A method and system for modeling and optimizing a cold chain transportation network for dairy targets

Abstract

The application discloses a kind of cold chain transport network modeling optimization method and system for dairy product target, it is related to dairy product transport field, the cold chain transport network modeling optimization method for dairy product target, comprising the following steps: S1, dairy product logistics data is preprocessed;S2, set cold chain decision variable and cold chain transport target to construct multi-objective mixed integer programming model;S3, extract constraint characteristic parameter, and root set cold chain constraint condition;S4, model is solved to obtain optimal solution set, and compromise solution is comprehensively evaluated and screened;S5, generate dairy product low-carbon logistics distribution scheme, and verify optimization;S6, scheme is output and is stored in real time record, the present application is realized in dairy product cold chain by three-objective mixed integer programming model, and all network resources Pareto optimal configuration is realized, and by the design of time window satisfaction maximization objective function, reduce the penalty cost generated due to distribution delay.

Description

Technical Field

[0001] This invention relates to the field of dairy product transportation, and more specifically, to a method and system for modeling and optimizing cold chain transportation networks for dairy products. Background Technology

[0002] Against the backdrop of the continuous development of global perishable goods logistics, dairy products, as typical perishable goods, face stringent challenges in the design of their logistics networks. Many studies have focused on optimizing the economic costs of dairy product logistics networks, such as minimizing transportation costs and inventory costs by constructing multi-objective mixed integer programming models.

[0003] However, these studies often overlook the environmental impact of logistics activities, especially carbon emissions. The continuous operation of cold chain equipment and transportation processes lead to a significant increase in carbon emissions, which violates the goals of sustainable development. Another group of studies has begun to focus on the low-carbon transformation of dairy logistics, optimizing the logistics network by introducing carbon emission constraints.

[0004] However, existing logistics networks often neglect social considerations, such as time constraints on customer demand and traffic congestion. Furthermore, traditional genetic algorithms are used to optimize dairy product logistics networks. While this classic heuristic search algorithm demonstrates good performance in solving single-objective problems, it suffers from limitations due to the complexity and multi-objective nature of dairy product logistics networks. Additionally, the lack of an effective diversity preservation mechanism during iteration can lead to population diversity loss and local optima. Currently, no effective solutions have been proposed to address these issues. Summary of the Invention

[0005] To address the problems in related technologies, this invention proposes a cold chain transportation network modeling and optimization method and system for dairy products, in order to overcome the aforementioned technical problems existing in the existing related technologies.

[0006] To achieve the above objectives, the specific technical solution adopted by the present invention is as follows:

[0007] According to one aspect of the present invention, a method for modeling and optimizing a cold chain transportation network for dairy products is provided, comprising the following steps:

[0008] S1. Collect dairy product logistics data from dairy companies and perform data preprocessing on the dairy product logistics data;

[0009] S2. Set cold chain decision variables and cold chain transportation objectives based on the preprocessed dairy product logistics data, and construct a multi-objective mixed integer programming model based on the cold chain decision variables and cold chain transportation objectives;

[0010] As a preferred embodiment, the step of setting cold chain decision variables and cold chain transportation objectives based on preprocessed dairy product logistics data, and constructing a multi-objective mixed-integer programming model based on the cold chain decision variables and cold chain transportation objectives includes the following steps:

[0011] S21. Preset decision variable rules, extract decision variable data from preprocessed dairy product logistics data based on decision variable rules, and set cold chain decision variables through decision variable data;

[0012] S22. Based on the pre-processed dairy product logistics data, set a total cost target generation strategy, a carbon emission target generation strategy, and a time window target generation strategy;

[0013] As a preferred embodiment, the step of setting the total cost target generation strategy, carbon emission target generation strategy, and time window target generation strategy based on the preprocessed dairy product logistics data includes the following steps:

[0014] S221. Based on the total cost data of dairy products in the dairy product logistics data, set a total cost minimization objective function;

[0015] As a preferred embodiment, the objective function for minimizing the total cost of dairy products, based on the total cost data of dairy products in the dairy product logistics data, is set as follows:

[0016]

[0017] in, Fixed costs for facilities;

[0018] For transportation costs;

[0019] Cost of refrigeration losses;

[0020] For procurement costs;

[0021] Inventory holding costs;

[0022] Penalty costs for time windows;

[0023] For economic objectives, the objective function value is minimized by reducing total cost.

[0024] For fixed facility costs, the one-time investment cost of establishing a distribution center at location i;

[0025] For the location decision variable, whether it is at position i is a binary variable {0, 1};

[0026] Let be the discount rate for period t;

[0027] The fixed operating costs for vehicle model k;

[0028] Let K be the unit cost of vehicle model K in period t.

[0029] For transportation distance, the distance from distribution center i to retailer r;

[0030] For the number of vehicles, from the distribution center i in period t to the retailer r, the number of products p transported by vehicle model k;

[0031] For the decision on the use of refrigeration equipment, whether the refrigeration equipment is turned on when the period t is from the distribution center i to the retailer r and the vehicle model k is used is a binary variable {0, 1}.

[0032] The energy consumption cost per unit time for vehicle model k when the refrigeration equipment is turned on during period t;

[0033] The travel time from distribution center i to retailer r under congestion-free conditions;

[0034] For the purchase volume, the quantity of product p purchased by distribution center i during period t;

[0035] The unit inventory cost of storing product p in distribution center i is the inventory holding cost.

[0036] For inventory level, the inventory level of product p in distribution center i during period t;

[0037] As a time window penalty, the retailer r violates the penalty for the total cost of time window t.

[0038] S222. Based on the carbon emission data in the dairy product logistics data, set a carbon emission minimization objective function;

[0039] As a preferred embodiment, the carbon emission minimization objective function in the carbon emission minimization objective function, based on the carbon emission data in the dairy product logistics data, is set as follows:

[0040]

[0041] in, The carbon emission value for refrigerated transport, and only when refrigeration is turned on. Carbon emissions are generated at that time; Values ​​for non-refrigerated transport; Damage to goods transported under refrigeration; Damage to goods transported without refrigeration; Storage losses at the distribution center; For environmental objectives, the objective function value is to minimize carbon emissions; It is a fuel conversion factor that converts fuel consumption into carbon emissions; As a product loss conversion factor, each lost product is converted into carbon emissions; As a storage loss conversion factor, each storage loss is converted into carbon emissions; It is an electricity conversion factor that converts electricity consumption into carbon emissions; For the decision on the use of refrigeration equipment, whether the refrigeration equipment is turned on when the period t is from the distribution center i to the retailer r and the vehicle model k is used is a binary variable {0, 1}. Let be the transportation distance, and be the distance from distribution center i to retailer r; Fuel consumption for cooling: Fuel consumption per unit distance when vehicle model k uses the cooling equipment; Let be the number of vehicles, and let be the number of products p transported from distribution center i to retailer r using vehicle type k during period t. This refers to non-cooled fuel consumption, specifically the fuel consumption per unit distance for vehicle model k when the cooling system is not in use. The power consumption of the refrigeration equipment for vehicle model k; The travel time from distribution center i to retailer r under congestion-free conditions; The refrigeration loss rate is the loss rate of product p when transported using refrigeration. For the quantity of products transported, the number of products p transported from distribution center i to retailer r using vehicle type during period t; Storage loss rate, the loss rate of product p per unit period in the distribution center; For inventory level, the inventory level of product p in distribution center i during period t.

[0042] S223. Based on the time window target in the dairy product logistics data, set a time window satisfaction maximization objective function;

[0043] As a preferred embodiment, the objective function for maximizing the satisfaction of the time window, based on the time window objective in the dairy product logistics data, is as follows:

[0044]

[0045]

[0046] in, For the social goal, the objective function value is maximized by satisfying the time window. The time window weight represents the importance weight of retailer r's time window t; To incur penalties for violations.

[0047] S224. Generate the total cost target generation strategy, carbon emission target generation strategy, and time window target generation strategy based on the total cost minimization objective function, the carbon emission minimization objective function, and the time window satisfaction maximization objective function, and output them.

[0048] S23. Summarize the total cost target generation strategy, carbon emission target generation strategy, and time window target generation strategy and use them as cold chain transportation targets.

[0049] S24. Define a decision variable vector based on cold chain decision variables and cold chain transportation objectives, and integrate the objective function with the injected constraints to form a multi-objective mixed integer programming model.

[0050] S3. Extract the constraint feature parameters of the preprocessed dairy product logistics data, and set cold chain constraint conditions based on the constraint feature parameters;

[0051] As a preferred embodiment, the step of extracting constraint feature parameters from the preprocessed dairy product logistics data and setting cold chain constraint conditions based on the constraint feature parameters includes the following steps:

[0052] S31. Preset the constraint feature extraction target, and extract constraint feature parameters from the preprocessed dairy product logistics data based on the constraint feature extraction target;

[0053] S32. Based on the extracted constraint feature parameters, set the demand rigidity parameters and spatiotemporal coupling parameters, and set the demand rigidity constraints and spatiotemporal coupling constraints.

[0054] As a preferred embodiment, the step involves setting demand rigidity constraints and spatiotemporal coupling constraints based on the extracted constraint feature parameters, including the demand rigidity parameter and the spatiotemporal coupling parameter.

[0055] Demand rigidity constraint:

[0056]

[0057]

[0058] Among them, the rigidity of demand ensures that retailer r's demand for product p in period t. The desired result is achieved, and considering transportation losses when using refrigeration equipment, the actual arrival quantity is: When not using cooling , This is a binary variable indicating whether cooling is enabled. Use refrigeration at times. Not used at times; through and Combined, the total arrival amount is accurately calculated under different refrigeration conditions;

[0059] Spatiotemporal coupling constraints:

[0060]

[0061] This constraint ensures that the vehicle arrives no earlier than the retailer's earliest expected delivery time. arrive;

[0062] When a vehicle goes to the retailer This constraint becomes When no vehicles are going The constraint is automatically applied;

[0063]

[0064]

[0065] This constraint ensures that the vehicle arrives no later than the retailer's latest expected delivery time. Arrival. When the distribution center opens. And when a vehicle is on its way, the constraint becomes When the distribution center is not open hour, Make the constraints automatically valid; M is a sufficiently large number to ensure the conditionality of the constraints;

[0066] Among them, transportation time Determined by the traffic congestion function:

[0067]

[0068] This formula simulates the impact of traffic congestion on travel time; Free-flow travel time (normal travel time under conditions of no congestion); Road congestion level (the ratio of actual traffic flow to road capacity); parameter A value of 4 represents the exponential impact of congestion on time; as congestion increases, transportation time increases non-linearly.

[0069] Total flow The calculation is as follows:

[0070]

[0071] This formula calculates the total traffic flow on the road; Basic traffic (background traffic from other sources); The additional traffic brought to model k; This is the weighting for vehicle traffic flow.

[0072] S33. Integrate the rigid demand constraints and the spatiotemporal coupling constraints, and output them as cold chain constraint conditions.

[0073] S4. Solve the multi-objective mixed integer programming model using a non-dominated sorting genetic algorithm to obtain the Pareto optimal solution set. Then, combine the entropy weight TOPSIS method to comprehensively evaluate the Pareto optimal solution set and select compromise solutions.

[0074] As a preferred approach, the method of solving the multi-objective mixed integer programming model using a non-dominated sorting genetic algorithm to obtain the Pareto optimal solution set, and then comprehensively evaluating the Pareto optimal solution set using the entropy-weighted TOPSIS method to select a compromise solution, includes the following steps:

[0075] S41. Set the population size and crossover mutation probability of the non-dominated sorting genetic algorithm, and randomly generate an initial feasible solution set that satisfies the cold chain constraint.

[0076] S42. Perform non-dominated sorting and crowding distance calculation, and iteratively evolve the population by selecting simulated binary crossover and polynomial mutation.

[0077] S43. Extract the three-dimensional objective function values ​​of economic cost, carbon emissions, and time window satisfaction from the final Pareto front solution set;

[0078] S44. The target value is standardized using the range method, and the objective weight is calculated based on the entropy weight method. The relative closeness between each solution and the positive and negative ideal solutions is calculated by Euclidean distance. The solution with the largest closeness is then selected as the final compromise solution for cold chain logistics distribution.

[0079] S5. Generate a low-carbon logistics and distribution solution for dairy products based on the selected compromise solution, and verify and optimize the low-carbon logistics and distribution solution for dairy products.

[0080] S6. Output the verified and optimized low-carbon logistics and distribution plan for dairy products, and record and store it in real time.

[0081] According to another aspect of the present invention, a cold chain transportation network modeling and optimization system for dairy products is provided. The system includes: a data acquisition and processing module, a variable target model module, a cold chain constraint module, a solution evaluation and screening module, a scheme verification and optimization module, and a scheme output storage module.

[0082] The data acquisition and processing module is used to collect dairy logistics data from dairy companies and to preprocess the dairy logistics data.

[0083] The variable objective model module is used to set cold chain decision variables and cold chain transportation objectives based on preprocessed dairy product logistics data, and to construct a multi-objective mixed integer programming model based on the cold chain decision variables and cold chain transportation objectives.

[0084] The cold chain constraint module is used to extract constraint feature parameters from the preprocessed dairy product logistics data and set cold chain constraint conditions based on the constraint feature parameters.

[0085] The solution evaluation and screening module is used to solve the multi-objective mixed integer programming model using a non-dominated sorting genetic algorithm to obtain the Pareto optimal solution set, and to comprehensively evaluate the Pareto optimal solution set by combining the entropy weight TOPSIS method to screen compromise solutions.

[0086] The solution verification and optimization module is used to generate a low-carbon logistics and distribution solution for dairy products based on the selected compromise solutions, and to verify and optimize the low-carbon logistics and distribution solution for dairy products.

[0087] The solution output storage module outputs the verified and optimized low-carbon logistics and distribution solution for dairy products and records and stores it in real time.

[0088] The beneficial effects of this invention are as follows:

[0089] 1. This invention achieves Pareto optimal allocation of network resources in the dairy cold chain by combining a three-objective mixed integer programming model with a hierarchical NSGA-II solution framework. This breaks through the resource mismatch bottleneck caused by traditional single-objective optimization. At the same time, it integrates multi-objective Pareto decision-making, quantification of the perishability of dairy products, and dynamic traffic coupling algorithm to form a technical triangle, which improves the accuracy of the solution. Furthermore, by designing an objective function that maximizes the satisfaction of time windows, the solution ensures that dairy products arrive at retailers on time, reducing the penalty costs caused by delivery delays. The fine management of time windows makes the transportation process more precise and efficient, improving customer satisfaction and optimizing transportation plans.

[0090] 2. This invention reduces various costs in logistics by constructing a comprehensive cold chain transportation optimization model. By optimizing these factors, operating expenses are reduced, and the economic efficiency of logistics is improved. A carbon emission minimization objective function is set, and carbon emissions during dairy product transportation are reduced by optimizing factors such as the use of refrigeration equipment, transportation methods, and fuel consumption. By reducing negative environmental impacts, the green image of dairy companies is enhanced, which meets the increasingly stringent global environmental protection policy requirements. Furthermore, by applying a multi-objective mixed integer programming model and combining it with a genetic algorithm to optimize the solution, the optimal solution can be found while comprehensively considering economic costs, carbon emissions, and time window satisfaction.

[0091] 3. This invention flexibly addresses uncertainties such as market demand and traffic conditions by setting rigid demand constraints and spatiotemporal coupling constraints. Through intelligent optimization schemes, it can respond quickly in different scenarios, ensuring efficient operation even in changing environments. Furthermore, it uses a non-dominated sorting genetic algorithm and the entropy weight TOPSIS method to comprehensively evaluate the Pareto front solution set, evaluating and selecting the optimal solution from multiple dimensions such as economy, environmental protection, and timeliness. The final selected compromise solution has good overall performance, thus providing optimal decision support and reducing the uncertainties encountered by enterprises in actual operation. Attached Figure Description

[0092] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0093] Figure 1 This is a flowchart of a cold chain transportation network modeling and optimization method for dairy products according to an embodiment of the present invention;

[0094] Figure 2 This is a system block diagram of a cold chain transportation network modeling and optimization system for dairy products according to an embodiment of the present invention.

[0095] In the picture:

[0096] 1. Data acquisition and processing module; 2. Variable target model module; 3. Cold chain constraint module; 4. Solution evaluation and screening module; 5. Scheme verification and optimization module; 6. Scheme output and storage module. Detailed Implementation

[0097] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0098] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0099] According to embodiments of the present invention, a method and system for modeling and optimizing cold chain transportation networks for dairy products are provided.

[0100] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments. According to one embodiment of the present invention, such as... Figure 1-2 As shown in the figure, the cold chain transportation network modeling and optimization method for dairy products according to an embodiment of the present invention includes the following steps:

[0101] S1. Collect dairy product logistics data from dairy companies and perform data preprocessing on the dairy product logistics data;

[0102] S2. Set cold chain decision variables and cold chain transportation objectives based on the preprocessed dairy product logistics data, and construct a multi-objective mixed integer programming model based on the cold chain decision variables and cold chain transportation objectives;

[0103] In this embodiment of the application, the step of setting cold chain decision variables and cold chain transportation objectives based on preprocessed dairy product logistics data, and constructing a multi-objective mixed integer programming model based on the cold chain decision variables and cold chain transportation objectives includes the following steps:

[0104] S21. Preset decision variable rules, extract decision variable data from preprocessed dairy product logistics data based on decision variable rules, and set cold chain decision variables through decision variable data;

[0105] S22. Based on the pre-processed dairy product logistics data, set a total cost target generation strategy, a carbon emission target generation strategy, and a time window target generation strategy;

[0106] As a preferred embodiment, the step of setting the total cost target generation strategy, carbon emission target generation strategy, and time window target generation strategy based on the preprocessed dairy product logistics data includes the following steps:

[0107] S221. Based on the total cost data of dairy products in the dairy product logistics data, set a total cost minimization objective function;

[0108] Specifically, the location and construction of distribution centers plays a primary role in the planning of dairy cold chain logistics. Choosing a suitable location for a distribution center not only helps to significantly reduce the overall logistics cost, but also takes into account the perishable nature and short shelf life of dairy products. The location selection process for distribution centers, which are randomly generated based on demand points, comprehensively considers the fixed costs, transportation costs, refrigeration costs, procurement costs, inventory costs, and time window penalty costs of the dairy distribution center.

[0109] In this embodiment of the application, the objective function for minimizing the total cost of dairy products, based on the total cost data of dairy products in the dairy product logistics data, is set as follows:

[0110]

[0111] in, Fixed costs for facilities;

[0112] For transportation costs;

[0113] Cost of refrigeration losses;

[0114] For procurement costs;

[0115] Inventory holding costs;

[0116] Penalty costs for time windows;

[0117] For economic objectives, the objective function value is minimized by reducing total cost.

[0118] For fixed facility costs, the one-time investment cost of establishing a distribution center at location i;

[0119] For the location decision variable, whether it is at position i is a binary variable {0, 1};

[0120] Let be the discount rate for period t;

[0121] The fixed operating costs for vehicle model k;

[0122] Let K be the unit cost of vehicle model K in period t.

[0123] For transportation distance, the distance from distribution center i to retailer r;

[0124] For the number of vehicles, from the distribution center i in period t to the retailer r, the number of products p transported by vehicle model k;

[0125] For the decision on the use of refrigeration equipment, whether the refrigeration equipment is turned on when the period t is from the distribution center i to the retailer r and the vehicle model k is used is a binary variable {0, 1}.

[0126] The energy consumption cost per unit time for vehicle model k when the refrigeration equipment is turned on during period t;

[0127] The travel time from distribution center i to retailer r under congestion-free conditions;

[0128] For the purchase volume, the quantity of product p purchased by distribution center i during period t;

[0129] The unit inventory cost of storing product p in distribution center i is the inventory holding cost.

[0130] For inventory level, the inventory level of product p in distribution center i during period t;

[0131] As a time window penalty, the retailer r violates the penalty for the total cost of time window t.

[0132] S222. Based on the carbon emission data in the dairy product logistics data, set a carbon emission minimization objective function;

[0133] In this embodiment of the application, the carbon emission minimization objective function in the carbon emission minimization objective function, based on the carbon emission data in the dairy product logistics data, is set as follows:

[0134]

[0135] in, The carbon emission value for refrigerated transport, and only when refrigeration is turned on. Carbon emissions are generated at that time; Values ​​for non-refrigerated transport; Damage to goods transported under refrigeration; Damage to goods transported without refrigeration; Storage losses at the distribution center; For environmental objectives, the objective function value is to minimize carbon emissions; It is a fuel conversion factor that converts fuel consumption into carbon emissions; As a product loss conversion factor, each lost product is converted into carbon emissions; As a storage loss conversion factor, each storage loss is converted into carbon emissions; It is an electricity conversion factor that converts electricity consumption into carbon emissions; For the decision on the use of refrigeration equipment, whether the refrigeration equipment is turned on when the period t is from the distribution center i to the retailer r and the vehicle model k is used is a binary variable {0, 1}. Let be the transportation distance, and be the distance from distribution center i to retailer r; Fuel consumption for cooling: Fuel consumption per unit distance when vehicle model k uses the cooling equipment; Let be the number of vehicles, and let be the number of products p transported from distribution center i to retailer r using vehicle type k during period t. This refers to non-cooled fuel consumption, specifically the fuel consumption per unit distance for vehicle model k when the cooling system is not in use. The power consumption of the refrigeration equipment for vehicle model k; The travel time from distribution center i to retailer r under congestion-free conditions; The refrigeration loss rate is the loss rate of product p when transported using refrigeration. For the quantity of products transported, the number of products p transported from distribution center i to retailer r using vehicle type during period t; Storage loss rate, the loss rate of product p per unit period in the distribution center; For inventory level, the inventory level of product p in distribution center i during period t.

[0136] S223. Based on the time window target in the dairy product logistics data, set a time window satisfaction maximization objective function;

[0137] Specifically, in dairy cold chain logistics, time window constraints are of particular importance due to two core requirements.

[0138] First, quality assurance requirements: Fresh milk, yogurt and other products are extremely sensitive to delivery time. Delivering too early (leading to a waste of refrigerated storage resources) or delivering late (causing the risk of product spoilage) will both result in losses.

[0139] Second, commercial contract requirements: Large supermarkets, convenience stores and other customers have strict agreements on delivery time. Breach of contract will result in penalties (such as deducting 5% of the payment for each hour of delay) and damage to reputation. Existing cost-oriented logistics models often regard the time window as a soft constraint, while this solution treats it as an independent optimization target (on par with cost and carbon emissions), highlighting the special operational requirements of the dairy industry.

[0140] In this embodiment of the application, the objective function for maximizing the satisfaction of a time window, based on the time window objective in the dairy product logistics data, is as follows:

[0141]

[0142]

[0143] in, For the social goal, the objective function value is maximized by satisfying the time window. The time window weight represents the importance weight of retailer r's time window t; To incur penalties for violations.

[0144] S224. Generate the total cost target generation strategy, carbon emission target generation strategy, and time window target generation strategy based on the total cost minimization objective function, the carbon emission minimization objective function, and the time window satisfaction maximization objective function, and output them.

[0145] S23. Summarize the total cost target generation strategy, carbon emission target generation strategy, and time window target generation strategy and use them as cold chain transportation targets.

[0146] Using MATLAB for coding and solving, a multi-objective mixed integer programming model of a dairy cold chain logistics distribution network was constructed, employing a hierarchical mixed coding approach to represent complex combinations of decision variables. In the coding scheme, a chromosome structure contains various types of decision variables, including binary, integer, and continuous types, and includes five dimensions: number of distribution centers (I), vehicle type (K), number of demand points (R), time period (T), and product type (P), forming four variable groups: location layer - distribution center location (yi), equipment layer - refrigeration equipment usage decision (gikrt), transportation layer - cargo transportation instructions (qipkrt) + allocation of vehicle quantity (zikrt) + transportation product quantity (xipkrt), and inventory layer - inventory quantity (Iipt) + purchase quantity (Hip), for a total of seven decision variables. This hierarchical mixed coding approach groups different variable types for processing, avoiding direct concatenation of high-dimensional vectors, while accurately reflecting various decisions in the cold chain logistics system and maintaining reasonable computational complexity.

[0147] The decoding process transforms the chromosome into an executable delivery plan. First, it analyzes the location decisions to determine which distribution centers are activated (locations with a yi value of 1 represent selected distribution centers). Then, it constructs a transportation network based on vehicle allocation variables (the qipkrt value determines whether vehicle dispatching is performed, and zikrt determines the number and type of vehicles dispatched to each distribution center). The transportation volume variable is converted into an actual product delivery plan (xipkrt determines the quantity of products transported), while also considering the weight characteristics of different products. The decision regarding the use of refrigeration equipment affects transportation costs and quality assurance (the gikrt value determines whether refrigeration is used). Inventory and procurement variables form a complete supply chain plan, ensuring that product supply matches demand (Iipt and Hipt determine inventory and procurement volumes for each period). Implicit constraint handling logic during the decoding process filters out unreasonable solutions, such as allocating vehicle resources only to selected distribution centers, ensuring the feasibility of the solution.

[0148] The initial solution is generated using the `generateInitialPopulation` function, employing a randomization-based heuristic to ensure basic feasibility while satisfying various constraints in the modeling, thus creating a diverse initial population. For distribution center selection, a random Bernoulli distribution is used for generation, with a minimum quantity guaranteed. Vehicle allocation is randomly generated within a preset quantity limit, maintaining logical consistency with the location decision. Transportation volume generation considers vehicle load limits and product weight characteristics, allocating transportation combinations of two products through random proportions. Inventory levels are set as a random proportion of distribution center capacity, while procurement volume is determined based on transportation demand plus a random buffer. This initial solution generation mechanism considers both the business characteristics of cold chain logistics and ensures population diversity through randomization, providing a good starting point for subsequent genetic algorithm evolution. The entire encoding and decoding process fully reflects the complexity and multi-objective nature of the cold chain logistics optimization problem, establishing an effective mapping relationship between the solution expression and the actual problem.

[0149] S24. Define a decision variable vector based on cold chain decision variables and cold chain transportation objectives, and integrate the objective function with the injected constraints to form a multi-objective mixed integer programming model.

[0150] S3. Extract the constraint feature parameters of the preprocessed dairy product logistics data, and set cold chain constraint conditions based on the constraint feature parameters;

[0151] In the distribution of dairy products, distribution centers are not limited to the delivery of a single type of product, but rather encompass the joint delivery of multiple types of dairy products. Refrigerated dairy products have stringent storage requirements; insufficient consideration of factors during delivery can affect product freshness. When jointly delivering multiple types of dairy products, it is necessary to rationally configure the distribution system based on the storage conditions of refrigerated dairy products, considering whether refrigeration is required. Therefore, this study considers establishing n distribution centers, dividing each center into delivery areas, and simultaneously delivering multiple types of dairy products to downstream retail customers. It fully considers the characteristics of refrigerated products, including temperature requirements and shelf life, to ensure product quality is maintained during delivery. Given a relatively fixed demand and the maximum service capacity of the distribution centers, this provides effective decision support for the company's cold chain distribution center services.

[0152] Specific information about retailers served by dairy companies should include the coordinates of demand points, the average daily demand at each demand point, the expected time window for refrigerated dairy products, and service hours. The locations of dairy companies' cold chain distribution centers and their sets of demand points form a logistics supply chain network comprised of a series of geographically dispersed points.

[0153] The combination of each alternative distribution center in the set of alternative points is compared in terms of cost and delivery time. The optimal combination is selected to form the final cold chain distribution center. At the same time, the routes of delivery vehicles are optimized to minimize transportation costs and reduce delivery time, thereby improving overall delivery efficiency. This results in a site selection and route optimization scheme for a multi-product dairy cold chain logistics distribution center with limited capacity.

[0154] A multi-product, multi-period, mixed-vehicle, multi-objective solution model was simulated, encompassing three time periods, two dairy products, and three vehicle types. Specific parameter data were substituted into the constructed model, and the NAGA-II algorithm was used with candidate distribution centers in Matlab R2023b, with 100 iterations. After multiple iterations, the optimal distribution center location and corresponding delivery route were obtained.

[0155] In this embodiment of the application, the step of extracting the constraint feature parameters of the preprocessed dairy product logistics data and setting cold chain constraint conditions based on the constraint feature parameters includes the following steps:

[0156] S31. Preset the constraint feature extraction target, and extract constraint feature parameters from the preprocessed dairy product logistics data based on the constraint feature extraction target;

[0157] S32. Based on the extracted constraint feature parameters, set the demand rigidity parameters and spatiotemporal coupling parameters, and set the demand rigidity constraints and spatiotemporal coupling constraints.

[0158] In the embodiments of this application,

[0159] The step involves setting demand rigidity constraints and spatiotemporal coupling constraints based on the extracted constraint feature parameters, including the demand rigidity parameter and spatiotemporal coupling parameter.

[0160] Demand rigidity constraint:

[0161]

[0162]

[0163] Among them, the rigidity of demand ensures that retailer r's demand for product p in period t. The desired result is achieved, and considering transportation losses when using refrigeration equipment, the actual arrival quantity is: When not using cooling , This is a binary variable indicating whether cooling is enabled. Use refrigeration at times. Not used at times; through and Combined, the total arrival amount is accurately calculated under different refrigeration conditions;

[0164] Spatiotemporal coupling constraints:

[0165]

[0166] This constraint ensures that the vehicle arrives no earlier than the retailer's earliest expected delivery time. arrive;

[0167] When a vehicle goes to the retailer This constraint becomes When no vehicles are going The constraint is automatically applied;

[0168]

[0169]

[0170] This constraint ensures that the vehicle arrives no later than the retailer's latest expected delivery time. Arrival. When the distribution center opens. And when a vehicle is on its way, the constraint becomes When the distribution center is not open hour, Make the constraints automatically valid; M is a sufficiently large number to ensure the conditionality of the constraints;

[0171] Among them, transportation time Determined by the traffic congestion function:

[0172]

[0173] This formula simulates the impact of traffic congestion on travel time; Free-flow travel time (normal travel time under conditions of no congestion); Road congestion level (the ratio of actual traffic flow to road capacity); parameter A value of 4 represents the exponential impact of congestion on time; as congestion increases, transportation time increases non-linearly.

[0174] Total flow The calculation is as follows:

[0175]

[0176] This formula calculates the total traffic flow on the road; Basic traffic (background traffic from other sources); The additional traffic brought to model k; This is the weighting for vehicle traffic flow.

[0177] Specifically, based on the extracted constraint feature parameters, including the demand rigidity parameter and the spatiotemporal coupling parameter, and setting demand rigidity constraints and spatiotemporal coupling constraints, the process also includes extracting constraint feature parameters such as vehicle capacity parameters, distribution center capacity parameters, inventory balance parameters, initial inventory parameters, ending inventory parameters, refrigeration equipment usage parameters, vehicle usage parameters, and perishable product shelf-life parameters, and setting constraints for vehicle capacity, distribution center capacity, inventory balance, initial inventory, ending inventory, refrigeration equipment usage, vehicle usage, and perishable product shelf-life.

[0178] Specifically, the vehicle capacity constraint is as follows:

[0179]

[0180] This constraint ensures that the total weight of the transported goods does not exceed the vehicle's load capacity. The total weight of transported product p; Given the total load capacity of all vehicle types k, ensure that the load on each vehicle does not exceed its physical limits.

[0181] The capacity constraints of the distribution center are:

[0182]

[0183] This constraint ensures that the storage capacity of the distribution center is not exceeded. For the inventory level of distribution center i, The sum of the quantities of goods shipped at the same time and the total quantity shipped cannot exceed the capacity of the distribution center. When the distribution center is not open At that time, the constraint forces all operations to be zero.

[0184] The inventory balance constraint is:

[0185]

[0186] This constraint describes the inventory balance relationship of the distribution center. Current inventory = Previous inventory × (1 - Loss rate) + Current purchases - Current shipments. Let p be the loss rate of product during storage, applicable to all periods t≥2.

[0187] The initial inventory constraint is:

[0188]

[0189] This constraint defines the inventory balance for the initial period (t=1). Since it is the initial period, there is no inventory from the previous period, so there are only purchases and shipments.

[0190] Ending inventory constraints are:

[0191]

[0192] This constraint requires all inventory to be cleared at the end of the planning period. This avoids inventory buildup and waste, and prepares for the next planning cycle.

[0193] The usage constraints for refrigeration equipment are as follows:

[0194]

[0195] This constraint establishes the relationship between the cargo indicator variable and the transport volume. When hour, It can be 1; when hour, It must be 0; and M must be greater than 0. This ensures the logical relationships of the constraints.

[0196]

[0197] This constraint ensures that refrigeration can only be activated when goods are being transported. When any products are being transported... At that time, the cooling can be turned on. Refrigeration is not permitted when there is no cargo transport. This avoids wasting energy by starting the refrigeration equipment when it is not in use.

[0198] Vehicle usage constraints are as follows:

[0199]

[0200] This constraint calculates the required number of vehicles based on the weight of the cargo. (Round up) Ensure sufficient vehicle capacity, denominator The numerator represents the single-vehicle capacity, and the numerator represents the total transport weight. The minimum vehicle requirement for each route is automatically calculated.

[0201] The shelf life of perishable goods is limited as follows:

[0202]

[0203] This constraint ensures that the transportation time does not exceed the product's shelf life. The maximum transportation time is calculated based on the Weibull distribution model, using parameters under refrigeration conditions. , and loss rate ; Under non-refrigeration conditions, use parameters , and loss rate ,pass Automatically selects appropriate parameter combinations; the inverse function of the Weibull distribution. Calculate the time required to reach a specific loss rate. , The parameters of the Weibull distribution of product p in time t when refrigeration is used. and These are the parameters when refrigeration is not used.

[0204] S33. Integrate the rigid demand constraints and the spatiotemporal coupling constraints, and output them as cold chain constraint conditions.

[0205] S4. Solve the multi-objective mixed integer programming model using a non-dominated sorting genetic algorithm to obtain the Pareto optimal solution set. Then, combine the entropy weight TOPSIS method to comprehensively evaluate the Pareto optimal solution set and select compromise solutions.

[0206] In this embodiment of the application, the step of solving the multi-objective mixed integer programming model using a non-dominated sorting genetic algorithm to obtain the Pareto optimal solution set, and then comprehensively evaluating the Pareto optimal solution set using the entropy-weighted TOPSIS method to select a compromise solution includes the following steps:

[0207] S41. Set the population size and crossover mutation probability of the non-dominated sorting genetic algorithm, and randomly generate an initial feasible solution set that satisfies the cold chain constraint.

[0208] S42. Perform non-dominated sorting and crowding distance calculation, and iteratively evolve the population by selecting simulated binary crossover and polynomial mutation.

[0209] Upon entering the main loop, the algorithm first performs a non-dominated sorting operation on the current population. This step divides the entire population into multiple Pareto levels by comparing the dominance relationships between individuals pairwise. The first front contains the current best non-dominated solution, and subsequent fronts contain solutions dominated by fewer individuals. This hierarchical mechanism ensures that the algorithm can identify and prioritize high-quality solutions. Next, the crowding distance of each individual in the target space is calculated, which is a crucial step in maintaining the diversity of the solution set. By measuring the distance between adjacent solutions in each target dimension, the algorithm ensures that the final solution set is evenly distributed on the Pareto fronts. The selection operation employs a binary tournament mechanism, comprehensively considering both the non-dominated level and the crowding distance. The genetic operation phase applies specially designed crossover and mutation operators, such as sequential crossover and exchange mutation, to ensure that the generated new solutions are both innovative and feasible. After each iteration, the algorithm merges the parent and offspring populations and selects a new generation of population through an environmental selection mechanism. This process continues until a preset termination condition is met, such as reaching the maximum number of iterations or the solution set converges. The final Pareto optimal solution set provides decision-makers with a variety of objective trade-offs, allowing them to choose the most suitable delivery route configuration based on actual business needs.

[0210] S43. Extract the three-dimensional objective function values ​​of economic cost, carbon emissions, and time window satisfaction from the final Pareto front solution set;

[0211] S44. The target value is standardized using the range method, and the objective weight is calculated based on the entropy weight method. The relative closeness between each solution and the positive and negative ideal solutions is calculated by Euclidean distance. The solution with the largest closeness is then selected as the final compromise solution for cold chain logistics distribution.

[0212] In practical use, it is necessary to ensure that the input data format matches and adjust the visualization details according to the problem. The output includes the TOPSIS score ranking and weight analysis for each solution, the index of the Pareto solution (its position in the population), and the final optimization results.

[0213] S5. Generate a low-carbon logistics and distribution solution for dairy products based on the selected compromise solution, and verify and optimize the low-carbon logistics and distribution solution for dairy products.

[0214] Through iteration, it was found that in the multi-distribution center model, in the simulated random data generation, the company ultimately selected distribution centers 5, 10, and 13 out of 15 simulated distribution centers, using a total of 550 vehicles and transporting a total of 1001.45 tons. Compared with the relatively disordered delivery vehicle management of dairy companies at present, scientifically optimizing delivery vehicle routes can complete the same number of delivery tasks in fewer delivery trips, thereby reducing the consumption of resources such as manpower, time, and vehicles, while improving economic, environmental, and social performance.

[0215] S6. Output the verified and optimized low-carbon logistics and distribution plan for dairy products, and record and store it in real time.

[0216] According to another aspect of the invention, such as Figure 2 The present invention provides a cold chain transportation network modeling and optimization system for dairy products. The system includes: a data acquisition and processing module 1, a variable target model module 2, a cold chain constraint module 3, a solution evaluation and screening module 4, a scheme verification and optimization module 5, and a scheme output storage module 6.

[0217] Among them, the data acquisition and processing module 1 is used to collect dairy logistics data from dairy companies and to preprocess the dairy logistics data.

[0218] Variable Objective Model Module 2 is used to set cold chain decision variables and cold chain transportation objectives based on preprocessed dairy product logistics data, and to construct a multi-objective mixed integer programming model based on the cold chain decision variables and cold chain transportation objectives;

[0219] The cold chain constraint module 3 is used to extract the constraint feature parameters of the preprocessed dairy product logistics data and set cold chain constraint conditions based on the constraint feature parameters.

[0220] The solution evaluation and screening module 4 is used to solve the multi-objective mixed integer programming model using a non-dominated sorting genetic algorithm to obtain the Pareto optimal solution set, and to comprehensively evaluate the Pareto optimal solution set by combining the entropy weight TOPSIS method to screen compromise solutions.

[0221] The scheme verification and optimization module 5 is used to generate a low-carbon logistics and distribution scheme for dairy products based on the selected compromise solutions, and to verify and optimize the low-carbon logistics and distribution scheme for dairy products.

[0222] The solution output storage module 6 outputs the verified and optimized low-carbon logistics and distribution solution for dairy products and records and stores it in real time.

[0223] In summary, by utilizing the above-mentioned technical solutions of this invention, the present invention achieves Pareto optimal allocation of network resources in the dairy cold chain through the combination of a three-objective mixed integer programming model and a hierarchical NSGA-II solution framework. This overcomes the resource mismatch bottleneck caused by traditional single-objective optimization. Furthermore, it integrates multi-objective Pareto decision-making, quantification of dairy product perishability characteristics, and dynamic traffic coupling algorithms to form a technical triangle, improving the accuracy of the solution. Moreover, by designing an objective function that maximizes the satisfaction of time windows, the solution ensures that dairy products arrive at retailers on time, reducing penalty costs caused by delivery delays. The meticulous management of time windows makes the transportation process more precise and efficient, improving customer satisfaction and optimizing transportation plans.

[0224] Furthermore, this invention reduces various costs in logistics by constructing a comprehensive cold chain transportation optimization model. By optimizing these factors, operating expenses are reduced, and the economic efficiency of logistics is improved. A carbon emission minimization objective function is set, and carbon emissions during dairy product transportation are reduced by optimizing factors such as the use of refrigeration equipment, transportation methods, and fuel consumption. By reducing negative environmental impacts, the green image of dairy companies is enhanced, which aligns with increasingly stringent global environmental policies. Moreover, by applying a multi-objective mixed integer programming model and combining it with a genetic algorithm to optimize the solution, the optimal solution can be found while comprehensively considering economic costs, carbon emissions, and time window satisfaction.

[0225] Furthermore, this invention flexibly addresses uncertainties such as market demand and traffic conditions by setting rigid demand constraints and spatiotemporal coupling constraints. Through intelligent optimization schemes, it can respond quickly in different scenarios, ensuring efficient operation even in changing environments. Moreover, it uses a non-dominated sorting genetic algorithm and the entropy weight TOPSIS method to comprehensively evaluate the Pareto front solution set, evaluating and selecting the optimal solution from multiple dimensions such as economy, environmental protection, and timeliness. The final selected compromise solution has good overall performance, thus providing optimal decision support and reducing the uncertainties encountered by enterprises in actual operation.

[0226] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for modeling and optimizing a cold chain transportation network for dairy targets, characterized in that, The method comprises the following steps: S1, collecting dairy product logistics data of a dairy product enterprise, and performing data preprocessing on the dairy product logistics data; S2, setting cold chain decision variables and cold chain transportation targets according to the preprocessed dairy product logistics data, and constructing a multi-objective mixed integer programming model based on the cold chain decision variables and the cold chain transportation targets; S3, extracting constraint characteristic parameters of the preprocessed dairy product logistics data, and setting cold chain constraint conditions according to the constraint characteristic parameters; S4, solving the multi-objective mixed integer programming model by using a non-dominated sorting genetic algorithm to obtain a Pareto optimal solution set, and comprehensively evaluating the Pareto optimal solution set by using an entropy weight TOPSIS method to screen a compromise solution; S5, generating a dairy product low-carbon logistics distribution scheme based on the screened compromise solution, and verifying and optimizing the dairy product low-carbon logistics distribution scheme; S6, outputting and recording the verified and optimized dairy product low-carbon logistics distribution scheme in real time; The extraction of the constraint characteristic parameters of the preprocessed dairy product logistics data and the setting of the cold chain constraint conditions according to the constraint characteristic parameters comprise the following steps: S31, presetting a constraint characteristic extraction target, and extracting constraint characteristic parameters from the preprocessed dairy product logistics data based on the constraint characteristic extraction target; S32, setting demand rigidity constraints and space-time coupling constraints according to demand rigidity parameters and space-time coupling parameters in the extracted constraint characteristic parameters; S33, integrating the demand rigidity constraints and the space-time coupling constraints, and outputting them as cold chain constraint conditions; The setting of the demand rigidity constraints and the space-time coupling constraints according to the demand rigidity parameters and the space-time coupling parameters in the extracted constraint characteristic parameters: Demand rigidity constraints: ； ； where the demand rigidity constraint ensures that the demand of retailer r for product p at period t is satisfied and taking into account transportation losses the actual amount arriving using refrigeration equipment is and without refrigeration is , where Yt is a binary variable indicating whether refrigeration is on or off, when refrigeration is used, when it is not; by combining and the total amount arriving under different refrigeration conditions is calculated exactly. Space-time coupling constraints: ； This constraint ensures that the vehicle does not arrive earlier than the earliest expected in-time of the retailer Arrival; When a vehicle is going to a retailer , the constraint becomes ; when no vehicle is going , the constraint is automatically established; ； ； This constraint ensures that the vehicle does not arrive later than the retailer's latest desired arrival time Arrival, when the distribution center i is open and a vehicle is going, the constraint becomes ; when the distribution center i is not open , the constraint is automatically satisfied; M is a number large enough to ensure the conditionality of the constraint; wherein the transport time is determined by the traffic congestion function: ； This formula simulates the impact of traffic congestion on transport time; is the free-flow travel time; is the road congestion level; parameter takes 4, indicating the exponential impact of congestion on time; as the congestion level increases, the transport time grows nonlinearly; And total flow Is calculated as follows: ； This formula calculates the total flow on the road; is the base flow; is the extra flow brought by vehicle type k; is the vehicle flow weight.

2. The method of claim 1, wherein, The setting of the cold chain decision variables and the cold chain transportation targets according to the preprocessed dairy product logistics data and the construction of the multi-objective mixed integer programming model based on the cold chain decision variables and the cold chain transportation targets comprise the following steps: S21, presetting a decision variable rule, extracting decision variable data from the preprocessed dairy product logistics data based on the decision variable rule, and setting cold chain decision variables through the decision variable data; S22, setting a total cost target generation strategy, a carbon emission target generation strategy, and a time window target generation strategy according to the preprocessed dairy product logistics data; S23, summarizing the total cost target generation strategy, the carbon emission target generation strategy, and the time window target generation strategy and taking them as cold chain transportation targets; S24, defining a decision variable vector based on the cold chain decision variables and the cold chain transportation targets, and integrating a target function to inject constraint conditions to form a multi-objective mixed integer programming model.

3. The method of claim 2, wherein, The setting of the total cost target generation strategy, the carbon emission target generation strategy, and the time window target generation strategy according to the preprocessed dairy product logistics data comprises the following steps: S221, setting a total cost minimization target function according to total dairy product cost data in the dairy product logistics data; S222, setting a carbon emission minimization target function according to carbon emission data in the dairy product logistics data; S223, set the time window satisfaction maximization objective function according to the time window target in the dairy product logistics data; S224, generate total cost target generation strategy, carbon emission target generation strategy and time window target generation strategy according to total cost minimization objective function, carbon emission minimization objective function and time window satisfaction maximization objective function.

4. The method for modeling and optimizing a cold chain transportation network for dairy targets according to claim 3, characterized in that, The total cost minimization objective function in the total cost minimization objective function is set according to the total cost data of dairy products in the dairy product logistics data. ； wherein, F is the facility fixed cost; for transportation costs; To refrigeration loss cost; To purchase cost; for inventory holding costs; Penalize cost for time windows; For the economic objective, the total cost is minimized. Ci is the fixed cost of establishing a distribution center at location i, which is a one-time investment cost for the facility; For the location decision variable, whether at location i is a binary variable {0, 1}; r is the discount rate for period t; Ck is the fixed operating cost for vehicle type k; uk(t) unit cost for vehicle model k at period t; for the transportation distance, the distance from distribution center i to retailer r; Vpikr(t) is the number of vehicles of vehicle type k used to transport product p from distribution center i to retailer r at time period t; Decision to use refrigeration equipment, period t, from distribution center i to retailer r using vehicle type k, whether refrigeration equipment is on or off, binary variable {0, 1}; Ck(t) is the cost per unit time of operating the refrigeration appliance for vehicle type k at period t; is the travel time from distribution center i to retailer r in uncongested conditions; Qp,ti quantity of product p purchased by distribution center i at time t; Cp, the unit inventory cost of product p at distribution center i for holding cost; Qp,t,i is the product p inventory at distribution center i at period t; The total cost of time window violation penalty for retailer r in period t.

5. The method of claim 3, wherein, The carbon emission minimization objective function in the carbon emission minimization objective function is set according to the carbon emission data in the dairy product logistics data. ； wherein, is the refrigerated transport carbon emission value, and only generated carbon emission when the refrigeration is on; is the non-refrigerated transport value; is the refrigerated transport product damage; is the non-refrigerated transport product damage; is the distribution center storage damage; is the environmental target, the carbon emission minimization objective function value; is the fuel conversion factor, converting fuel consumption to carbon emission; is the product damage conversion factor, converting each damaged product to carbon emission; is the storage damage conversion factor, converting each damaged product to carbon emission; is the electricity conversion factor, converting electricity consumption to carbon emission; is the refrigeration equipment usage decision, whether the refrigeration equipment is on for vehicle type k from distribution center i to retailer r in period t, is a binary variable {0, 1}; is the transportation distance, the distance from distribution center i to retailer r; is the refrigerated fuel consumption, the fuel consumption per distance when the refrigeration equipment is used for vehicle type k; is the vehicle quantity, the quantity of vehicle type k used for transporting product p from distribution center i to retailer r in period t; is the non-refrigerated fuel consumption, the fuel consumption per distance when the refrigeration equipment is not used for vehicle type k; is the power consumption of the refrigeration equipment of vehicle type k; is the travel time without congestion from distribution center i to retailer r; is the refrigeration damage rate, the damage rate when product p is transported using refrigeration; is the transportation product quantity, the quantity of product p transported using vehicle type from distribution center i to retailer r in period t; is the storage damage rate, the unit period damage rate of product p in the distribution center; is the inventory, the inventory of product p in distribution center i in period t.​ 6. The method of claim 3, wherein, The time window satisfaction maximization objective function in the time window satisfaction maximization objective function is set according to the time window target in the dairy product logistics data. ； ； wherein, is the social objective, the time window satisfaction maximization objective function value; is the time window weight, representing the importance weight of the time window t of the retailer r; is the time window penalty cost.

7. The method of claim 1, wherein, The multi-objective mixed integer programming model is solved by the non-dominated sorting genetic algorithm to obtain a Pareto optimal solution set, and the Pareto optimal solution set is comprehensively evaluated by combining the entropy weight TOPSIS method to screen the compromise solution, including the following steps: S41, set the population size and crossover mutation probability of the non-dominated sorting genetic algorithm, and randomly generate an initial feasible solution set that satisfies the cold chain constraint; S42, execute non-dominated sorting and crowded distance calculation, and iterate the population by selecting simulated binary crossover and polynomial mutation; S43, extract the three-dimensional objective function values of economic cost, carbon emission and time window satisfaction from the final Pareto front solution set; S44, standardize the objective values by range method, calculate the objective weights based on entropy weight method, solve the relative closeness of each solution and ideal solution by Euclidean distance, and select the maximum closeness solution as the final compromise solution of cold chain logistics distribution.

8. A dairy-targeted cold chain transportation network modeling optimization system for implementing the dairy-targeted cold chain transportation network modeling optimization method of any one of claims 1-7, wherein, The system comprises a data acquisition and processing module, a variable target model module, a cold chain constraint condition module, a solution evaluation and screening module, a scheme verification and optimization module, and a scheme output and storage module; The data acquisition and processing module is used to collect the dairy product logistics data of dairy product enterprises and pre-process the dairy product logistics data; The variable target model module is used to set cold chain decision variables and cold chain transportation targets according to the pre-processed dairy product logistics data, and to construct a multi-objective mixed integer programming model based on the cold chain decision variables and the cold chain transportation targets; The cold chain constraint condition module is used to extract constraint characteristic parameters of the pre-processed dairy product logistics data, and to set cold chain constraint conditions according to the constraint characteristic parameters; The solution evaluation and screening module is used to solve the multi-objective mixed integer programming model by the non-dominated sorting genetic algorithm to obtain a Pareto optimal solution set, and to comprehensively evaluate the Pareto optimal solution set by combining the entropy weight TOPSIS method to screen the compromise solution; The scheme verification and optimization module is used to generate a dairy product low-carbon logistics distribution scheme based on the screened compromise solution, and to verify and optimize the dairy product low-carbon logistics distribution scheme; The scheme output and storage module outputs and stores in real time the dairy product low-carbon logistics distribution scheme after verification and optimization.

Images