Method for collaborative optimization of corn raw grain procurement and processing and finished grain transportation under carbon constraints

Abstract

The application discloses a kind of corn raw grain procurement processing and finished product grain transportation collaborative optimization methods under carbon constraint, the method includes: corn processing enterprise supply chain cost and different transportation mode carbon emission accounting mode;Comprehensively consider the seasonality fluctuation of raw grain supply, inventory loss, and different transportation mode and its packaging form in direct transport and reconditioning link timeliness, cost, damage, carbon footprint and other factors, construct mathematical model with corn processing enterprise supply chain cost minimization as target;To the constraint rules of purchase, processing, inventory, shortage, transportation and carbon emission management in target mathematical model, and linearization is handled to constraint condition;Based on the branch and price algorithm of fusion dynamic programming, obtain the collaborative optimization scheme of corn raw grain procurement processing and finished product grain transportation under carbon constraint.The present application solves the technical problems that the existing method is difficult to coordinate for the inventory loss of finished product grain, the transportation loss of multimodal transport and the carbon emission in transportation link.

Description

Technical Field

[0001] This invention relates to the field of grain supply chain management and logistics optimization technology, and in particular to a method for the coordinated optimization of corn raw material procurement, processing and finished grain transportation under carbon constraints. Background Technology

[0002] Existing research on grain supply chain management and logistics optimization mainly falls into three categories: First, logistics optimization along the "producer-hub-customer" chain. A mixed-integer nonlinear programming model was constructed for grain silo location selection, aiming to minimize total cost and delivery time, and the NCRO algorithm was used to solve the problem. A mixed-integer linear programming model was constructed to analyze transportation problems, finding that replacing traditional vehicles with electric vehicles could reduce transportation costs by about one-third and reduce carbon emissions by 70%. Second, multi-objective transportation allocation and inventory management problems. A joint optimization model for grain loading and transportation routes considering cargo damage was developed. Heuristic genetic algorithm analysis showed that containerized multimodal transport (rail-road-water) is significantly superior to bulk or packaged grain transport in terms of total cost and losses. Simultaneously, a multi-objective optimization was constructed using the damage factor, capacity utilization rate, and weekly penalty value, and a customized optimization method was developed using the MOSA and NSGA-II algorithms to solve for the optimal solution. Third, uncertainty and supply chain sustainability issues. These include an agricultural supply chain risk management model based on fuzzy logic, a multi-objective mixed-integer linear programming model that robustly optimizes multi-dimensional parameter uncertainties, and a mixed-integer nonlinear programming model for India's public allocation system combined with an improved max-min ant colony algorithm to optimize bulk grain transportation costs. Another study constructs a hybrid optimization-simulation model to address the supply chain complexity brought about by Canada's newly implemented wheat classification system, aiming to determine the optimal wheat hybridization strategy.

[0003] However, existing research has gaps in the following areas: a lack of research on finished grain inventory losses, multimodal transport losses, and the difficulty of coordinating carbon emissions in the transportation process. Furthermore, existing research often focuses on single-stage optimization or joint optimization of inventory and transportation, lacking research on coordinated optimization from the perspective of the entire "production, storage, and transportation" chain. In addition, there is a lack of research on the cyclical patterns of seasonal fluctuations in raw grain supply using procurement prices. Finally, research on carbon emissions in grain transportation mainly focuses on direct land transport such as roads and railways, lacking research on transportation modes suitable for China's national conditions, including multiple carbon emission management models, diversified packaging, bulk-to-container conversion, river-sea intermodal transport, and return transport of grain and agricultural inputs. Summary of the Invention

[0004] This invention provides a method for the coordinated optimization of corn raw material procurement, processing, and finished grain transportation under carbon constraints, in order to overcome the above-mentioned technical problems.

[0005] To achieve the above objectives, the technical solution of this invention is: a method for collaborative optimization of corn raw material procurement, processing, and finished grain transportation under carbon constraints, comprising the following steps: S1: Considering raw material procurement costs, raw material processing costs, finished grain inventory costs, finished grain inventory loss costs, basic finished grain transportation costs, finished grain loss costs during transportation, finished grain transportation time costs, and delayed delivery costs, a supply chain cost accounting model for corn processing enterprises is constructed; S2: Based on the order demand, delivery time, processing, and inventory status of user enterprises, a transportation plan is formulated to construct a target mathematical model that minimizes the total delivery cost of the corn processing enterprise supply chain cost accounting model, based on the characteristics of different transportation modes; the target mathematical model includes an objective function and constraints that minimize the total delivery cost; S3: The constraints in the target mathematical model are linearized, and the optimized target model is obtained by combining the objective function; S4: A carbon constraint strategy is constructed based on carbon emission accounting under different transportation modes; S5: Based on the branch pricing algorithm framework, a collaborative optimization scheme for corn raw material procurement, processing, and finished grain transportation under carbon constraints is obtained according to the optimized target model and carbon constraint strategy.

[0006] This invention provides a method for the coordinated optimization of corn raw material procurement, processing, and finished grain transportation under carbon constraints, with the following beneficial effects: (1) A carbon constraint strategy based on carbon emission accounting under different transportation modes was constructed. Specifically, a collaborative optimization framework for corn supply chain loss reduction was proposed under periodic carbon constraint, cumulative carbon constraint, full-cycle carbon constraint and fluctuating carbon constraint strategies. This framework integrates grain saving and loss reduction with carbon emission reduction into a unified decision-making system, which can more comprehensively reflect the characteristics of supply chain operation under different carbon management scenarios. It solves the problems of separation of loss reduction and emission reduction targets and single carbon constraint form in existing technologies, thereby improving the scientificity, adaptability and practicality of corn supply chain collaborative optimization, and providing technical support for low-carbon loss reduction decision-making in actual operation.

[0007] (2) A target mathematical model was constructed to minimize the total delivery cost of the corn processing enterprise supply chain cost accounting model. This model incorporates the seasonality of raw grain supply, finished grain categories, inventory losses, and different transportation modes into a unified decision-making framework, achieving coordinated optimization of procurement, processing, and transportation. Compared to existing technologies, this invention more accurately reflects the actual operating characteristics of the corn supply chain, reduces inventory losses and transportation costs, improves the supply and demand matching level and transportation organization efficiency of multiple types of finished grain, thereby enhancing the overall operating efficiency and collaborative decision-making capabilities of the corn supply chain.

[0008] (3) Based on the branch pricing algorithm framework, according to the optimization target model and carbon constraint strategy, a collaborative optimization scheme for corn raw material procurement and processing and finished grain transportation under carbon constraints is obtained. That is, by designing a customized branch pricing algorithm embedded in dynamic programming, dynamic programming and branch pricing are combined, which effectively improves the solution efficiency of complex corn supply chain collaborative optimization problems while taking into account the solution accuracy. Based on the branch pricing algorithm, this invention can obtain the basis for transportation mode selection under different carbon constraint strategies, identify the applicable conditions of bulk-to-container conversion in direct and transshipment situations, and provide quantitative and implementable decision support for enterprises' actual transportation organization and carbon strategy selection, which has strong engineering application value. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a flowchart of the method described in this invention; Figure 2 This is the framework for the relationships between the participants in the corn supply chain in this embodiment; Figure 3 This is a technical roadmap diagram of the branch pricing algorithm framework in this embodiment; Figure 4 This is a diagram showing the relationship between carbon emission limits and transportation mode selection under periodic carbon constraints in this embodiment; Figure 5 This is a diagram showing the relationship between carbon emission limits and transportation mode selection under cumulative carbon constraints in this embodiment; Figure 6 This is a diagram showing the relationship between carbon emission limits and transportation mode selection under full-cycle carbon constraints in this embodiment; Figure 7 This is a diagram showing the relationship between carbon emission limits and transportation mode selection under fluctuating carbon constraints in this embodiment; Figure 8 This is a diagram showing the relationship between carbon emission limits, transportation modes, and transportation categories under periodic carbon constraints in this embodiment. Figure 9 This is a diagram showing the relationship between carbon emission limits, transportation modes, and transportation categories under the cumulative carbon constraint in this embodiment; Figure 10 This is a diagram showing the relationship between carbon emission limits, transportation modes, and transportation categories under the full-cycle carbon constraint in this embodiment; Figure 11 This is a diagram showing the relationship between carbon emission limits, transportation modes, and transportation categories under fluctuating carbon constraints in this embodiment. Figure 12 This is a diagram showing the relationship between carbon emission limits and cost composition under periodic carbon constraints in this embodiment; Figure 13 This is a diagram showing the relationship between carbon emission limits and cost composition under cumulative carbon constraints in this embodiment; Figure 14 This is a diagram showing the relationship between carbon emission limits and cost composition under full-cycle carbon constraints in this embodiment; Figure 15 This is a diagram showing the relationship between carbon emission limits and cost composition under fluctuating carbon constraints in this embodiment; Figure 16 This is a graph showing the impact of the proportion of low-toxin enhanced feed corn on the choice of transportation mode in this embodiment; Figure 17 This is a graph showing the impact of the proportion of low-toxin enhanced feed corn on the choice of transportation mode for different corn varieties in this embodiment; Figure 18 This is a graph showing the impact of the proportion of low-toxin enhanced feed corn on costs in this embodiment; Figure 19 This is a diagram illustrating the impact of the total demand of feed companies on the choice of transportation mode throughout the entire planning period in this embodiment. Figure 20 This is a diagram illustrating the impact of the total demand of feed companies during the entire planning period on the selection of transportation modes for different varieties of corn in this embodiment. Figure 21 This is a diagram illustrating the impact of total demand on costs for feed companies throughout the entire planning period in this embodiment. Figure 22 This is a diagram illustrating the impact of seasonal fluctuations in raw grain procurement prices on transportation mode selection in this embodiment. Figure 23 This is a graph showing the impact of seasonal fluctuations in raw grain procurement prices on costs in this embodiment; Figure 24 This diagram illustrates the impact of return freight transportation cost sharing on the selection of transportation modes 7 and 12 in this embodiment. Figure 25 This diagram illustrates the impact of the cost sharing of return freight transportation for different types of feed corn on the selection of transportation modes 7 and 12 in this embodiment. Detailed Implementation

[0011] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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 only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0012] This embodiment provides a method for the coordinated optimization of corn raw material procurement, processing, and finished grain transportation under carbon constraints, such as... Figures 1 to 2 As shown, the specific steps include: S1: Considering the raw grain procurement cost, raw grain processing cost, finished grain inventory cost, finished grain inventory loss cost, finished grain basic transportation cost, finished grain loss cost during transportation, finished grain transportation time cost, and delayed delivery cost, construct a supply chain cost accounting model for corn processing enterprises. The model also includes the following assumptions: (1) Farmers can provide relatively stable grain sources for processing enterprises, but the grain procurement decisions of processing enterprises are still affected by seasonal market fluctuations. (2) At the beginning of each service cycle, farmers deliver grain to processing enterprises, and the grain is put into processing immediately after arrival. All grain must be processed within the current period, and the finished grain formed is stored in silos at the beginning of the next service cycle. Different grades of finished grain are stored in different silos to accurately identify their storage time. At the end of each service cycle, the loss of the finished grain in stock is calculated, and the outbound follows the first-in-first-out (FIFO) principle. (3) Both processing enterprises and feed enterprises have dedicated railway lines and storage facilities, and each node along the transportation line has loading, unloading and transshipment conditions. Transportation is organized in the form of whole trucks, whole ships and whole containers, without considering consolidation transportation. The transportation speed and unit transportation cost of the same type of vehicles and ships on different lines are considered to be the same. (4) Carbon constraint strategies include periodic carbon constraints, cumulative carbon constraints, full-cycle carbon constraints and volatile carbon constraints. (5) Corn processing enterprises have the business of processing raw corn and distributing fertilizers, so they can reduce the empty load rate and spread costs through the north-to-south grain transport and the south-to-north fertilizer transport. (6) At the beginning of each service cycle, feed enterprises determine the demand for various types of feed corn products based on supply and demand, and submit orders to corn processing enterprises accordingly; logistics enterprises formulate transportation arrangements at the same time. The transportation time of each transportation mode does not exceed a single service cycle. If delivery is not made on time, late delivery costs will be incurred.

[0013] Specifically, in this embodiment, the total cost of a corn processing enterprise consists of raw grain procurement cost, raw grain processing cost, finished grain inventory cost, finished grain inventory loss cost, finished grain basic transportation cost, finished grain loss cost during transportation, finished grain transportation time cost, and delayed delivery cost, as shown in equation (1): (1) The raw material procurement cost of corn processing enterprises mainly includes the product cost of raw materials and the transportation cost of raw materials, as shown in formula (2): (2) The raw grain processing costs of corn processing enterprises mainly include raw grain processing costs, depreciation and maintenance costs of processing equipment, energy consumption costs, labor costs, processing material and auxiliary material costs, waste disposal and environmental protection costs, and quality control and testing costs, as shown in formula (3): (3) The finished grain inventory costs of corn processing enterprises mainly include storage fees, insurance fees, and capital occupation costs, as shown in formula (4): (4) The inventory loss cost of finished corn products in corn processing enterprises refers to the cost incurred during the storage of finished corn products due to natural losses, pests, rodents, mold, etc., which lead to a decrease in the quantity or quality of finished corn products, as shown in formula (5): (5) The basic transportation cost of finished corn products for corn processing enterprises refers to the direct transportation costs incurred during the transportation of finished corn products, mainly including outbound operation costs, inbound operation costs, and costs incurred during transit. This cost is primarily affected by transportation distance, mode of transport, loading method, freight volume, market freight rates, and policy subsidies (such as railway transport discounts and bulk-to-container subsidies). Different transportation modes are shown in formulas (6) to (22): Direct rail transport cost of bulk cargo ,Right now hour.

[0014] (6) Direct rail transport cost of bagged goods ,Right now hour.

[0015] (7) Direct rail container transport ,Right now hour.

[0016] (8) Direct road transport of bulk cargo ,Right now hour.

[0017] (9) Direct road transport of bagged goods ,Right now hour.

[0018] (10) Direct road container transport ,Right now hour.

[0019] (11) "Iron-Water-Iron" Bulk Transport (River-Sea Transshipment) ,Right now hour.

[0020] (12) "Rail-Water-Rail" Bulk Transport (Direct River-Sea Transport) ,Right now hour.

[0021] (13) "Rail-Water-Rail" Container Transport (River-Sea Transshipment) ,Right now hour.

[0022] (14) "Rail-Water-Rail" Bulk-to-Container (River-Sea Transshipment) Intermodal Transport ,Right now hour.

[0023] (15) Rail-Water-Rail Container Transport (Direct River-Sea Shipping) ,Right now hour, (16) Container transport (river-sea transshipment) on the "North Grain to South and South Fertilizer to North Dedicated Line (Rail-Water-Rail)" ,Right now hour, (17) Container transport (direct river-sea) on dedicated rail lines for transporting grain from the north to the south and fertilizer from the south to the north. ,Right now hour.

[0024] (18) "Road-Rail-Road" Bulk Transportation ,Right now hour.

[0025] (19) "Road-Rail-Road" Container Transportation ,Right now hour.

[0026] (20) Bulk transport via road-water-road (river-sea transshipment) ,Right now hour.

[0027] (twenty one) "Road-Water-Road" Container Transport (River-Sea Transshipment) ,Right now hour.

[0028] (twenty two) The loss cost of finished grain during transportation for corn processing enterprises includes two parts: transportation and transshipment. The transportation loss rate and transshipment loss rate are based on historical experience and industry averages. The comprehensive loss rate during the entire transportation process includes natural loss rate, breakage rate, and spoilage rate, as shown in formula (23): (twenty three) In this embodiment, considering that transportation losses mainly come from the loading and unloading process at the starting point and the transshipment process in the middle, the formula (23) is set as follows: Calculated using formula (24): (twenty four) The time cost of transporting finished corn products under different transportation modes for corn processing enterprises refers to the cost incurred due to time consumption during the process from the completion of corn processing to its delivery destination under different transportation modes. This cost is affected by the combined factors of transportation distance, transportation mode efficiency, traffic congestion, transshipment and loading / unloading efficiency, and force majeure factors such as extreme weather. The time cost of each transportation mode is calculated according to the transportation mode, as shown in formulas (25) to (34): when hour, The calculation formula is: (25) when hour, The calculation formula is: (26) when hour, The calculation formula is: (27) when hour, The calculation formula is: (28) when hour, The calculation formula is: (29) when hour, The calculation formula is: (30) when hour, The calculation formula is: (31) when hour, The calculation formula is: (32) when hour, The calculation formula is: (33) when hour, The calculation formula is: (34) in: This indicates that corn processing enterprises adopt In the transportation mode, at the place of shipment Method of delivery Unit outbound operation cost of packaged corn products; Indicates during the service period Initially, corn processing enterprises adopted In the transportation mode, Freight section transportation The unit turnover transportation cost of packaged corn products; This indicates that corn processing enterprises adopt In the transportation mode, at the destination Method of delivery The cost of warehousing finished corn products after packaging; This indicates the cost of bagging each ton of finished corn grain over a service cycle; This represents the usage cost of each container of finished corn products over a service cycle, with the container containing bulk cargo. This indicates that corn processing enterprises adopt Transportation modes in Freight transshipment section The cost of repackaging finished corn products for the packaging unit; This indicates that corn processing enterprises adopt Total transport distance of each mode of transport; This indicates that corn processing enterprises adopt Transportation modes in Freight Section The transportation distance of packaged corn products; This indicates the railway freight subsidy per ton of cargo under different transportation modes (calculated based on cargo weight). This indicates the rail freight subsidy per TEU container (calculated based on the number of containers) under different transport modes. This represents the subsidy for bulk-to-container conversion of corn products to corn processing enterprises per TEU; This indicates that corn processing enterprises process raw corn into finished corn products. Conversion factor that leads to a reduction in cargo weight; Indicates the average speed of road transport; This indicates the average speed of railway transportation; Indicates the average speed of waterway transportation; This indicates that corn processing enterprises adopt The non-transportation time, namely loading and unloading time and transshipment time, is used for a single complete transport of each ton of finished corn grain. This represents the average value per unit of time for railway transportation. This represents the average value of road transport per unit of time. This represents the average value per unit time of waterway transportation; This indicates that corn processing enterprises adopt The transportation mode involves sharing the transportation costs of finished corn products generated by the return trip of the North-to-South Grain Transport Special Line for transporting fertilizer. This indicates the variety of finished corn products per ton for corn processing enterprises. The selling price; This indicates that corn processing enterprises are Freight section completed The cargo loss coefficient corresponding to direct transportation of packaged finished corn grain; This indicates that corn processing enterprises are The freight transshipment section will Goods packaging changed Cargo loss coefficient during the transshipment of packaged corn products; This indicates that logistics companies are Freight section transportation Carbon emissions per unit of packaged corn products: for bulk or bagged goods, the unit is kg / (ton·km); for containers, the unit is kg / (TEU·km). This indicates that corn processing enterprises are The freight transshipment section will Goods packaging changed Carbon emissions per unit during the transshipment of packaged corn products (kg / ton for bulk or bagged goods; kg / TEU for containers). The delayed delivery cost of corn processing enterprises refers to the various direct and indirect economic losses caused by failure to deliver products on time. This cost has an important impact on the enterprise's market competitiveness and customer relationships, as shown in formula (35): (35) In the formula: This represents the total delivery cost for corn processing enterprises; This indicates the service cycle of corn processing enterprises. The resulting procurement costs; This indicates the service cycle of corn processing enterprises. The resulting processing costs; This indicates the service cycle of corn processing enterprises. The resulting inventory costs; This represents the inventory loss cost incurred by corn processing enterprises throughout the entire planning period; This indicates the service cycle of corn processing enterprises. use The basic transportation cost of the transportation mode; This indicates the service cycle of corn processing enterprises. use Costs of cargo loss during the transportation process in different transportation modes; This indicates the service cycle of corn processing enterprises. use Transportation time cost of transportation modes; This indicates the service cycle of corn processing enterprises. The cost of delayed delivery; Indicates the total service period. ; Represents a set of transportation modes. Indicates different modes of transportation; This indicates that corn processing enterprises are The unit price of raw grain at the beginning of the service period; This indicates that corn processing enterprises in the first The amount of corn raw materials purchased periodically; This indicates that corn processing enterprises process it into... The unit processing cost of graded feed corn; Indicates service period Of the initial batch of raw corn stored, a significant portion is processed into finished corn products. The amount; Indicates the varieties of finished grain products of corn processing enterprises ( In order, they are low-toxin enhanced type, low-moisture long-storage type, and high-standard general-purpose feed corn varieties. This indicates the variety of finished corn products per ton for corn processing enterprises. Inventory costs over a service cycle; Indicates service period Initial inventory The service life of finished corn varieties Beginning inventory ; This represents the inventory loss cost per ton of different varieties of finished corn products for corn processing enterprises. This indicates that the contents were stored in the silo. Inventory loss rate of finished corn products after one service cycle; This indicates that corn processing enterprises adopt Cargo loss coefficient for transportation modes; This indicates the variety of finished corn products per ton for corn processing enterprises. The selling price; This indicates that corn processing enterprises in the first Periodic use Corn finished grain varieties in different transportation modes Transportation volume; This indicates the variety of finished corn products per ton for corn processing enterprises. The cost of delay penalties; Indicates corn processing enterprises Corn finished grain varieties at the end of the cycle The amount of stock shortage; These represent different freight transport sections, in order: railway transport section, highway transport section, ocean transport section, and inland waterway transport section. S2: Based on the user company's order demand, delivery time, processing and inventory status, formulate a transportation plan to construct a target mathematical model that minimizes the total delivery cost of the corn processing enterprise's supply chain cost accounting model, based on the characteristics of different transportation modes; the target mathematical model includes the objective function and constraints that minimize the total delivery cost. Specifically, in this embodiment, the logistics company formulates a transportation plan based on the feed company's order demand, delivery time, and processing and inventory status. The aim is to select a suitable transportation mode based on the characteristics of different transportation modes, thereby minimizing the total delivery cost. The objective function is: (36) The constraints are as follows: (37) (38) (39) (40) (41) (42) (43) (44) (45) (46) (47) (48) (49) (50) (51) (52) (53) (54) Among them, constraint (37) represents the weight correspondence between the raw grain purchased by the corn processing enterprise and the finished corn product it has processed; constraint (38) represents that the sum of the raw grain allocation is equal to the raw grain purchase quantity; constraint (39) represents the transportation volume. The number of containers required for containerized packaging; constraints (40) and (41) indicate and Only both can be greater than 0 or equal to 0; constraint (42) indicates that each corn variety can only choose one transportation mode or not transport in each service cycle; constraints (43) and (44) indicate that, based on the first-in-first-out (FIFO) principle, after deducting the quantity already transported at the beginning of the service cycle, the remaining inventory of newly received grain is calculated recursively; constraints (45) and (46) indicate that, based on the first-in-first-out (FIFO) principle, after deducting the transportation volume at the beginning of each service cycle, the remaining inventory of stored old grain is calculated; constraint (47) indicates the recursive relationship of the shortage quantity; constraint (48) sets the first term of the shortage quantity recursive sequence to 0 and sets the last term of the shortage quantity to 0, indicating that the final demand of the feed processing enterprise needs to be met within the entire planning cycle; constraint (49) indicates that the transportation volume must be higher than the minimum delivery volume; constraints (50) to (52) indicate the non-negativity constraints of the variables; constraint (53) indicates As a zero-one variable, corn processing enterprises during the service cycle ,use Transportation mode for transporting corn varieties If the value is 1, then the value is 1; otherwise, the value is 0. Constraint (54) indicates that... The integers are non-negative, and the mathematical symbols used in constraints (44) to (46) are also non-negative. Its meaning is This embodiment considers seasonal corn raw material purchase price parameters. It can be calculated using equation (56): (55) In the formula: This indicates that corn processing enterprises in the first Cyclical Corn Finished Grain Varieties Production volume; This indicates that corn processing enterprises process raw corn into finished corn products. Conversion factor that leads to a reduction in cargo weight; This indicates that corn processing enterprises adopt Transportation mode, in the Cyclical corn finished grain varieties Number of containers required for transportation; Indicates corn finished grain varieties The density; Indicates container volume under different transportation modes; Represents positive real numbers; This indicates the service cycle of corn processing enterprises. Whether to adopt Transportation mode for transporting corn varieties ; This indicates the goods received at the beginning of the first service cycle. Inventory of finished corn products at the beginning of the first service cycle; Indicates the product type in the first service cycle. Production volume of finished corn products; This indicates that the first service cycle adopts... Corn finished grain varieties in different transportation modes Transportation volume; Indicates the first The initial inventory at the beginning of each service cycle The finished corn varieties in the first Inventory at the beginning of each service cycle; This indicates that corn processing enterprises in the first Each service cycle product Production volume of finished corn products; Indicates the first Each service cycle adopts a transportation mode. Varieties The volume of finished corn products transported; This indicates that the contents were stored in the silo. Inventory loss rate of finished corn products after one service cycle; This indicates the varieties that entered the warehouse at the beginning of the first service cycle. Corn finished grain in the first Inventory at the beginning of each service cycle; This indicates the varieties that entered the warehouse at the beginning of the first service cycle. Corn finished grain in the first Inventory at the beginning of each service cycle. This indicates that the contents were stored in the silo. Inventory loss rate of finished corn products after one service cycle This indicates the varieties that entered the warehouse at the beginning of the first service cycle. Corn finished grain in the first Inventory at the beginning of each service cycle; Indicates the first Varieties entering the warehouse at the beginning of each service cycle Corn finished grain in the first Inventory at the beginning of each service cycle; This indicates that the contents were stored in the silo. Inventory loss rate of finished corn products after one service cycle; Indicates the first End of each service cycle The shortage of finished corn products; This indicates that feed companies in the first Cyclical Corn Finished Grain Varieties Demand; express Initial shortage of finished corn products of various varieties. Indicates the first End of each service period Shortage of finished corn varieties; This indicates the minimum order quantity for each mode of transport; This indicates that corn processing enterprises adopt Transportation mode, in the Cyclical corn finished grain varieties Number of containers required for transportation; Represents the set of non-negative integers; This indicates the highest price per unit of raw corn purchased by corn processing enterprises within the planning period; This indicates the lowest price per unit of raw corn purchased by corn processing enterprises within the planning period; This indicates the service period with the lowest corn raw material procurement price within the planning period.

[0029] S3: Linearize the constraints in the target mathematical model and combine them with the objective function to obtain the optimized target model; in this embodiment, the linearization of the constraints in the target mathematical model includes: The constraint (39) is linearized because it is a nonlinear constraint in the model. The constraint uses an up-rounding function, and considering that... Since it is a non-negative integer, the constraint can be linearized as shown in equation (56): (56) Constraint (44) is clearly also a nonlinear constraint, which is addressed by introducing an auxiliary continuous variable. With auxiliary zero-one variables ,make By linearizing constraint (44), the constraint can be linearized as shown in equations (57) to (63): (57) (58) (59) (60) (61) (62) (63) In the formula: This represents an auxiliary zero-one variable, taking the value 0 or 1; By introducing auxiliary zero-one variables By linearizing constraints (44) and (45), constraint (45) can be linearized as shown in equations (64) to (69): (64) (65) (66) (67) (68) (69) By introducing auxiliary continuous variables With auxiliary zero-one variables , The constraint (46) is linearized using a two-step method: Step 1: Order We can obtain: (70) (71) (72) (73) (74) (75) Step Two: Order We can obtain: (76) (77) (78) (79) (80) (81) Therefore, the 12 linear inequalities from equation (70) to (81) are equivalent to the nonlinear mathematical relationship expressed by constraint (46); furthermore, note that constraint (51) has already... Since the constraints are non-negative real numbers, equations (67) and (79) can be deleted. S4: Construct carbon emission accounting models for different transportation modes. Logistics companies select the carbon emission amounts for different transportation modes, as shown in formulas (82) to (91): when hour, The calculation formula is: (82) when hour, The calculation formula is: (83) when hour, The calculation formula is: (84) when hour, The calculation formula is: (85) when hour, The calculation formula is: (86) when hour, The calculation formula is: (87) when hour, The calculation formula is: (88) when hour, The calculation is as follows: (89) when hour, The calculation formula is: (90) when hour, The calculation formula is: (91) Construct carbon constraint strategies based on carbon emission accounting for different transportation modes, including periodic carbon constraints, cumulative carbon constraints, full-cycle carbon constraints, and fluctuating carbon constraints. Periodic carbon constraints: Carbon allowances are allocated according to the planned period, distributed across different cycles. The goal is to ensure that carbon emissions in each cycle do not exceed the allowance for that cycle, and any remaining carbon allowances from the previous cycle cannot be carried over to the next cycle. Indicates service period The maximum permissible carbon emissions are: (92) After simplification, we get: (93) In the formula: Indicates logistics companies Maximum unit carbon dioxide emissions during the cycle; Indicates logistics companies Periodic selection Carbon emissions from transportation modes; The cumulative carbon constraint involves reducing the intensity of the periodic carbon constraint, allowing unused carbon emissions from the previous service cycle to be used in the next service cycle, while ensuring that the total accumulated carbon emissions do not exceed a given total carbon emission limit. (94) In the formula: Indicates logistics companies Maximum unit carbon dioxide emissions during the cycle; Indicates logistics companies Periodic selection Carbon emissions from transportation modes; The full-cycle carbon constraint means that the carbon emissions from delivering corn do not exceed the maximum allowable carbon emissions for the entire cycle; therefore, the maximum allowable emissions per unit of emissions for the environment. It no longer depends on the planning cycle and is fixed, that is , ,get: (95) In the formula: This indicates the total carbon emission limit per unit for logistics companies throughout the entire planning period; The aforementioned fluctuating carbon constraint: if there is If carbon emissions fluctuate over a given period, then the scope of the constraint no longer depends on the entire planning period, resulting in: (96) In the formula: This indicates that carbon emissions are a cyclical quantity with fluctuations, when When the condition is met, the model is equivalent to the periodic constraint case; when... In this case, the model is equivalent to the full-cycle scenario, meaning that periodicity and full-cycle are special forms of volatility.

[0030] The established mixed integer programming model has the following structural characteristics that make it difficult to solve directly: (1) Multi-stage strong coupling: The three stages of procurement-production, inventory management and transportation decision-making form a close spatiotemporal coupling relationship through FIFO (first-in, first-out) inventory constraints (43)-(46) and intertemporal carbon emission constraints (92)-(96). This coupling leads to complex dependencies between decision variables, making it difficult to apply traditional decomposition methods effectively.

[0031] (2) Numerous auxiliary variables and constraints: In order to achieve model linearization, a large number of zero- and one-auxiliary variables were introduced. and continuous auxiliary variables And the corresponding Big M constraint. For a typical instance containing 48 service cycles, 3 maize varieties, and 17 transportation modes, the total number of decision and auxiliary variables in the model exceeds 14,440, and the number of constraints exceeds 42,000. Given the above characteristics, directly using commercial solvers to solve large-scale instances faces computational bottlenecks. Preliminary numerical experiments show that even for medium-sized instances ( Even with a time limit of 3600 seconds, Gurobi struggles to find a solution with a gap value below 20%. To address the multi-stage and strongly coupled nature of this problem, a branch-pricing algorithm is proposed, where a greedy algorithm is used to generate the initial column, i.e., the integrated service plan sequence.

[0032] S5: Based on the branch-pricing algorithm framework, and according to the optimization objective model and carbon constraint strategy, obtain a collaborative optimization scheme for the procurement, processing, and transportation of corn raw materials and finished products under carbon constraints, such as... Figure 3 As shown, it includes: S51: Define the integrated service plan. Based on the integrated service plan, reconstruct the service plan-oriented model according to the optimization objective model and carbon constraint strategy. Specific steps include: S511: Define Integrated Services Plan ,in This refers to a collection of integrated service plans. This represents a feasible complete decision within the planning period, specifically expressed mathematically as follows: (97) In the formula: Indicates integrated service plan During the service period The amount of corn raw materials purchased; Indicates integrated service plan During the service period The following are used for production varieties The allocation of corn raw materials; , Indicates integrated service plan The transportation mode selection and transportation volume decision variables; in addition, each service plan All constraints of the original problem, except for the carbon constraint, must be satisfied. Simultaneously define the integrated service plan Plan characteristic parameters for: Indicates integrated service plan During the service period For varieties Total transport volume and ; Indicates integrated service plan During the service period Total carbon emissions and ; Indicates integrated service plan During the service period choose Carbon emissions from transportation modes; Indicates integrated service plan The total delivery cost is calculated according to formula (1); S512: Based on step S511, reconstruct the service planning-oriented model according to the optimization objective model and carbon constraint strategy, wherein the service planning-oriented model includes the master problem model. With constraints: (98) Constraints: (99) (100) (101) In the formula: Indicate whether to select the integrated service plan Decision variables; This refers to the integrated service plan under cyclical carbon constraints, cumulative carbon constraints, full-cycle carbon constraints, or fluctuating carbon constraints. The corresponding negative average carbon emission surplus; constraint (99) represents the carbon constraint, constraint (100) guarantees that only one complete service plan can be selected, and constraint (101) represents For zero or one variables; Indicates integrated service plan Total delivery cost of corn processing enterprises ; Under periodic carbon constraints, cumulative carbon constraints, full-cycle carbon constraints, and fluctuating carbon constraints The specific calculation methods are shown in formulas (102)-(104): Under periodic carbon constraints: (102) Under cumulative carbon constraints: (103) Under full-cycle carbon constraints: (104) Under the constraint of fluctuating carbon: (105) Furthermore, in the column generation framework, the restricted master problem (RMP) of the master problem MP is defined as the problem in the current column set. The main problem on the above, its linearly relaxed form is as follows: (106) Where constraints (99) and (100) remain unchanged, the decision variables are relaxed to continuous variables: (107) S52: Based on the branch pricing algorithm framework, a pricing sub-problem model is constructed according to the reconstructed service plan-oriented model; and the pricing sub-problem model is solved based on the enhanced hybrid dynamic programming-MILP pricing algorithm to obtain a coordinated optimization scheme for corn raw material procurement and processing and finished grain transportation under carbon constraints. Specific steps include: S521: The pricing subproblem aims to generate feasible procurement-inventory-transportation plans that satisfy global demand balance constraints. Let... The dual variable corresponding to the carbon constraint of the main problem. To constrain the dual variables of (100), a pricing subproblem is constructed based on the reconstructed service plan-oriented model. The model is: (108) In the formula: Representing the pricing sub-problem Model The model function; Constraints: (109) (110) (111) (112) (113) (114) (115) (116) (117) In the formula: Indicates integrated service plan The corresponding negative average carbon emission surplus is ; express The abbreviated form; This represents the dual variable corresponding to the carbon constraint in the main problem model; Represents the dual variable of constraint (100); Represents the recursive relationship of inventory under the first-in, first-out (FIFO) strategy; This represents the conversion coefficient that reduces the weight of corn as a processing enterprise processes raw corn into finished corn products; constraint (109) is a procurement and production allocation constraint, and constraint (110) is expressed through the function The first-in-first-out (FIFO) strategy represents the inventory recursion relationship. Constraints (111)-(114) are transportation mode selection constraints, constraint (115) is the cumulative shortage calculation formula, constraint (116) ensures that the total transportation volume is consistent with the total demand volume, and constraint (117) is the non-negativity constraint for continuous variables and the value constraint for integer variables.

[0033] In this embodiment, the pricing subproblem is divided into three strongly coupled stages: procurement and production, inventory management, and transportation decisions. Noting that the evolution of inventory states has discrete, stage-based characteristics (i.e., periodic transitions), and that procurement, production, and transportation decisions under a given inventory state can be precisely solved through continuous optimization, this paper proposes a hybrid dynamic programming-MILP solution framework. This framework decomposes the problem into a dynamic programming (DP) layer and a MILP optimization layer. The former handles discrete state transitions and temporal dependencies, while the latter precisely optimizes continuous decision variables based on the DP solution. S522: Solving the pricing subproblem model based on the enhanced hybrid dynamic programming-MILP pricing algorithm, which includes: Augmented state vectors are defined based on dynamic programming layers. and ,in This represents the inventory level vector for each service cycle; Batch representing the storage duration distribution vector; Define state transition function And based on the pricing sub-problem The constraints corresponding to the model, based on the previous augmented state vector, result in the current state vector as follows: (118) In the formula: This represents the vector of the previous augmented state; They represent the corresponding corn processing enterprises in the first The quantity of corn raw materials procured periodically, and the service period. Of the initial batch of raw corn stored, a significant portion is processed into finished corn products. The quantity and corn processing enterprises in the first Periodic use Corn finished grain varieties in different transportation modes Transportation volume; for the first service period (t=1), the current state vector is determined by the initial value; for subsequent periods (t≥2), the current state vector is determined based on the augmented state vector of the previous period and the current decision variables; Based on the defined adaptive discretization strategy and state domination and pruning strategy, the solution obtained by the discrete dynamic programming is obtained according to step S521 and formula (80). Specifically, this embodiment considers that traditional dynamic programming has difficulty directly handling continuous decision spaces. In order to handle continuous variables in the DP framework, their value space is divided into a discrete grid. The defined adaptive discretization strategy includes: (1) Price sensitivity principle: Based on experience, dynamically adjust according to the time series characteristics of the purchase price. The discrete density indicates that if the price is low, companies tend to purchase in large quantities. A setting is established where the maximum grid value is increased when the price falls below 80% of the average, and the maximum grid value and grid density are scaled proportionally based on the distance between the current price and the lowest price.

[0034] (2) Inventory and Stockout Balance Principle: Set appropriate discrete decision points based on the current inventory balance and accumulated stockout: that is, for This increases the decision points for meeting the demand in the next period, meeting the current period's accumulated stock shortage, meeting the sum of the current period's accumulated stock shortage and the next period's demand, and meeting the sum of the current period's accumulated stock shortage and the demand in the next two periods; for Then you need to first select The decision quantity is determined by the current inventory balance and the accumulated shortage to determine the maximum value. ,in express The corresponding inventory loss rate vector, and under the constraint of this maximum value, based on empirical values, increases the decision points to meet the demand of the next period, meet the cumulative shortage of the current period, and meet the sum of the cumulative shortage of the current period and the demand of the next period.

[0035] (3) Transportation capacity constraint principle: Adjust decision based on transportation volume constraint (77) The discrete grid is used. For the minimum delivery quantity constraint, the minimum value of the discrete grid for the transportation quantity is determined, and discrete points below the minimum value are deleted; for the container transportation mode... Add an appropriate number of discrete points, whose values ​​are set to integer multiples of container capacity; and for the constraint (116) that requires the total transportation volume to remain consistent with the total demand, it is necessary to forcibly set the transportation volume decision value for the last period to be... ,in Indicates the first End of each service cycle The shortage of finished corn products. Indicates the first End of each service cycle Demand for finished corn products.

[0036] In the state space of dynamic programming, many states are unlikely to produce better solutions in subsequent optimizations. Identifying and pruning these redundant states is key to improving algorithm efficiency. To identify redundant states, strong dominance relationships and weak dominance relationships are defined, i.e., the state dominance and pruning strategies include: (1) Strong dominance relation: requires that the dominant state is not inferior to the dominated state in all dimensions, i.e., the augmented state vector Strong dominance state It is required if and only if the following conditions are met: (i) inventory level (ii) Total cumulative cost (iii) Average inventory duration (iv) Accumulate stockout costs. ;in, Indicates the variety of corn processed by enterprises per ton The penalty cost for delays in finished corn products; Representing state The corresponding inventory level vector; Representing state The corresponding cumulative total cost; Representing state The corresponding average inventory duration; (2) Weak dominance: This allows weaknesses in certain dimensions to be compensated for by strengths in others. Specifically, dominance is determined using the following weighted scoring function: (122) In the formula: This represents the weighting coefficient, which can be dynamically adjusted based on the characteristics of the problem and the current stage. Indicates the variety of corn processed by enterprises per ton The penalty cost for delays in finished corn products; (3) Strong dominance pruning: By traversing the state pairs, identify and delete the strongly dominated states confirmed by the strong dominance relationship. In this embodiment, since strong dominance ensures that the deleted states cannot participate in the optimal path, this pruning does not affect the optimality.

[0037] (4) Weak Domination Pruning: Further compressing the state space based on strong domination pruning. That is, based on strong domination pruning, when the cost difference between two states is less than a preset threshold (5%) and a weak domination relationship exists, only the dominant state is retained and the dominated state is deleted. This pruning may sacrifice theoretical optimality, so it is necessary to reasonably set the size of the cost difference threshold.

[0038] This embodiment also includes an analysis of the algorithm's theoretical properties: Theorem 1 (Discretization Error Bound): Let the Lipschitz constant of the cost function with respect to the decision variables be... Service period The maximum grid spacing is Then the upper bound of the error introduced by discretization is This theorem shows that the optimal solution can be arbitrarily approximated by controlling the grid spacing. In practice, the adaptive discretization strategy ensures that the solution is accurate in critical regions. Small enough, while maintaining a large spacing in non-critical areas to control computational complexity.

[0039] Theorem 2 (Optimality Preservation in Strong Dominated Pruning): Strong dominated pruning does not remove states that participate in the optimal dynamic programming path.

[0040] Proof: Proof by contradiction. Assume the optimal dynamic programming path passes through states that have been strongly dominated and deleted. Due to the existence of state Strong Domination ,from The cost of the optimal dynamic programming path starting from [the starting point] must not be higher than that starting from [the point where] The starting dynamic programming path, which is... There is a contradiction regarding the optimal dynamic programming path.

[0041] S523: Based on the MILP optimization layer, a two-stage optimization mechanism is constructed according to steps S521 to S522. Although discretization makes dynamic programming feasible, it inevitably introduces suboptimal behavior. This paper reduces this accuracy loss through a two-stage optimization mechanism. Phase 1: Extracting the structural features of the discrete solution. Let the solution obtained from the discrete dynamic programming be... The structure extraction operator is defined as follows: (123) In the formula: Indicates the actual production strategy and ; Indicates actual transportation decisions and ; This represents the amount of raw grain processed and allocated in the solution obtained from discrete dynamic programming. ; This represents the zero- or one-variable choice of transportation mode in the solution obtained from discrete dynamic programming. ; The second stage involves constructing a continuous optimization model (COM) based on the structure extraction operator. This model forces some decision variables of the original problem to be defined as 0 according to the structure extraction operator, thereby significantly reducing the solution space of the decision variables. Its expression is as follows: (124) The constraints of the continuous optimization model COM consist of constraints (112)-(120) and constraints (125)-(126): (125) (126) Among them, constraints (125)-(126) are the extraction operators. Structural constraints.

[0042] The enhanced hybrid dynamic programming-MILP pricing algorithm determines the combinatorial structure of the decision, i.e., feasible procurement-inventory-transportation schemes, through the discretization stage in step S522; and determines the precise numerical values ​​of the decisions through a two-stage optimization mechanism. The key to this process lies in utilizing the decomposition characteristics of the problem, enabling the model to operate in multinomial time. We can obtain high-quality solutions and then obtain a collaborative optimization scheme for the procurement, processing and transportation of corn raw materials and finished products under carbon constraints.

[0043] S53: Step S51 reconstructed the service-oriented model, and step S52 constructed the pricing sub-problem and designed the corresponding solution algorithm. This embodiment will introduce a column generation process and subsequent branching strategy on this basis to improve the branch pricing algorithm structure, so as to solve the final collaborative optimization scheme of corn raw material procurement and processing and finished grain transportation under carbon constraints. Specifically: S531: In the column generation part of this embodiment, firstly, a linear relaxation problem (RMLP) containing only the restricted master problem of an initially feasible integrated service plan needs to be constructed and solved to obtain the dual variable; then, a pricing subproblem (PSP) is constructed and solved based on the dual variable. If the obtained objective function value is negative, a new service plan is constructed based on the solution result, added to the RMLP, and solved again. After updating the dual variable, the above process of solving the pricing subproblem and adding the service plan is repeated until the objective function value obtained by solving the PSP is non-negative, at which point the column generation process ends. S532: This embodiment also includes the fact that after the column generation process is completed, the linear relaxation problem (RMLP) that restricts the main problem often produces non-integer solutions. In this embodiment, in order to obtain integer solutions, a suitable branching strategy needs to be designed. The traditional branch and bound method directly performs 0-1 branches on the fractional variables in the obtained solution. However, under the main problem and subproblems constructed in this embodiment, this strategy can only force the disabling / enabling of one solution of the subproblem, resulting in an extreme imbalance between the left and right branches and relatively low solution efficiency. Therefore, this embodiment adopts an indirect branching strategy based on transportation decision variables: Defining transportation mode decision-making The score contribution is: (127) Then select That is, select the decision pair closest to 0.5 and create branch child nodes: (1) Left branch child node: All integrated service plans must include decision pairs For RMLP, the decision variables corresponding to integrated service plans that do not contain decision pairs are forced to 0; for PSP, in the phase Under these circumstances, mandatory transportation decision variables =1; (2) Right branch child node: All integrated service plans cannot contain decision pairs For RMLP, the decision variables corresponding to the integrated service plan containing decision pairs are forced to 0; for PSP, in the phase... Under these circumstances, transportation modes are excluded from transportation decisions. ; Regarding the branch node selection rules, a hybrid strategy combining depth-first and best-bound approaches is adopted, defining node priorities as follows: (128) In the formula: Used to balance depth exploration and boundary improvement Represents a node The depth of the branch tree, To avoid issues with division by zero, this hybrid strategy is preferred. By exploring the largest node, we can maintain the efficiency and solvability of subproblems while balancing the left and right branches, systematically explore the integer solution space, and ensure convergence to the global optimum or near-optimal solution.

[0044] This embodiment also includes the following case study: Using the Chinese corn supply chain as a case study, Jilin Yuntianhua Agricultural Development Co., Ltd. is selected as the processing enterprise, and Wuhan Zhengda Co., Ltd. as the downstream feed enterprise. The cost of transporting and operating finished grain is calculated. Both enterprises have access to railway facilities, with commonly used transit railway stations being Tongliao Station and Hankou Station, providing convenient hub nodes for railway transportation. The study considers the processing enterprise's raw material procurement, processing, and finished grain inventory management, while the logistics enterprise is responsible for finished grain transportation planning and carbon emission accounting, to simulate the entire chain's collaborative optimization under carbon constraints in the transportation process. This scenario can realistically reflect the transportation, inventory, and carbon emission constraints of the inter-provincial corn supply chain, providing an important foundation for optimizing transportation mode selection and reducing supply chain losses.

[0045] (1) Analysis of carbon constraint strategies: 1) When This involves comparing the number of times a corn processing company chooses different transportation modes under various carbon constraint strategies—periodic, cumulative, full-cycle, and fluctuating—with a fixed total carbon emission limit per unit over the entire planning period. The experimental results are as follows: Figure 4-7 As shown. ① Under the four carbon constraint strategies, the choice of transportation modes mainly focuses on five modes, with "road-rail-road" intermodal bulk cargo transportation being the most common, followed by direct rail bulk cargo transportation, "rail-water-rail" intermodal river-sea direct bulk cargo transportation, "north-to-south grain and south-to-north fertilizer" dedicated river-sea direct container transportation, and direct road container transportation. ② Among the "road-rail-road" intermodal bulk cargo transportation mode, fluctuating carbon constraint strategies are more frequently adopted. ③ Among the "rail direct" bulk cargo transportation mode, full-cycle carbon constraint strategies are more frequently adopted. ④ Among the "rail-water-rail" intermodal river-sea direct bulk cargo transportation mode, cyclical carbon constraint strategies are more frequently adopted. ⑤ Among the "north-to-south grain and south-to-north fertilizer" dedicated river-sea direct container transportation mode, cyclical carbon constraint strategies are more frequently adopted. ⑥ Among the "direct road container transportation mode," full-cycle carbon constraint strategies are more frequently adopted. ⑦ Under the four carbon emission strategies, the total number of transportation mode selections, arranged in descending order, is: volatile carbon constraint strategy > full-cycle carbon constraint strategy > cyclical carbon constraint strategy > cumulative carbon constraint strategy. It is evident that the volatile carbon constraint strategy results in the least stockpiling, while the cumulative carbon constraint strategy results in the most stockpiling. 2) When This involves comparing the number of times corn processing enterprises choose different transportation modes for different varieties under various carbon constraint strategies—periodic, cumulative, full-cycle, and fluctuating—with a fixed total carbon emission limit per unit throughout the entire planning period. The experimental results are as follows: Figure 8-11As shown. ① Under the full-cycle carbon constraint strategy, if the corn variety is a low-toxin enhanced variety, the direct rail transport and "road-rail-road" intermodal transport modes are more commonly chosen; if the corn variety is a low-moisture long-term storage variety, the direct rail transport mode is more commonly chosen; if the corn variety is a high-standard general-purpose variety, the "road-rail-road" intermodal transport mode is more commonly chosen. ② Under the periodic carbon constraint strategy, cumulative carbon constraint strategy, and volatile carbon constraint strategy, if the corn variety is a low-toxin enhanced variety, the direct rail transport mode is more commonly chosen; if the corn variety is a low-moisture long-term storage variety, the direct rail transport mode is more commonly chosen; if the corn variety is a high-standard general-purpose variety, the "road-rail-road" intermodal transport mode is more commonly chosen. 3) When When the total carbon emission limit per unit is fixed throughout the entire planning cycle for logistics companies, this study compares the changes in total supply chain costs, procurement costs, processing costs, basic transportation costs, transportation time costs, transportation loss costs, inventory costs, inventory loss costs, and delayed delivery costs for corn processing companies under cyclical carbon constraint strategies, cumulative carbon constraint strategies, full-cycle carbon constraint strategies, and fluctuating carbon constraint strategies. The experimental results are as follows: Figure 12-15 As shown in the figure: ① The total supply chain costs of corn processing enterprises under the four carbon emission strategies, ranked in descending order, are: volatile carbon constraint strategy > full-cycle carbon constraint strategy > cyclical carbon constraint strategy > cumulative carbon constraint strategy. ② In the volatile carbon constraint strategy, as the total carbon emission limit per unit increases throughout the planning period, the procurement costs and transportation time costs of corn processing enterprises show a significant decrease, and the total supply chain cost shows a volatile decrease. ③ In the full-cycle carbon constraint strategy, as the total carbon emission limit per unit increases throughout the planning period, the total supply chain cost, basic transportation cost, and transportation time cost of corn processing enterprises show a significant decrease. ④ In the cyclical carbon constraint strategy, as the total carbon emission limit per unit increases throughout the planning period, the procurement costs and transportation time costs of corn processing enterprises show a significant decrease. ⑤ In the cumulative carbon constraint strategy, as the total carbon emission limit per unit increases throughout the planning period, the inventory costs of corn processing enterprises show a significant decrease, while the processing costs and inventory loss costs show a volatile decrease.

[0046] (2) Analysis of finished grain demand under the full-cycle carbon constraint strategy: Considering the differences in control methods and applicable scope of the four carbon emission strategies, the full-cycle carbon constraint strategy can systematically reflect the carbon emission characteristics of enterprises throughout the entire supply chain life cycle, and has higher representativeness and research value. Therefore, this study focuses on the full-cycle carbon constraint strategy to reveal its key role in comprehensive cost control and sustainable optimization: 1) If feed enterprises need low-toxin fortified and high-standard general-purpose feed corn, and the proportion and total amount of demand in each service cycle remain unchanged, the proportion of low-toxin fortified and high-standard general-purpose feed corn is changed, and the proportion of transportation mode selection is analyzed. The experimental results are as follows: Figure 16-18As shown in the diagram: ① When the proportion of low-toxin-enhanced feed corn reaches 100%, the choice of container transport mode reaches its peak. ② With the increase in the proportion of low-toxin-enhanced feed corn, the direct rail bulk transport mode, the direct road container transport mode, and the "North Grain to South and South Fertilizer to North" dedicated river-sea direct container transport mode show a fluctuating upward trend, while the "rail-water-rail" intermodal river-sea direct bulk transport mode and the "road-rail-road" intermodal bulk transport mode show a fluctuating downward trend. ③ With the increase in the proportion of low-toxin-enhanced feed corn, the impact of direct rail bulk transport of low-toxin-enhanced corn and "road-rail-road" intermodal bulk transport of high-standard general-purpose corn fluctuates more significantly compared to other modes. ④ With the increase in the proportion of low-toxin-enhanced feed corn, the total supply chain cost and inventory loss cost of corn processing enterprises show a fluctuating downward trend, while the transportation time cost, transportation loss cost, and delayed delivery cost show a downward trend. 2) If feed companies require low-toxin fortified, low-moisture long-storage, and high-standard general-purpose feed corn varieties, with each variety accounting for 1 / 3 of the total, the proportion of the total demand in each service cycle to the total demand in the entire planning cycle remains unchanged. The experimental results are as follows: Figure 19-21 As shown: ① With the increase in total demand from feed companies throughout the entire planning cycle, the direct road container transport mode shows a fluctuating upward trend, while the "North Grain to South and South Fertilizer to North" dedicated container river-sea direct transport mode shows a fluctuating downward trend. ② With the increase in total demand from feed companies throughout the entire planning cycle, the impact of direct rail bulk transport for low-toxin fortified feed, low-moisture long-term storage feed, and "road-rail-road" multimodal bulk transport for high-standard general-purpose feed fluctuates more significantly compared to other modes. ③ With the increase in total demand from feed companies throughout the entire planning cycle, the overall cost of each link in the supply chain increases significantly, but inventory costs decrease significantly, which is due to less stockpiling.

[0047] (3) Analysis of raw grain procurement prices under the full-cycle carbon constraint strategy: Considering the seasonal changes in raw grain procurement prices, the impact of raw grain procurement price fluctuations on the choice of transportation mode for finished corn products is analyzed. The experimental results are as follows: Figure 22-23 As shown, with the increase in the amplitude of raw grain procurement prices, the main transportation modes selected are direct rail bulk transport, direct road container transport, "rail-water-rail" intermodal river-sea direct bulk transport, "north grain to south and south fertilizer to north" dedicated river-sea direct container transport, and "road-rail-road" intermodal bulk transport. Among them, with the increase in the amplitude of raw grain procurement prices, the direct rail bulk transport mode and the "road-rail-road" intermodal bulk transport mode will gradually decrease, while the direct road container transport, "rail-water-rail" intermodal river-sea direct bulk transport, and "north grain to south and south fertilizer to north" dedicated river-sea direct container transport will gradually increase. At the same time, with the increase in the amplitude of raw grain procurement prices, supply chain costs and procurement costs decrease significantly, with more costs being invested in transportation loss costs and delayed delivery costs.

[0048] (4) Analysis of the "Bulk to Container" Policy under the Full-Cycle Carbon Constraint Strategy: From the perspective of the overall grain market, bulk and bagged transportation still dominate. Therefore, to promote the implementation of the "Bulk to Container" policy, this study explores the acceptable range for processing enterprises to adopt containerized transport of important agricultural inputs and fertilizers to reduce empty load rates and flatten transportation costs, as well as the range of policy subsidies. Experimental results are as follows: Figure 24 As shown. Comparing the return trip cost sharing of Mode 7 (rail-water-rail) bulk river-sea transshipment and Mode 12 ("North Grain to South and South Fertilizer to North") dedicated line container river-sea transshipment on the same route, the cost intersects at 125.417 yuan / TEU. This indicates that if the return trip cost sharing or container transport policy subsidies reach 125.417 yuan / TEU, cargo owners will be willing to use container transport. For different types of feed corn, such as... Figure 25 As shown, firstly, for low-moisture, long-term storage cargo, if the cost-sharing of return cargo transportation or the policy subsidy for container transportation reaches 186.67 yuan / TEU, cargo owners will be willing to use container transportation. Secondly, for high-standard, general-purpose cargo, if the cost-sharing of return cargo transportation or the policy subsidy for container transportation reaches 177.34 yuan / TEU, cargo owners will be willing to use container transportation.

[0049] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for collaborative optimization of corn raw material procurement, processing, and finished grain transportation under carbon constraints, characterized in that, Specifically, the following steps are included: S1: Considering the raw grain procurement cost, raw grain processing cost, finished grain inventory cost, finished grain inventory loss cost, finished grain basic transportation cost, finished grain loss cost during transportation, finished grain transportation time cost, and delayed delivery cost, construct a supply chain cost accounting model for corn processing enterprises. S2: Based on the user company's order demand, delivery time, processing and inventory status, formulate a transportation plan to construct a target mathematical model that minimizes the total delivery cost of the corn processing enterprise's supply chain cost accounting model, based on the characteristics of different transportation modes. The target mathematical model includes an objective function and constraints that minimize the total delivery cost; S3: Linearize the constraints in the target mathematical model and combine them with the objective function to obtain the optimized target model; S4: Construct carbon constraint strategies based on carbon emission accounting for different transportation modes; S5: Based on the branch pricing algorithm framework, obtain a collaborative optimization scheme for corn raw material procurement and processing and finished grain transportation under carbon constraints, according to the optimization objective model and carbon constraint strategy.

2. The method for coordinated optimization of corn raw material procurement, processing, and finished grain transportation under carbon constraints according to claim 1, characterized in that, The supply chain cost accounting model for corn processing enterprises constructed in S1 is as follows: （1） （2） （3） （4） （5） （6） （7） In the formula: This represents the total delivery cost for corn processing enterprises; This indicates the service cycle of corn processing enterprises. The resulting procurement costs; This indicates the service cycle of corn processing enterprises. The resulting processing costs; This indicates the service cycle of corn processing enterprises. The resulting inventory costs; This represents the inventory loss cost incurred by corn processing enterprises throughout the entire planning period; This indicates the service cycle of corn processing enterprises. use The basic transportation cost of the transportation mode; This indicates the service cycle of corn processing enterprises. use Costs of cargo loss during the transportation process in different transportation modes; This indicates the service cycle of corn processing enterprises. use Transportation time cost of transportation modes; This indicates the service cycle of corn processing enterprises. The cost of delayed delivery; Indicates the total service period; Represents a set of transportation modes; This indicates that corn processing enterprises are The unit price of raw grain at the beginning of the service period; This indicates that corn processing enterprises in the first The amount of corn raw materials purchased periodically; This indicates that corn processing enterprises process it into... The unit processing cost of graded feed corn; Indicates service period Of the initial batch of raw corn stored, a significant portion is processed into finished corn products. The amount; This indicates the types of finished grain products produced by corn processing enterprises; This indicates the variety of finished corn products per ton for corn processing enterprises. Inventory costs over a service cycle; Indicates service period Initial inventory The service life of finished corn varieties Beginning inventory ; This represents the inventory loss cost per ton of different varieties of finished corn products for corn processing enterprises. This indicates that the contents were stored in the silo. Inventory loss rate of finished corn products after one service cycle; This indicates that corn processing enterprises adopt Cargo loss coefficient for transportation modes; This indicates the variety of finished corn products per ton for corn processing enterprises. The selling price; This indicates that corn processing enterprises in the first Periodic use Corn finished grain varieties in different transportation modes Transportation volume; This indicates the variety of finished corn products per ton for corn processing enterprises. The cost of delay penalties; Indicates corn processing enterprises Corn finished grain varieties at the end of the cycle The amount of stock shortage.

3. The method for coordinated optimization of corn raw material procurement, processing, and finished grain transportation under carbon constraints according to claim 2, characterized in that, The objective function in S2 is: （8） The constraints are as follows: （9） （10） （11） （12） （13） （14） （15） （16） （17） （18） （19） （20） （21） （22） （23） （24） （25） （26） （27） In the formula: This indicates that corn processing enterprises in the first Cyclical Corn Finished Grain Varieties Production volume; This indicates that corn processing enterprises process raw corn into finished corn products. Conversion factor that leads to a reduction in cargo weight; This indicates that corn processing enterprises adopt Transportation mode, in the Cyclical corn finished grain varieties Number of containers required for transportation; Indicates corn finished grain varieties The density; Indicates container volume under different transportation modes; Represents positive real numbers; This indicates the service cycle of corn processing enterprises. Whether to adopt Transportation mode for transporting corn varieties ; This indicates the goods received at the beginning of the first service cycle. Inventory of finished corn products at the beginning of the first service cycle; Indicates the product type in the first service cycle. Production volume of finished corn products; This indicates that the first service cycle adopts... Corn finished grain varieties in different transportation modes Transportation volume; Indicates the first The initial inventory at the beginning of each service cycle The finished corn varieties in the first Inventory at the beginning of each service cycle; This indicates that corn processing enterprises in the first Each service cycle product Production volume of finished corn products; Indicates the first Each service cycle adopts a transportation mode. Varieties The volume of finished corn products transported; This indicates that the contents were stored in the silo. Inventory loss rate of finished corn products after one service cycle; This indicates the varieties that entered the warehouse at the beginning of the first service cycle. Corn finished grain in the first Inventory at the beginning of each service cycle; This indicates the varieties that entered the warehouse at the beginning of the first service cycle. Corn finished grain in the first Inventory at the beginning of each service cycle. This indicates that the contents were stored in the silo. Inventory loss rate of finished corn products after one service cycle This indicates the varieties that entered the warehouse at the beginning of the first service cycle. Corn finished grain in the first Inventory at the beginning of each service cycle; Indicates the first Varieties entering the warehouse at the beginning of each service cycle Corn finished grain in the first Inventory at the beginning of each service cycle; This indicates that the contents were stored in the silo. Inventory loss rate of finished corn products after one service cycle; Indicates the first End of each service cycle The shortage of finished corn products; This indicates that feed companies in the first Cyclical Corn Finished Grain Varieties Demand; express Initial shortage of finished corn products of various varieties. Indicates the first End of each service period Shortage of finished corn varieties; This indicates the minimum order quantity for each mode of transport; This indicates that corn processing enterprises adopt Transportation mode, in the Cyclical corn finished grain varieties Number of containers required for transportation; Represents the set of non-negative integers; This indicates the highest price per unit of raw corn purchased by corn processing enterprises within the planning period; This indicates the lowest price per unit of raw corn purchased by corn processing enterprises within the planning period; This indicates the service period with the lowest corn raw material procurement price within the planning period.

4. The method for coordinated optimization of corn raw material procurement, processing, and finished grain transportation under carbon constraints according to claim 3, characterized in that, S3 performs linearization processing on the constraints in the target mathematical model, including: Linearize constraint (11): (28) By introducing auxiliary continuous variables With auxiliary zero-one variables , make Linearize constraint (16): （29） （30） （31） （32） （33） （34） （35） In the formula: This represents an auxiliary zero-one variable, taking the value 0 or 1; By introducing auxiliary zero-one variables Linearization is performed on constraints (16) and (17): （36） （37） （38） （39） （40） （41） By introducing auxiliary continuous variables With auxiliary zero-one variables , The constraint (18) is linearized using a two-step method: Step 1: Order We can obtain: （42） （43） （44） （45） （46） （47） Step Two: Order We can obtain: （48） （49） （50） （51） （52） （53）。 5. The method for coordinated optimization of corn raw material procurement, processing, and finished grain transportation under carbon constraints according to claim 4, characterized in that, The carbon constraint strategies constructed in S4 based on carbon emission accounting for different transportation modes include periodic carbon constraints, cumulative carbon constraints, full-cycle carbon constraints, and fluctuating carbon constraints. The expression for the periodic carbon constraint is: （54） In the formula: Indicates logistics companies Maximum unit carbon dioxide emissions during the cycle; Indicates logistics companies Periodic selection Carbon emissions from transportation modes; The expression for the cumulative carbon constraint is: （55） In the formula: Indicates logistics companies Maximum unit carbon dioxide emissions during the cycle; Indicates logistics companies Periodic selection Carbon emissions from transportation modes; The expression for the full-cycle carbon constraint is: （56） In the formula: This indicates the total carbon emission limit per unit for logistics companies throughout the entire planning period; The expression for the fluctuating carbon constraint is: （57） In the formula: This indicates that carbon emissions are fluctuating in a cyclical manner.

6. The method for coordinated optimization of corn raw material procurement, processing, and finished grain transportation under carbon constraints according to claim 5, characterized in that, S5 specifically includes the following steps: S51: Define the integrated service plan, and based on the integrated service plan, reconstruct the service plan-oriented model according to the optimization target model and carbon constraint strategy; S52: Based on the branch pricing algorithm framework, a pricing sub-problem model is constructed according to the reconstructed service plan-oriented model; and the pricing sub-problem model is solved based on the enhanced hybrid dynamic programming-MILP pricing algorithm to obtain a collaborative optimization scheme for corn raw material procurement and processing and finished grain transportation under carbon constraints.

7. The method for coordinated optimization of corn raw material procurement, processing, and finished grain transportation under carbon constraints according to claim 6, characterized in that, S51 specifically includes the following steps: S511: Define Integrated Services Plan ,in Indicates a collection of integrated service plans: （58） In the formula: Indicates integrated service plan During the service period The amount of corn raw materials purchased; Indicates integrated service plan During the service period The following are used for production varieties The allocation of corn raw materials; , Indicates integrated service plan Transportation mode selection and transportation volume decision variables; At the same time, define the integrated service plan Plan characteristic parameters for: Indicates integrated service plan During the service period For varieties Total transport volume and ; Indicates integrated service plan During the service period Total carbon emissions and ; Indicates integrated service plan During the service period choose Carbon emissions from transportation modes; Indicates integrated service plan Total delivery cost; S512: Based on step S511, reconstruct the service planning-oriented model according to the optimization objective model and carbon constraint strategy, wherein the service planning-oriented model includes the master problem model. With constraints: （59） Constraints: （60） （61） （62） In the formula: Indicate whether to select the integrated service plan The decision variables, and Represents zero and one variables; This refers to the integrated service plan under cyclical carbon constraints, cumulative carbon constraints, full-cycle carbon constraints, or fluctuating carbon constraints. The corresponding negative average carbon emission remaining amount; Indicates integrated service plan Total delivery cost of corn processing enterprises ; Under periodic carbon constraints: (63) Under cumulative carbon constraints: (64) Under full-cycle carbon constraints: (65) Under the constraint of fluctuating carbon: (66).

8. The method for coordinated optimization of corn raw material procurement, processing, and finished grain transportation under carbon constraints according to claim 7, characterized in that, S52 specifically includes the following steps: S521: Based on the branch-pricing algorithm framework, construct pricing sub-problems according to the reconstructed service plan-oriented model. The model is: （67） In the formula: Representing the pricing sub-problem Model The model function; Constraints: （68） （69） （70） （71） （72） （73） （74） （75） （76） In the formula: Indicates integrated service plan The corresponding negative average carbon emission surplus is ; This represents the dual variable corresponding to the carbon constraint in the main problem model; Represents the dual variable of constraint (61); Represents the recursive relationship of inventory under the first-in, first-out (FIFO) strategy; This represents the conversion factor that indicates the reduction in the weight of corn products when corn processing enterprises process raw corn into finished corn products. S522: Solving the pricing subproblem model based on the enhanced hybrid dynamic programming-MILP pricing algorithm, which includes: Augmented state vectors are defined based on dynamic programming layers. and ,in This represents the inventory level vector for each service cycle; Batch representing the storage duration distribution vector; Define state transition function And based on the pricing sub-problem The constraints corresponding to the model, based on the previous augmented state vector, result in the current state vector as follows: （77） In the formula: This represents the vector of the previous augmented state; They represent the corresponding corn processing enterprises in the first The quantity of corn raw materials procured periodically, and the service period. Of the initial batch of raw corn stored, a significant portion is processed into finished corn products. The quantity and corn processing enterprises in the first Periodic use Corn finished grain varieties in different transportation modes Transportation volume; Based on the defined adaptive discretization strategy and state domination and pruning strategy, the solution obtained by the discrete dynamic programming is obtained according to step S521 and formula (77); S523: Based on the MILP optimization layer, construct a two-stage optimization mechanism according to steps S521 to S522: Phase 1: Let the solution obtained from the discrete dynamic programming be... The structure extraction operator is defined as follows: （78） In the formula: Indicates the actual production strategy and ; Indicates actual transportation decisions and ; This represents the amount of raw grain processed and allocated in the solution obtained from discrete dynamic programming. ; This represents the zero- or one-variable choice of transportation mode in the solution obtained from discrete dynamic programming. ; Phase Two: Constructing a continuous optimization model COM based on the structure extraction operator: （79） The constraints of the continuous optimization model COM consist of constraints (68)-(76) and constraints (80)-(81): （80） （81） The enhanced hybrid dynamic programming-MILP pricing algorithm determines the combinatorial structure of the decision, i.e., feasible procurement-inventory-transportation schemes, through the discretization stage in step S522; and uses a two-stage optimization mechanism to determine the precise numerical values ​​of the decisions, enabling the model to operate in multinomial time. The solution is obtained internally, and then a collaborative optimization scheme for the procurement, processing and transportation of corn raw materials and finished grains under carbon constraints is obtained.

9. The method for coordinated optimization of corn raw material procurement, processing, and finished grain transportation under carbon constraints according to claim 8, characterized in that, The adaptive discretization strategy and state domination and pruning strategy defined in S522 are as follows: The adaptive discretization strategy includes: Price sensitivity principle: Dynamically adjust based on experience values ​​and the time-series characteristics of procurement prices. Discrete density; The principle of balancing inventory and stockouts: Set appropriate discrete decision points based on the current inventory balance and accumulated stockout: that is, for This increases the decision points for meeting the demand in the next period, meeting the current period's accumulated stock shortage, meeting the sum of the current period's accumulated stock shortage and the next period's demand, and meeting the sum of the current period's accumulated stock shortage and the demand in the next two periods; for Then you need to first select The decision quantity is determined by the current inventory balance and the accumulated shortage to determine the maximum value. ,in express The corresponding inventory loss rate vector, and under the constraint of this maximum value, based on empirical values, decision points are added to meet the demand of the next period, meet the cumulative stockout of the current period, and meet the sum of the cumulative stockout of the current period and the demand of the next period; Transportation capacity constraint principle: Adjusting decisions based on transportation volume constraints (75) Discrete grid; The state domination and pruning strategies include: Strong dominance requires that the dominant state is not inferior to the dominated state in all dimensions, i.e., the augmented state vector... Strong dominance state It is required if and only if the following conditions are met: inventory level Total Cumulative Costs Average inventory duration and accumulated stockout costs ;in, Indicates the variety per ton of corn processed by enterprises The penalty cost for delays in finished corn products; Representing state The corresponding inventory level vector; Representing state The corresponding cumulative total cost; Representing state The corresponding average inventory duration; Weak dominance: Dominance determined by a constructed weighted scoring function. （82） In the formula: Indicates the weighting coefficient; Indicates the variety per ton of corn processed by enterprises The penalty cost for delays in finished corn products; Strong dominance pruning: Identify and delete strongly dominated states as determined by the strong dominance relationship; Weak domination pruning: Based on strong domination pruning, when the cost difference between two states is less than a preset threshold and a weak domination relationship exists, only the dominant state is retained and the dominated state is deleted.

Images