Eelist ant colony search method for multi-factory capacity and inventory integrated distribution problem

Abstract

The invention aims to provide an elite colony search method for a multi-factory capacity and inventory integrated allocation problem. The method comprises the following steps: obtaining order data and production resource data to be allocated; calculating an order allocation scheme according to the order data to be allocated and the production resource data by adopting an elite ant colony algorithm; and calculating the productivity and inventory distribution result of each order according to the order distribution scheme. In a large-scale multi-factory order allocation scene, the idea of the ant colony algorithm is adopted, an elite ant colony search idea of directional production line search is introduced, convergence of the algorithm to an optimal solution is accelerated, calculation logic is optimized, the result goodness is improved, and therefore the limitation of manual order allocation decision making is made up.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an elite ant colony search method for multi-factory capacity and inventory integrated allocation problem. Background Art

[0002] In recent years, as the home appliance industry market has shifted from incremental to stock, the production and manufacturing model of enterprises has gradually changed from production-driven sales to sales-driven production. In the customer-centric order-based production and delivery environment, large home appliance manufacturers often adopt a regional manufacturing and distribution model, setting up factories in different geographical locations for production. However, with the segmentation of sales channels caused by competition in the industry, the orders received by enterprises have gradually shown a trend of uneven distribution of customer addresses and fragmented quantities. The regional independent operation model of each factory is no longer suitable for the highly agile and changing order environment due to the lack of collaboration between factories, forcing enterprises to shift to a model of unified group order management. Therefore, research on multi-factory collaborative production in multi-address order scenarios has theoretical and practical significance. Summary of the invention

[0003] The purpose of the present invention is to provide an elite ant colony search method for the problem of integrated allocation of multi-factory capacity and inventory. This method adopts the idea of ​​ant colony algorithm for large-scale multi-factory order allocation scenarios, introduces the elite ant colony search idea of ​​directional production line search, accelerates the algorithm to converge to the optimal solution, optimizes the calculation logic, and improves the result quality, thereby making up for the limitations of manual order splitting decision-making.

[0004] An elite ant colony search method for multi-factory capacity and inventory integrated allocation problem, including:

[0005] Obtain order data and production resource data to be allocated;

[0006] Using an elite ant colony algorithm to calculate an order allocation plan based on the order data to be allocated and the production resource data;

[0007] The production capacity and inventory allocation results for each order are calculated according to the order allocation plan.

[0008] Preferably, the obtaining of order data and production resource data to be allocated comprises:

[0009] The order data to be allocated includes: order line number, order quantity, product model, address information, order delivery date, and unit delay cost;

[0010] The production resource data includes: the number of factories, the factory production line opening and closing calendar, the factory production line daily capacity information, the unit production cost of the factory corresponding to the product model, and the unit logistics cost of the factory corresponding to the order address.

[0011] Preferably, the method of using the elite ant colony algorithm to calculate the order allocation scheme according to the order data to be allocated and the production resource data comprises:

[0012] Set the parameters of the elite ant colony algorithm;

[0013] Generate ant colony code, initialize resource capability generation and pheromone;

[0014] Set up the normal ant colony roaming and the elite ant colony roaming respectively;

[0015] Decode the fitness value of each ant's coded information after the tour, and the ant's coded information with the highest fitness value is the optimal order allocation plan;

[0016] Update the pheromone matrix according to the fitness value;

[0017] Repeat the iteration until the preset requirements are met.

[0018] Preferably, the generating of ant colony code, initializing resource capability generation and pheromone comprises:

[0019] Get the number of order information lines, order quantity, order product model, factory production line opening and working calendar information, and obtain the relationship between the factory and production line process route and the production line-factory mapping relationship through the product model;

[0020] Assume that i and j represent the current decision node ant selecting order i to process on production line j, and the probability that ant k moves to the adjacent node is:

[0021]

[0022] τ ij represents the trajectory strength of ant k passing through path (i, j), η ij is a heuristic function, which is set to induce ants to choose the next state according to the heuristic rule. The inverse of the critical process is used as the heuristic function, where cr i is the number of production lines that can be selected for the model corresponding to order i, α is the pheromone weight of ant movement, β is the decision weight of ant heuristic rule, which is used to control the influence of pheromone on ant decision movement, and N(k) is the set of state nodes that ants can choose at the current decision moment, that is, N(k)∈[0,n-1]-tabu k , of which tabu k is the set of nodes that the ant has walked through.

[0023] Preferably, the respectively setting of ordinary ant roaming and elite ant colony roaming comprises:

[0024] During the state transition process of the elite ant colony, order selection is subject to the state transition probability, and orders are drawn in a roulette wheel probability manner;

[0025] When selecting production lines, a greedy search strategy is used to accelerate the ants' exploration of the optimal path. The state transfer function of the search process of the improved elite ant colony is derived into two parts: order selection and production line directional selection:

[0026]

[0027] It represents the state transition probability that the elite ant colony obeys when selecting orders. At this time, the action space of the ants changes from two dimensions to one dimension, and then the directional production line selection is carried out. Represents the set of optional production lines L in the model process corresponding to the current order i i The production line with the best total cost.

[0028] Preferably, the method for selecting the optimal production line specifically includes:

[0029] Get the production line information that the order can go to in the factory;

[0030] Perform production check based on the work calendar to obtain the required completion time for the order on the current production line;

[0031] After the production verification is passed, the production cost evaluation is carried out and the total cost is expressed as:

[0032] Cost Total =Cost l +Cost m +Cost t

[0033] Cost Total Represents the total cost of logistics, Cost l Cost m Cost t They represent the logistics cost, manufacturing cost and delay cost generated by the order respectively, and each cost is expressed as:

[0034] Cost l =TC*R i

[0035] Cost m =PC*R i -PC*z i

[0036] Cost t =DC*d i

[0037] Where TC represents the unit logistics cost from the factory to the order address, R irepresents the order quantity, PC represents the unit production cost of the order in the selected factory, z i represents the inventory occupancy of the order, DC represents the unit delay cost of the order, and d i Indicates the number of days it takes to complete the order;

[0038] Select the production line occupancy plan with the lowest total cost, record the occupancy information, and complete the current order occupancy.

[0039] Preferably, the fitness value of each ant encoding information after decoding the tour, the ant encoding information with the highest fitness value is the optimal order allocation solution, including:

[0040] Obtain information about the order with the highest priority currently waiting to be assigned and information about the production line to which the order with the highest priority currently waiting to be assigned is assigned from the code;

[0041] Determine whether the factory where the order allocation production line is located has inventory of the current order. If so, calculate the production capacity and inventory allocation results of each order according to the order allocation plan;

[0042] If it does not exist, the order will start to reduce the inventory quantity of the corresponding model in the factory. If the full amount can be reduced, the order allocation is marked as completed, the corresponding inventory quantity is deducted to get the inventory balance, and the logistics and delay costs generated by the order are recorded;

[0043] If the inventory is insufficient for full reduction, the corresponding inventory is reduced, the inventory capacity is updated, and after obtaining the remaining quantity of the order, the multi-level capacity verification rules of the order are referred to, and the capacity is locked in sequence from the front to the back according to the work calendar to obtain the completion time of the order;

[0044] According to the order completion time, obtain the logistics time from the selected factory to the order address, get the final delivery time of the order, and calculate the production cost, logistics cost and delay cost incurred in the order completion process;

[0045] Calculate all costs incurred by all orders in this order allocation task, obtain the total cost information of order allocation production, complete chromosome decoding, and obtain the fitness value of each ant colony encoding, that is, the total cost information.

[0046] Preferably, the updating of the pheromone matrix according to the fitness value comprises:

[0047] Set the upper and lower bounds of pheromone and set τ ij (t)∈[τ min ,τ max ], once the pheromone concentration exceeds the upper and lower limit thresholds, it is forced to be assigned the upper or lower limit;

[0048] Locally update pheromone concentrations for all ants:

[0049] τ ij (t+1)=(1-ρ)τ ij (t)+Δτ ij (k)#

[0050] Δτ ij (k) = γQ#

[0051] where τ ij (t+1) represents the next node to which the ant moves, ρ is the volatility of the pheromone, where 0≤ρ<1, τ ij (t) represents the current node where the ant is located, Δτ ij (k) represents the pheromone increment of the ant on the local optimal path, Q is the pheromone constant preset by the algorithm parameters, and γ is the local pheromone increment weight required by the algorithm;

[0052] Perform global pheromone update for all ants:

[0053] τ ij (t+1)=(1-ρ)τ ij (t)+Δτ' ij (k)#

[0054] Δτ' ij (k) = w k *Q#

[0055]

[0056] where w k It represents the contribution of the path quality solution found by the current ant tour to the solution quality, C n represents the solution obtained after the current ant travels around and decodes, C w is the solution with the worst decoding quality in the current ant colony, C * It is the optimal solution information sought in the current ant colony's roaming decoding.

[0057] Preferably, generating an ant colony code comprises:

[0058] Obtain the order number information of all orders, and calculate the selection probability of all ant colony decision spaces according to the probability function

[0059] According to the selection probability, the action nodes of the ant colony are randomly determined to obtain the order allocation order and production line selection, which are then compiled into the production line code of the corresponding order;

[0060] Obtain the number of order lines and the quantity information of all production lines in the factory involved in the inventory and capacity allocation decision, and generate a two-dimensional ant action state space matrix based on the number of order lines and the number of production lines;

[0061] The order-production line integrated pheromone matrix is ​​used to represent the ant colony pheromone concentration, that is, the ant colony pheromone is composed of an i×j matrix, where i represents the number of order rows and j represents all production lines of all factories.

[0062] An elite ant colony search system for multi-factory capacity and inventory integrated allocation problem, including:

[0063] A data acquisition module, used to acquire order data and production resource data to be allocated;

[0064] A data processing module, used for calculating an order allocation plan according to the order data to be allocated and the production resource data by using an elite ant colony algorithm;

[0065] The capacity and inventory allocation module is used to calculate the capacity and inventory allocation results of each order according to the order allocation plan.

[0066] The beneficial effects of the present invention are as follows: 1. The present invention effectively shortens the solution time in solving the problem of multi-factory order capacity and inventory allocation, so that the order can achieve the optimal allocation effect under the trade-off of manufacturing cost, delivery punctuality and logistics distance; 2. The present invention adopts the idea of ​​ant colony algorithm and introduces the elite ant colony search idea of ​​directional production line search, which accelerates the algorithm to converge to the optimal solution, optimizes the calculation logic, improves the result quality, and thus makes up for the limitations of manual order splitting decision-making. Provide guidance for the reasonable formulation of order allocation plan and the improvement of order allocation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0068] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0069] Figure 1 A flow chart of an elite ant colony search method for the problem of integrated allocation of multi-factory production capacity and inventory according to the present invention;

[0070] Figure 2 It is a schematic diagram of ant colony coding of the elite ant colony search method of the present invention. DETAILED DESCRIPTION

[0071] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0072] It should be noted that all directional indications in the embodiments of the present invention (such as up, down, left, right, front, back, etc.) are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0073] In addition, the descriptions of "first", "second", etc. in the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0074] In the customer-centric order-based production and delivery environment, large household appliance manufacturers often adopt a regional manufacturing and distribution model, setting up factories in different geographical locations for production. However, with the segmentation of sales channels caused by competition in the industry, the orders received by enterprises have gradually shown a trend of uneven distribution of customer addresses and fragmented quantity. The regional independent operation model of each factory is no longer suitable for the highly agile and changing order environment due to the lack of collaboration between factories, forcing enterprises to shift to a model of unified group order management. Therefore, the study of multi-factory collaborative production in multi-address order scenarios has theoretical and practical significance.

[0075] The present invention effectively shortens the solution time in solving the problem of multi-factory order capacity and inventory allocation, so that orders can achieve the optimal allocation effect under the trade-off of manufacturing cost, delivery punctuality, and logistics distance; the present invention adopts the idea of ​​ant colony algorithm and introduces the elite ant colony search idea of ​​directional production line search, which accelerates the algorithm to converge to the optimal solution, optimizes the calculation logic, and improves the result quality, thereby making up for the limitations of manual order allocation decision-making. Provide guidance for the reasonable formulation of order allocation plan and the improvement of order allocation efficiency.

[0076] Example 1

[0077] An elite ant colony search method for multi-factory capacity and inventory integrated allocation problem, reference Figure 1 ,include:

[0078] S100, obtaining order data and production resource data to be allocated;

[0079] S200, uses the elite ant colony algorithm to calculate the order allocation plan based on the order data to be allocated and the production resource data;

[0080] S300, calculating the production capacity and inventory allocation results of each order according to the order allocation plan.

[0081] The present invention adopts the search framework of ant colony algorithm and provides an elite ant colony based on greedy directional production line search to accelerate the convergence of the algorithm. The solution obtained achieves an effective balance in the three dimensions of order production, transportation and delivery, effectively reducing the total cost incurred in the order production and delivery process, and providing help for enterprises to reduce costs and increase efficiency at the production decision-making level.

[0082] Preferably, S100, obtaining order data and production resource data to be allocated includes:

[0083] The order data to be allocated includes: number of order lines, order quantity, product model, address information, order delivery date, and unit delay cost;

[0084] Production resource data includes: the number of factories, the factory production line opening and closing calendar, the factory production line daily capacity information, the unit production cost of the factory corresponding to the product model, and the unit logistics cost of the factory corresponding to the order address.

[0085] With order allocation sequence, order factory selection, order production line selection, and inventory occupancy selection as the main decision actions, the relationship between the factory and the production line process route and the production line-factory mapping relationship can be determined through the product model, and the problem can be abstractly simplified to: determine the order allocation sequence and the order selection for the production line. This can avoid infeasible solutions and improve search efficiency.

[0086] Preferably, S200, using the elite ant colony algorithm to calculate the order allocation plan according to the order data to be allocated and the production resource data includes:

[0087] S210, setting parameters of the elite ant colony algorithm;

[0088] Parameter setting initialization: Parameters are set for the number of ants, the number of elite ants, the initial pheromone concentration, pheromone volatilization, the previous pheromone, the next pheromone, the pheromone increase constant, the local pheromone increase weight, the movement probability selection weight, the heuristic probability selection weight, and the number of iterations, as shown in Table 1.

[0089] Table 1: Parameter information

[0090]

[0091] S220, generating ant colony code, initializing resource capability generation and pheromone;

[0092] Ant colony code generation, resource capacity generation and pheromone initialization: Generate a daily production resource capacity matrix based on the work calendar of each factory production line. And generate an initial state space matrix.

[0093] S230, setting the ordinary ant colony roaming and the elite ant colony roaming respectively;

[0094] S240, decoding the fitness value of each ant's coded information after the tour, and the ant's coded information with the highest fitness value is the optimal order allocation plan;

[0095] S250, updating the pheromone matrix according to the fitness value;

[0096] S260, repeat the iteration until the preset requirements are met.

[0097] Preferably, S220, generating ant colony code, initializing resource capability generation and pheromone includes:

[0098] Get the number of order information lines, order quantity, order product model, factory production line opening and working calendar information, and obtain the relationship between the factory and production line process route and the production line-factory mapping relationship through the product model;

[0099] Get the number of order information lines, order quantity, order product model, and factory production line opening and working calendar information. The product model can be used to find the relationship between the factory and the production line process route and the production line-factory mapping relationship. The problem can be abstracted and simplified to: determine the order of order allocation and the selection of production lines for orders. The order layer consists of a set of order sequence vectors (OV) and the line assignment vectors (LV) corresponding to the order. The OV layer can map the corresponding product model layer coding information according to the binding relationship between its order and its own product model. Similarly, the LV layer can also obtain the factory information corresponding to the selected production line according to the factory-production line mapping relationship. After obtaining the order allocation sequence and the assigned factory and corresponding production line, the orders with the front allocation order will use the factory inventory first, and then use the production capacity. Due to the restriction of the process route from the product model to the specified production line, there will be a large number of infeasible solutions in the scheduling space. In order to avoid infeasible solutions, it is necessary to refer to the product model process route of the order at the time of the order selection production line decision action, that is, all generated solutions must be feasible, so as to improve the solution efficiency.

[0100] Assuming that i and j represent the current decision node ant selecting order i to be processed on production line k, the probability that ant k moves to the adjacent node is:

[0101]

[0102] τ ij represents the trajectory strength of ant k passing through path (i, j), η ij is a heuristic function, which is set to induce ants to choose the next state according to the heuristic rule. The inverse of the critical process is used as the heuristic function, where cr i is the number of production lines that can be selected for the model corresponding to order i, α is the pheromone weight of ant movement, β is the decision weight of ant heuristic rule, which is used to control the influence of pheromone on ant decision movement, and N(k) is the set of state nodes that ants can choose at the current decision moment, that is, N(k)∈[0,n-1]-tabu k , of which tabu k is the set of nodes that the ant has walked through.

[0103] Preferably, S230, respectively setting the common ant colony roaming and the elite ant colony roaming includes:

[0104] During the state transition process of the elite ant colony, order selection is subject to the state transition probability, and orders are drawn in a roulette wheel probability manner;

[0105] When selecting production lines, a greedy search strategy is used to accelerate the ants' exploration of the optimal path. The state transfer function of the search process of the improved elite ant colony is derived into two parts: order selection and production line directional selection:

[0106]

[0107] It represents the state transition probability that the elite ant colony obeys when selecting orders. At this time, the action space of the ants changes from two dimensions to one dimension, and then the directional production line selection is carried out. Represents the set of optional production lines L in the model process corresponding to the current order i i The production line with the best total cost.

[0108] Elite ants use a directional node selection method to explore during the roaming process, and the exploration method during the roaming will be more aggressive. These ants will frequently select the optimal state node in the current action space, thereby improving the efficiency of convergence to the optimal solution target during the search process. During the state transition process of the elite ant group, order selection still obeys the state transition probability and extracts orders in a roulette probability manner, but when selecting the production line, a greedy search strategy is used to accelerate the ants' exploration of the optimal path.

[0109] Preferably, the method for selecting the optimal production line specifically includes:

[0110] Get the production line information that the order can go to in the factory;

[0111] Perform production check based on the work calendar to obtain the required completion time for the order on the current production line;

[0112] After the production verification is passed, the production cost evaluation is carried out and the total cost is expressed as:

[0113] Cost Total =Cost l +Cos m +Cost t

[0114] Among them, Cos Total Represents the total cost of logistics, Cost l Cost m Cost t They represent the logistics cost, manufacturing cost and delay cost generated by the order respectively, and each cost is expressed as:

[0115] Cost l =TC*R i

[0116] Cost m =PC*R i -PC*z i

[0117] Cost t =DC*d i

[0118] Where TC represents the unit logistics cost from the factory to the order address, R i represents the order quantity, PC represents the unit production cost of the order in the selected factory, z i represents the inventory occupancy of the order, DC represents the unit delay cost of the order, and d i Indicates the number of days it takes to complete the order;

[0119] Select the production line occupancy plan with the lowest total cost, record the occupancy information, and complete the current order occupancy.

[0120] Preferably, S240, decoding the fitness value of each ant's coded information after the tour, the ant's coded information with the highest fitness value is the optimal order allocation solution, including:

[0121] Obtain information about the order with the highest priority currently waiting to be assigned and information about the production line to which the order with the highest priority currently waiting to be assigned is assigned from the code;

[0122] Determine whether the factory where the order allocation production line is located has inventory for the current order. If so, calculate the production capacity and inventory allocation results for each order based on the order allocation plan;

[0123] If it does not exist, the order will start to reduce the inventory quantity of the corresponding model in the factory. If the full amount can be reduced, the order allocation is marked as completed, the corresponding inventory quantity is deducted to get the inventory balance, and the logistics and delay costs generated by the order are recorded;

[0124] If the inventory is insufficient for full reduction, the corresponding inventory is reduced, the inventory capacity is updated, and after obtaining the remaining quantity of the order, the multi-level capacity verification rules of the order are referred to, and the capacity is locked in sequence from the front to the back according to the work calendar to obtain the completion time of the order;

[0125] According to the order completion time, obtain the logistics time from the selected factory to the order address, get the final delivery time of the order, and calculate the production cost, logistics cost and delay cost incurred in the order completion process;

[0126] Calculate all costs incurred by all orders in this order allocation task, obtain the total cost information of order allocation production, complete chromosome decoding, and obtain the fitness value of each ant colony encoding, that is, the total cost information.

[0127] Preferably, S250, updating the pheromone matrix according to the fitness value includes:

[0128] Set the upper and lower bounds of pheromone and set τ ij (t)∈[τ min ,τ max ], once the pheromone concentration exceeds the upper and lower limit thresholds, it is forced to be assigned the upper or lower limit;

[0129] Locally update pheromone concentrations for all ants:

[0130] τ ij (t+1)=(1-ρ)τ ij (t)+Δτ ij (k)#

[0131] Δτ ij (k) = γQ#

[0132] where τ ij (t+1) represents the next node to which the ant moves, ρ is the volatility of the pheromone, where 0≤ρ<1, τ ij (t) represents the current node where the ant is located, Δτ ij (k) represents the pheromone increment of the ant on the local optimal path, Q is the pheromone constant preset by the algorithm parameters, and γ is the local pheromone increment weight required by the algorithm;

[0133] Perform global pheromone update for all ants:

[0134] τ ij (t+1)=(1-ρ)τ ij (t)+Δτ' ij (k)#

[0135] Δτ' ij (k) = w k *Q#

[0136]

[0137] where w k It represents the contribution of the path quality solution found by the current ant tour to the solution quality, C n represents the solution obtained after the current ant travels around and decodes, C w is the solution with the worst decoding quality in the current ant colony, C * It is the optimal solution information sought in the current ant colony's roaming decoding.

[0138] Preferably, generating the ant colony code comprises:

[0139] Obtain the order number information of all orders, and calculate the selection probability of all ant colony decision spaces according to the probability function

[0140] According to the selection probability, the action nodes of the ant colony are randomly determined to obtain the order allocation order and production line selection, which are then compiled into the production line code of the corresponding order;

[0141] Obtain the number of order lines and the quantity information of all production lines in the factory involved in the inventory and capacity allocation decision, and generate a two-dimensional ant action state space matrix based on the number of order lines and the number of production lines;

[0142] The order-production line integrated pheromone matrix is ​​used to represent the ant colony pheromone concentration, that is, the ant colony pheromone is composed of an i×j matrix, where i represents the number of order rows and j represents all production lines of all factories.

[0143] Since the factory and production line to which the order is assigned, as well as the order allocation order, have been determined in the coding, the work to be done in the chromosome decoding stage is to occupy inventory for each order according to the order allocation priority order and to occupy production capacity according to the work calendar from front to back rules, and output the order capacity and inventory allocation result information.

[0144] Example 2

[0145] An elite ant colony search system for multi-factory capacity and inventory integrated allocation problem, including:

[0146] A data acquisition module, used to acquire order data and production resource data to be allocated;

[0147] A data processing module, used to calculate an order allocation plan based on the order data to be allocated and the production resource data by using an elite ant colony algorithm;

[0148] The capacity and inventory allocation module is used to calculate the capacity and inventory allocation results for each order based on the order allocation plan.

[0149] The present invention effectively shortens the solution time in solving the problem of multi-factory order capacity and inventory allocation, so that orders can achieve the optimal allocation effect under the trade-off of manufacturing cost, delivery punctuality, and logistics distance; the present invention adopts the idea of ​​ant colony algorithm and introduces the elite ant colony search idea of ​​directional production line search, which accelerates the algorithm to converge to the optimal solution, optimizes the calculation logic, and improves the result quality, thereby making up for the limitations of manual order allocation decision-making. Provide guidance for the reasonable formulation of order allocation plan and the improvement of order allocation efficiency.

[0150] The foregoing is merely a specific embodiment of the present invention, which enables those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features claimed herein.

Claims

1. An elite ant colony search method for the integrated allocation problem of multi-factory capacity and inventory, characterized in that: include: Obtain order data and production resource data to be allocated; Using an elite ant colony algorithm to calculate an order allocation plan based on the order data to be allocated and the production resource data; The production capacity and inventory allocation results for each order are calculated according to the order allocation plan.

2. The elite ant colony search method for the multi-factory capacity and inventory integrated allocation problem according to claim 1 is characterized in that: The obtaining of order data and production resource data to be allocated includes: The order data to be allocated includes: order line number, order quantity, product model, address information, order delivery date, and unit delay cost; The production resource data includes: the number of factories, the factory production line opening and closing calendar, the factory production line daily capacity information, the unit production cost of the factory corresponding to the product model, and the unit logistics cost of the factory corresponding to the order address.

3. The elite ant colony search method for the multi-factory capacity and inventory integrated allocation problem according to claim 1 is characterized in that: The method of using the elite ant colony algorithm to calculate the order allocation scheme according to the order data to be allocated and the production resource data includes: Set the parameters of the elite ant colony algorithm; Generate ant colony code, initialize resource capability generation and pheromone; Set up the normal ant colony roaming and the elite ant colony roaming respectively; Decode the fitness value of each ant's coded information after the tour, and the ant's coded information with the highest fitness value is the optimal order allocation plan; Update the pheromone matrix according to the fitness value; Repeat the iteration until the preset requirements are met.

4. The elite ant colony search method for the multi-factory capacity and inventory integrated allocation problem according to claim 3 is characterized in that: The generating of ant colony code, initialization of resource capability generation and pheromone includes: Get the number of order information lines, order quantity, order product model, factory production line opening and working calendar information, and obtain the relationship between the factory and production line process route and the production line-factory mapping relationship through the product model; Assume that i and j represent the current decision node ant selecting order i to process on production line j, and the probability that ant k moves to the adjacent node is: τ ij represents the trajectory strength of ant k passing through path (i, j), η ij is a heuristic function, which is set to induce ants to choose the next state according to the heuristic rule. The inverse of the critical process is used as the heuristic function, where cr i is the number of production lines that can be selected for the model corresponding to order i, α is the pheromone weight of ant movement, β is the decision weight of ant heuristic rule, which is used to control the influence of pheromone on ant decision movement, and N(k) is the set of state nodes that ants can choose at the current decision moment, that is, N(k)∈[0,n-1]-tabu k , of which tabu k is the set of nodes that the ant has walked through.

5. The elite ant colony search method for the multi-factory capacity and inventory integrated allocation problem according to claim 3 is characterized in that: The setting of ordinary ant colony roaming and elite ant colony roaming respectively includes: During the state transition process of the elite ant colony, order selection is subject to the state transition probability, and orders are drawn in a roulette wheel probability manner; When selecting production lines, a greedy search strategy is used to accelerate the ants' exploration of the optimal path. The state transfer function of the search process of the improved elite ant colony is derived into two parts: order selection and production line directional selection: It represents the state transition probability that the elite ant colony obeys when selecting orders. At this time, the action space of the ants changes from two dimensions to one dimension, and then the directional production line selection is carried out. Represents the set of optional production lines L in the model process corresponding to the current order i i The production line with the best total cost.

6. The elite ant colony search method for the multi-factory capacity and inventory integrated allocation problem according to claim 5 is characterized in that: The method for selecting the optimal production line specifically includes: Get the production line information that the order can go to in the factory; Perform production check based on the work calendar to obtain the required completion time for the order on the current production line; After the production verification is passed, the production cost evaluation is carried out and the total cost is expressed as: Cos Total =Cost l +Cost m +Cos t Cost Total Represents the total cost of logistics, Cost l 、Cos m 、Cos t They represent the logistics cost, manufacturing cost and delay cost generated by the order respectively, and each cost is expressed as: Cost l =TC*R i Cos m =PC*R i -PC*z i Cost t =DC*d i Where TC represents the unit logistics cost from the factory to the order address, R i represents the order quantity, PC represents the unit production cost of the order in the selected factory, z i represents the inventory occupancy of the order, DC represents the unit delay cost of the order, and d i Indicates the number of days it takes to complete the order; Select the production line occupancy plan with the lowest total cost, record the occupancy information, and complete the current order occupancy.

7. The elite ant colony search method for the multi-factory capacity and inventory integrated allocation problem according to claim 3 is characterized in that: The fitness value of each ant's coded information after decoding the tour, the ant's coded information with the highest fitness value is the optimal order allocation solution, including: Obtain information about the order with the highest priority currently waiting to be assigned and information about the production line to which the order with the highest priority currently waiting to be assigned is assigned from the code; Determine whether the factory where the order allocation production line is located has inventory of the current order. If so, calculate the production capacity and inventory allocation results of each order according to the order allocation plan; If it does not exist, the order will start to reduce the inventory quantity of the corresponding model in the factory. If the full amount can be reduced, the order allocation is marked as completed, the corresponding inventory quantity is deducted to get the inventory balance, and the logistics and delay costs generated by the order are recorded; If the inventory is insufficient for full reduction, the corresponding inventory is reduced, the inventory capacity is updated, and after obtaining the remaining quantity of the order, the multi-level capacity verification rules of the order are referred to, and the capacity is locked in sequence from the front to the back according to the work calendar to obtain the completion time of the order; According to the order completion time, obtain the logistics time from the selected factory to the order address, get the final delivery time of the order, and calculate the production cost, logistics cost and delay cost incurred in the order completion process; Calculate all costs incurred by all orders in this order allocation task, obtain the total cost information of order allocation production, complete chromosome decoding, and obtain the fitness value of each ant colony encoding, that is, the total cost information.

8. The elite ant colony search method for the multi-factory capacity and inventory integrated allocation problem according to claim 3 is characterized in that: The updating of the pheromone matrix according to the fitness value comprises: Set the upper and lower bounds of pheromone and set τ ij (t)∈[τ min ,τ max ], once the pheromone concentration exceeds the upper and lower limit thresholds, it is forced to be assigned the upper or lower limit; Locally update pheromone concentrations for all ants: t ij (t+1)=(1-ρ)τ ij (t)+Δτ ij (k)# Dt ij (k)=γQ# where τ ij (t+1) represents the next node to which the ant moves, ρ is the volatility of the pheromone, where 0≤ρ<1, τ ij (t) represents the current node where the ant is located, Δτ ij (k) represents the pheromone increment of the ant on the local optimal path, Q is the pheromone constant preset by the algorithm parameters, and γ is the local pheromone increment weight required by the algorithm; Perform global pheromone update for all ants: t ij (t+1)=(1-ρ)τ ij (t)+Δt' ij (k)# Dt' ij (k)=w k *Q# where w k It represents the contribution of the path quality solution found by the current ant tour to the solution quality, C n represents the solution obtained after the current ant travels around and decodes, C w is the solution with the worst decoding quality in the current ant colony, C * It is the optimal solution information sought in the current ant colony's roaming decoding.

9. The elite ant colony search method for the multi-factory capacity and inventory integrated allocation problem according to claim 3 is characterized in that: The generating of ant colony code comprises: Obtain the order number information of all orders, and calculate the selection probability of all ant colony decision spaces according to the probability function According to the selection probability, the action nodes of the ant colony are randomly determined to obtain the order allocation order and production line selection, which are then compiled into the production line code of the corresponding order; Obtain the number of order lines and the quantity information of all production lines in the factory involved in the inventory and capacity allocation decision, and generate a two-dimensional ant action state space matrix based on the number of order lines and the number of production lines; The order-production line integrated pheromone matrix is ​​used to represent the ant colony pheromone concentration, that is, the ant colony pheromone is composed of an i×j matrix, where i represents the number of order rows and j represents all production lines of all factories.

10. An elite ant colony search system for the problem of integrated allocation of multi-factory capacity and inventory, characterized in that: include: A data acquisition module, used to acquire order data and production resource data to be allocated; A data processing module, used for calculating an order allocation plan according to the order data to be allocated and the production resource data by using an elite ant colony algorithm; The capacity and inventory allocation module is used to calculate the capacity and inventory allocation results of each order according to the order allocation plan.

Images