Tobacco raw material warehouse-out multi-objective scheduling method based on dynamic programming

Abstract

The present application relates to intelligent logistics scheduling technology, aiming at providing a tobacco raw material warehouse-out multi-objective scheduling method based on dynamic programming. It includes: establishing raw material attribute model and raw material demand estimation model; according to the production characteristics of tobacco raw material aging and tobacco substitution, an optimization index suitable for tobacco raw material warehouse-out distribution is proposed, and a warehouse-out scheduling multi-objective optimization model is established; based on the raw material attribute model and the raw material demand estimation model, the dynamic programming algorithm with constraints is used to solve the warehouse-out scheduling multi-objective optimization model; according to the evaluation index, the path planning of warehouse-out scheduling is evaluated and optimized, and the corresponding tobacco variety and the corresponding tobacco variety are selected for the tobacco demand task. The present application can select the targeted warehouse-out scheduling optimization index according to the demand, improve the decision efficiency, effectively guarantee the supply quality and warehouse-out efficiency of tobacco raw material, and has important significance for improving the accurate supply and supply quality of tobacco raw material and optimizing the warehouse-out operation efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the warehouse logistics process of tobacco raw materials, and belongs to the intelligent scheduling technology in the intelligent logistics technology. BACKGROUND

[0002] The warehouse-out link of tobacco raw materials is a crucial link in the tobacco logistics process, which is closely related to the planning and demand of cigarette production and processing. The warehouse-out plan of tobacco raw materials needs to calculate the demand of corresponding tobacco raw materials through cigarette production planning and formula calculation, and select tobacco substitutes according to the actual inventory and quality of the warehouse, while optimizing the aging progress in time. The warehouse-out scheduling of tobacco raw materials also needs to consider the production task completion rate, tobacco selection cost, warehouse-out efficiency, and warehouse-out tobacco aging quality. According to different needs of different warehouse status, targeted warehouse-out scheduling strategies with goal-oriented are developed, which has great significance for the warehouse-out benefit of tobacco raw materials.

[0003] Currently, the material measurement, substitute tobacco selection, and scheduling decision of tobacco raw materials warehouse-out mainly rely on manual operation and historical experience, which has low intelligence and is difficult to achieve efficient optimization. Therefore, by applying intelligent optimization algorithm to develop multi-objective scheduling strategy of tobacco raw materials warehouse-out, the supply efficiency and aging quality of tobacco raw materials can be greatly improved. Reasonable use of algorithm can automatically optimize the selection strategy of warehouse-out tobacco by analyzing the aging quality and stack position of tobacco raw materials, reduce the factors of manual intervention, and improve work efficiency. In addition, the intelligent scheduling strategy of tobacco raw materials warehouse-out can also play an important role in inventory management and warehouse-out efficiency. Through reasonable warehouse-out strategy, the loss and waste of tobacco raw materials can be minimized, and the quality and quantity of tobacco can be fully ensured.

[0004] At present, for the warehouse task scheduling problem, the academic circle has had a large number of researches. The literature 《W company multi-variety small-batch production planning and scheduling optimization》 constructs specific optimization strategies and organizes implementation from main production planning, material requirement planning, job shop planning and scheduling and other aspects; mainly adopts rolling planning method, material requirement planning theory (MRP), improved genetic algorithm, material ABC classification method and smoothing index method and other theories to carry out strategy optimization on production planning and scheduling, and improve overall production efficiency. The literature 《Modeling and algorithm research of supply chain planning and scheduling in military manufacturing industry》 analyzes the process and characteristics of the supply chain of the military manufacturing industry, establishes the corresponding scheduling model combined with the supply chain characteristics of this type of enterprise, and analyzes and compares a plurality of optimization solving methods, and finally proposes the idea of using ant colony algorithm to solve the optimal solution. The literature 《Multi-objective optimization problem research of production and transportation collaborative scheduling for order-oriented manufacturing enterprises》 establishes a mathematical model of multi-objective optimization problem of production and transportation collaborative scheduling for order-oriented manufacturing enterprises, takes minimizing transportation cost and minimizing total order delivery time as optimization objectives, and uses a non-dominated sorting genetic algorithm with an elite strategy to solve. The literature 《Modeling of intelligent scheduling of whole vehicle logistics service supply chain delivery》 establishes a three-layer decision model of whole vehicle logistics service supply chain delivery logistics order distribution, the first layer is a transportation mode selection model taking minimizing delayed delivery time and logistics cost as objectives, the second layer is a vehicle loading model taking maximizing loading value as an objective, and the third layer is a logistics service provider selection model taking maximizing logistics service provider satisfaction as an objective, and based on the hierarchical optimization and comprehensive integration idea, a hierarchical solving idea is proposed, and a heuristic algorithm is proposed for large-scale solving of transportation mode selection, and simulation experiments show that, compared with GUROBI, NSGA-2 and MOPSO algorithms, the heuristic algorithm has good solving quality and solving efficiency.

[0005] However, the existing research work is usually still based on the traditional commercial logistics scene, and the characteristics and evaluation indexes of the incoming goods are relatively convergent. In view of different warehouse layouts and tobacco-specific evaluation indexes, the classic model is difficult to cover all scenes, and cannot be effectively applied to the relatively characteristic warehouse scheduling scene. At the same time, due to the particularity of the goods location code and the scheduling task, the commonly used algorithm also has the problem of poor local optimal search ability. SUMMARY

[0006] The technical problem to be solved by the present application is to overcome the shortcomings of the existing tobacco raw material warehouse delivery scheduling strategy, and to provide a tobacco raw material delivery multi-objective scheduling method based on dynamic programming.

[0007] To solve the technical problem, the solution of the present application is:

[0008] A multi-objective scheduling strategy for tobacco raw material outbound delivery based on dynamic programming is provided, which includes the following steps:

[0009] (1) Establish a raw material attribute model based on the raw material attributes of tobacco leaves in the warehouse;

[0010] (2) Based on the number of cigarette brands produced in a single month as planned in the production plan and the tobacco leaf formula of the corresponding brand, calculate the demand for tobacco leaf raw materials and establish a raw material demand calculation model;

[0011] (3) Based on the production characteristics of tobacco leaf raw material aging and tobacco leaf substitution, the optimization indicators suitable for tobacco leaf raw material outbound allocation are proposed, and a multi-objective optimization model for outbound scheduling is established;

[0012] (4) Based on the raw material attribute model and the raw material demand estimation model, a constrained dynamic programming algorithm is used to solve the multi-objective optimization model of warehouse dispatch scheduling; the path planning of warehouse dispatch scheduling is evaluated and optimized according to the evaluation indicators, and the corresponding tobacco leaf type and stack location are selected for the tobacco leaf demand task.

[0013] As a preferred embodiment of the present invention, in step (1), the raw material attributes of the tobacco leaves stored in the warehouse include: location, quantity, type, origin, year, grade, standard aging quality, aging rate and aging degree; wherein,

[0014] (1.1) Location refers to the geographical location of the warehouse where the tobacco raw materials are stored and the location of the stack within the warehouse;

[0015] (1.2) Quantity refers to the number of pieces of tobacco leaf raw materials in stock;

[0016] (1.3) Category refers to the classification based on the origin, year, and part of the tobacco leaf;

[0017] (1.4) Origin refers to the specific place where the tobacco leaves are produced;

[0018] (1.5) Year refers to the year of production of tobacco leaf raw materials;

[0019] (1.6) Grades refer to the three grades of tobacco leaves, namely upper, middle and lower, based on the position, color, aroma and texture of the raw materials;

[0020] (1.7) Standard aging quality. Different types of tobacco have different optimal aging times. The optimal aging degree is recorded as A. std ;

[0021] (1.8) Alcoholization rate. Different types of tobacco have different alcoholization speeds. The alcoholization rate is recorded as a r ;

[0022] (1.9) Degree of alcoholization, the stock tobacco raw material has different degree of alcoholization after being alcoholized for different time, the degree of alcoholization A is calculated by storage time d :

[0023] A d，t = A d，0 + a r ·t f ·t

[0024] Wherein, A d，t is the degree of alcoholization of the tobacco raw material after t time, A d，0 is the initial degree of alcoholization of the tobacco raw material, a r is the self-alcoholization rate of the tobacco raw material, a f is the alcoholization coefficient of the environment where the stack is located, and t is the storage time in the stack with the alcoholization coefficient a f .

[0025] As a preferred scheme of the present application, the location of the stock tobacco raw material is recorded according to the following method:

[0026] The stack where the tobacco raw material is located is marked as a three-dimensional coordinate (f, r, c), the row and column coordinates of the stack are mapped to the plane number q on the layer based on the warehouse plane space layout, and each stack in the warehouse is mapped to a three-dimensional space code e with uniqueness by the stack number Q of a single layer and the floor information f of the stack;

[0027] e = (f-1) × Q + q, q ∈ [1, Q]

[0028] Wherein, f is the floor where the stack is located, r is the row number of the stack on the floor, c is the column number of the stack on the floor, e is the three-dimensional number of the stack in the warehouse, Q is the total number of stacks in a single layer of warehouse, and q is the floor plane number of the stack.

[0029] As a preferred scheme of the present application, in the step (2), the raw material demand calculation model is established by the following way:

[0030] (2.1) According to the production plan, the production amount B i of each cigarette brand this month is planned, and the feeding amount F i of tobacco raw material is calculated according to the output ratio of the corresponding cigarette brand:

[0031]

[0032] Wherein, F i is the feeding amount of the brand, unit is batch; i is the index of cigarette brand; B i is the production amount of the brand, R o，i is the output ratio of the brand, and d cpbThe number of boxes corresponding to the batch of feeding amount;

[0033] (2.2) According to the production formula of each cigarette brand, the demand amount W of each cut tobacco type in the formula is calculated ij :

[0034]

[0035] Wherein, W ij is the demand amount of this kind of cut tobacco in the formula of i-th cigarette brand, unit: piece; j is the index of this kind of cut tobacco; F i is the feeding amount of i-th brand, R f，ij is the feeding proportion of this kind of cut tobacco in the formula of i-th cigarette brand, d ppb is the total number of raw materials in a batch of feeding amount;

[0036] (2.3) The demand amount of each cut tobacco in n cigarette brands is summed up to obtain the total demand amount W a，j of the original formula of each cut tobacco:

[0037]

[0038] Wherein, W a，j is the total planned demand amount of j-th tobacco leaf, W ij is the demand amount of j-th cut tobacco in the formula of i-th cigarette brand; n is the total number of cigarette brands;

[0039] (2.4) When the cut tobacco has problems such as insufficient inventory, pests, and mold, a substitute cut tobacco needs to be selected; after substitution, according to the type and quantity of the substitute cut tobacco and the substituted cut tobacco, the expected total quantity of each cut tobacco is calculated:

[0040] W r，jk = W a，k

[0041] W r，kj = -W a，k

[0042]

[0043] Wherein, W r，jk is the tobacco leaf quantity of j-th tobacco leaf material replacing k-th tobacco leaf material; W a，k is the total planned demand amount of k-th tobacco leaf material; O j is the expected total quantity of j-th tobacco leaf material, W a，j is the total planned demand amount of j-th tobacco leaf material; m is the total number of tobacco leaf materials.

[0044] As a preferred scheme of the present application, in the step (3), evaluation indexes suitable for the distribution of tobacco raw materials in the warehouse are proposed according to the production demand shortage rate, the aging quality of tobacco raw materials, the warehouse operation efficiency and the tobacco leaf rejection cost, and a warehouse scheduling multi-objective optimization model is established; specifically including:

[0045] (3.1) The evaluation index f1 of the production demand shortage rate is calculated according to the following formula:

[0046]

[0047] Wherein, is the amount of tobacco raw materials required by the production task, is the amount of tobacco raw materials expected to be discharged from the warehouse, is the amount of tobacco raw materials required by the production task;

[0048] (3.2) The evaluation index f2 of the overall aging quality of tobacco raw materials in the warehouse task is calculated according to the following formula:

[0049]

[0050] Wherein, m is the total number of types of tobacco raw materials, A d，j is the aging degree of the jth type of tobacco raw material that can be reached at the expected discharge time; A std，j is the standard value of the aging degree that the jth type of tobacco raw material should reach as much as possible when used;

[0051] (3.3) The evaluation index f3 of the warehouse operation efficiency of tobacco raw materials in the warehouse task is calculated according to the following formula:

[0052]

[0053]

[0054] Wherein, T e is the time of discharging tobacco raw materials from the stack position in the warehouse; x q is the in-out warehouse distance of the stack position corresponding to the plane number q in the x direction, y q is the in-out warehouse distance of the stack position corresponding to the plane number q in the y direction, f is the floor where the stack position is located; v x , v y are the in-out warehouse speed of horizontal transportation equipment, v z is the transportation speed of vertical elevator; l is the total number of warehouse stack positions required to be called in the warehouse task, T e，i is the time of discharging the i th stack position from the warehouse, T w，i is the time required for the i th stack position to be transferred from the warehouse to the cigarette factory or the material allocation center;

[0055] (3.4) Calculate the evaluation index f4 of the tobacco leaf cost existing in the abandonment according to the following formula:

[0056]

[0057]

[0058] Wherein, alpha j is the abandonment coefficient of tobacco leaf, N j is the total inventory of the jth tobacco raw material, W a，j is the total demand of the original formula of the jth tobacco raw material, Delta t j is the production year difference of the jth tobacco raw material and its substitute tobacco, O j is the predicted quantity of the jth tobacco raw material, c p，j is the purchase and processing cost of the unit quantity of the jth tobacco;

[0059] (3.5) The evaluation indexes f1 to f4 are normalized, different weights are set for the four evaluation indexes by using the analytic hierarchy process, a multi-objective optimization model is established, among the four evaluation indexes, the production demand shortage rate (production task completion degree) and the overall aging quality of tobacco leaf determine whether the tobacco raw material supply can be completed and the supply quality of tobacco raw material, and play a dominant role in the optimization target; on this basis, according to the requirements of warehouse operation efficiency and tobacco abandonment cost index, an optimization strategy with target tendency is developed.

[0060] As a preferred scheme of the present application, in the step (3.5), the warehouse discharge scheduling optimization model weight with high efficiency discharge as the target is set with the discharge efficiency as the important optimization target.

[0061] As a preferred scheme of the present application, in the step (3.5), the warehouse discharge scheduling optimization model weight with low cost production as the target is set with the tobacco abandonment cost as the important optimization target.

[0062] Compared with the prior art, the technical effect of the present application is:

[0063] 1、The present application establishes the inventory tobacco raw material attribute model and raw material demand calculation model for the warehouse inventory of tobacco raw material for the supply of production plan demand this specific scene, can select the warehouse discharge scheduling optimization index with target tendency according to the demand, adopts the dynamic programming algorithm with constraint to optimize, to improve the decision efficiency.

[0064] 2、The application is based on the research on the actual scene, and proposes a stock tobacco raw material attribute model, a raw material demand calculation model and a tobacco raw material warehouse-out multi-objective optimization model, which conforms to the characteristics and laws of tobacco raw material warehouse logistics. Considering the different warehouse demands and characteristic performances between tobacco raw material and ordinary goods, and the influence of the warehouse environment on the preservation and aging quality of tobacco raw material, the proposed model can effectively ensure the supply quality and warehouse-out efficiency of tobacco raw material.

[0065] 3、The application adopts a dynamic programming algorithm with constraints to evaluate and plan the path by the multi-objective optimization model, thereby improving the calculation efficiency. Meanwhile, the task model of tobacco raw material supply is abstracted as a path planning problem, so that the problem is solved by using an exact algorithm, which has higher precision than a heuristic algorithm and better solving efficiency on the multi-objective optimization problem of tobacco raw material supply.

[0066] 4、The warehouse-out multi-objective scheduling method of the application has important significance for improving the accurate supply and supply quality of tobacco raw material, optimizing the warehouse-out operation efficiency, and provides a new idea for further research and application. BRIEF DESCRIPTION OF DRAWINGS

[0067] Figure 1 It is a warehouse single-layer stack position plane distribution map.

[0068] Figure 2 It is a path selection schematic diagram. DETAILED DESCRIPTION

[0069] The implementation mode of the application will be described in detail below.

[0070] The warehouse-out multi-objective scheduling strategy based on dynamic programming of the application comprises the following steps:

[0071] 1. A raw material attribute model is established according to the raw material attributes of warehouse stock tobacco.

[0072] The raw material attributes of warehouse stock tobacco include: location, quantity, type, origin, year, large grade, standard aging quality, aging rate and aging degree; wherein,

[0073] (1) The location refers to the geographical position of the warehouse where the stock tobacco raw material is located and the in-warehouse position of the stack position; the location of the stock tobacco raw material is recorded according to the following method:

[0074] The stack position of the tobacco raw material is marked as a three-dimensional coordinate (f, r, c), the row and column coordinates of the stack position are mapped to the plane number q on the layer based on the warehouse plane space layout, and each stack position in the warehouse is mapped to a three-dimensional space code e with uniqueness by the stack position number Q of the single layer and the floor information f of the stack position;

[0075] e = (f - 1) x Q + q, q e [1, Q]

[0076] Wherein, f is the floor of the stack, r is the floor of the stack plane number, c is the floor of the stack plane column number, e is the warehouse three-dimensional number of the stack, Q is the total number of single-layer warehouse stack, q is the floor of the stack plane number.

[0077] (2) Quantity, refers to the number of pieces of inventory tobacco raw materials;

[0078] (3) Type, refers to the division according to the information of the origin, year, and tobacco leaf position;

[0079] (4) Origin, refers to the specific origin of tobacco raw materials;

[0080] (5) Year, refers to the production year of tobacco raw materials;

[0081] (6) Large grade, refers to the upper, middle and lower three grades divided according to the position, color, aroma and texture of tobacco raw materials;

[0082] (7) Standard alcohol quality, different types of tobacco have different optimal aging time, and the optimal aging degree is recorded as A std ;

[0083] (8) Aging rate, different types of tobacco have different aging rates, and the aging rate of itself is recorded as a r ;

[0084] (9) Aging degree, the inventory tobacco raw materials have different aging degrees after aging for different times, and the change of aging degree A d is calculated by storage time:

[0085] A d，t = A d，0 + a r · a f · t

[0086] Wherein, A d，t is the aging degree of tobacco raw materials after t time, A d，0 is the initial aging degree of tobacco raw materials, a r is the aging rate of tobacco raw materials, a f is the aging coefficient of the stack, and t is the storage time in the stack with aging coefficient a f .

[0087] 2. According to the monthly production of cigarette brand quantity and the corresponding tobacco formula planned by the production plan, the demand of tobacco raw materials is calculated, and a raw material demand calculation model is established.

[0088] (1) According to the production plan, plan the production volume of each cigarette brand this month. i , the unit is box, and the output ratio of each cigarette brand is corresponding to calculate the input amount of tobacco raw materials.

[0089]

[0090] Among them, i is the index of cigarette brand, F i is the feed amount of this brand, in batches, B i is the production volume of the brand, R o，i is the output ratio of the brand, d cpb The number of boxes corresponding to a batch of material.

[0091] (2) Based on the production formula of each cigarette brand, calculate the demand for each type of tobacco in its formula.

[0092]

[0093] Among them, j is the index of the tobacco leaf, W ij is the demand for this type of tobacco in the formula of cigarette brand i, in pieces, F i is the feed amount of brand i, R f，ij is the proportion of this type of tobacco in the formula of No. i cigarette brand, d ppb The total number of raw materials in a batch of material.

[0094] (3) Sum the demand for each type of cigarette in n cigarette brands to obtain the total demand for the original formula of each type of cigarette.

[0095]

[0096] Among them, W a，j is the original planned total demand for the jth type of tobacco raw material, W ij is the demand for the jth type of cigarette in the formulation of cigarette brand i; n is the total number of cigarette brands.

[0097] (4) Tobacco strips may need to be substituted due to insufficient inventory, pests and diseases, mildew and other problems. After the tobacco leaves are substituted, the total expected output quantity of each type of tobacco strip is calculated based on the type and quantity of the substitute tobacco strips and the substituted tobacco strips.

[0098] W r，jk =W a，k

[0099] W r，kj =-W a，k

[0100]

[0101] Wherein, W r，jk is the amount of tobacco leaves of the jth tobacco material instead of the kth tobacco material, W a，k is the total planned demand of the kth tobacco material; O j is the total estimated quantity of the jth tobacco material to be delivered; W a，j is the total planned demand of the jth tobacco material, and m is the total number of tobacco materials.

[0102] 3. According to the production characteristics of tobacco material aging and tobacco substitution, an optimization index suitable for the delivery allocation of tobacco materials is proposed, and a multi-objective optimization model for delivery scheduling is established.

[0103] (1) The evaluation index f1 of the production demand shortage rate is calculated:

[0104] The tobacco materials required for cigarette production and processing are the primary goal of the delivery scheduling task. The shortage value of the production demand is calculated by the difference between the required tobacco material quantity for the production task and the estimated quantity of the tobacco material to be delivered, and the ratio of the shortage value to the required tobacco material quantity for the production task is calculated to evaluate the completion degree of the delivery tobacco material quantity to the production task demand.

[0105]

[0106] Wherein, m is the total number of tobacco materials, O j is the estimated quantity of the jth tobacco material to be delivered, W a，j is the total planned demand of the jth tobacco material.

[0107] (2) The evaluation index f2 of the overall aging quality of tobacco materials in the delivery task is calculated:

[0108] The aging quality is an important monitoring index of tobacco materials in the storage process. In order to make the quality of the processed cigarettes reach the expected level, the tobacco materials used for production should as far as possible reach the standard aging value A std，j Therefore, the aging degree that the tobacco materials can reach at the planned delivery time is predicted, and the absolute difference between the aging degree and the standard aging value is calculated to evaluate the overall aging quality of the tobacco materials in the delivery task.

[0109] A d，t = A d，0 + a r · a f · t j

[0110]

[0111] Wherein, m is the total number of tobacco materials, A d，j is the aging degree that the jth tobacco material can reach at the estimated delivery time.

[0112] (3) Calculate the evaluation index f3 of the tobacco raw material unloading operation efficiency in the task:

[0113] With the unloading speed of the task order as the target, the tobacco in the stack position with high unloading efficiency is preferentially used. The time required for the tobacco raw material in the stack position to enter and exit the warehouse can be calculated from the horizontal distance x q , y q and floor f of the stack position from the vertical elevator or the entrance and exit, and the time required for the unloading of all tobacco in the stack position is weighted and averaged, while considering the shipping time cost of the warehouse location where the stack position is located.

[0114]

[0115]

[0116] wherein x q is the entry and exit distance of the stack position corresponding to the plane number q in the x direction, y q is the entry and exit distance of the stack position corresponding to the plane number q in the y direction, f is the floor where the stack position is located, v x , v y is the entry and exit speed of the forklift and other horizontal transportation equipment, v z is the transportation speed of the vertical elevator, l is the total number of warehouse stack positions required to be called for the unloading task, T e，i is the time for the i-th stack position to unload from the warehouse, T w，i is the time required for the i-th stack position to be transferred from the warehouse to the cigarette factory or the material allocation center.

[0117] (4) Calculate the evaluation index f4 of the cost of tobacco with abandonment according to the following formula:

[0118] Under the condition that the inventory is sufficient to guarantee the supply of production raw materials, generally according to the first-in-first-out principle of tobacco raw materials, the target is to improve the utilization rate of purchased tobacco, and the tobacco with longer years is preferentially used. The inventory of the original formula tobacco can meet the production demand, and the substitute tobacco is the newer tobacco. The abandonment of old tobacco is regarded as an old tobacco abandonment phenomenon, and the abandonment cost of tobacco is calculated according to the quantity, and the cost of tobacco with abandonment is weighted and averaged.

[0119]

[0120]

[0121] wherein N j is the total inventory of the j-th tobacco raw material, W a，j is the total demand of the j-th tobacco raw material, Δt j is the production year difference between the j-th tobacco raw material and its substitute tobacco, and αj is the abandonment coefficient of tobacco leaf, O j is the predicted quantity of the jth tobacco leaf material, c p，j is the purchasing and processing cost of the unit quantity of the tobacco leaf.

[0122] (3.5) Normalizing the evaluation indexes f1 to f4, and formulating an optimization strategy with a target tendency.

[0123] Each evaluation index is normalized, different weights are set for the four evaluation indexes by using the analytic hierarchy process, and a multi-objective optimization model is established. Among the evaluation indexes, the production task completion degree and the predicted aging quality of tobacco leaf determine whether the supply of tobacco leaf material can be completed and the quality of the supply of tobacco leaf material, which plays a dominant role in the optimization target. On this basis, an optimization strategy with a target tendency is formulated according to the requirements of the warehouse in terms of operation efficiency and tobacco leaf abandonment cost indexes.

[0124] 4. Dynamic programming algorithm solution:

[0125] For the dispatching of tobacco leaf materials, it can be abstracted as a path planning problem to search for the optimal solution. The specific operation includes: generating a tobacco leaf substitution path according to the tobacco leaf substitution rules, generating a stack position selection path according to the warehouse inventory information, evaluating the length of each path according to the optimization index, and abstracting the optimization model as a path planning problem with multiple starting points and multiple ending points. Based on the material attribute model and the material demand estimation model, a dynamic programming algorithm with constraints is used to solve the multi-objective optimization model of the dispatching; the path planning of the dispatching is evaluated and optimized according to the evaluation indexes, and finally the corresponding tobacco leaf type and the stack position are selected for the tobacco leaf demand task.

[0126] The following will provide a specific application example combined with the accompanying drawings:

[0127] (1) Warehouse stack position coding

[0128] Taking a tobacco company's raw material warehouse as an example, there are 8 layers of warehouses in the building, and there are 24 stack positions in a single layer of warehouse. The stack position plane distribution of a single layer is as Figure 1 shown. The row and column coordinates of the stack position are mapped to the plane number q on the layer, and according to the floor information f of the stack position, each stack position in the warehouse is mapped to a unique three-dimensional space coding e by the following formula.

[0129] e = (f-1) x 24 + q, q ∈ [1, 24], f ∈ [1, 8]

[0130] (2) Material demand estimation

[0131] Take a production task as an example, the current plan to export a cigarette brand processing raw materials, the production capacity is 10000 boxes, the output ratio is 2.4, and the output of a batch of raw materials is 100 boxes, calculate the required tobacco raw materials.

[0132]

[0133] According to its production formula, for example, 60% of No. 1 tobacco, 40% of No. 2 tobacco, and 20 pieces of tobacco in a batch of raw materials, calculate the required amount respectively.

[0134]

[0135]

[0136] Finally, the demand of each tobacco in all cigarette brands is summed up to get the total demand of each tobacco.

[0137] W a，1 = W 11 = 504

[0138] W a，2 = W 12 = 336

[0139] (3) Tobacco selection path generation

[0140] According to the tobacco substitute rule, generate the tobacco selection path, each demand tobacco has at least one tobacco selection path, that is, itself, through expert evaluation to determine other types of tobacco as the optional substitute tobacco of demand tobacco, so as to generate the tobacco selection path from demand tobacco to various optional substitute tobaccos, as shown in Figure 2 No. 1 demand tobacco can choose its own tobacco species for export, or choose No. 3 tobacco for substitution. Then calculate the evaluation index according to the year, aging degree, inventory quantity, etc. of the substitute tobacco, to evaluate the pros and cons of tobacco selection.

[0141] (4) Pile position selection path generation

[0142] According to the warehouse inventory information, generate the pile position selection path, each tobacco raw material is stored in multiple different pile positions, according to query its specific pile position code, generate the pile position selection path from the substitute tobacco to the pile position code, as shown in Figure 2 No. 2 tobacco has two storage points of No. 8 and No. 10 in the warehouse, which can be combined with the tobacco demand quantity for pile position selection. Then calculate the evaluation index according to the export efficiency of the pile position, the inventory quantity of tobacco on the pile position, etc. to evaluate the pros and cons of pile position selection.

[0143] (5) Evaluation function

[0144] After normalizing each evaluation index, the analytic hierarchy process is used to set different weights for the four evaluation indexes, and a multi-objective optimization model is established. In the evaluation index, the production task completion degree and the predicted tobacco aging quality determine whether the tobacco raw material supply can be completed and the supply quality of the tobacco raw material, which plays a dominant role in the optimization objective. In addition, according to the requirements of the warehouse in the operation efficiency and the tobacco discard cost index, an objective-oriented optimization strategy is formulated.

[0145] The warehouse-out efficiency is an important optimization objective, and the warehouse-out scheduling optimization model weight with high efficiency as the target is set.

[0146] min f(x,y,z) = 0.447f1 + 0.370f2 + 0.139f3 + 0.043f4

[0147] The tobacco discard cost is an important optimization objective, and the warehouse-out scheduling optimization model weight with low cost as the target is set.

[0148] min f(x,y,z) = 0.470f1 + 0.358f2 + 0.043f3 + 0.129f4

[0149] (6) Constraints

[0150] In the present application, there is a maximum inventory constraint according to the inventory condition, and the demand of the tobacco raw material cannot exceed the inventory quantity of the selected tobacco leaf type. Especially when a kind of tobacco raw material is used as a substitute for multiple demand tobaccos, in order to ensure that the quantity and stack position selection do not conflict, the network flow augmentation path idea is used. When the tobacco raw material on a stack position is selected for warehouse-out, the inventory quantity of the stack position itself and the corresponding tobacco raw material on the path is deducted, so as to meet the maximum inventory constraint of the tobacco raw material.

Claims

1. A multi-objective scheduling method for tobacco leaf raw material delivery based on dynamic programming, characterized in that: The steps include: (1) Establish a raw material attribute model based on the raw material attributes of tobacco leaves in the warehouse; The raw material attributes include: location, quantity, type, origin, year, grade, standard aging quality, aging rate and aging degree; (2) Calculate the demand for tobacco raw materials based on the number of cigarette brands produced in a single month and the tobacco leaf formula of the corresponding brand as planned in the production plan; establish a raw material demand calculation model; the specific steps are: (a) According to the production plan, plan the production volume of each cigarette brand this month, and calculate the input amount of tobacco raw materials corresponding to the output ratio of each cigarette brand; (b) According to the production formula of each cigarette brand, calculate the demand for each type of tobacco leaf in its formula; (c) Sum the demand for each type of tobacco leaf in n cigarette brands to obtain the total demand for the original formula of each tobacco leaf; (d) When tobacco leaf is insufficient in inventory, pests or mildew problems occur, substitute tobacco leaves are selected instead, and the total expected outbound quantity of each type of tobacco leaf is calculated based on the type and quantity of the substitute tobacco leaf and the substitute tobacco leaf; (3) Based on the production characteristics of tobacco leaf raw material aging and tobacco leaf substitution, evaluation indicators suitable for tobacco leaf raw material outbound allocation are proposed. After normalizing each evaluation indicator, the hierarchical analysis method is used to set different weights for each evaluation indicator and establish a multi-objective optimization model for outbound scheduling. The evaluation indicators include: production demand shortage rate, overall aging quality of tobacco leaf raw materials, tobacco leaf raw material outbound operation efficiency in outbound tasks, and the cost of abandoned tobacco leaves; (4) For the dispatch scheduling of tobacco leaf raw materials, it can be abstracted into a path planning problem to realize the search for the optimal solution. The specific operations include: generating tobacco leaf substitution paths according to tobacco leaf substitution rules, generating stack selection paths according to warehouse inventory information, evaluating the length of each path by combining evaluation indicators, and abstracting the optimization model into a path planning problem with multiple starting points and multiple end points. Based on the raw material attribute model and the raw material demand estimation model, a dynamic programming algorithm with maximum inventory constraints is used to solve the multi-objective optimization model of outbound scheduling. According to the evaluation indicators, the tobacco leaf substitute paths and stack selection path planning of outbound scheduling are evaluated and optimized, and the corresponding tobacco leaf types and stack locations are selected for the tobacco leaf demand tasks.

2. The method according to claim 1, characterized in that In the raw material properties of step (1), (1.1) Location refers to the geographical location of the warehouse where the tobacco leaf raw materials are stored and the location of the stack within the warehouse; (1.2) Quantity refers to the number of pieces of tobacco leaf raw materials in stock; (1.3) Category refers to the classification based on the origin, year, and part of the tobacco leaf; (1.4) Place of origin refers to the specific place of origin of the tobacco leaves; (1.5) Year refers to the year of production of the tobacco raw materials; (1.6) Grade refers to the three grades of tobacco leaves, namely upper, middle and lower, based on their location, color, aroma and texture; (1.7) Standard aging quality. Different types of tobacco have different optimal aging times. The optimal aging degree is recorded as ; (1.8) Aging rate. Different types of tobacco have different aging speeds. The aging rate of each type is recorded as ; (1.9) Degree of aging. The degree of aging of tobacco leaves in stock varies after aging for different periods of time. The degree of aging can be calculated based on the storage time. Changes: ； in, Tobacco leaf raw materials The degree of aging after time, is the initial alcoholization degree of tobacco raw materials, is the alcoholization rate of tobacco raw materials themselves, is the environmental aging coefficient of the stack location, For the alcoholization coefficient The storage time of the stack.

3. The method according to claim 2, characterized in that Record the location of tobacco leaf stock in the following manner: Mark the stack location of tobacco leaves as three-dimensional coordinates Based on the warehouse floor plan layout, the row and column coordinates of the stack are mapped to the plane number on the floor where the stack is located. , and by the number of stacks of single layer and stack floor information , mapping each stack in the warehouse into a unique three-dimensional spatial code ; ； in, The floor where the stack is located. The row number of the floor where the stack is located. The floor plan number where the stack is located. It is the three-dimensional number of the warehouse where the stack is located. is the total number of stacking locations in a single-layer warehouse, The floor plan number where the stack is located.

4. The method according to claim 1, wherein In step (2), a raw material demand estimation model is established in the following manner: (2.1) According to the production plan, plan the production volume of each cigarette brand this month , the unit is box; corresponding to the output ratio of each cigarette brand, calculate the input amount of tobacco raw materials : ； in, is the material feeding amount of the brand, in batches; is the index of cigarette brand; is the production volume of the brand, is the output ratio of the brand, The number of boxes corresponding to a batch of material; (2.2) Based on the production formula of each cigarette brand, calculate the demand for each type of tobacco in its formula : ； in, For some kind of cigarette The quantity demanded in the formula of the cigarette brand No. 1, in pieces; is the index of the tobacco leaf; for The feeding amount of brand No. For this type of tobacco The proportion of raw materials in the formula of No. 1 cigarette brand, The total number of raw materials in a batch of material; (2.3) Place each piece of tobacco in Sum up the demand for each cigarette brand to get the total demand for each original formula of cigarette : ； in, For the The original planned total demand for tobacco leaf raw materials, For the Planting tobacco in The demand for the formula of the cigarette brand is n; n is the total number of cigarette brands; (2.4) When tobacco strips are in short supply, infested with pests or mildew, substitute strips are needed. After substitution, the estimated total shipment quantity of each type of tobacco strip is calculated based on the type and quantity of the substitute and substituted strips: ； ； ； in, For the The tobacco raw material replaces the The quantity of tobacco leaves used as raw material for growing tobacco leaves; For the The original planned total demand for tobacco leaf raw materials; For the The estimated total outbound quantity of tobacco leaf raw materials, For the The original planned total demand for tobacco leaf raw materials; The total number of types of tobacco leaf raw materials.

5. The method according to claim 1, wherein In the step (3), evaluation indicators suitable for tobacco leaf raw material outbound allocation are proposed based on the production demand shortage rate, tobacco leaf raw material aging quality, outbound operation efficiency and tobacco leaf abandonment cost, and an outbound scheduling multi-objective optimization model is established; specifically, the following are included: (3.1) Calculate the evaluation index of production demand shortage rate according to the following formula: : ； in, Refers to the quantity of tobacco leaves required for production tasks. Refers to the quantity of tobacco leaf raw materials expected to be shipped out of the warehouse; (3.2) Calculate the evaluation index of the overall aging quality of tobacco leaf raw materials in the outbound task according to the following formula: : ； in, is the total number of types of tobacco leaf raw materials, For the The degree of aging that the tobacco leaf raw materials can achieve at the estimated time of shipment; For the The standard value of the degree of alcoholization that the tobacco leaf raw materials should reach when used; (3.3) Calculate the evaluation index of tobacco leaf raw material outbound operation efficiency in the outbound task according to the following formula: : ； ； in, The time when tobacco leaves are shipped out from the stack in the warehouse; The plane number corresponding to the stack position exist The distance in and out of the warehouse in the direction, The plane number corresponding to the stack position exist The distance in and out of the warehouse in the direction, The floor where the stack is located; 、 is the storage and retrieval speed of the horizontal transport equipment, is the transport speed of the vertical lift; The total number of warehouse locations required for outbound tasks, for The time when stack position No. is shipped out from the warehouse, for The time required to transfer the stack from the warehouse to the cigarette factory or material distribution center; (3.4) Calculate the evaluation index of the cost of abandoned tobacco leaves according to the following formula: : ； ； in, For the The total inventory of tobacco leaf raw materials, For the The total demand for the original formula of tobacco raw materials, For the The difference in production years between the tobacco leaf raw material and its substitute tobacco strips is For the The expected quantity of tobacco leaf raw materials shipped out, For the The purchasing and processing cost of each unit of tobacco strips; (3.5) Evaluation indicators to Normalization processing was carried out, and the hierarchical analysis method was used to set different weights for the four evaluation indicators to establish a multi-objective optimization model; among the four evaluation indicators, the production demand shortage rate and the overall aging quality of tobacco leaves determine whether the supply of tobacco raw materials can be completed and the supply quality of tobacco raw materials, and play a dominant role in the optimization objectives; on this basis, according to the requirements of warehouse operation efficiency and tobacco abandonment cost indicators, a target-oriented optimization strategy was formulated.

6. The method according to claim 5, characterized in that In the step (3.5), the outbound delivery efficiency is taken as an important optimization target, and the weight of the outbound delivery scheduling optimization model with high-efficiency outbound delivery as the target is set.

7. The method according to claim 5, characterized in that In the step (3.5), the cost of tobacco leaf abandonment is taken as an important optimization target, and the weight of the delivery scheduling optimization model with low-cost production as the goal is set.

Images