption Index, and its formula is:

[0042] DBFCI=ω3×(e HAT )+ω4×(e HDT )+ω5×SDR,

[0043] ω3, ω4, and ω5 are weighting coefficients, set to default values: ω3 = 0.4, ω4 = 0.5, and ω5 = 0.1. HAT represents the number of rapid accelerations, i.e., the number of times the vehicle accelerates rapidly per unit time; HDT represents the number of rapid decelerations, i.e., the number of times the vehicle decelerates rapidly per unit time; and SDR represents the smooth driving rate, reflecting the smoothness of vehicle speed changes during driving.

[0044] Wherein, RCFCI represents the fuel consumption index under road conditions, and its formula is:

[0045] RCFCI=ω6×(1-CI)+ω7×AS

[0046] ω6 and ω7 are weighting coefficients, set to default values: ω6 = 0.4 and ω7 = 0.6. Cl represents the degree of congestion, an indicator used to reflect the degree of road congestion. AS represents the average vehicle speed, which represents the average speed of a vehicle during its journey.

[0047] Wherein, LCFCI represents the fuel consumption index under load conditions, and its formula is:

[0048]

[0049] T M Indicates the actual load. Indicates the full load capacity, C M This indicates fuel consumption.

[0050] As a method for selecting a traffic splitting center considering vehicle fuel consumption according to the present invention, the traffic splitting center selection model includes:

[0051] The objective function is:

[0052] minC = C1 + C2 + C3 + C4 + C5

[0053] Where C1 represents the basic cost, including the construction cost of the distribution center, the purchase cost of transport vehicles, the operating cost of the distribution center, and the depreciation cost of the vehicles. K represents the numerical value of the basic cost, which remains constant. The expression for the basic cost is as follows:

[0054]

[0055] Where C2 represents the cost of the penalty mechanism, including the cost of waiting for vehicles to be recycled at the recycling station and the cost of delay in using the shared parcel box, i.e., the cost of delay in recycling the vehicles. Let p a P represents the on-site waiting cost for vehicle recycling. ap represents the maximum non-penalty waiting cost within the vehicle recovery area. b P represents the delay cost of recycling station vehicles per unit of time. b Let t represent the maximum non-penalized delay cost within the area where the vehicle is recovered. a This represents the time it takes for the recycling vehicle to arrive at parcel station a, which is waiting to collect the shared parcel boxes to be recycled. The cost of the penalty mechanism is as follows:

[0056]

[0057] C3 represents the cost of recycling shared parcel boxes, which includes the cost of damage to the boxes, the cost of recycling the boxes, and the cost of loading and unloading the boxes. K a Let L represent the number of shared parcel boxes to be recycled at parcel station a, L represent the loss rate of shared parcel boxes during delivery, M represent the cost of a single shared parcel box, and H represent the sum of the recycling cost and loading / unloading cost of the shared parcel boxes. The expression for the recycling cost is as follows:

[0058]

[0059] Where C4 represents the transportation cost during the forward and reverse logistics of the shared express box, and S... a K represents the distance from courier station a to the shared parcel box distribution center. a Z represents the amount of shared parcel boxes to be recycled at parcel station a, which is also the amount of shared parcel boxes to be transported at each station. a This represents the demand of courier station a for shared parcel boxes that have already been recycled, and c a This represents the logistics cost required to transport shared express boxes within a unit distance. The expression for the transportation cost is as follows:

[0060]

[0061] Where C5 represents the sorting cost of the shared parcel box, and F represents the sorting time cost of a single shared parcel box. The sorting cost is as follows:

[0062] C5 = K a *F

[0063] As a method for selecting a traffic splitting center considering vehicle fuel consumption as described in this invention, the constraint condition is as follows:

[0064] Condition 1: The total weight of the shared parcel boxes loaded on the transport vehicle cannot exceed the vehicle's load capacity, as shown below:

[0065] ∑ a＝1 X a ≤X all ,

[0066] Condition 2: The volume of goods loaded on the vehicle cannot exceed the vehicle's capacity, expressed as:

[0067]

[0068] Condition 3: Must comply with local environmental regulations and urban planning strategies;

[0069] Condition 4: The amount of shared parcel boxes recycled should always be less than the capacity of the recycling center;

[0070]

[0071] As described in this invention, a method for selecting a traffic splitting center that considers vehicle fuel consumption, wherein:

[0072] To mitigate the risk of the whale optimization algorithm getting stuck in local optima in the later stages, an improved whale optimization algorithm (ZWOA) is designed based on the established distribution center location model. The improved ZWOA algorithm is used to refine the target mathematical model, thereby obtaining the final location scheme for the shared parcel box distribution center. The improved ZWOA algorithm includes the following steps:

[0073] A local search phase is added after each main loop iteration. The local search is achieved by generating multiple candidate solutions near the leader's position and evaluating these solutions. In the local search phase, a candidate solution is generated, and its fitness value is checked to see if it is better than the current leader. If so, the leader's position and score are updated. This helps the algorithm perform a more refined search in local regions of the search space, thereby increasing the likelihood of finding the global optimum.

[0074] Candidate_pos = X leader +(u b -l b )*(r-0.5),

[0075] Where r represents a random number between [0, 1], u b and l b These represent the upper and lower bounds of the search space, respectively.

[0076] As described in this invention, a method for selecting a traffic splitting center that considers vehicle fuel consumption, wherein:

[0077] An adaptive step size adjustment factor is introduced to make the algorithm more flexible. The specific steps are as follows:

[0078] Step 1: Introduce an algorithm parameter 'a', representing the spiral shape and shrinking encirclement mechanism that influence the prey's encirclement. This parameter changes non-linearly with the number of iterations, rather than linearly as in standard WOA, helping the algorithm better balance global exploration and local exploitation capabilities during the search process. The formula for α is:

[0079]

[0080] Where, α initial and α final These are the initial and final values ​​of α, respectively; t is the current iteration number; and Max_iter is the maximum iteration number.

[0081] Step Two: Introduce an algorithm parameter 'a2' to calculate the spiral shape. This parameter changes non-linearly with the number of iterations, unlike the linear change in standard WOA, helping the algorithm better balance global exploration and local exploitation capabilities during the search process. The formula for α2 is:

[0082]

[0083] Where, β initial and β final These are the initial and final values ​​of α2, respectively.

[0084] Step 3: Introduce a dynamically adjusted algorithm parameter 'b' to influence the shrinkage rate of the spiral shape, which changes with the number of iterations to alter the tightness of the spiral shape. This helps improve global search capabilities.

[0085] The formula for expressing b:

[0086]

[0087] Where b initial and b final These are the initial and final values ​​of b, respectively.

[0088] As described in this invention, a method for selecting a traffic splitting center that considers vehicle fuel consumption, wherein:

[0089] For cases where the position exceeds the search space boundary, the improved whale optimization algorithm adjusts the position back within the boundary. This boundary handling mechanism ensures that the algorithm always remains within the effective search space during the search process.

[0090] The whale's position in a certain dimension is x. i The lower bound of the search space in this dimension is L. i The upper bound is U i If the whale's new position x′ i If the search exceeds the boundaries of the search space, adjustments are needed. The adjusted position is x″. i for:

[0091]

[0092] If the whale's new location x′ i Less than the lower bound Li Then adjust its position to the lower bound L. i If the whale's new position x′ i In the lower bound L i and the upper realm U i Within the area (including the boundary), its position remains unchanged; if the whale's new position x′ i Greater than the upper bound U i Then adjust its position to the upper bound U. i .

[0093] The improved whale optimization algorithm ZWOA also records the leader's fitness value (i.e., the quality of the optimal solution) after each iteration and stores these values ​​in the Convergence_curve array. This helps in analyzing the algorithm's convergence performance and stability.

[0094] As described in this invention, a method for selecting a traffic splitting center that considers vehicle fuel consumption, wherein:

[0095] The objective function of the diversion center location method model is solved using the improved whale optimization algorithm. The steps include the following:

[0096] 1) Algorithm initialization and parameter definition

[0097] During the initialization phase of the algorithm, the location of each whale is set to a random value for the decision variable, which should be within the given search space (i.e., the range of possible locations).

[0098] 2) Fitness evaluation: For each whale's location (i.e., the location of each candidate diversion center), the fitness value is calculated using an objective function.

[0099] 3) Whale location update

[0100] Based on the principle of the whale optimization algorithm, the whale will search around the current optimal position (i.e. the current optimal diversion center position), randomly select the positions of other whales as reference points, and update its own position in a spiral manner.

[0101] 4) Local search and evaluation

[0102] Generate candidate solutions within a certain range of the current optimal position and evaluate their fitness values ​​to determine whether to exit the loop. When the algorithm converges or reaches the maximum number of iterations, output the optimal splitting center position and its corresponding fitness value. Otherwise, return to step 2 above.

[0103] 5) Location Decision: Based on the optimal distribution center location output by the algorithm, the final location decision and minimum value C are determined. MIN .

[0104] Beneficial effects:

[0105] 1. This invention uses an improved whale optimization algorithm, which optimizes the algorithm to a certain extent at each stage, so that the algorithm always stays within the effective search space during the search process, and can better balance global exploration and local development capabilities during the search process, and can provide a more reasonable distribution center location scheme.

[0106] 2. This invention uses an improved whale optimization algorithm to solve the distribution center location model that considers the fuel consumption of transport vehicles, so as to obtain a distribution center location method that meets the requirements of low fuel consumption of transport vehicles and low total logistics cost. Attached Figure Description

[0107] Figure 1 This is a flowchart of the forward and reverse logistics distribution process proposed in this invention.

[0108] Figure 2 This is a detailed flowchart of the overall recycling process proposed in this invention.

[0109] Figure 3 A comparison chart showing the number of iterations of the improved Whale Optimization Algorithm (ZWOA) with BWO, DBO, BOA, SCA, and WOA algorithms.

[0110] Figure 4 For specific implementation of the map.

[0111] Figure 5 This is a comparison chart of the location selection and cost of the improved whale optimization algorithm of this invention.

[0112] Figure 6 This is a schematic diagram of the reverse logistics planning scheme based on the improved whale optimization algorithm of this invention. Detailed Implementation

[0113] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are merely illustrative and are not intended to limit the present invention.

[0114] A method for selecting a distribution center that considers the fuel consumption of transport vehicles, the method comprising the following steps:

[0115] Step S1: Obtain information on express delivery stations, shared express box recycling information, and vehicle operation information.

[0116] Step S2: Obtain vehicle fuel consumption information and establish a total transportation cost evaluation mechanism that takes vehicle fuel consumption into account.

[0117] Step S3: Design a diversion center location model that takes into account the fuel consumption of transport vehicles by combining factors such as express station information, shared express box recycling information, vehicle operation information, and vehicle fuel consumption information under different conditions.

[0118] Step S4: Based on the established diversion center location model, design an improved whale optimization algorithm to optimize the location method, and combine it with the total transportation cost evaluation mechanism oriented towards vehicle fuel consumption to obtain the optimal total cost diversion center location result.

[0119] The following is a detailed description of the specific implementation of the above-mentioned multimodal transport scheme for the Yangtze River Economic Belt provided by the present invention through a specific embodiment.

[0120] like Figure 1 As shown, the present invention provides a method for selecting a diversion center location considering the fuel consumption of transport vehicles, comprising the following steps:

[0121] Step 1: Collect information on express delivery stations, share information on express box recycling, and information on vehicle operation.

[0122] The information on express delivery stations includes the longitude and latitude of the station's location, the demand for shared express boxes at each station, and the amount to be recycled.

[0123] Information on shared parcel box recycling includes the shared parcel box recycling rate, the shared parcel box loss rate during transportation, the shared parcel box loss rate for customers, and the shared parcel box loss rate at parcel stations.

[0124] Vehicle operation information includes the maximum capacity of transport vehicles, the basic rental fee for vehicles, and the vehicle speed.

[0125] Step 2: Collect vehicle fuel consumption information.

[0126] The vehicle fuel consumption information includes factors such as road fuel consumption when transporting shared express boxes from various express stations to the distribution center, fuel consumption during internal transportation within the distribution center, and fuel consumption during loading and unloading of shared express boxes.

[0127] The total transportation cost evaluation mechanism B for vehicle fuel consumption refers to the evaluation based on the description of shared express boxes used in vehicle transportation, the freight charges to be paid, and the penalties imposed on shared express boxes. The comprehensive vehicle fuel consumption score also incorporates the vehicle fuel consumption evaluation index S, which is determined by vehicle performance, driving behavior, road conditions, and load conditions.

[0128] Step 3: Establish a total transportation cost evaluation mechanism B for vehicle fuel consumption, and vehicle fuel consumption assessment indicators S. The evaluation mechanism B consists of:

[0129] B = (B1, B2, B3),

[0130] Where B1 is the description of the shared parcel box, scored according to the actual situation; B2 is the unpaid shipping fee; and B3 is the penalty situation for the shared parcel box, scored according to the actual situation. Let the comprehensive score set of each factor in the factor set B = (B1, B2, B3) be B = (B′1, B′2, B′3), B′n The maximum estimated transportation cost is defined by the shared parcel box recycling company itself, and the total transportation cost evaluation mechanism is b. h ,

[0131] b h =B′1+B′2+(B′ n -B′3)

[0132]

[0133] Where F i D represents the fuel consumption rate from the i-th express delivery station to the distribution center. i G represents the length of the transportation path from the i-th express delivery station to the distribution center. fuel H represents the unit price of fuel. i Q represents the road condition coefficient from the i-th express delivery station to the distribution center. i Let Q represent the loading and unloading cost coefficient of the i-th express station. maintenance L represents the unit loading and unloading cost. i L represents the internal transportation coefficient of the vehicle at the i-th express station within the distribution center. maintenance This represents the internal transportation cost within the distribution center.

[0134] After each transport operation, the distribution center is evaluated according to the evaluation mechanism to select the route with the lowest total fuel consumption cost, thus ensuring the stability of fuel consumption for transport vehicles.

[0135] The vehicle fuel consumption assessment index S is determined by vehicle performance, driving behavior, road conditions, and load conditions, and is used to assist the vehicle fuel consumption evaluation mechanism.

[0136] S=γ×VPFCI+δ×DBFCI+μ×RCFCI+ρ×LCFCI

[0137] Where S is the total score, γ, δ, μ, and ρ represent the weighting coefficients of each indicator, and their sum is 1, i.e., γ + δ + μ + ρ = 1. VPFCI represents the vehicle performance fuel consumption index, and its formula is:

[0138]

[0139] FCR represents the fuel consumption rate per 100 kilometers, and its formula is: FCR' represents the baseline fuel consumption per 100 kilometers, and EEI represents the engine efficiency index, which reflects the engine's operating condition. It is the ratio of the engine's actual power to its rated power, and its formula is: ω1 and ω2 are weighting coefficients, set to default values ​​of ω1 = 0.6 and ω2 = 0.4.

[0140] DBFCI stands for Driving Behavior Fuel Consumption Index, and its formula is:

[0141] DBFCI=ω3×(e HAT )+ω4×(e HDT )+ω5×SDR,

[0142] ω3, ω4, and ω5 are weighting coefficients, set to default values: ω3 = 0.4, ω4 = 0.5, ω5 = 0.1. HAT represents the number of rapid accelerations (the number of times the vehicle accelerates rapidly per unit time), HDT represents the number of rapid decelerations (the number of times the vehicle decelerates rapidly per unit time), and SDR represents the smooth driving rate, reflecting the smoothness of vehicle speed changes during driving.

[0143] Wherein, RCFCI represents the fuel consumption index under road conditions, and its formula is:

[0144] RCFCI=ω6×(1-CI)+ω7×AS

[0145] ω6 and ω7 are weighting coefficients, set to default values: ω6 = 0.4 and ω7 = 0.6. Cl represents the degree of congestion, an indicator used to reflect the degree of road congestion. AS represents the average vehicle speed, which represents the average speed of a vehicle during its journey.

[0146] Wherein, LCFCI represents the fuel consumption index under load conditions, and its formula is:

[0147]

[0148] T M Indicates the actual load. Indicates the full load capacity, C M This indicates fuel consumption.

[0149] Step 4: Establish a location model for the diversion center.

[0150] The objective function is:

[0151] minC = C1 + C2 + C3 + C4 + C5

[0152] Where C1 represents the basic cost, including the construction cost of the distribution center, the purchase cost of transport vehicles, the operating cost of the distribution center, and the depreciation cost of the vehicles. K represents the numerical value of the basic cost, which remains constant. The expression for the basic cost is as follows:

[0153]

[0154] Where C2 represents the cost of the penalty mechanism, including the cost of waiting for vehicles to be recycled at the recycling station and the cost of delay in using the shared parcel box, i.e., the cost of delay in recycling the vehicles. Let p a P represents the on-site waiting cost for vehicle recycling. ap represents the maximum non-penalty waiting cost within the vehicle recovery area. b P represents the delay cost of recycling station vehicles per unit of time. b Let t represent the maximum non-penalized delay cost within the area where the vehicle is recovered. a This represents the time it takes for the recycling vehicle to arrive at parcel station a, which is waiting to collect the shared parcel boxes to be recycled. The cost of the penalty mechanism is as follows:

[0155]

[0156] C3 represents the cost of recycling shared parcel boxes, which includes the cost of damage to the boxes, the cost of recycling the boxes, and the cost of loading and unloading the boxes. K a Let L represent the number of shared parcel boxes to be recycled at parcel station a, L represent the loss rate of shared parcel boxes during delivery, M represent the cost of a single shared parcel box, and H represent the sum of the recycling cost and loading / unloading cost of the shared parcel boxes. The expression for the recycling cost is as follows:

[0157]

[0158] Where C4 represents the transportation cost during the forward and reverse logistics of the shared express box, and S... a K represents the distance from courier station a to the shared parcel box distribution center. a Z represents the amount of shared parcel boxes to be recycled at parcel station a, which is also the amount of shared parcel boxes to be transported at each station. a This represents the demand of courier station a for shared parcel boxes that have already been recycled, and c a This represents the logistics cost required to transport shared express boxes within a unit distance. The expression for the transportation cost is as follows:

[0159]

[0160] Where C5 represents the sorting cost of the shared parcel box, and F represents the sorting time cost of a single shared parcel box. The sorting cost is as follows:

[0161] C5 = K a *F

[0162] Compared to previous site selection models, this model adds a penalty mechanism for classification discussions. At the same time, the site selection model for distribution centers also incorporates the classification cost C5 of shared express boxes, which helps to alleviate the clutter in the process.

[0163] The constraints of the diversion center location model are:

[0164] Condition 1: The total weight of the shared parcel boxes loaded on the transport vehicle cannot exceed the vehicle's load capacity, as shown below:

[0165] ∑a＝1 X a ≤X all ,

[0166] Condition 2: The volume of goods loaded on the vehicle cannot exceed the vehicle's capacity, expressed as:

[0167]

[0168] Condition 3: Must comply with local environmental regulations and urban planning strategies;

[0169] Condition 4: The amount of shared parcel boxes recycled should always be less than the capacity of the recycling center;

[0170]

[0171] Step 5: Solve the objective function using the improved whale optimization algorithm.

[0172] The specific steps include the following:

[0173] 1) Algorithm initialization and parameter definition

[0174] During the initialization phase of the algorithm, the location of each whale is set to a random value for the decision variable, which should be within the given search space (i.e., the range of possible locations).

[0175] 2) Fitness assessment

[0176] For each whale's location (i.e., the location of each candidate split center), its fitness value is calculated using an objective function.

[0177] 3) Whale location update

[0178] Based on the principle of the whale optimization algorithm, the whale will search around the current optimal position (i.e. the current optimal diversion center position), randomly select the positions of other whales as reference points, and update its own position in a spiral manner.

[0179] An adaptive step size adjustment factor is introduced to make the algorithm more flexible. The specific steps are as follows:

[0180] Step 1: Introduce an algorithm parameter 'a', representing the spiral shape and shrinking encirclement mechanism that influence the prey's encirclement. This parameter changes non-linearly with the number of iterations, rather than linearly as in standard WOA, helping the algorithm better balance global exploration and local exploitation capabilities during the search process. The formula for α is:

[0181]

[0182] Where, α initial and α finalThese are the initial and final values ​​of α, respectively; t is the current iteration number; and Max_iter is the maximum iteration number.

[0183] Step 2: Introduce an algorithm parameter a2 to calculate the parameters of the spiral shape. It changes non-linearly with the number of iterations, rather than linearly as in the standard WOA. This helps the algorithm better balance global exploration and local exploitation capabilities during the search process.

[0184] The formula for α2 is:

[0185]

[0186] Where, β initial and β final These are the initial and final values ​​of α2, respectively.

[0187] Step 3: Introduce a dynamically adjusted algorithm parameter b to influence the shrinkage rate of the spiral shape. This parameter changes with the number of iterations to alter the tightness of the spiral shape, which helps improve the global search capability.

[0188] The formula for expressing b:

[0189]

[0190] Where b initial and b final These are the initial and final values ​​of b, respectively.

[0191] 4) Local search and evaluation

[0192] Generate candidate solutions within a certain range of the current optimal position and evaluate their fitness values ​​to determine whether to exit the loop. When the algorithm converges or reaches the maximum number of iterations, output the optimal splitting center position and its corresponding fitness value. Otherwise, return to step 2 above.

[0193] A local search phase is added after each main loop iteration. The local search is achieved by generating multiple candidate solutions near the leader's position and evaluating these solutions. In the local search phase, a candidate solution is generated, and its fitness value is checked to see if it is better than the current leader. If so, the leader's position and score are updated. This helps the algorithm perform a more refined search in local regions of the search space, thereby increasing the likelihood of finding the global optimum.

[0194] Candidate_pos = X leader +(ub-lb)*(r-0.5),

[0195] Where r represents a random number between [0, 1], and ub and lb represent the upper and lower bounds of the search space, respectively.

[0196] For cases where the position exceeds the search space boundary, the improved whale optimization algorithm adjusts the position back within the boundary. This boundary handling mechanism ensures that the algorithm always remains within the effective search space during the search process.

[0197] The whale's position in a certain dimension is x. i The lower bound of the search space in this dimension is L. i The upper bound is U i If the whale's new position x′ i If the search exceeds the boundaries of the search space, adjustments are needed. The adjusted position is x″. i for:

[0198]

[0199] If the whale's new location x′ i At the lower bound L i Then adjust its position to the lower bound L. i If the whale's new position x′ i In the lower bound L i and the upper realm U i Within the area (including the boundary), its position remains unchanged; if the whale's new position x′ i Greater than the upper bound U i Then adjust its position to the upper bound U. i .

[0200] The improved whale optimization algorithm ZWOA also records the fitness value of the leader (i.e. the quality of the optimal solution) after each iteration and stores these values ​​in the Convergence_curve array.

[0201] 5) Site selection decision

[0202] Based on the optimal distribution center location output by the algorithm, the final location decision and the minimum value C are determined. MIN .

[0203] This invention proposes a method for selecting a distribution center location that considers the fuel consumption of transport vehicles. Based on existing location models, it fully considers vehicle fuel consumption information under different conditions and proposes a distribution center location model that considers vehicle fuel consumption. This method selects the most suitable distribution center location, thereby achieving the lowest total transportation cost for shared express boxes, alleviating the complexity of recycling center processes, and reducing the fuel consumption of transport vehicles, effectively reducing time costs and logistics pressure.

[0204] like Figure 1As shown, this invention provides a method for establishing a distribution center to optimize reverse logistics transportation routes, specifically including obtaining the location information of express stations; in this implementation case, based on the actual situation, the latitude and longitude information of 15 express stations and one warehouse station in a certain area was obtained from Baidu Maps, as shown in Table 1.

[0205] Table 1: Latitude and Longitude Coordinates of Each Express Delivery Station on the Detailed Implementation Map

[0206]

[0207]

[0208] The simulation experimental data of this invention includes the following:

[0209] In this experiment, there were many constraints and management parameters, all of which needed to be preset manually. These included factors such as the construction cost of the distribution center, the cost of transport vehicles, handling time, and the loss rate of shared parcel boxes during transportation. All data was preset by the researcher and set to default values. The coordinates of each node are as follows: 118.88844, 31.691516, 2;

[0210] 118.8663158,31.953849,2;

[0211] 118.74873,31.883382,2;

[0212] 118.79523,31.943797,2;

[0213] 118.927347,31.910045,2;

[0214] 118.912142,31.901284,2;

[0215] 118.787558,31.774961,2;

[0216] 118.847503,31.849141,2;

[0217] 118.725652,31.637603,2;

[0218] 118.783823,31.780662,2;

[0219] 119.134461,32.225449,2;

[0220] 118.835123,32.327436,2;

[0221] 118.842206,31.964061,2;

[0222] 118.89277,31.710644,2;

[0223] 118.611232,32.063585,2;

[0224] 118.825168,31.764453,20

[0225] This invention is implemented in MATLAB, based on Figure 3 The comparison chart of the improved whale optimization algorithm (ZWOA) with BWO, DBO, BOA, SCA and WOA algorithms shows that the ZWOA algorithm is superior to these algorithms.

[0226] This invention compares the value of the location selection for the distribution center using the Improved Whale Optimization Algorithm (ZWOA) with other algorithms, based on... Figure 5 and Figure 3 It can be seen that the improved algorithm has significantly better convergence speed and addressing strategy than the traditional algorithm.

[0227] This invention employs an improved whale optimization algorithm, which optimizes the algorithm to a certain extent at each stage, ensuring that the algorithm remains within an effective search space throughout the search process. It also better balances global exploration and local development capabilities during the search process, and provides a more reasonable distribution center location scheme.

[0228] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for selecting a distribution center location considering the fuel consumption of transport vehicles, characterized in that, The method includes the following steps: Step S1: Obtain information on express delivery stations, shared express box recycling, vehicle operation, and vehicle fuel consumption under different conditions; Step S2: Establish a total transportation cost evaluation mechanism that takes into account vehicle fuel consumption, and add a penalty mechanism to this evaluation mechanism to improve efficiency; Establish a total transportation cost evaluation mechanism that considers vehicle fuel consumption. The vehicle fuel consumption evaluation index is S, and the evaluation mechanism set is... B=( , , )， in For descriptions of shared parcel boxes, please rate based on actual conditions. For unpaid shipping fees, To determine penalties for shared parcel boxes, a scoring system is established based on the actual situation. Let the factor set B = ( , , The comprehensive score set B of each factor in ) = ( , , ), The maximum estimated transportation cost is defined by the shared parcel box recycling company itself, and the total transportation cost evaluation mechanism is as follows. , = in Indicates the fuel consumption rate from the i-th express delivery station to the distribution center. Indicates the length of the transportation path from the i-th express delivery station to the distribution center. This indicates the unit price of fuel. This represents the road condition coefficient from the i-th express delivery station to the distribution center. This represents the loading and unloading cost coefficient of the i-th express station. This indicates the unit loading and unloading cost. This represents the internal transportation coefficient of the vehicle at the i-th express station within the distribution center. This indicates the internal transportation costs within the distribution center; Step S3: Combining information from express delivery stations, shared express box recycling, vehicle operation, and fuel consumption under different conditions, design a diversion center location model that considers the fuel consumption of transport vehicles. The method includes a diversion center location model, the objective function of which is: in, The basic cost represents the construction cost of the distribution center, the purchase cost of transport vehicles, the operating cost of the distribution center, and the depreciation cost of the vehicles. K represents the numerical value of the basic cost, which remains constant. The expression for the basic cost is as follows: = in, The cost of the penalty mechanism includes the cost of waiting for vehicles to be recycled at the recycling station and the cost of delay in using shared parcel boxes, i.e., the cost of delay in recycling vehicles. This indicates the cost of waiting at the recycling site for the vehicle. This represents the maximum non-penalty waiting cost within the area where the vehicle can be recycled. This represents the delay cost of vehicles at the recycling station per unit of time. Let represent the maximum non-penalized delay cost within the area where the vehicle is recovered. This represents the time it takes for the recycling vehicle to arrive at parcel station a, which is waiting to collect the shared parcel boxes to be recycled. The cost of the penalty mechanism is as follows: = in The cost of recycling shared parcel boxes includes the cost of damage to the boxes, the cost of recycling the boxes, and the cost of loading and unloading the boxes. This represents the number of shared parcel boxes awaiting recycling at parcel station a. Let M represent the loss rate of the shared parcel box during delivery, M represent the cost of a single shared parcel box, and H represent the total cost of recycling and loading / unloading of the shared parcel box. The expression for the recycling cost is as follows: = + in To reduce transportation costs during the forward and reverse logistics processes of shared express boxes. This represents the distance from courier station 'a' to the shared parcel box distribution center. This indicates the number of shared parcel boxes to be recycled at parcel station a, which is also the number of shared parcel boxes to be transported at each station. This indicates the demand of courier station A for shared courier boxes that have already been recycled. This represents the logistics cost required to transport shared express boxes within a unit distance. The expression for the transportation cost is as follows: = + in The sorting cost for shared parcel boxes is given by F, where F represents the time cost of sorting a single shared parcel box. The sorting cost is as follows: = *F； Step S4: Based on the established diversion center location model, design an improved whale optimization algorithm to optimize the location method, and combine it with the total transportation cost evaluation mechanism that considers vehicle fuel consumption to obtain the optimal total cost diversion center location result. To improve the situation where whale optimization algorithms get stuck in local optima in the later stages, an improved whale optimization algorithm is designed based on the established distribution center location model. The improved ZWOA (Zootral Whale Optimization Algorithm) is used to refine the target mathematical model, thereby obtaining the final location scheme for the shared parcel box distribution center. The improved ZWOA algorithm includes the following improvements: a local search phase is added after each main loop iteration. This local search is achieved by generating multiple candidate solutions near the leader's position and evaluating these solutions. During the local search phase, a candidate solution is generated, and its fitness value is checked to see if it is better than the current leader. If so, the leader's position and score are updated. This helps the algorithm to perform a more detailed search in the local region of the search space, thereby increasing the probability of finding the global optimum. = +( - )*(r-0.5)， Where r represents a random number between [0, 1], and These represent the upper and lower bounds of the search space, respectively.

2. The method for selecting a distribution center considering the fuel consumption of transport vehicles according to claim 1, characterized in that: The method includes express station location information; the express station location information includes the express station's geographical location information, the demand for shared express boxes at each express station, and the amount to be recycled, and the express station location information is collected as follows: , , , ),in Indicates longitude. Indicates latitude, This indicates the demand for shared parcel boxes. This indicates the number of shared parcel boxes awaiting recycling. The shared express box recycling information includes: shared express box recycling rate, shared express box loss rate during transportation, customer shared express box loss rate, and express station shared express box loss rate. The vehicle operation information includes the maximum capacity of the transport vehicle, the basic rental fee for the vehicle, and the vehicle speed. The vehicle fuel consumption information for different situations includes road condition fuel consumption when transporting shared express boxes from various express stations to the distribution center, internal transportation fuel consumption within the distribution center, and fuel consumption when loading and unloading shared express boxes. After each transport operation, the distribution center is evaluated according to the evaluation mechanism to select the route with the lowest total fuel consumption cost, thus ensuring the stability of fuel consumption for transport vehicles.

3. The method for selecting a distribution center considering the fuel consumption of transport vehicles according to claim 1, characterized in that, The location model for the diversion center must meet the following constraints, including: Condition 1: The total weight of the shared parcel boxes loaded on the transport vehicle cannot exceed the vehicle's load capacity, as shown below: , Condition 2: The volume of goods loaded on the vehicle cannot exceed the vehicle's capacity, expressed as: Condition 3: Must comply with local environmental regulations and urban planning strategies; Condition 4: The amount of shared parcel boxes recycled should always be less than the capacity of the recycling center; 。 4. The method for selecting a distribution center considering the fuel consumption of transport vehicles according to claim 1, characterized in that, The method is used to assist in the vehicle fuel consumption evaluation mechanism, namely: Where S is the total score. , , , This represents the weighting coefficients of each indicator, and their sum is 1. =1, in, The formula for representing a vehicle's fuel consumption performance index is: × ， The formula for expressing fuel consumption per 100 kilometers is: 100, This represents the baseline value for fuel consumption per 100 kilometers. The engine efficiency index reflects the engine's operating condition and is the ratio of the engine's actual power to its rated power. Its formula is: , and These are the weighting coefficients, set to the default value. =0.6, =0.4; in, The formula for representing fuel consumption based on driving behavior is: ， , These are the weighting coefficients, set to the default value: , =0.1, HAT represents the number of rapid accelerations, i.e., the number of times the vehicle accelerates rapidly per unit time; HDT represents the number of rapid decelerations, i.e., the number of times the vehicle decelerates rapidly per unit time; SDR represents the smooth driving rate, which reflects the smoothness of the vehicle's speed changes during driving. in, The formula for fuel consumption under road conditions is: These are the weighting coefficients, set to the default value: 0.6, CI represents the degree of congestion, an indicator used to reflect the degree of road congestion, and AS represents the average speed of vehicles during travel. in, The formula for indicating fuel consumption under load conditions is: Indicates the actual load. Indicates the full load weight. This indicates fuel consumption.

5. The method for selecting a distribution center considering the fuel consumption of transport vehicles according to claim 1, characterized in that, An adaptive step size adjustment factor is introduced to make the algorithm more flexible. The specific steps are as follows: Step 1: Introduce an algorithm parameter This indicates that the spiral shape and shrinking encirclement mechanism that affect the prey change non-linearly with the number of iterations, rather than linearly as in the standard WOA, which helps the algorithm to better balance global exploration and local exploitation capabilities during the search process; The formula is: )* ; in, and They are The initial and final values, where t is the current iteration number and Max_iter is the maximum iteration number; Step 2: Introduce an algorithm parameter The parameters used to calculate the spiral shape change non-linearly with the number of iterations, rather than linearly as in standard WOA, which helps the algorithm better balance global exploration and local exploitation capabilities during the search process; The formula is: )* ; in, and They are The initial and final values; Step 3: Introduce a dynamically adjusted algorithm parameter b to influence the shrinkage rate of the spiral shape. This parameter changes with the number of iterations to alter the tightness of the spiral shape, which helps improve the global search capability. The formula for expressing b is: )* ; in and These are the initial and final values ​​of b, respectively.

6. The improved ZWOA (Zoo Whale Optimization Algorithm) for a diversion center location method considering transport vehicle fuel consumption according to claim 1, is characterized in that... For cases where the position exceeds the search space boundary, the improved whale optimization algorithm adjusts the position back within the boundary. This boundary handling mechanism ensures that the algorithm always stays within the effective search space during the search process. The whale's position in a certain dimension is The lower bound of the search space in this dimension is The upper boundary is If the whale's new location If the position exceeds the boundaries of the search space, adjustments are needed, and the adjusted position will be determined by the search space. for: If the whale's new location Less than the lower bound Then adjust its position to the lower bound. If the whale's new location In the Nether and the Upper Realm If the boundary is included, its position remains unchanged; if the whale's new position... Greater than the upper bound Then adjust its position to the upper bound. The improved whale optimization algorithm ZWOA also records the fitness value of the leader after each iteration, i.e. the quality of the optimal solution, and stores these values ​​in the Convergence_curve array, which helps to analyze the convergence performance and stability of the algorithm.

7. The method for selecting a distribution center considering the fuel consumption of transport vehicles according to claim 1, characterized in that, The objective function of the diversion center location model is solved using the improved whale optimization algorithm, and the steps are as follows: Algorithm initialization and parameter definition; During the initialization phase of the algorithm, the position of each whale is set to a random value for the decision variable, and these values ​​should be within the given search space, i.e., the range of possible site selections; 2) Fitness assessment; For each whale's location, i.e., each candidate split center location, the fitness value is calculated using the objective function; 3) Whale location update; According to the principle of the whale optimization algorithm, the whale will search around the current optimal position, that is, the current optimal diversion center position, randomly select the positions of other whales as reference points, and update its own position in a spiral manner. 4) Local search and evaluation; Generate candidate solutions within a certain range of the current optimal position and evaluate their fitness values ​​to determine whether to exit the loop. When the algorithm converges or reaches the maximum number of iterations, output the optimal splitting center position and its corresponding fitness value; otherwise, return to step 2 above. 5) Site selection decision; Based on the optimal distribution center location output by the algorithm, the final location decision and minimum value are determined. .

Images