Discrete material set management method, system, computer equipment and storage medium

Abstract

The present invention belongs to the technical field of material management, and relates to a discrete material complete set management method, system, computer equipment and storage medium. The method comprises the following steps: constructing a material demand forecast data set and a total inventory data set, preprocessing the material demand forecast data set to obtain a preprocessed material demand forecast data set; extracting features from the preprocessed material demand forecast data set to obtain feature parameters; inputting the feature parameters into the constructed material demand forecast model to obtain the demand for various materials in the future production process, forming a material demand forecast result data set; comparing the total inventory data set with the material demand forecast result data set, if the total inventory data set ≥ the material demand forecast result data set, then the materials are complete; if the total inventory data set < the material demand forecast result data set, then the materials are not complete. The problem of large demand forecast error and unreasonable inventory management existing in the existing complete set management method is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of material management, and in particular relates to a discrete material set management method, system, computer equipment and storage medium. Background Art

[0002] With the rapid development and intensified market competition, material set management has become one of the important factors affecting production efficiency and enterprise economic benefits. In the existing technology, material demand forecasting, inventory management and supply chain optimization are the core links of material set management. However, there are many problems in the actual application of these links, which makes it difficult to achieve the expected accuracy and efficiency of material set management.

[0003] The accuracy of material demand forecasting is crucial to material set management. Existing technologies usually use traditional statistical methods for demand forecasting, such as time series analysis and regression analysis. Although these methods can reflect the changing trends of historical data to a certain extent, they are difficult to accurately capture the complex and changing market demand and production environment. For example, traditional methods often ignore the combined impact of multiple factors such as production plans, historical material usage, order data, seasonal factors, and changes in market demand, resulting in large errors in demand forecasting and failure to meet the dynamic needs of actual production.

[0004] Inventory management also has many shortcomings in the existing technology. Traditional inventory management methods mainly rely on experience and simple mathematical models, which lack scientificity and systematicity. Although the economic order quantity model commonly used in the existing technology can optimize the order quantity, it fails to fully consider the situation of materials in transit and ordered but not delivered, making it difficult to reasonably control the inventory level.

[0005] In summary, the existing complete set management methods have problems such as large demand forecasting errors and unreasonable inventory management, which ultimately leads to chaos in complete set management. Summary of the invention

[0006] The purpose of the present invention is to provide a discrete material set management method, system, computer equipment and storage medium, which solves the problems of large demand forecasting errors and unreasonable inventory management in existing set management methods.

[0007] The present invention is achieved through the following technical solutions:

[0008] A discrete material set management method includes the following steps:

[0009] S1. Construct a material demand forecasting data set, which includes production plan data, historical material usage data, order data, seasonal factors and market demand change data;

[0010] Constructing a total inventory data set, wherein the total inventory data set includes the current inventory, the in-transit inventory, and the ordered but undelivered material data;

[0011] S2. preprocessing the material demand forecast data set to obtain a preprocessed material demand forecast data set;

[0012] S3, extracting features from the preprocessed material demand forecast data set to obtain feature parameters;

[0013] S4. Input the characteristic parameters into the constructed material demand forecasting model to obtain the demand for various materials in the future production process, forming a material demand forecasting result data set;

[0014] S5. Compare the total inventory data set in S1 with the material demand forecast result data set in S4 to obtain the discrete material complete set management solution:

[0015] If the total inventory data set is ≥ the material demand forecast result data set, the materials are complete; if the total inventory data set is < the material demand forecast result data set, the materials are not complete.

[0016] Further, in S1, the building of the material demand forecasting data set specifically includes the following process:

[0017] Obtain production plan data, which includes the type, quantity and production time of the planned products, recorded as a data set ;

[0018] ;

[0019] in, Indicates i The production load quantity of each product; Indicates i The planned production quantity of the product; represents the production time of the i-th product; T represents the production cycle constant; f Indicates the number of product types. i =1,2,…, f , exp () represents an exponential function with base e;

[0020] Obtain historical material usage data, which includes the usage records of various materials over a period of time in the past and is recorded as a data set ;

[0021] ;

[0022] in, Indicates jThe quantity of material required, Indicates j The amount of material used, Indicates j The use time of the material, represents the historical time period constant, n Indicates the number of material types. j =1,2,…, n , cos () is the cosine function;

[0023] Obtain order data, which includes the confirmed order product type, quantity and delivery time, recorded as a data set ;

[0024] ;

[0025] in, Indicates k The contract value of the order, Indicates k The number of products in an order, Indicates k Delivery time for orders, represents the order cycle constant, p Indicates the total order quantity; k =1,2,3... p ; sin () is the sine function; π is the circumference of a circle;

[0026] Obtain seasonal factor data, which includes the impact of seasonal changes on material demand, and is recorded as a data set ;

[0027] ;

[0028] in, Indicates l The correction coefficient of seasonal factors on material production efficiency, Indicates the time of seasonal changes. represents the seasonal cycle constant, q Indicates the number of seasonal factors; l =1,2,3... q ; cos() is the cosine function; π is the circumference of a circle; exp () represents an exponential function with base e;

[0029] Obtain market demand change data, which includes fluctuations and trends in market demand and is recorded as a data set ;

[0030] ;

[0031] in, It represents the adjustment coefficient of material demand under the rth market demand condition; Indicates the time when market demand changes; represents the market demand cycle constant; s Indicates the number of types of market demand changes; r =1,2,3... s ; sin() is the sine function; π is the pi; exp () represents an exponential function with base e;

[0032] The production plan data, historical material usage data, order data, seasonal factor data and market demand change data are aggregated to construct a material demand forecast data set, denoted as ;

[0033] .

[0034] Furthermore, in S3, feature extraction is performed on the preprocessed material demand forecast data set to obtain the feature matrix X ,in ;

[0035] Each feature represents a specific attribute in the material demand forecast dataset.

[0036] Further, in S4, the characteristic parameters are input into the constructed material demand forecasting model to obtain the demand for various materials in the future production process, forming a material demand forecasting result data set, which specifically includes:

[0037] S41, the constructed material demand forecasting model is based on the feature matrix X As input, along the optimized final path, first output the demand in the future production process and t Moment a The demand forecast value of the material;

[0038] ; The expanded form of this formula is:

[0039] ;

[0040] in, Indicated in t Moment a The demand forecast value of the material. , and are model parameters, is the feature matrix X Theb Features, is the period constant, is the attenuation constant, m is the number of features; represents the material demand forecasting model, Represents the material demand forecasting model With feature matrix X As input quantity; Indicates the demand in the future production process;

[0041] S42, process the demand output by S41 to obtain the demand for various materials in the future production process, expressed as:

[0042] ;

[0043] ;

[0044] in, Indicates a The total demand for the material, Indicates a Materials in t The demand at the moment; , , and To adjust the parameters; n is the number of material types; T is the production cycle constant; π is the circumference of a circle; exp () represents an exponential function with base e; c =1,2,3..., n ; c is the influencing factor, indicating that the future demand for the a-th material is multiple different influencing factors c;

[0045] S43. Summarize the demand for all materials in the future production cycle to form a material demand forecast result data set ,in:

[0046] ;

[0047] ;

[0048] in, and is the weight parameter, For materials a is the periodic constant; π is the circumference of a circle; cos () is the cosine function; sin () is a sine function.

[0049] Further, in S41, the optimization process of the optimized final path is specifically as follows:

[0050] S411, firstly optimize each individual using a genetic algorithm to obtain an optimized population; each individual represents a material demand forecasting scheme;

[0051] S412, based on the optimized population, optimizing using a particle swarm optimization algorithm to obtain an optimized particle swarm;

[0052] S413. Based on the optimized particle swarm, an ant colony algorithm is used to perform optimization to obtain an optimized final path.

[0053] Furthermore, when the material demand forecasting model is constructed, after obtaining the material demand forecasting result data set according to steps S41-S43, the material demand forecasting result data set is cross-validated, specifically including the following steps:

[0054] S44. Material demand forecast result data set Divide into training set and validation set, and use cross-validation method for evaluation;

[0055] S45, calculating the prediction error of each cross validation, the prediction error is expressed as:

[0056] ;

[0057] in, For the x The average prediction error of cross validation, Indicates the real demand. represents the forecast demand, z is the number of samples in the validation set; y =1,2...3, z ;

[0058] S46. Perform statistical analysis on all cross-validated prediction errors and calculate the total error :

[0059] ;

[0060] in, v is the number of cross-validation folds; x =1,2,3... v ;

[0061] S47, using the error analysis method to analyze the prediction error, identify the error source and its influencing factors, and obtain the error analysis result;

[0062] S48. Based on the error analysis results, adjust the initial parameters of the material demand forecasting model. The initial parameters include , the optimized parameters are expressed as:

[0063] ;

[0064] in, is the optimized parameter, is the initial parameter, is the learning rate, is the gradient of the total error;

[0065] S49, repeat S44 to S48 until the total error Converge to the predetermined threshold and obtain the final optimized material demand forecasting model.

[0066] Further, in S5, the construction of the total inventory data set specifically includes the following process:

[0067] Get the current inventory data, which is the current inventory of each material, recorded as a data set ;

[0068] ;

[0069] in, Indicates n The current inventory of the material, n is the number of material types;

[0070] Get the in-transit inventory data, which is the quantity of materials that have been shipped but have not yet arrived at the warehouse, recorded as a data set ;

[0071] ;

[0072] in, Indicates n The in-transit inventory of the material, n is the number of material types;

[0073] Get the data of ordered but undelivered materials. The ordered but undelivered materials data is the quantity of materials that have been ordered but not yet delivered, recorded as a data set ;

[0074] ;

[0075] in, Indicates n The quantity of the material that has been ordered but not delivered, n is the number of material types;

[0076] Aggregate the current inventory data, in-transit inventory data, and ordered but undelivered material data to form a total inventory data set ,in:

[0077] .

[0078] The present invention also discloses a discrete material complete set management system, comprising:

[0079] A first data construction module is used to construct a material demand forecast data set, wherein the material demand forecast data set includes production plan data, historical material usage data, order data, seasonal factors and market demand change data;

[0080] A data preprocessing module is used to preprocess the material demand forecast data set to obtain a preprocessed material demand forecast data set;

[0081] The feature extraction module is used to extract features from the preprocessed material demand forecast data set to obtain feature parameters;

[0082] The material demand forecasting module is used to input characteristic parameters into the constructed material demand forecasting model to obtain the demand for various materials in the future production process and form a material demand forecasting result data set;

[0083] A second data construction module is used to construct a total inventory data set, wherein the total inventory data set includes the current inventory, the in-transit inventory, and the ordered but undelivered material data;

[0084] The complete set comparison module is used to compare the total inventory data set with the material demand forecast result data set to obtain the discrete material complete set management solution:

[0085] If the total inventory data set is ≥ the material demand forecast result data set, the materials are complete; if the total inventory data set is < the material demand forecast result data set, the materials are not complete.

[0086] The present invention also discloses a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the discrete material set management method are implemented.

[0087] The present invention also discloses a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the discrete material set management method are implemented.

[0088] Compared with the prior art, the present invention has the following beneficial technical effects:

[0089] The present invention discloses a discrete material set management method. By combining multi-dimensional data such as production plan, historical material usage, order data, seasonal factors and market demand changes, a material demand forecast data set is constructed. By utilizing feature extraction and material demand forecasting models, the demand for various materials in future production can be more accurately predicted to form a material demand forecast result data set, thereby significantly reducing the prediction error, reducing the situation of material surplus or shortage, and improving the accuracy of material demand forecasting. With the help of a comprehensive analysis of market demand changes and seasonal factors, the method can quickly respond to market changes, adjust material demand forecasts, optimize material ordering and reserve strategies, and help enterprises respond to market fluctuations more flexibly and enhance market competitiveness. The present invention also constructs a total inventory data set by acquiring data on existing inventory, in-transit inventory, and materials that have been ordered but not delivered. This not only avoids capital occupation and increased storage costs caused by excessive inventory, but also ensures a stable supply of materials in the production process, reduces the risk of production interruption or stagnation due to material shortages, and optimizes inventory management. Finally, by comparing the material demand forecast result data set with the total inventory data set, it is determined whether the materials are complete. Through complete management, the timely supply of materials required for production is ensured, and delays or adjustments in production plans due to material shortages are reduced, thereby improving the continuity and efficiency of production. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Figure 1 A flow chart of a discrete material set management method of the present invention;

[0091] Figure 2 The present invention is a block diagram of a discrete material complete set management system. DETAILED DESCRIPTION

[0092] In order to make the purpose, technical solution and advantages of the present invention more clear, the following is further described in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention, that is, the embodiments described are only part of the embodiments of the present invention, not all embodiments.

[0093] The detailed description of the embodiments of the present invention provided in the following drawings is not intended to limit the scope of the invention claimed for protection, but merely represents a selected embodiment of the present invention. Based on the drawings and embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0094] The features and performances of the present invention are further described in detail below in conjunction with the embodiments.

[0095] Example 1

[0096] like Figure 1As shown, the present invention discloses a discrete material set management method, comprising the following steps:

[0097] S1. Construct a material demand forecasting data set, which includes production plan data, historical material usage data, order data, seasonal factors and market demand change data;

[0098] S2. preprocessing the material demand forecast data set to obtain a preprocessed material demand forecast data set;

[0099] Perform feature extraction on the preprocessed material demand forecast data set to obtain feature parameters;

[0100] S4. Input the characteristic parameters into the constructed material demand forecasting model to obtain the demand for various materials in the future production process, forming a material demand forecasting result data set;

[0101] S5. Construct a total inventory data set, wherein the total inventory data set includes existing inventory, in-transit inventory, and ordered but undelivered material data;

[0102] S5. Compare the total inventory data set obtained in S5 with the material demand forecast result data set obtained in S4. If the total inventory data set is ≥ the material demand forecast result data set, the materials are complete; if the total inventory data set is < the material demand forecast result data set, the materials are incomplete.

[0103] The present invention constructs a material demand forecasting data set by acquiring multi-dimensional data such as production plans, historical material usage, order data, seasonal factors, and market demand changes; after preprocessing the material demand forecasting data set, feature extraction is performed to obtain feature parameters; the feature parameters are input into the constructed material demand forecasting model. Through the prediction of the material demand forecasting model, the demand for various materials in the future production process can be more accurately predicted, which significantly reduces the prediction error and avoids the problem of material surplus or shortage due to inaccurate prediction.

[0104] Example 2

[0105] Based on Example 1, the process of constructing a material demand forecasting data set is specifically introduced, which specifically includes the following processes:

[0106] S11. Obtain production plan data, which includes the type, quantity and production time of the planned products, recorded as a data set :

[0107] ;

[0108] in, Indicates i The production load quantity of each product; Indicates i The planned production quantity of the product; represents the production time of the i-th product; T represents the production cycle constant; f Indicates the number of product types. i =1,2,…, f , exp () represents an exponential function with base e;

[0109] S12. Obtain historical material usage data, which includes the usage records of various materials in the past period of time, and is recorded as a data set. :

[0110] ;

[0111] in, Indicates j The quantity of material required, Indicates j The amount of material used, Indicates j The use time of the material, represents the historical time period constant, n Indicates the number of material types. j =1,2,…, n , cos () is the cosine function;

[0112] S13. Obtain order data, which includes confirmed customer order information. The confirmed customer order information includes order product type, quantity and delivery time, recorded as a data set :

[0113] ;

[0114] in, Indicates k The contract value of the order, Indicates k The number of products in an order, Indicates k Delivery time for orders, represents the order cycle constant, p Indicates the total order quantity; k =1,2,3... p ; sin () is the sine function; π is the circumference of a circle;

[0115] S14. Obtain seasonal factor data, which includes the impact of seasonal changes on material demand, and record it as a data set :

[0116] ;

[0117] in, Indicates l The correction coefficient of seasonal factors on material production efficiency, Indicates the time of seasonal changes. represents the seasonal cycle constant, q Indicates the number of seasonal factors; l =1,2,3... q ; cos() is the cosine function; π is the circumference of a circle; exp () represents an exponential function with base e;

[0118] S15. Obtain market demand change data, which includes fluctuations and trends in market demand and is recorded as a data set. :

[0119] ;

[0120] in, Indicated in r Adjustment coefficient of material demand under different market demand conditions; Indicates the time when market demand changes; represents the market demand cycle constant; s Indicates the number of types of market demand changes; r =1,2,3... s ; sin() is the sine function; π is the pi; exp () represents an exponential function with base e;

[0121] S16, summarizing the production plan data, historical material usage data, order data, seasonal factor data and market demand change data to construct a material demand forecast data set:

[0122] .

[0123] S2 is specifically:

[0124] Preprocess the constructed material demand forecast data set by using data cleaning, normalization and denoising techniques to obtain the preprocessed material demand forecast data set;

[0125] Perform feature extraction on the preprocessed material demand forecast data set to obtain the feature matrix X ,in

[0126] ;

[0127] Each feature represents a specific attribute in the material demand forecast dataset.

[0128] The feature extraction process is as follows:

[0129] Data preprocessing:

[0130] Data cleaning: First, the material demand forecast data set Dpd is cleaned to delete invalid or missing data points to ensure the integrity and consistency of the data.

[0131] Data standardization: Standardize or normalize the data to eliminate the impact of dimensional differences on feature extraction. For example, scale all features to the same numerical range.

[0132] Frequency analysis: Based on the data of each material, calculate the demand frequency of each material in a specific time period (such as quarter, month), and identify the peak and trough periods of demand.

[0133] Through the above steps, the material demand forecasting data set Dpd is converted into a structured feature matrix X. Each feature corresponds to a specific attribute in Dpd and provides basic data support for subsequent material set management.

[0134] Example 3

[0135] Based on Example 2, the construction process of the material demand forecasting model is mainly introduced. The material demand forecasting model is based on the final path of the hybrid intelligent optimization algorithm optimization output, and the hybrid intelligent optimization algorithm includes a genetic algorithm, a particle swarm optimization algorithm and an ant colony algorithm.

[0136] Similar to Example 2, firstly, a material demand forecasting data set is constructed, and the material demand forecasting data set is preprocessed to obtain a preprocessed material demand forecasting data set; and feature extraction is performed on the preprocessed material demand forecasting data set to obtain a feature matrix X.

[0137] In S4, the construction process of the material demand forecasting model is as follows:

[0138] 1. First, optimize the final output path, which includes the following three parts:

[0139] A. Use genetic algorithm to Optimize and obtain the optimized population; specifically include the following processes:

[0140] A1. Initialize the population of the genetic algorithm and set the population size to N. Each individual Represents a material demand forecasting scheme;

[0141] A2. Calculate each individual The fitness value of ,The fitness value is defined according to the prediction error, and by introducing adaptive crossover and mutation probabilities, the crossover and mutation probabilities are larger in the early stage of the algorithm to increase diversity, and gradually decrease in the later stage to ensure convergence;

[0142] A3. Perform selection, crossover and mutation operations on the population of the genetic algorithm to generate a new population. The selection operation adopts the tournament selection method to improve the selection efficiency and convergence speed.

[0143] A4. Introduce the elite retention strategy to directly retain the individuals with the highest fitness to the next generation to avoid the loss of high-quality solutions;

[0144] A5. Repeat steps A1 to A4 until the termination condition is met and the optimized population is obtained. ;

[0145] B. Based on the optimized population , further optimize using the particle swarm optimization algorithm to obtain the optimized particle swarm; specifically, the following steps are included:

[0146] B1. Initialize the particle swarm of the particle swarm optimization algorithm, set the number of particles, and each particle represents an optimization scheme;

[0147] B2. Calculate each particle The fitness value of , according to the prediction error definition, and the introduction of adaptive inertia weight mechanism, a larger inertia weight is used in the early stage to enhance the global search capability, and then gradually reduced in the later stage to improve the local search accuracy;

[0148] B3. Introduce a local search mechanism. When the particle swarm converges to a certain area, use the local search algorithm to further optimize and improve the accuracy of the solution;

[0149] B4. Update the speed and position of each particle to make it tend to the global optimal position and the individual optimal position. The speed update formula is:

[0150] ;

[0151] in, is the inertia weight, , is the acceleration coefficient, , is a random number, For particles The best historical position, is the global optimal position; It represents the velocity of a particle in the particle group at time t; Represents the particle position of a particle in the particle swarm at time t.

[0152] B5. Repeat steps B2 to B4 until the termination condition is met to obtain the optimized particle swarm.

[0153] C. Based on the optimized particle swarm, the ant colony algorithm is used for final optimization to obtain the optimized final path; specifically, the following steps are included:

[0154] C1. Initialize the ant colony of the ant colony algorithm, set the number of ants to Q, and each ant represents an optimization path;

[0155] C2. Determine the location of each ant based on the pheromone concentration and the heuristic function. The moving probability is calculated, and by introducing a dynamic pheromone update strategy, the pheromone concentration is dynamically adjusted according to the path quality, the pheromone concentration of high-quality paths increases, and the pheromone concentration of low-quality paths decreases;

[0156] C3. In the process of path selection, global information and local information are combined to improve the globality and stability of the solution through multi-objective optimization methods;

[0157] C4. The pheromone update formula is:

[0158] ;

[0159] in, Indicates the path d arrive g The pheromone concentration, is the pheromone volatility coefficient, To increase the amount of pheromones; Indicates the path at a certain moment d arrive g pheromone concentration;

[0160] C5. Repeat steps C2 to C4 until the termination condition is met to obtain the final optimized path. .

[0161] 2. Taking the feature matrix X as input, the feature matrix X is output along the final optimized path to build the initial material demand forecasting model .

[0162] Material Demand Forecasting Model First, calculate the demand in the future production process and the forecast demand value of the a-th material at time t;

[0163] ; The expanded form of this formula is:

[0164] ;

[0165] in, Indicated in t Moment a The demand forecast value of the material. , and are model parameters, is the feature matrix X The b Features, is the period constant, is the attenuation constant, m is the number of features; represents the material demand forecasting model, Represents the material demand forecasting model With feature matrix X As input quantity; Indicates the demand in the future production process;

[0166] The output demand is processed to obtain the demand for various materials in the future production process, expressed as:

[0167] ;

[0168] ;

[0169] in, Indicates a The total demand for the material, Indicates a Materials in t The demand at the moment; , , and To adjust the parameters; n is the number of material types; T is the production cycle constant; π is the circumference of a circle; exp () represents an exponential function with base e; c =1,2,3..., n , c is the influencing factor, indicating that the future demand for the a-th material is multiple different influencing factors c.

[0170] 3. Summarize the demand for all materials in the future production cycle to form a material demand forecast result data set ,in:

[0171] ;

[0172] ;

[0173] in, and is the weight parameter, For materials a is the periodic constant; π is the circumference of a circle; cos () is the cosine function; sin () is a sine function.

[0174] Material demand forecast result dataset Conduct evaluation, use cross-validation and error analysis methods to adjust the parameters of the material demand forecasting model, and obtain the final optimized material demand forecasting model .

[0175] The material demand forecasting model of the present invention significantly improves the accuracy and efficiency of the material demand forecasting model and the accuracy of the material demand forecasting by comprehensively applying the genetic algorithm, the particle swarm optimization algorithm and the ant colony algorithm when optimizing the final path. The details are as follows:

[0176] Genetic algorithm, particle swarm optimization algorithm and ant colony algorithm each have their own advantages. Genetic algorithm is good at global search and population diversity maintenance, particle swarm optimization algorithm has the characteristics of fast convergence speed and strong adaptability, and ant colony algorithm is suitable for global optimization of path optimization problems. The combination of the three can effectively avoid the local optimal problem of a single algorithm, improve the global search ability and optimization effect of the material demand forecasting model, and thus significantly improve the accuracy of material demand forecasting.

[0177] By combining the mutation and crossover operations of the genetic algorithm with the group collaboration mechanism of the particle swarm optimization, and supplemented by the distributed optimization strategy of the ant colony algorithm, the material demand forecasting model of the present invention can converge to the optimal solution more quickly when performing large-scale complex data calculations. The comprehensive application of this optimization path not only makes the material demand forecasting model have higher prediction accuracy, but also significantly improves the calculation efficiency and reduces the time cost in the prediction process.

[0178] Through the global optimization of model parameters by genetic algorithm, the rapid exploration of local optimal solutions by particle swarm algorithm, and the fine optimization of ant colony algorithm in complex paths, this method can more accurately simulate the multi-dimensional influencing factors of material demand, effectively reduce prediction errors, further improve the accuracy of material demand forecasting, and avoid material surplus or shortage problems caused by inaccurate predictions.

[0179] Since the three optimization algorithms each have different working mechanisms, the combination of the three makes the model more robust and adaptable to different types of data. Even in the face of highly complex production plans, market changes and historical data, the material demand forecasting model can maintain high stability and accuracy, ensuring accurate forecasting of material demand.

[0180] Through the integration of genetic algorithm, particle swarm optimization algorithm and ant colony algorithm, the material demand forecasting model of the present invention can dynamically adjust and optimize the path, adapt to the changes in material demand in the production process, quickly respond to fluctuations in market demand, enhance the flexibility and elasticity of the material management system, and effectively improve the resilience of the production supply chain.

[0181] Example 4

[0182] In this embodiment, the process of cross-validating the material demand forecast result data set is mainly introduced, which specifically includes the following steps:

[0183] Material demand forecast result dataset Divide into training set and validation set, and use cross-validation method for evaluation;

[0184] Calculate the prediction error for each cross validation, the prediction error is expressed as:

[0185] ;

[0186] in, For the x The average prediction error of cross validation, Indicates the real demand. represents the forecast demand, z is the number of samples in the validation set; y =1,2...3, z ;

[0187] Perform statistical analysis on all cross-validation prediction errors and calculate the total error :

[0188] ;

[0189] in, v is the number of cross-validation folds; x =1,2,3... v ;

[0190] The error analysis method is used to analyze the prediction error, identify the error source and its influencing factors, and obtain the error analysis results;

[0191] Adjust the material demand forecast model based on the error analysis results The initial parameters include , the optimized parameters are expressed as:

[0192] ;

[0193] in, is the optimized parameter, is the initial parameter, is the learning rate, is the gradient of the total error;

[0194] Repeat the above steps until the total error Converge to the predetermined threshold and obtain the final optimized material demand forecasting model.

[0195] The error analysis method of the present invention adopts the component error decomposition method to decompose the forecast error into components such as trend error, seasonal error and random error. For a time series data with obvious seasonality and trend, if the forecast model cannot capture the seasonal law well, seasonal error will be generated; if the long-term growth or decline trend of the data is not accurately predicted, trend error will be generated; and random error is caused by unpredictable factors.

[0196] Example 5

[0197] Based on Example 1, the construction process of the total inventory data set is mainly introduced, which specifically includes the following processes:

[0198] S51. Obtain existing inventory data, which is the current inventory of each material, recorded as a data set ;

[0199] ;

[0200] in, Indicates n The current inventory of the material, n is the number of material types;

[0201] S52. Obtain in-transit inventory data, which is the quantity of materials that have been shipped but have not yet arrived at the warehouse, and is recorded as a data set. ;

[0202] ;

[0203] in, Indicates n The in-transit inventory of the material, n is the number of material types;

[0204] S53. Obtain the data of materials that have been ordered but not delivered. The data of materials that have been ordered but not delivered is the quantity of materials that have been ordered but not delivered, recorded as a data set. ;

[0205] ;

[0206] in, Indicates nThe quantity of the material that has been ordered but not delivered, n is the number of material types;

[0207] S54. Aggregate the existing inventory data, in-transit inventory data, and ordered but undelivered material data to form a total inventory data set ,in:

[0208] .

[0209] The present invention constructs a total inventory data set by acquiring the existing inventory, in-transit inventory, and ordered but undelivered material data. This not only avoids capital occupation and increased storage costs caused by excessive inventory, but also ensures a stable supply of materials during the production process, reducing the risk of production interruption or stagnation caused by material shortages; while effectively balancing the supply and demand of materials, through accurate demand forecasting and inventory management, it avoids the waste of funds caused by excessive inventory backlogs and emergency purchases caused by insufficient inventory, thereby reducing operating costs.

[0210] Example 6

[0211] like Figure 2 As shown, the present invention discloses a discrete material complete set management system, comprising:

[0212] A first data construction module is used to construct a material demand forecast data set, wherein the material demand forecast data set includes production plan data, historical material usage data, order data, seasonal factors and market demand change data;

[0213] A data preprocessing module is used to preprocess the material demand forecast data set to obtain a preprocessed material demand forecast data set;

[0214] The feature extraction module is used to extract features from the preprocessed material demand forecast data set to obtain feature parameters;

[0215] The material demand forecasting module is used to input characteristic parameters into the constructed material demand forecasting model to obtain the demand for various materials in the future production process and form a material demand forecasting result data set;

[0216] The material demand forecasting model is constructed based on a hybrid intelligent optimization algorithm, which includes a genetic algorithm, a particle swarm optimization algorithm, and an ant colony algorithm;

[0217] A second data construction module is used to construct a total inventory data set, wherein the total inventory data set includes the current inventory, the in-transit inventory, and the ordered but undelivered material data;

[0218] The complete set comparison module is used to compare the total inventory data set with the material demand forecast result data set. If the total inventory data set ≥ the material demand forecast result data set, the materials are complete; if the total inventory data set < the material demand forecast result data set, the materials are not complete.

[0219] Example 7

[0220] The present invention also discloses a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the discrete material set management method are implemented. The memory may include a memory, such as a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk memory, etc. The processor, the network interface, and the memory are interconnected through an internal bus. The internal bus may be an industrial standard architecture bus, a peripheral component interconnection standard bus, an extended industrial standard structure bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The memory is used to store programs. Specifically, the program may include a program code, and the program code includes computer operation instructions. The memory may include a memory and a non-volatile memory, and provide instructions and data to the processor.

[0221] Example 8

[0222] The present invention also discloses a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the discrete material set management method are implemented. Specifically, the computer-readable storage medium includes, but is not limited to, for example, a volatile memory and / or a non-volatile memory. The volatile memory may include a random access memory and / or a cache memory, etc. The non-volatile memory may include a read-only memory, a hard disk, a flash memory, an optical disk, a magnetic disk, etc.

[0223] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, optical storage, etc.) containing computer-usable program code.

[0224] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0225] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0226] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0227] Application Examples

[0228] In an equipment manufacturing company, the company's production involves various types of mechanical parts and raw materials, and its material management has always faced many challenges. Specifically, the company has great difficulties in material demand forecasting and inventory management, and often has excessive or insufficient inventory, resulting in low production efficiency and increased operating costs.

[0229] In this scenario, first, the enterprise begins to obtain material demand forecasting data sets, including production plan data, historical material usage data, order data, seasonal factors, and market demand change data. For example, the enterprise obtains production plan data for the past two years, including the production quantity and time of each product, historical material usage data records the usage of various materials, order data includes detailed information on all customer orders, seasonal factor data reflects the impact of seasonal changes on material demand, and market demand change data shows the fluctuations and trends of market demand. These data are processed through preprocessing steps, including data cleaning, normalization, and denoising techniques, to construct a clean and standardized forecasting data set.

[0230] Next, the preprocessed material demand forecast data set is extracted for features and input into the material demand forecast model to calculate the demand for various materials in the future production process. For example, the material demand forecast model calculates that the demand for a certain mechanical part in the next quarter is 1,500 pieces, while the demand for a certain raw material is 3,000 kilograms. These forecast results are evaluated and adjusted through cross-validation and error analysis techniques to ensure the forecast accuracy of the material demand forecast model.

[0231] Subsequently, the company obtained the data on existing inventory, in-transit inventory, and ordered but undelivered materials. These data include the current inventory of each material, the number of materials that have been shipped but have not yet arrived at the warehouse, and the number of materials that have been ordered but not yet delivered. By aggregating these data, the company formed a total inventory data set, and found that the total inventory of a certain mechanical part is 2,000 pieces, and the inventory of a certain raw material is 4,000 kilograms. In the next quarter, the result of the complete set of materials for a certain mechanical part is complete, and the result of the complete set of materials for a certain raw material is complete.

[0232] After implementing the new method, the company has achieved remarkable results in improving material set management. The specific data are as follows:

[0233] 1. Accuracy of material demand forecast: Before implementation, the error rate of material demand forecast was 20%. After implementation, the error rate was reduced to 8% through the material demand forecast model, which significantly improved the accuracy of the forecast.

[0234] 2. Inventory management optimization: Before implementation, the average monthly capital occupation caused by excessive inventory was about 1 million yuan, and the loss of shutdown caused by inventory shortage was about 500,000 yuan. After implementation, the capital occupation caused by excessive inventory was reduced to 300,000 yuan, and the loss of shutdown was reduced to 100,000 yuan.

[0235] 3. Improved production efficiency: Before implementation, production downtime due to material shortages or supply delays was about 15 hours per month. After implementation, production downtime was reduced to 2 hours, and production efficiency was greatly improved.

[0236] Implementation location: The production base of a certain equipment manufacturing enterprise, located in a certain city in a certain province.

[0237] Implementation time: The implementation period is from January 2023 to December 2023.

[0238] It can be seen from the above specific data that the discrete material set management method disclosed in the present invention has significant beneficial effects in improving the accuracy of material demand forecasting and optimizing inventory management level. This not only improves the production efficiency of the enterprise, but also effectively reduces the operating cost, bringing significant economic benefits and competitive advantages to the enterprise.

[0239] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A discrete material set management method, characterized in that: The following steps are involved: S1. Construct a material demand forecasting data set, which includes production plan data, historical material usage data, order data, seasonal factors and market demand change data; Constructing a total inventory data set, wherein the total inventory data set includes the current inventory, the in-transit inventory, and the ordered but undelivered material data; S2. preprocessing the material demand forecast data set to obtain a preprocessed material demand forecast data set; S3, extracting features from the preprocessed material demand forecast data set to obtain feature parameters; S4. Input the characteristic parameters into the constructed material demand forecasting model to obtain the demand for various materials in the future production process, forming a material demand forecasting result data set; S5. Compare the total inventory data set in S1 with the material demand forecast result data set in S4 to obtain the discrete material complete set management solution: If the total inventory data set is ≥ the material demand forecast result data set, the materials are complete; if the total inventory data set is < the material demand forecast result data set, the materials are not complete; In S3, feature extraction is performed on the preprocessed material demand forecast data set to obtain the feature matrix X ,in ; Each feature represents a specific attribute in the material demand forecast dataset; In S4, the characteristic parameters are input into the constructed material demand forecasting model to obtain the demand for various materials in the future production process, forming a material demand forecasting result data set, which specifically includes: S41, the constructed material demand forecasting model is based on the feature matrix X As input, along the optimized final path, the demand in the future production process and the t Moment a The demand forecast value of the material; ; The expanded form of this formula is: ； in, Indicated in t Moment a The demand forecast value of the material. , and are model parameters, is the feature matrix X The b Features, is the period constant, is the attenuation constant, m is the number of features; represents the material demand forecasting model, Represents the material demand forecasting model With feature matrix X As input quantity; Indicates the demand in the future production process; S42, process the demand output by S41 to obtain the demand for various materials in the future production process, expressed as: ； ； in, Indicates a The total demand for the material, Indicates a Materials in t The demand at the moment; , , and To adjust the parameters; n is the number of material types; T is the production cycle constant; π is the circumference of a circle; exp () represents an exponential function with base e; c =1,2,3..., n ; c is the influencing factor, indicating that the future demand for the a-th material is multiple different influencing factors c; S43. Summarize the demand for all materials in the future production cycle to form a material demand forecast result data set ,in: ； ； in, and is the weight parameter, For materials a is the periodic constant; π is the circumference of a circle; cos () is the cosine function; sin () is a sine function; In S41, the optimization process of the optimized final path is specifically as follows: S411, firstly optimize each individual using a genetic algorithm to obtain an optimized population; each individual represents a material demand forecasting scheme; S412, based on the optimized population, optimizing using a particle swarm optimization algorithm to obtain an optimized particle swarm; S413, based on the optimized particle swarm, using the ant colony algorithm to perform optimization to obtain the optimized final path; In S5, the construction of the total inventory data set specifically includes the following processes: Get the current inventory data, which is the current inventory of each material, recorded as a data set ; ； in, Indicates n The current inventory of the material, n is the number of material types; Get the in-transit inventory data, which is the quantity of materials that have been shipped but have not yet arrived at the warehouse, recorded as a data set ; ； in, Indicates n The in-transit inventory of the material, n is the number of material types; Get the data of ordered but undelivered materials. The ordered but undelivered materials data is the quantity of materials that have been ordered but not yet delivered, recorded as a data set ; ； in, Indicates n The quantity of the material that has been ordered but not delivered, n is the number of material types; Aggregate the current inventory data, in-transit inventory data, and ordered but undelivered material data to form a total inventory data set ,in: 。 2. A discrete material set management method according to claim 1, characterized in that: In S1, the material demand forecasting data set is constructed. The process includes: Obtain production plan data, which includes the type, quantity and production time of the planned products, recorded as a data set ; ； in, Indicates i The production load quantity of each product; Indicates i The planned production quantity of the product; represents the production time of the i-th product; T Represents the production cycle constant; f Indicates the number of product types. i =1,2,…, f , exp () represents an exponential function with base e; Obtain historical material usage data, which includes the usage records of various materials over a period of time in the past and is recorded as a data set ; ； in, Indicates j The quantity of material required, Indicates j The amount of material used, Indicates j The use time of the material, represents the historical time period constant, n Indicates the number of material types. j =1,2,…, n , cos () is the cosine function; Obtain order data, which includes the confirmed order product type, quantity and delivery time, recorded as a data set ; ； in, Indicates k The contract value of the order, Indicates k The number of products in an order, Indicates k Delivery time for orders, represents the order cycle constant, p Indicates the total order quantity; k =1,2,3... p ; sin () is the sine function; π is the circumference of a circle; Obtain seasonal factor data, which includes the impact of seasonal changes on material demand, and is recorded as a data set ; ； in, Indicates l The correction coefficient of seasonal factors on material production efficiency, Indicates the time of seasonal changes. represents the seasonal cycle constant, q Indicates the number of seasonal factors; l =1,2,3... q ; cos() is the cosine function; π is the circumference of a circle; exp () represents an exponential function with base e; Obtain market demand change data, which includes fluctuations and trends in market demand and is recorded as a data set ; ； in, It represents the adjustment coefficient of material demand under the rth market demand condition; Indicates the time when market demand changes; represents the market demand cycle constant; s Indicates the number of types of market demand changes; r =1,2,3... s ; sin() is the sine function; π is the pi; exp () represents an exponential function with base e; The production plan data, historical material usage data, order data, seasonal factor data and market demand change data are aggregated to construct a material demand forecast data set, denoted as ; 。 3. A discrete material complete set management method according to claim 1, characterized in that: When the material demand forecasting model is constructed, after obtaining the material demand forecasting result data set according to steps S41-S43, the material demand forecasting result data set is cross-validated, specifically including the following steps: S44. Material demand forecast result data set Divide into training set and validation set, and use cross-validation method for evaluation; S45, calculating the prediction error of each cross validation, the prediction error is expressed as: ； in, For the x The average prediction error of cross validation, Indicates the real demand. represents the forecast demand, z is the number of samples in the validation set; y =1,2...3, z ; S46. Perform statistical analysis on all cross-validated prediction errors and calculate the total error : ； in, v is the number of cross-validation folds; x =1,2,3... v ; S47, using the error analysis method to analyze the prediction error, identify the error source and its influencing factors, and obtain the error analysis result; S48. Based on the error analysis results, adjust the initial parameters of the material demand forecasting model. The initial parameters include , the optimized parameters are expressed as: ； in, is the optimized parameter, is the initial parameter, is the learning rate, is the gradient of the total error; S49, repeat S44 to S48 until the total error Converge to the predetermined threshold and obtain the final optimized material demand forecasting model.

4. A discrete material set management system based on the discrete material set management method according to any one of claims 1 to 3, characterized in that: include: A first data construction module is used to construct a material demand forecast data set, wherein the material demand forecast data set includes production plan data, historical material usage data, order data, seasonal factors and market demand change data; A data preprocessing module is used to preprocess the material demand forecast data set to obtain a preprocessed material demand forecast data set; The feature extraction module is used to extract features from the preprocessed material demand forecast data set to obtain feature parameters; The material demand forecasting module is used to input characteristic parameters into the constructed material demand forecasting model to obtain the demand for various materials in the future production process and form a material demand forecasting result data set; A second data construction module is used to construct a total inventory data set, wherein the total inventory data set includes the existing inventory, the in-transit inventory, and the ordered but undelivered material data; The complete set comparison module is used to compare the total inventory data set with the material demand forecast result data set to obtain the discrete material complete set management solution: If the total inventory data set is ≥ the material demand forecast result data set, the materials are complete; if the total inventory data set is < the material demand forecast result data set, the materials are not complete.

5. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the discrete material set management method as described in any one of claims 1 to 3 are implemented.

6. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the discrete material set management method as described in any one of claims 1 to 3 are implemented.

Images