CN108280538A - Based on distributed logistics inventory's optimization method under cloud computing environment - Google Patents

Based on distributed logistics inventory's optimization method under cloud computing environment Download PDF

Info

Publication number
CN108280538A
CN108280538A CN201810011879.8A CN201810011879A CN108280538A CN 108280538 A CN108280538 A CN 108280538A CN 201810011879 A CN201810011879 A CN 201810011879A CN 108280538 A CN108280538 A CN 108280538A
Authority
CN
China
Prior art keywords
inventory
goods
warehouse
distributed
solution vector
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201810011879.8A
Other languages
Chinese (zh)
Inventor
李文敬
唐玮峰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangxi Teachers College
Original Assignee
Guangxi Teachers College
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guangxi Teachers College filed Critical Guangxi Teachers College
Priority to CN201810011879.8A priority Critical patent/CN108280538A/en
Publication of CN108280538A publication Critical patent/CN108280538A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/04Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/08Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/083Shipping

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Physics & Mathematics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Development Economics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Game Theory and Decision Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses one kind based on distributed logistics inventory's optimization method under cloud computing environment, includes the following steps:The algebraic model and its constraints function of S1, structure evaluation function f (U);S2, the value for calculating each component establish the optimization equation matrix H M of harmony inventory;S3, with probability P (Xi) replace harmony search memory probability HMCR generate new solution vector according to harmonic search algorithm;S4, the target function value f (U that new solution vector is calculated by evaluation function f (U)b), the target function value f (U with script solution vectorb) relatively and update inventory;S5, step S2 S4 are repeated, until stopping when predetermined maximum iteration, exports optimal solution.The present invention solves the problems, such as the early warning of distributed logistics inventory and cost optimization under cloud computing environment, realizes the dynamic management of distributed intelligence logistics inventory, has and respond, multiple-objection optimization computational efficiency high advantageous effect fast with speed of searching optimization.

Description

Distributed logistics inventory optimization method based on cloud computing environment
Technical Field
The invention relates to the technical field of cloud computing, intelligent logistics and distributed databases. More specifically, the invention relates to a distributed logistics inventory optimization method based on a cloud computing environment.
Background
The high inventory cost seriously restricts the benefit and development of the logistics enterprise, and the inventory management and optimization are a key and necessary problem for the logistics enterprise. With the wide application of high and new technologies such as internet, cloud computing, internet of things, GPS/beidou navigation and the like, the e-commerce logistics are rapidly expanded, particularly, cross-border e-commerce and rural e-commerce are rapidly developed, the storage distribution of e-commerce logistics enterprises is more and more, the requirement on inventory management is higher and higher, and optimizing distributed logistics inventory management to reduce logistics cost becomes a hot problem of research of people.
Currently, inventory management has been developed from traditional quantitative ordering and regular ordering into various modern logistics management models with modern features, such as MRP (material resource planning), MRP2 (material resource planning), VMI (supplier managed inventory), CMI (customer managed inventory), JMI (joint inventory management), and so on. In recent years, the study of the method is widely conducted by domestic and foreign scholars, and a plurality of distributed inventory management models are proposed. Rao and Krisman firstly use a newsstand model to process the allocation and order quantity, and then Zhang improves the newsstand model to research the inventory sharing problem under the condition of harmonious demand. Rudi and Robinson propose analysis strategies for two inventory models, and Robinson also expands the analysis strategies into an analysis method for a multi-inventory model, and obtains an approximate optimal solution by using large-scale linear programming under decentralized control. Cohen and Tagars consider the advance period of replenishment in the earliest process of modeling distributed inventory, and adopt a partial sharing strategy to deal with the uncertainty of the demand in the advance period of allocation. Zhao Chilobong et al put forward a GA-BP distributed inventory model on the basis of a traditional VMI model through research on a BP neural network and a legacy algorithm, utilize the advantage that the BP neural network has fewer fitting iteration steps and stable fitting effect for solving the nonlinear problem, add the legacy algorithm to improve the BP neural network, and overcome the problem of slow convergence rate of the traditional BP neural network. Yoran et al studied a centralized control distributed inventory management model, introduced a genetic simulated annealing algorithm, avoided the problem of the genetic algorithm in local search, made up the weakness of the simulated annealing algorithm that the global search capability is not strong, and better solved the distributed inventory ordering and allocation model under centralized control.
The model solution of distributed inventory management is substantially a nonlinear programming multi-objective optimization problem, and the solution methods of the problems comprise accurate algorithms such as a tangent plane method, a branch boundary method, a dynamic programming method and the like, but the problems are large in target quantity, complex in constraint conditions, low in accurate algorithm solution speed and unsatisfactory in effect. With the development of cloud computing, internet of things and electronic commerce, the design and the deployment of logistics warehouses are more and more decentralized and miniaturized. The distributed inventory management model and the optimization method have the advantages of low response and optimization speed and low multi-objective optimization calculation efficiency.
Disclosure of Invention
An object of the present invention is to solve at least the above problems and to provide at least the advantages described later.
The invention also aims to provide a distributed logistics inventory optimization method based on the cloud computing environment, which has the advantages of high multi-objective optimization computing accuracy and efficiency, high optimization speed, easy global optimal solution obtaining and capability of realizing quick early warning of logistics inventory.
To achieve these objects and other advantages in accordance with the purpose of the invention, there is provided a distributed logistics inventory optimization method in a cloud-based computing environment, comprising the steps of:
s1, constructing an algebraic model of an evaluation function f (U) and a constraint condition function thereof according to the distributed logistics inventory algebraic model;
s2, calculating the value of each component according to the safety stock of the goods, the algebraic expression of the order point, the algebraic model of the evaluation function f (U) and the constraint conditional function thereof, and establishing an optimal equation matrix HM of the harmony stock;
s3, probability P (X) of quantity required by cargo i in period Ti) Replacing the memory probability HMCR of the harmony search, and generating a new solution vector according to the harmony search algorithm;
s4, calculating objective function value f (U) of new solution vector according to evaluation function f (U)b) And the objective function value f (U) of the original solution vectorb) Comparing and updating the inventory;
and S5, repeating the steps S2-S4, stopping until the preset maximum iteration times, and outputting the optimal solution.
Preferably, the distributed logistics inventory algebraic model in step S1 is a distributed hierarchical logistics inventory algebraic model based on a cloud platform, and is:
the constraint function is:
safe stock quantity QQ of goods i in step S2iThe algebraic expression of (A) is:
order point Q for goods iiThe algebraic expression of (A) is:
wherein the objective function Y is the total inventory cost, Y2iIndicating the stock holding fee, Y, of the goods i during the time T3iIndicating the out-of-stock cost, Y, of the individual goods i5iRepresenting a transaction fee, Y, for a good i during a time T6iIndicating purchase fee, Y, of individual goods i7iIndicating the freight i per unit distance, Y8iIndicating a warehousing charge, Y, for the goods i9iIndicating a delivery fee, X, for goods iiRepresenting the demand, x, of goods i during time TkDenotes the distance, x, from the supplier to the warehouse kj_kDenotes the distance between warehouse j and warehouse k, HiIndicating lead time, U, of goods i1_iIndicating an initial stock of goods i, U2_iRepresents the order quantity, U, of the goods i within the time Tj_i_kRepresents the allocation amount of goods i between the warehouse j and the warehouse k within the time T,UkDenotes an initial stock total amount of the warehouse k, W denotes a maximum cargo capacity of the warehouse, N denotes a number of warehouses, θ (U) denotes a cargo amount judgment function, αiIndicating a safety factor, mu, for cargo iiDesired for demand of goods i, σiIs the demand standard deviation for cargo i.
Preferably, the evaluation function f (u) of the acoustic search algorithm in step S1 is:
wherein, P (X)i)=Xi/(U1_i+U2_i)。
Preferably, the method for establishing the optimal equation matrix HM for harmonic inventory in step S2 specifically includes:
a1, definition and size of acoustic inventory HMS, then
Wherein n represents the number of decision variables;
is the ith component of the mth solution vector, i is 1,2, … n, b is 1,2, … HMS;
f(Ub) Function value of the b-th solution vector;
a2, decomposing the algebraic model of the evaluation function f (U) into order costCost of inventory holdingExpense of lack of goodsAdjusted costOther costsFive formulas corresponding to 5 Map functions, and taking the algebraic expression of the safe stock quantity of the goods i and the algebraic expression of the ordering point of the goods i as 2 Map functions, and submitting 7 Map functions to 7 different CPUs for calculation to obtain each componentAnd an objective function value f (U) is obtainedb);
Wherein,
preferably, in step S3, performing harmony search algorithm to generate a new solution vector specifically includes:
a1, memory harmony: probability P (X) of demand for cargo i in period Ti) Equal to HMCR, one of the original solution vectors is obtainedA new componentThe distributed inventory memory and acoustic parameter conditions are:
If random(0,1)≤P(Xi);
a2, pitch trimming: for new components retained by memory and acoustic processesAnd carrying out pitch fine adjustment on the pitch fine adjustment probability PAR according to the pitch fine adjustment probability PAR to obtain a fine-adjusted new solution, and combining the fine-adjusted new solution to form a new solution vector, wherein the new solution vector specifically comprises the following steps:
PAR=PARmax·exp(1-α-1);
wherein α is a dynamic balance factor, α ═ iter/ImaxIter is the current iteration number, ImaxIs a predetermined maximum number of iterations, PARmaxIs the maximum perturbation probability;
bw is the bandwidth value of the cargo, wherein, the bandwidth value bw (i) of the cargo i is calculated by the following formula:
in the above formula, max (U)i) And min (U)i) Maximum and minimum of the demand for goods i, respectively, upper (U)i) And lower (U)i) Is the global maximum solution for cargo i.
Preferably, the updating of the inventory in step S4 specifically includes:calculating the objective function value f (U) corresponding to the new solution vectorb) If the new solution vector corresponds to the objective function value f (U)b) Objective function value f (U) corresponding to the worst solution in HMb) And if the sum is small, replacing the worst solution in the HM by the new solution vector to obtain a new HM.
The invention at least comprises the following beneficial effects:
the method solves the problems of early warning and cost optimization of the distributed logistics inventory in the cloud computing environment, realizes dynamic management of the distributed intelligent logistics inventory, has the advantages of high multi-objective optimization calculation accuracy and efficiency, high optimization speed, easy calculation of global optimal solution and capability of realizing quick early warning of the logistics inventory.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention.
Drawings
FIG. 1 is a distributed hierarchical material inventory model based on a cloud platform according to the present invention;
fig. 2 is a schematic flow chart of the distributed logistics inventory optimization method in the cloud-based computing environment according to the present invention.
Detailed Description
The present invention is described in further detail below with reference to the attached drawings so that those skilled in the art can implement the invention by referring to the description.
As shown in fig. 1-2, the present invention provides a distributed logistics inventory optimization method based on a cloud computing environment, which includes the following steps:
s1, constructing an algebraic model of an evaluation function f (U) and a constraint condition function thereof according to the distributed logistics inventory algebraic model;
s2, calculating the value of each component according to the safety stock of the goods, the algebraic expression of the order point, the algebraic model of the evaluation function f (U) and the constraint conditional function thereof, and establishing an optimal equation matrix HM of the harmony stock;
the establishing of the optimization equation matrix HM of the harmonic inventory may specifically be:
Begin:
step 1: initializing various parameters of the distributed logistics inventory, and inputting various initial parameters of the distributed logistics inventory, including the number M of branch control centers of the warehouse, the number N of the warehouse, the parameters of various goods and the like;
step 2: initializing distributed logistics inventory;
safe stock quantity QQ of goods i according to formulaiAlgebraic type ordering point QiThe algebraic expression of (1) is an algebraic expression of the total inventory cost corresponding to the harmony search algorithm evaluation function f (U) and a corresponding constraint condition function, each solution vector in the inventory is initialized, and the value of each component is calculated;
s3, probability P (X) of quantity required by cargo i in period Ti) Replacing the memory probability HMCR of the harmony search, and generating a new solution vector according to the harmony search algorithm;
s4, calculating objective function value f (U) of new solution vector according to evaluation function f (U)b) And the objective function value f (U) of the original solution vectorb) Comparing and updating the inventory;
and S5, repeating the steps S2-S4, stopping until the preset maximum iteration times, and outputting the optimal solution to obtain the optimal inventory cost.
In the above technical solution, assuming that the distributed logistics inventory consists of 1 supplier, one warehousing information center, M subcontrol centers and N warehouses, analyzing the characteristics of the existing distributed inventory model and the advantages of cloud computing, and combining the hierarchical control inventory model with the cloud computing system to construct a distributed hierarchical material inventory model (specifically shown in fig. 1) and an algebraic model based on a cloud platform, fig. 1 shows the structure of the entire distributed logistics inventory management model from the perspective of the cloud platform, the distributed management system can be divided into four module layers according to the transmission flow direction of the warehousing data on the cloud platform, wherein a data access layer is responsible for collecting inventory information of each warehouse and leading the inventory information into an HDFS of a data storage layer, and the HDFS can simultaneously receive task information and configuration parameters required by each subcontrol center (DataNode) transmitted from a warehousing information center (NameNode), the data node is used for reading, the data node carries out data communication with the NameNode through a heartbeat feedback mechanism, the NameNode carries out algorithm scheduling according to information fed back by the data node, and real-time inventory state information is displayed on terminal equipment of an interaction layer;
updating the stock specifically comprises comparing a new stock solution vector with the worst solution vector in the original stock, and if the adaptive value corresponding to the new stock solution vector is superior to the adaptive value corresponding to the worst solution vector, replacing the worst solution vector by the new stock solution vector and storing the worst solution vector in the stock, otherwise, still storing the worst solution vector in the original stock;
the predetermined maximum number of iterations is a number set by a person skilled in the art according to actual operation, and when the predetermined maximum number of iterations is ended, the optimal solution is output.
In another technical solution, the distributed logistics inventory algebraic model in step S1 is:
the constraint function is:
the constraint function comprises 5 constraint conditions, and the conditions represented by each row of algebraic expressions are respectively as follows:
(1) in the whole warehouse management system, the sum of the total ordered quantity of all the goods and the initial stock quantity of the goods in all the warehouses is smaller than the sum of the capacities of all the warehouses;
(2) the sum of the initial inventory total amount of the goods i and the ordering amount is greater than the total ordering point of the goods;
(3) the sum of the goods allocation amount (positive when the allocation amount is greater than the allocation amount, negative when the allocation amount is less than the allocation amount) of the warehouse k and the goods capacity of the warehouse k is less than the maximum capacity W of the warehouse;
(4) the ordering amount of goods i is more than or equal to 0;
(5) the cargo capacity of warehouse k is less than the maximum cargo capacity;
safe stock quantity QQ of goods i in step S2iThe algebraic expression of (A) is:
order point Q for goods iiThe algebraic expression of (A) is:
wherein the objective function Y is the total inventory cost, Y2iIndicating the stock holding fee, Y, of the goods i during the time T3iIndicating the out-of-stock cost, Y, of the individual goods i5iRepresenting a transaction fee, Y, for a good i during a time T6iIndicating purchase fee, Y, of individual goods i7iIndicating the freight i per unit distance, Y8iIndicating a warehousing charge, Y, for the goods i9iIndicating a delivery fee, X, for goods iiRepresenting the demand, x, of goods i during time TkDenotes the distance, x, from the supplier to the warehouse kj_kDenotes the distance between warehouse j and warehouse k, HiIndicating lead time, U, of goods i1_iIndicating an initial stock of goods i, U2_iRepresents the order quantity, U, of the goods i within the time Tj_i_kIndicating the quantity of transfer of goods i, U, between warehouse j and warehouse k within time TkRepresenting warehouse kInitial inventory amount, W represents maximum cargo capacity of the warehouse, N represents warehouse quantity, theta (U) represents cargo quantity judgment function, αiIndicating a safety factor, mu, for cargo iiDesired for demand of goods i, σiIs the demand standard deviation for cargo i. By adopting the technical scheme, the distributed logistics inventory of cloud computing is modeled, and the modeling method adoptsContinuous inventory replenishment strategy, when the inventory of goods i in a warehouse is reduced to the average order pointAnd then, immediately carrying out goods allocation between warehouses or ordering by the order quantity U, wherein the specific model assumes the following conditions:
the transaction fee is independent of the goods variety;
the demand for goods i is continuous, its demand Xi being obeyedNormal distribution (i ═ 1,2, 3 …, n);
the same kind of goods i in each warehouse are provided with the same amount of safety stock QQi
The goods accommodating quantity of all the warehouses is the same;
according to the above assumptions, the cost factors to be considered for the distributed logistics inventory include: (1) cost of ordering; (2) a cost of inventory holding; (3) a cost of goods allocation; (4) a stock out fee; (5) other costs;
wherein, the calculation of the order cost is as follows: the ordering cost mainly includes the charge of ordering commission, communication charge, travelling charge, goods inspection charge and goods checking charge, which are related to the ordered amount and transaction charge, but not the ordered variety. Thus, the algebraic equation for the order cost is:
calculation of inventory holding costs: the hold amount of stock is the sum of the initial stock and the order amount minus the demand amount of goods. In time T, the demand X of the goods iiObey a normal distribution. Thus, the algebraic expression of the distributed inventory holding cost is:
and (4) the stock shortage fee: when the initial stock U of goods i1_iAnd order quantity U2_iSum of less than the demand X for goods iiIn the meantime, the shortage of goods occurs. The algebraic formula of the stock shortage fee is as follows:
allocating cost: the allocation cost comprises transportation fees for the supplier to send the goods to each warehouse, transportation fees for the goods among the warehouses and transaction fees generated in the allocation process. The algebraic formula is as follows:
other costs mainly include the labor cost of goods warehouse-in and warehouse-out, the number of transfers between warehouses, transaction fees, manual shipping fees, platform operation fees, electricity fees of warehouses and the like. Other cost algebraic expressions are:
summing the above equations (1) - (5) to obtain the total inventory cost Y of the objective function of the algebraic model of the distributed logistics inventory in the period T*
Whether goods i need to be ordered to the supplier or not needs to be judged according to safety stock and an ordering point.
In another technical solution, the evaluation function f (u) of the acoustic search algorithm in step S1 is:
wherein, P (X)i)=Xi/(U1_i+U2_i). By adopting the technical scheme, both the distributed inventory algebraic model and the harmony search algorithm belong to the mathematical problem with constrained optimization, and the objective function of the distributed inventory algebraic model is equivalent to the evaluation function f (x) of the harmony search algorithm.
In another technical solution, the method for establishing the optimal equation matrix HM for harmonic inventory in step S2 specifically includes:
a1, definition and size of acoustic inventory HMS, then
Wherein n represents the number of decision variables;
is the ith component of the mth solution vector, i is 1,2, … n, b is 1,2, … HMS;
f(Ub) Function value of the b-th solution vector;
a2, decomposing the algebraic model of the evaluation function f (U) into order costCost of inventory holdingExpense of lack of goodsAdjusted costOther costsFive formulas corresponding to 5 Map functions, and taking the algebraic expression of the safe stock quantity of the goods i and the algebraic expression of the ordering point of the goods i as 2 Map functions, and submitting 7 Map functions to 7 different CPUs for calculation to obtain each componentAnd an objective function value f (U) is obtainedb);
Wherein,
by adopting the technical scheme, the solution vector is constructed into a Harmony memory bank in the Harmony search algorithm, and the memory bank value probability (HMCR) and the Pitch fine tuning probability (Pitch tuning rate) are introducedPAR) to retain and transform harmony in the harmony library, continuously updating the harmony library until a maximum number of iterations is reached.
The algorithm flow is as follows:
c 1: defining problems and parameters, setting an evaluation function f (x) according to the problems, and initializing parameters HMCR, PAR and bw in an algorithm;
c 2: initialization and Acoustic memory Bank size HMS
Wherein n represents the number of decision variables, HMS is the size of harmony memory bank, i.e. the number of storable harmony sounds, xjFor the jth solution vector, the j-th solution vector,is the i component of the j solution vector, f (x)j) Is the function value of the jth solution vector,
note that: as described aboveThe symbols x, i and j in the above description are different from the corresponding symbols x, i and j in the claims of the present invention, and they are applicable only to the above symbols x, i and jjxj
c 3: generate a new solution
The process by which the algorithm generates a new solution follows two mechanisms: memorize harmony and fine tuning of tone.
1) Memory and sound
Defining memory probability HMCR belonged to (0,1), and a new solutionFirst decision variable of new solutionProbability of HMCR being selected from the memory poolNamely, it is
If random(0,1)≤HMCR;
Since the generation of each new solution depends on the HMCR, in order to control the global optimization of the harmonic search algorithm, the HMCR should take a larger value to ensure that a larger probability of searching in the memory base, usually 0.85 or 0.9 is taken as the equivalent.
2) Tone fine tuning
For new solutions retained by memory and acoustic processesFine tuning the pitch according to the probability PARmax·exp(1-α-1),α=iter/ImaxWhere α is the dynamic trade-off factor, PARmaxFor maximum perturbation probability, iter is the current iteration number, ImaxIs the maximum number of iterations. The specific operation is as follows:
where P is a random number in the range of [0,1], PAR is a smaller number in the range of [0,1], typically 0.1 or 0.2, etc. bw is an arbitrary bandwidth value, calculated by the following formula:
wherein, max (x)k) And min (x)k) The maximum value and the minimum value of the k-th dimension and the sound component in the sub-library, that is, when the component difference is larger, the distribution of the solution is more dispersed, and therefore the corresponding fine tuning amplitude is also larger. However, at the later stage of the algorithm operation, the maximum value and the minimum value of the harmonic component of the sub-library may be the same, and when the difference value is equal to 0, the fine tuning will lose meaning. For this reason, when max (x)k)-min(xk) When 0, the k-th dimension and the acoustic global maximum component upper (x)k) With the minimum component lower (x)k) Respectively replace max (x)k) And min (x)k) Harmonic fine adjustment bandwidth bw (k) upper (x)k)-lower(xk);
According to the harmony search algorithm principle, the holding cost of the stored goods, the shortage cost of the goods, the total cost of the shortage, the transfer assembly cost, the transaction cost, the purchase cost, the distance transportation cost, the warehousing cost, the ex-warehouse cost, the demand, the distance from a supplier to a warehouse, the distance between the warehouse and a goods ordering point and the safe stock of the goods are specific known values, relevant values are input, various parameters of the cloud platform distributed logistics stock are initialized, and the U is used for calculating the distribution of the cloud platform distributed logistics stockj_i_k(quantity of transfer of goods i between warehouse j and warehouse k within time T), and U2_i(order quantity of goods i within time T) as a decision variable and calculating each componentTo construct an optimal equation matrix HM that establishes harmonic inventory.
In another technical solution, performing harmony search algorithm in step S3 to generate a new solution vector specifically includes:
a1, memory harmony: the probability of the harmony search is HMCR, and the probability P (X) of the demand of the cargo i in the period Ti) Equal to HMCR, a new component of the original solution vector is obtainedDistributed inventory memory harmonyThe parameter conditions are as follows:
If random(0,1)≤P(Xi);
a2, pitch trimming: for new components retained by memory and acoustic processesAnd carrying out pitch fine adjustment on the pitch fine adjustment probability PAR according to the pitch fine adjustment probability PAR to obtain a fine-adjusted new solution, and combining the fine-adjusted new solution to form a new solution vector, wherein the new solution vector specifically comprises the following steps:
PAR=PARmax·exp(1-α-1);
wherein α is a dynamic balance factor, α ═ iter/ImaxIter is the current iteration number, ImaxIs a predetermined maximum number of iterations, PARmaxIs the maximum perturbation probability;
bw is the bandwidth value of the cargo, wherein, the bandwidth value bw (i) of the cargo i is calculated by the following formula:
in the above formula, max (U)i) And min (U)i) The maximum and minimum values of the demand for goods i, respectively, i.e. the larger the difference, the more distributed the solution is, and therefore the larger the corresponding magnitude of the fine adjustment, but the fine adjustment will be meaningless when the maximum and minimum values may be the same, i.e. max (U)i)-min(Ui) When the value is 0, solving the global maximum of the ith cargo to upper (U)i) And minimum solution lower (U)i) Respectively replace max (U)i) And min (U)i) Harmonic fine tuning bandwidth bw (i) upper (U)i)-lower(Ui),upper(Ui) And lower (U)i) Is the global maximum solution for cargo i. By adopting the technical scheme, the distributed logistics inventory constraint condition function is equivalent to the constraint of the components by the formulas (6), (7) and (8) in the harmony search algorithm, the objective function of the algebraic model of the distributed logistics inventory and the relevant parameters in the constraint condition are fused with the functions in the formulas (6), (7) and (8), and the dynamic setting formula of the distributed logistics inventory and the harmony search parameter is provided.
In another technical solution, the updating the inventory in step S4 specifically includes: calculating the objective function value f (U) corresponding to the new solution vectorb) If the new solution vector corresponds to the objective function value f (U)b) Objective function value f (U) corresponding to the worst solution in HMb) If the difference is small, the worst solution in the HM is replaced by the new solution vector to obtain the new HM, otherwise, the worst solution vector is still stored in the original library. By adopting the technical scheme, the optimal solution is obtained.
Example 1:
a logistics company has N-9 parts warehouses in a certain area, four parts with different prices are selected as research objects, the maximum capacity W of each warehouse is set to 300, and the related cost of each part is shown in table 1:
table 1: each item of the cost of the accessory
The initial inventory of parts for each warehouse is shown in table 2:
TABLE 2 stock of initial parts from each warehouse
The required quantity of the parts in each warehouse and the required parameters of each part in one order period T (T ═ 10 days) are shown in table 3.
TABLE 3 Accessory demand related parameters
Demand (Unit: piece) Accessory 1 Accessory 2 Fitting 3 Accessory 4
Warehouse 1 8 6 19 15
Warehouse 2 18 21 9 13
Warehouse 3 11 7 13 16
Warehouse 4 9 17 22 17
Warehouse 5 13 10 11 12
Warehouse 6 17 8 19 18
Warehouse 7 7 9 13 14
Warehouse 8 12 12 8 19
Warehouse 9 15 10 14 9
Total demand 110 100 128 133
Safety factor αi 1.29 1.65 0.53 0.84
Total demand expectation 131 81 142 123
Total demand standard deviation σi 18.8 14.5 11.9 10.1
Order period H in advancei 24(2.4T) 20(2.0T) 35(3.5T) 26(2.6T)
Safe stock quantity QQ using goods iiAlgebraic expression of (a) and order point Q of goods iiCalculating the safe stock QQ of four accessoriesiAnd a total order point QiSpecifically, the results are shown in Table 4.
TABLE 4 safe inventory QQ of accessoriesiAnd a total order point Qi
Inventory (Unit: piece) Accessory 1 Accessory 2 Fitting 3 Accessory 4
Safe stock 38 34 12 14
Order point 352 196 509 334
The distances between the warehouses and from the supplier are shown in table 5.
TABLE 5 distances between warehouses and suppliers
The harmony search is carried out on four goods in the warehouses 1-3, real matrix coding is adopted, and harmony photon library HM is established as follows: ,
as shown in the harmonic library, a set of solutions for the model is the quantity of parts allocated between each warehouse and the number of parts ordered on the supply by the entire inventory management system.
The model solution of distributed inventory management is substantially a nonlinear programming multi-objective optimization problem, and the solution methods of the problems comprise precise algorithms such as a tangent plane method, a branch boundary method, a dynamic programming method and the like, but the problem has large target quantity, complex constraint conditions and unsatisfactory solution effect of the precise algorithms. Heuristic algorithms such as genetic algorithm, neural network, harmony search algorithm, simulated annealing algorithm and the like become effective methods for solving the problems. The harmony search algorithm has low calculation requirements on a complex multi-target optimization problem, has the advantages of simple structure, fewer parameters, good robustness and the like, and is applied to optimization of distributed logistics inventory;
let PARmax=0.9,ImaxThe method is characterized in that 1000 times of simulation experiment solving and comparison are respectively carried out through a distributed logistics inventory harmony search optimization algorithm (DLI-HS), a common Genetic Algorithm (GA), a simulated annealing algorithm (SA), a basic harmony search algorithm (HS) and an MR-DHS algorithm, wherein the MR-DHS algorithm is a dynamic parameter harmony search algorithm based on MapRedece in the research of the subject group, and specifically, a academic thesis is shown as follows: distributed inventory allocation model construction and algorithm research chapter 3 based on cloud computing; the iteration result shows that the DLI-HS algorithm obtains the best optimal solution, and the overall demand of the whole logistics warehousing system is considered in the invention, unlike the traditional logistics management which considers the scheduling strategy from the aspect of the goods demand of an individual warehouse. In this strategy, the entire inventory of the parts 1 cannot meet the total demand, and therefore the suppliers are collectively ordered, and corresponding goods are distributed according to the inventory of each warehouseThe multiple ordering cost generated by separate ordering is reduced; while the overall inventory of the fittings 2 can meet the demand. Therefore, the inter-warehouse transfer strategy is adopted, and goods are transferred from the warehouse with high inventory to the warehouse with low inventory, so that the ordering cost is reduced, and the held inventory is reduced; the total stock amounts of the parts 3 and 4 cannot satisfy the total demand, but a warehouse with a high stock amount exists in the holding warehouse. Therefore, the ordering and allocation combined strategy is adopted, the ordering cost is reduced, the goods holding cost of a warehouse with higher inventory is reduced, and the effect is better compared with that of an MR-DHS algorithm.
In the aspect of the performance of algorithm optimization, the convergence rate of the traditional harmony search algorithm is obviously higher than that of the genetic algorithm and the simulated annealing algorithm, but the traditional harmony search algorithm still falls into local optimization earlier, and the final result of the optimization is equivalent to that of the genetic algorithm. The DLI-HS algorithm makes up the deficiency, and the parallel optimization of multiple matrixes enables the algorithm to jump out of local optimization, so that a better optimal solution is obtained in the later period of iteration.
In sum, the distributed inventory model under the hierarchical control is improved in the cloud computing environment, the distributed logistics inventory model based on the cloud computing is provided, and the advantages of simple structure, few parameters and high optimizing speed of the harmony search algorithm are utilized to optimize the distributed logistics inventory of the multi-objective problem, so that a relatively ideal effect is achieved.
The number of apparatuses and the scale of the process described herein are intended to simplify the description of the present invention. Applications, modifications and variations of the recycled monochrome cartridge of the present invention will be apparent to those skilled in the art.
While embodiments of the invention have been disclosed above, it is not limited to the applications set forth in the description and the embodiments, which are fully applicable in a variety of fields of endeavor to which the invention pertains, and further modifications may readily be made by those skilled in the art, it being understood that the invention is not limited to the details shown and described herein without departing from the general concept defined by the appended claims and their equivalents.

Claims (6)

1. A distributed logistics inventory optimization method based on a cloud computing environment is characterized by comprising the following steps:
s1, constructing an algebraic model of an evaluation function f (U) and a constraint condition function thereof according to the distributed logistics inventory algebraic model;
s2, calculating the value of each component according to the safe stock of goods, the algebraic expression of the order point, the algebraic model of the evaluation function f (U) and the constraint condition function thereof, and establishing an optimal equation matrix HM of harmonic stock;
s3, demand with goods i in period TProbability of quantity P (X)i) Replacing the memory probability HMCR of the harmony search, and generating a new solution vector according to the harmony search algorithm;
s4, calculating objective function value f (U) of new solution vector according to evaluation function f (U)b) And the worst objective function value f (U) in harmonic inventoryb) Comparing and updating the inventory;
and S5, repeating the steps S2-S4, stopping until the preset maximum iteration times, and outputting the optimal solution.
2. The distributed logistics inventory optimization method based on the cloud computing environment as claimed in claim 1, wherein the distributed logistics inventory algebraic model in step S1 is a distributed hierarchical logistics inventory algebraic model based on a cloud platform, which is:
the constraint function is:
safe stock quantity QQ of goods i in step S2iThe algebraic expression of (A) is:
order point Q for goods iiThe algebraic expression of (A) is:
wherein the objective function Y is the total inventory cost, Y2iIndicating the stock holding fee, Y, of the goods i during the time T3iIndicating the out-of-stock cost, Y, of the individual goods i5iRepresenting a transaction fee, Y, for a good i during a time T6iIndicating purchase fee, Y, of individual goods i7iIndicating the freight i per unit distance, Y8iIndicating a warehousing charge, Y, for the goods i9iIndicating a delivery fee, X, for goods iiShows the time T of the cargoRequirement of object i, xkDenotes the distance, x, from the supplier to the warehouse kj_kDenotes the distance between warehouse j and warehouse k, HiIndicating lead time, U, of goods i1_iIndicating an initial stock of goods i, U2_iRepresents the order quantity, U, of the goods i within the time Tj_i_kIndicating the quantity of transfer of goods i, U, between warehouse j and warehouse k within time TkDenotes an initial stock total amount of the warehouse k, W denotes a maximum cargo capacity of the warehouse, N denotes a warehouse number, θ (U) denotes a cargo amount judgment function, αiIndicating a safety factor, mu, for cargo iiDesired for demand of goods i, σiIs the demand standard deviation for cargo i.
3. The distributed logistics inventory optimization method based on cloud computing environment as claimed in claim 2, wherein the evaluation function f (u) of the acoustic search algorithm in step S1 is:
wherein, P (X)i)=Xi/(U1_i+U2_i)。
4. The distributed logistics inventory optimization method based on the cloud computing environment as claimed in claim 3, wherein the method for establishing the optimization equation matrix HM of harmonic inventory in the step S2 is specifically as follows:
a1, definition and size of acoustic inventory HMS, then
Wherein n represents the number of decision variables;
is the ith component of the mth solution vector, i is 1,2, … n, b is 1,2, … HMS;
f(Ub) Function value of the b-th solution vector;
a2, decomposing the algebraic model of the evaluation function f (U) into order costCost of inventory holdingFee for lack of goodsAdjusted costOther costsFive formulas corresponding to 5 functions and matching with related constraint condition functions, and simultaneously taking the algebraic expression of the safe stock quantity of the goods i and the algebraic expression of the ordering point of the goods i as 2 functions, and submitting 7 functions to 7 different CPUs for calculation to obtain each componentAnd an objective function value f (U) is obtainedb);
Wherein,
5. the distributed logistics inventory optimization method based on the cloud computing environment as claimed in claim 4, wherein the harmony search algorithm is performed in step S3, and the generation of the new solution vector specifically comprises:
a1, memory harmony: probability P (X) of demand for cargo i in period Ti) Equal to HMCR, a new component of the original solution vector is obtainedThe distributed inventory memory and acoustic parameter conditions are:
If random(0,1)≤P(Xi);
a2, pitch trimming: for new components retained by memory and acoustic processesAnd carrying out pitch fine adjustment on the pitch fine adjustment probability PAR according to the pitch fine adjustment probability PAR to obtain a fine-adjusted new solution, and combining the fine-adjusted new solution to form a new solution vector, wherein the new solution vector specifically comprises the following steps:
PAR=PARmax·exp(1-α-1);
wherein α is a dynamic balance factor, α ═ iter/ImaxIter is the current iteration number, ImaxFor a predetermined maximum number of iterations, PARmaxIs the maximum perturbation probability;
bw is the bandwidth value of the cargo, wherein, the bandwidth value bw (i) of the cargo i is calculated by the following formula:
in the above formula, max (U)i) And min (U)i) Maximum and minimum of the demand for goods i, respectively, upper (U)i) And lower (U)i) Is the global maximum solution for cargo i.
6. The distributed logistics inventory optimization method based on cloud computing environment of claim 5,
the step S4 of updating the inventory specifically includes: calculating the objective function value f (U) corresponding to the new solution vectorb) If the new solution vector corresponds to the objective function value f (U)b) Objective function value f (U) corresponding to the worst solution in HMb) And if the sum is small, replacing the worst solution in the HM by the new solution vector to obtain a new HM.
CN201810011879.8A 2018-01-05 2018-01-05 Based on distributed logistics inventory's optimization method under cloud computing environment Pending CN108280538A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810011879.8A CN108280538A (en) 2018-01-05 2018-01-05 Based on distributed logistics inventory's optimization method under cloud computing environment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810011879.8A CN108280538A (en) 2018-01-05 2018-01-05 Based on distributed logistics inventory's optimization method under cloud computing environment

Publications (1)

Publication Number Publication Date
CN108280538A true CN108280538A (en) 2018-07-13

Family

ID=62803346

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810011879.8A Pending CN108280538A (en) 2018-01-05 2018-01-05 Based on distributed logistics inventory's optimization method under cloud computing environment

Country Status (1)

Country Link
CN (1) CN108280538A (en)

Cited By (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109255481A (en) * 2018-08-31 2019-01-22 重庆大学 A kind of distribution inventory dispatches system and improved method
CN109460954A (en) * 2018-10-25 2019-03-12 河南科技大学 A kind of supplier's direct-furnish line side material allocation method based on JIT-VMI
CN109670705A (en) * 2018-12-19 2019-04-23 连云港师范高等专科学校 A kind of construction method for optimization system Visualization Platform of storing in a warehouse
CN109741083A (en) * 2018-11-29 2019-05-10 杭州览众数据科技有限公司 A kind of material requirement weight predicting method based on enterprise MRP
CN110046865A (en) * 2019-05-20 2019-07-23 重庆大学 Distributed inventory's dispatching method
CN111222815A (en) * 2018-11-27 2020-06-02 顺丰科技有限公司 Method, device and equipment for determining single item replenishment and storage medium
CN111260119A (en) * 2020-01-10 2020-06-09 浙江工商大学 Product inventory control and distribution path planning method
CN111325490A (en) * 2018-12-14 2020-06-23 顺丰科技有限公司 Replenishment method and device
CN111401626A (en) * 2020-03-12 2020-07-10 东北石油大学 Social network numerical optimization method, system and medium based on six-degree separation theory
CN112102040A (en) * 2020-09-27 2020-12-18 中国建设银行股份有限公司 Global inventory control method and system in distributed environment
CN112633541A (en) * 2019-09-24 2021-04-09 北京沃东天骏信息技术有限公司 Inventory network optimization method and device based on single commodity flow
CN113420464A (en) * 2021-07-22 2021-09-21 西南交通大学 Aisle arrangement method considering robustness
CN113505908A (en) * 2021-05-01 2021-10-15 合肥食里挑一网络科技有限公司 Dynamic inventory optimization method
CN113537859A (en) * 2020-04-16 2021-10-22 顺丰科技有限公司 Bin net construction method and device, computer equipment and storage medium
CN113762859A (en) * 2020-11-25 2021-12-07 北京沃东天骏信息技术有限公司 Stock warehouse dividing method, device, system and storage medium
CN115345554A (en) * 2022-08-15 2022-11-15 上海东普信息科技有限公司 Storage model based transceiving control method, device, equipment and storage medium
CN117875189A (en) * 2024-03-06 2024-04-12 安徽建筑大学 Three-dimensional warehouse space layout method based on GA optimization GRO

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102594909A (en) * 2012-03-14 2012-07-18 西安电子科技大学 Multi-objective community detection method based on spectrum information of common neighbour matrix
CN102820662A (en) * 2012-08-17 2012-12-12 华北电力大学 Distributed power source contained power system multi-target reactive-power optimization method
CN103631758A (en) * 2013-11-21 2014-03-12 陕西理工学院 Method for solving non-linear programming and absolute value equation through improved harmony search algorithm
CN103955864A (en) * 2014-04-14 2014-07-30 武汉大学 Power system multi-target differentiation planning method based on improved harmony search algorithm
CN104007478A (en) * 2014-06-09 2014-08-27 桂林电子科技大学 Ground nuclear magnetic resonance retrieval method based on harmony search algorithm
CN104881871A (en) * 2015-05-22 2015-09-02 湖南大学 Traffic image segmentation method based on improved multi-object harmony search algorithm
US20170185933A1 (en) * 2015-06-14 2017-06-29 Jda Software Group, Inc. Distribution-Independent Inventory Approach under Multiple Service Level Targets

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102594909A (en) * 2012-03-14 2012-07-18 西安电子科技大学 Multi-objective community detection method based on spectrum information of common neighbour matrix
CN102820662A (en) * 2012-08-17 2012-12-12 华北电力大学 Distributed power source contained power system multi-target reactive-power optimization method
CN103631758A (en) * 2013-11-21 2014-03-12 陕西理工学院 Method for solving non-linear programming and absolute value equation through improved harmony search algorithm
CN103955864A (en) * 2014-04-14 2014-07-30 武汉大学 Power system multi-target differentiation planning method based on improved harmony search algorithm
CN104007478A (en) * 2014-06-09 2014-08-27 桂林电子科技大学 Ground nuclear magnetic resonance retrieval method based on harmony search algorithm
CN104881871A (en) * 2015-05-22 2015-09-02 湖南大学 Traffic image segmentation method based on improved multi-object harmony search algorithm
US20170185933A1 (en) * 2015-06-14 2017-06-29 Jda Software Group, Inc. Distribution-Independent Inventory Approach under Multiple Service Level Targets

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
WENJING LI: "Research on Distributed Logistics Inventory Model Based on Cloud Computing", 《2016 12TH INTERNATIONAL CONFERENCE ON COMPUTATIONAL INTELLIGENCE AND SECURITY (CIS)》 *
唐玮峰: "基于云计算的分布式库存调拨模型构建与算法研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 *
李永林: "轮盘赌选择自适应和声搜索算法", 《计算机应用研究》 *

Cited By (28)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109255481A (en) * 2018-08-31 2019-01-22 重庆大学 A kind of distribution inventory dispatches system and improved method
CN109255481B (en) * 2018-08-31 2021-09-24 重庆大学 Distributed inventory scheduling system and improvement method
CN109460954A (en) * 2018-10-25 2019-03-12 河南科技大学 A kind of supplier's direct-furnish line side material allocation method based on JIT-VMI
CN109460954B (en) * 2018-10-25 2021-07-30 河南科技大学 Supplier straight supply line side material distribution method based on JIT-VMI
CN111222815A (en) * 2018-11-27 2020-06-02 顺丰科技有限公司 Method, device and equipment for determining single item replenishment and storage medium
CN109741083A (en) * 2018-11-29 2019-05-10 杭州览众数据科技有限公司 A kind of material requirement weight predicting method based on enterprise MRP
CN109741083B (en) * 2018-11-29 2020-09-11 杭州览众数据科技有限公司 Material demand weighted prediction method based on enterprise MRP
CN111325490A (en) * 2018-12-14 2020-06-23 顺丰科技有限公司 Replenishment method and device
CN111325490B (en) * 2018-12-14 2024-04-16 顺丰科技有限公司 Goods supplementing method and device
CN109670705A (en) * 2018-12-19 2019-04-23 连云港师范高等专科学校 A kind of construction method for optimization system Visualization Platform of storing in a warehouse
CN110046865A (en) * 2019-05-20 2019-07-23 重庆大学 Distributed inventory's dispatching method
CN112633541A (en) * 2019-09-24 2021-04-09 北京沃东天骏信息技术有限公司 Inventory network optimization method and device based on single commodity flow
CN112633541B (en) * 2019-09-24 2024-06-18 北京沃东天骏信息技术有限公司 Inventory network optimization method and device based on single commodity flow
CN111260119A (en) * 2020-01-10 2020-06-09 浙江工商大学 Product inventory control and distribution path planning method
CN111260119B (en) * 2020-01-10 2024-01-19 浙江工商大学 Product inventory control and distribution path planning method
CN111401626A (en) * 2020-03-12 2020-07-10 东北石油大学 Social network numerical optimization method, system and medium based on six-degree separation theory
CN113537859B (en) * 2020-04-16 2024-02-06 顺丰科技有限公司 Bin net construction method, device, computer equipment and storage medium
CN113537859A (en) * 2020-04-16 2021-10-22 顺丰科技有限公司 Bin net construction method and device, computer equipment and storage medium
CN112102040A (en) * 2020-09-27 2020-12-18 中国建设银行股份有限公司 Global inventory control method and system in distributed environment
CN113762859A (en) * 2020-11-25 2021-12-07 北京沃东天骏信息技术有限公司 Stock warehouse dividing method, device, system and storage medium
CN113505908A (en) * 2021-05-01 2021-10-15 合肥食里挑一网络科技有限公司 Dynamic inventory optimization method
CN113505908B (en) * 2021-05-01 2024-01-12 合肥食里挑一网络科技有限公司 Dynamic inventory optimization method
CN113420464B (en) * 2021-07-22 2022-04-19 西南交通大学 Aisle arrangement method considering robustness
CN113420464A (en) * 2021-07-22 2021-09-21 西南交通大学 Aisle arrangement method considering robustness
CN115345554A (en) * 2022-08-15 2022-11-15 上海东普信息科技有限公司 Storage model based transceiving control method, device, equipment and storage medium
CN115345554B (en) * 2022-08-15 2024-05-14 上海韵达高新技术有限公司 Transmit-receive control method, device, equipment and storage medium based on storage model
CN117875189A (en) * 2024-03-06 2024-04-12 安徽建筑大学 Three-dimensional warehouse space layout method based on GA optimization GRO
CN117875189B (en) * 2024-03-06 2024-05-14 安徽建筑大学 Three-dimensional warehouse space layout method based on GA optimization GRO

Similar Documents

Publication Publication Date Title
CN108280538A (en) Based on distributed logistics inventory's optimization method under cloud computing environment
Nădăban et al. Fuzzy TOPSIS: a general view
Peng et al. Deep reinforcement learning approach for capacitated supply chain optimization under demand uncertainty
Beskorovainyi Combined method of ranking options in project decision support systems
Tenney et al. Structures for distributed decisionmaking
JP2002521756A (en) Computer-implemented value management tools for asset-intensive manufacturers
CN108092320A (en) The method and system for planning of the grid-connected allowed capacity of distributed photovoltaic
CN109086900B (en) Electric power material guarantee and allocation platform based on multi-target particle swarm optimization algorithm
CN111313957B (en) Hybrid satellite communication system resource allocation method based on classification multi-objective optimization
Benjaafar et al. Dynamic inventory repositioning in on-demand rental networks
CN112884236B (en) Short-term load prediction method and system based on VDM decomposition and LSTM improvement
CN112990958A (en) Data processing method, data processing device, storage medium and computer equipment
Panda et al. Long term electrical load forecasting: an empirical study across techniques and domains
Lai et al. Mechanism design for collaborative production-distribution planning with shipment consolidation
Möller et al. Airline network revenue management by multistage stochastic programming
FRIDMAN et al. Gradient coordination technique for controlling hierarchical and network systems
CN113780913B (en) Method and device for generating safety stock information
Xiong et al. Supplier selection model based on D numbers and transformation function
CN113935673A (en) Warehouse network optimization method and device based on inventory, computer equipment and storage medium
CN111027709B (en) Information recommendation method and device, server and storage medium
Zhang et al. A deep reinforcement learning algorithm for order acceptance decision of individualized product assembling
Schwind et al. Dynamic pricing of information products based on reinforcement learning: A yield-management approach
Maiti et al. Two storage inventory model in a mixed environment
CN114912647A (en) Apparatus, method and machine-readable storage medium for decision making
Lu et al. Corrigendum to “The Fourth-Party Logistics Routing Problem Using Ant Colony System-Improved Grey Wolf Optimization”

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20180713

RJ01 Rejection of invention patent application after publication