CN113962416A - Engineering machinery part stock prediction method, management method and system - Google Patents
Engineering machinery part stock prediction method, management method and system Download PDFInfo
- Publication number
- CN113962416A CN113962416A CN202010633731.5A CN202010633731A CN113962416A CN 113962416 A CN113962416 A CN 113962416A CN 202010633731 A CN202010633731 A CN 202010633731A CN 113962416 A CN113962416 A CN 113962416A
- Authority
- CN
- China
- Prior art keywords
- demand
- data
- prediction
- model
- accessory
- 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
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION 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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06315—Needs-based resource requirements planning or analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION 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
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/087—Inventory or stock management, e.g. order filling, procurement or balancing against orders
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Marketing (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Software Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Mathematical Physics (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Educational Administration (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention provides a method for predicting the stock of engineering machinery parts, which comprises the following steps: constructing data characteristics from historical sales data of the accessories; matching a corresponding demand prediction model according to the data characteristics; and obtaining the predicted demand of the accessory based on the historical sales data and the demand prediction model of the accessory. Meanwhile, a corresponding engineering machinery part stock management method and system are also disclosed. The implementation mode of the invention solves the problem that a single prediction model has poor adaptability to different types of engineering machinery parts, and improves the efficiency in the stock management of the engineering machinery parts and the economy and accuracy of stock.
Description
Technical Field
The invention relates to the field of inventory management, in particular to a method for predicting the stock of engineering mechanical parts, two methods for managing the stock of the engineering mechanical parts and a system for managing the stock of the engineering mechanical parts.
Background
The accessory requirements of the engineering machinery industry mainly come from the maintenance and repair requirements of corresponding main machines. When the host computer is out of schedule and stops, the construction period of a client can be greatly influenced, extra economic loss is brought, the accessory requirement has the characteristics of time urgency, multiple varieties, small single material demand and the like, and most materials cannot form stable logistics.
In the industry, a weighted average or a moving average of historical demand of parts in a recent period is mainly used as future demand of the parts, and a certain multiple is used as stock safety of the parts, so that a part stock plan is formed. Under different host computer operating rate, the volume of keeping the circumstances, the accessory demand is different, and prior art does not consider the influence of host computer factor to the accessory demand, can only carry out comparatively effectual stock usually to forming the material of stabilizing the commodity circulation less part, can't form more accurate prediction to the material of most lower demand.
Aiming at the conditions of multiple varieties of engineering machinery parts, unstable demand and the like, the prior art does not fully combine the requirements of a total bin and a branch bin to carry out modeling stock, and does not carry out targeted modeling on safety stock and turnover stock of different types of parts.
Disclosure of Invention
The invention aims to provide a method for predicting the stock of engineering machinery parts, a method for managing the stock of the engineering machinery parts and a system, and aims to solve the problem of poor adaptability of the existing single prediction model applied to the stock management of different types of engineering machinery parts.
In order to achieve the above object, in a first aspect of the present invention, there is provided a method for predicting a stock of work machine parts, the method including:
constructing data characteristics from historical sales data of the accessories; matching a corresponding demand prediction model according to the data characteristics; and obtaining the predicted demand of the accessory based on the historical sales data and the demand prediction model of the accessory.
Optionally, the constructing data features from historical sales data of the accessories includes: taking each sales index of N sales indexes in historical sales data of the accessories as a dimension; extracting sales data of each sales index of the accessories in a period of historical time, and processing the sales data into a characteristic value by adopting a sliding window algorithm; and obtaining a feature vector with N dimensions corresponding to the accessory as the data feature.
Optionally, the matching of the corresponding demand prediction model according to the data features includes: inputting the data characteristics of the accessories into a spectral clustering model to obtain a classification result; the classification result is one of M preset results, and the M preset results all have corresponding demand prediction models; and obtaining a demand prediction model corresponding to the classification result.
Optionally, the demand prediction model is a machine learning prediction model, including: a random forest model, a GBDT model and an XGboost model; the demand forecasting model is trained by using the data features constructed from historical sales data of the accessories.
Optionally, the method further includes: based on the prediction demand, distributing and obtaining the bin-dividing prediction demand of each bin: according to the demand data of the sub-bins, an additive regression model is adopted to predict the demand proportion coefficient of the accessory in the sub-bins; and obtaining the sub-bin prediction demand corresponding to the sub-bin based on the prediction demand and the demand proportion coefficient.
In a second aspect of the present invention, there is also provided a method for managing spare parts of a construction machine, the method including:
obtaining the predicted demand of the accessory based on the prediction method; obtaining a turnover storage inventory based on the predicted demand and delivery cycle of the part; and obtaining the stock quantity of the parts based on the turnover stock and the safety stock.
Optionally, the safety stock is obtained by:
fitting a probability distribution shape of historical sales data of the accessory;
determining a safety stock coefficient of the part based on the probability distribution form and a preset spot goods satisfaction rate target;
and obtaining the safety stock based on the safety stock coefficient and the mean square error of the historical sales data.
Optionally, the fitting a probability distribution shape of historical sales data of the accessory includes:
respectively fitting a plurality of different probability distribution forms according to the historical sales data;
measuring the distance between the actual distribution of the historical sales data and each probability distribution form through KL divergence;
and taking the probability distribution form with the minimum distance as the probability distribution form of the accessory.
In a third aspect of the present invention, there is provided a method for managing spare parts of construction machinery, the method comprising:
obtaining the predicted demand of the accessory based on the prediction method; obtaining turnover inventory based on the predicted demand and stock cycle of the parts; and obtaining the warehouse-dividing spare goods quantity of the accessories in the warehouse dividing based on the turnover inventory and the demand proportional coefficient of the warehouse dividing.
Optionally, the method further includes: monitoring inventory data for the parts, and generating an early warning when one of the following conditions is met: 1) the storage age of the accessories reaches a set storage age threshold value; 2) the demand of the fittings of the sub-bins is larger than the prediction demand of the sub-bins.
In a fourth aspect of the present invention, there is provided a work machine accessory stock management system, the system including: a storage module and a calculation module, wherein,
the memory module includes:
the source data warehouse submodule is used for storing historical sales data and current inventory data of the accessories;
the characteristic and result data warehouse submodule is used for storing the processed data characteristics, the intermediate results and the final results of the accessory data;
the model and parameter data warehouse submodule is used for storing a demand prediction model, a probability distribution morphological model, a spectral clustering model, an additive regression model and parameters of the model;
the calculation module comprises: the demand forecasting submodule is used for obtaining a forecasting demand according to historical sales data of the accessories; the accessory stocking submodule is used for determining the stocking amount and actual inventory management of accessories; the accessory early warning submodule is used for monitoring inventory data of accessories and generating early warning when the preset conditions are met; the total warehouse forecast demand submodule is used for obtaining the stock quantity of the total warehouse; and the sub-warehouse prediction demand sub-module is used for obtaining the sub-warehouse prediction demand and the sub-warehouse reserve quantity of the sub-warehouse.
In a fifth aspect of the present invention, there is also provided a computer-readable storage medium having stored therein instructions, which when run on a computer, cause the computer to execute the aforementioned work machine part prediction method or work machine part stocking method.
The technical scheme provided by the invention has the following beneficial effects:
(1) in consideration of obvious difference of data distribution of the total warehouse and the sub-warehouse, the method adopts a machine learning algorithm to respectively predict the total warehouse demand for various accessories, adopts an additive regression model to predict the proportion coefficient of the sub-warehouse demand, and combines the two, so that the accuracy of the total warehouse demand prediction and the sub-warehouse demand prediction is higher.
(2) The automatic method is used for classifying the accessories, different machine learning models are trained aiming at different types of accessories, and the accuracy of accessory demand prediction is high. And for different types of accessories, more reasonable safety stock coefficients are calculated respectively, so that the accuracy of stock preparation is further improved, and the stock shortage rate is reduced.
(3) The early warning module can utilize the inventory information of the accessories to give an early warning when the inventory age of a certain accessory reaches a threshold value or give an early warning for replenishment when a certain warehouse is about to be out of stock.
Additional features and advantages of embodiments of the invention will be set forth in the detailed description which follows.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the embodiments of the invention without limiting the embodiments of the invention. In the drawings:
FIG. 1 is a flow chart illustrating a method for predicting an engineering machine component according to an embodiment of the present disclosure;
FIG. 2 is a flowchart illustrating a method for managing accessories of a construction machine according to an embodiment of the present disclosure;
fig. 3 is a schematic structural diagram of a work machine accessory management system according to an embodiment of the present invention.
Detailed Description
In addition, the embodiments of the present invention and the features of the embodiments may be combined with each other without conflict.
The following detailed description of embodiments of the invention refers to the accompanying drawings. It should be understood that the detailed description and specific examples, while indicating embodiments of the invention, are given by way of illustration and explanation only, not limitation.
Fig. 1 is a flowchart illustrating a method for predicting a work machine component according to an embodiment of the present disclosure. As shown in fig. 1, the present invention provides a method for predicting a stock of engineering machinery parts, including:
constructing data characteristics from historical sales data of the accessories; matching a corresponding demand prediction model according to the data characteristics; and obtaining the predicted demand of the accessory based on the historical sales data and the demand prediction model of the accessory.
So, through the characteristic analysis to the historical sales volume data of every kind of accessory, extract the service characteristics of every kind of accessory, select different demand prediction models for use through different characteristics, avoid different kinds of accessories to adopt same demand prediction model and the unsafe problem of demand prediction that brings, the reality of the corresponding model of utensil that this embodiment adopted is laminated the accessory more to this accuracy and the efficiency of stocking up of promotion demand prediction.
In one embodiment of the present invention, the constructing the data feature from the historical sales data of the accessory includes: taking each sales index of N sales indexes in historical sales data of the accessories as a dimension; extracting sales data of each sales index of the accessories in a period of historical time, and processing the sales data into a characteristic value by adopting a sliding window algorithm; and obtaining a feature vector with N dimensions corresponding to the accessory as the data feature. The sales index can include the following three parts: the part of characteristics comprise monthly sales volume characteristics, sales proportion characteristics, change rate characteristics and the like; the host sales volume characteristics comprise host ownership volume, host volume in a three-pack period, host volume in different ages and the like; the host startup characteristics comprise host startup rate, host startup rates in different years, host working time and the like. Wherein the period of history may be 6 months, 12 months or 18 months. A characteristic construction link: constructing a sample window data sequence based on a sliding window with a fixed size, calculating high-order statistics (1 order to 4 orders) of each basic feature class according to the high-order statistics, constructing a multi-class multi-dimensional feature time sequence by combining original data, using the multi-class multi-dimensional feature time sequence as an input sample of a feature extraction unit, and obtaining a corresponding output value according to the input sample. And obtaining a characteristic value by each index through the same method, combining the N characteristic values into a characteristic vector, and finally obtaining the characteristic vector of the accessory, namely the data characteristic of the accessory.
In an embodiment provided by the present invention, the matching a corresponding demand prediction model according to the data features includes: inputting the data characteristics of the accessories into a spectral clustering model to obtain a classification result; the classification result is one of M preset results, and the M preset results all have corresponding demand prediction models; and obtaining a demand prediction model corresponding to the classification result. And classifying the accessories into M classes by adopting a spectral clustering method based on the data characteristics of the historical sales data of the accessories, and adding the class labels as the characteristics into the historical sales characteristics of the accessories. The clustering model training steps are as follows: (1) clustering the historical sales characteristics of the accessories of all historical months; (2) and saving the clustering model under the specified path. The clustering model is used as follows: (1) loading a model under a specified path; (2) loading historical sales features of the accessory; (3) giving the category of each accessory according to the clustering model; (4) and updating the model and saving the model to the specified path. The general method of spectral clustering is as follows, with specific details referring to the prior art: 1) selecting a proper similarity function and a proper clustering number M; 2) constructing a similar graph and an assigned adjacent matrix thereof; 3) calculating a Laplacian matrix of the similarity graph; 4) calculating eigenvectors corresponding to a plurality of first eigenvalues of the Laplacian matrix, and spelling out an eigenvector matrix U by taking the eigenvectors as columns; 5) according to one sample of each behavior of the matrix U, performing k-means clustering on the input sample to obtain M clusters; 6) and outputting clustering results and classifying.
In an embodiment provided by the present invention, the demand prediction model is a machine learning prediction model, including: a random forest model, a GBDT model and an XGboost model; the demand forecasting model is trained by using the data features constructed from historical sales data of the accessories. The demand forecasting algorithm integrates machine learning algorithms such as random forest, GBDT, XGboost and the like, and a Bayesian optimization method is adopted to automatically select an optimal model and parameters thereof. The training process is as follows: (1) loading historical sales characteristics of accessories, historical sales characteristics of a host and start-up characteristics of the host; (2) selecting a model and model parameters by using a Bayesian optimization algorithm according to the types of the accessories, wherein the selected model comprises the following steps: random forests, GBDT, XGboost, etc.; (3) and saving the optimal model to the directory named by the corresponding category. The process of predicting the next-period accessory demand by using the trained accessory demand prediction algorithm is as follows: (1) loading historical sales characteristics of accessories, historical sales characteristics of a host and start-up characteristics of the host; (2) loading a corresponding prediction model according to the accessory category; (3) and outputting the next-period prediction quantity of each accessory.
In one embodiment provided by the present invention, the method further comprises: based on the prediction demand, distributing and obtaining the bin-dividing prediction demand of each bin: according to the demand data of the sub-bins, an additive regression model is adopted to predict the demand proportion coefficient of the accessory in the sub-bins; and obtaining the sub-bin prediction demand corresponding to the sub-bin based on the prediction demand and the demand proportion coefficient. The main bins and the sub bins distributed at different places exist in an actual scene, the main bins and the sub bins have different characteristics, and the quantity, regularity and characteristics of the accessory demands distributed in the sub bins of each region are obviously weakened, so that the accessory demand of the sub bins is difficult to directly and accurately predict. Thus, after the total bin is predicted, the predicted demand of the total bin needs to be allocated to each of the sub-bins. According to the method, according to the warehouse-dividing historical warehouse-out data and the regional construction data, warehouse-dividing warehouse-out characteristics and regional construction heat characteristics are extracted, and then the requirement proportion coefficient of the accessory in each warehouse is predicted by adopting an additive regression model. An Additive Model (Additive Model) is a non-parametric Model, and if the two-dimensional scatter diagram is generalized as a simple linear regression Model, the Additive Model is generalized as a multiple regression Model. Additive models are very flexible because they do not require the assumption of some functional form as do parametric models, as long as the effect of the predictor variables on the response variables is independent, also referred to as additive assumptions. The additive regression model is adopted to reduce the linearity in the prediction and improve the accuracy of the prediction. In this embodiment, the calculation method for obtaining the bin-based prediction demand corresponding to the bin based on the prediction demand and the demand proportional coefficient is as follows: and dividing the bin to predict the demand, namely predicting the demand multiplied by a demand proportional coefficient.
Fig. 2 is a schematic flow chart of a method for managing engineering machine parts according to an embodiment of the present invention, and as shown in fig. 2, in an embodiment of the present invention, a method for managing spare parts of an engineering machine is further provided, where the method includes: obtaining the predicted demand of the accessory based on the prediction method; obtaining a turnover storage inventory based on the predicted demand and delivery cycle of the part; and obtaining the stock quantity of the parts based on the turnover stock and the safety stock. In the prior art, a three-box inventory management model is mostly adopted in the management of the inventory, namely, turnover inventory, turnover reserve inventory and safety inventory are adopted for inventory management in the inventory management. Wherein the turnover inventory meets daily requirements, the turnover inventory guarantees delivery date requirements, and the safety inventory is used to offset out-of-stock risks caused by demand forecast deviations and delivery delays. The calculation steps are as follows: forecasting the demand quantity (per period) of the total warehouse accessories based on a total warehouse demand forecasting module, multiplying the demand quantity of the total warehouse accessories per period by the delivery cycle, and giving an accessory turnover storage stock; the formula for calculating the spare capacity is as follows: spare part A is spare part A turnover stock + part A safety stock.
On the basis of the above embodiment, the safety stock is obtained by the following steps: fitting a probability distribution shape of historical sales data of the accessory; determining a safety stock coefficient of the part based on the probability distribution form and a preset spot goods satisfaction rate target; and obtaining the safety stock based on the safety stock coefficient and the mean square error of the historical sales data. The calculation steps are as follows: that is, for each type of part, based on the fitted distribution and the specified stock-in-stock satisfaction rate target (preset value, e.g., 95%), a safety stock coefficient for the corresponding part is determined, and then multiplied by the mean square error of the historical sales feature as a safety stock value.
On the basis of the foregoing embodiment, the fitting a probability distribution shape of the historical sales data of the accessory includes: respectively fitting a plurality of different probability distribution forms according to the historical sales data; measuring the distance between the actual distribution of the historical sales data and each probability distribution form through KL divergence; and taking the probability distribution form with the minimum distance as the probability distribution form of the accessory. The different probability distribution forms here include: binomial distribution, gaussian distribution, normal distribution, gaussian mixture distribution, and the like. And fitting the historical sales data into a plurality of possible probability distribution forms, and evaluating by adopting KL divergence to obtain the best fitting probability distribution form so as to improve the accuracy of data prediction.
In the method for managing spare parts of engineering machinery provided in the embodiment, the forecast demand of the parts is obtained based on the forecast method; obtaining turnover inventory based on the predicted demand and stock cycle of the parts; and obtaining the warehouse-dividing spare goods quantity of the accessories in the warehouse dividing based on the turnover inventory and the demand proportional coefficient of the warehouse dividing. The present embodiment is mainly applied to binning, and is related to the aforementioned demand proportionality coefficient. The calculation process is as follows: multiplying the demand quantity of the parts in the total warehouse per period by the stock cycle, and giving out turnover and inventory of the parts; all stock in the turnover inventory of accessory arrives each branch storehouse, and the accessory stock volume of concrete each branch storehouse is: the spare goods quantity of each sub-bin of the part A is equal to the part A turnover inventory multiplied by the part demand proportion coefficient of each sub-bin of the part A, wherein the demand proportion coefficient is determined by the method.
The invention further provides early warning management of the inventory parts in one implementation mode. The method further comprises the following steps: monitoring inventory data for the parts, and generating an early warning when one of the following conditions is met: 1) the storage age of the accessories reaches a set storage age threshold value; for preventing long-term stagnation of certain components in inventory; 2) the demand of the fittings of the sub-warehouse is larger than the forecast demand of the sub-warehouse, which is used for preventing the shortage of goods caused by the conventional forecast.
Fig. 3 is a schematic structural diagram of a work machine accessory management system according to an embodiment of the present invention. As shown in fig. 3, in an embodiment provided by the present disclosure, a system for managing spare parts of a construction machine includes: a storage module and a calculation module, wherein,
the memory module includes: the source data warehouse submodule is used for storing historical sales data and current inventory data of the accessories; the characteristic and result data warehouse submodule is used for storing the processed data characteristics, the intermediate results and the final results of the accessory data; the model and parameter data warehouse submodule is used for storing a demand prediction model, a probability distribution morphological model, a spectral clustering model, an additive regression model and parameters of the model; the calculation module comprises: the demand forecasting submodule is used for obtaining a forecasting demand according to historical sales data of the accessories; the accessory stocking submodule is used for determining the stocking amount and actual inventory management of accessories; the accessory early warning submodule is used for monitoring inventory data of accessories and generating early warning when the preset conditions are met; or the system also comprises a sub-module for predicting the demand of the separated bins, which is used for obtaining the predicted demand of the separated bins and the spare goods amount of the separated bins.
The system mainly comprises an accessory demand prediction module, an accessory stock module, an accessory early warning module, a data storage module and the like. The accessory demand prediction module, the accessory stock module and the accessory early warning module are calculation modules.
The memory module includes: a source data warehouse submodule, a feature and result data warehouse submodule, and a model and parameter data warehouse submodule. The source data warehouse submodule is used for storing historical sales data, host startup data, host archive data and host sales data of accessories; the characteristic and result data warehouse submodule is used for storing part prediction characteristic engineering data, prediction structure data, warehouse-dividing prediction data, early warning result data, safety stock coefficients, safety stock and the like; the model and parameter warehouse (i.e. the model library in the graph) stores data such as model paths, model index data, model parameters, etc.
The calculation module comprises: and the accessory demand prediction submodule is used for clustering the accessories by using a spectral clustering method and performing demand prediction on the accessories of each category by adopting a corresponding accessory demand prediction algorithm. And the total warehouse sales volume prediction sub-module predicts the sales volume of the next-period accessories by using the corresponding optimal model according to the data construction characteristics of the accessory historical sales data, the host startup data and the like. The sub-warehouse demand forecasting sub-module is used for constructing characteristics of historical sales data corresponding to each accessory through warehouse distribution and adopting a warehouse distribution additive regression forecasting model to forecast the coefficient of each warehouse distribution demand in the total demand at the next period; and combining the total bin prediction demand and the bin division prediction demand coefficient to give the prediction demand of each bin division. And the accessory stock module is used for giving the final stock quantity by combining the forecast quantity of the accessory demand forecasting module and an improved three-box inventory management strategy. The accessory early warning module comprises a stay stock early warning and out-of-stock early warning submodule: the stay inventory early warning submodule is used for carrying out inventory age calculation on accessories in an inventory list and giving an early warning when the inventory age reaches a specified threshold value; and the out-of-stock early warning sub-module is used for early warning when the demand of the sub-warehouse accessories is greater than the forecast quantity.
In an embodiment of the present invention, a computer-readable storage medium is further provided, which stores instructions that, when executed on a computer, cause the computer to perform the aforementioned method for predicting a work machine component or the method for stocking work machine components. The running entity of the method is a server.
The embodiment of the invention establishes an engineering machinery accessory stock model and a system, the system module mainly comprises an accessory demand prediction module, an accessory stock module, an accessory early warning module, a data storage module and other functional modules, and simultaneously provides a stock model combining a machine learning prediction method with a three-box stock management strategy, thereby effectively improving the stock accuracy, improving the spot stock satisfaction rate, providing a linear weighted proportion prediction model of a total bin accessory distribution center bin, and effectively improving the stock accuracy of each center bin. The method has the advantages of high accuracy, convenient realization and wide application scene.
Although the embodiments of the present invention have been described in detail with reference to the accompanying drawings, the embodiments of the present invention are not limited to the details of the above embodiments, and various simple modifications can be made to the technical solutions of the embodiments of the present invention within the technical idea of the embodiments of the present invention, and the simple modifications all belong to the protection scope of the embodiments of the present invention.
It should be noted that the various features described in the above embodiments may be combined in any suitable manner without departing from the scope of the invention. In order to avoid unnecessary repetition, the embodiments of the present invention do not describe every possible combination.
Those skilled in the art will understand that all or part of the steps in the method according to the above embodiments may be implemented by a program, which is stored in a storage medium and includes several instructions to enable a single chip, a chip, or a processor (processor) to execute all or part of the steps in the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In addition, any combination of various different implementation manners of the embodiments of the present invention is also possible, and the embodiments of the present invention should be considered as disclosed in the embodiments of the present invention as long as the combination does not depart from the spirit of the embodiments of the present invention.
Claims (11)
1. A method for predicting the stock of engineering machinery parts is characterized by comprising the following steps:
constructing data characteristics from historical sales data of the accessories;
matching a corresponding demand prediction model according to the data characteristics;
and obtaining the predicted demand of the accessory based on the historical sales data and the demand prediction model of the accessory.
2. The prediction method of claim 1, wherein the constructing data features from historical sales data for accessories comprises:
taking each sales index of N sales indexes in historical sales data of the accessories as a dimension;
extracting sales data of each sales index of the accessories in a period of historical time, and processing the sales data into a characteristic value by adopting a sliding window algorithm;
and obtaining an N-dimensional feature vector corresponding to the accessory as the data feature.
3. The prediction method of claim 2, wherein said matching a corresponding demand prediction model based on said data characteristics comprises:
inputting the data characteristics of the accessories into a spectral clustering model to obtain a classification result; the classification result is one of M preset results, and the M preset results all have corresponding demand prediction models;
and obtaining a demand prediction model corresponding to the classification result.
4. The prediction method of claim 3, wherein the demand prediction model is a machine learning prediction model comprising: a random forest model, a GBDT model and an XGboost model;
the demand forecasting model is trained by using the data features constructed from historical sales data of the accessories.
5. The prediction method according to any one of claims 1 to 4, wherein the prediction method further comprises: based on the prediction demand, distributing and obtaining the bin-dividing prediction demand of each bin:
according to the demand data of each sub-bin, an additive regression model is adopted to predict the demand proportion coefficient of the accessory in each sub-bin;
and obtaining the sub-bin prediction demand corresponding to each sub-bin based on the prediction demand and the demand proportion coefficient of the accessories in each sub-bin.
6. A method for managing spare parts of engineering machinery, which is characterized by comprising the following steps:
obtaining a predicted demand of the accessory based on the prediction method of any one of claims 1 to 4;
obtaining a turnover storage inventory based on the predicted demand and delivery cycle of the part;
and obtaining the stock quantity of the parts based on the turnover stock and the safety stock.
7. The method of managing according to claim 6, wherein the safety stock is obtained by:
fitting a probability distribution shape of historical sales data of the accessory;
determining a safety stock coefficient of the part based on the probability distribution form and a preset spot goods satisfaction rate target;
and obtaining the safety stock based on the safety stock coefficient and the mean square error of the historical sales data.
8. The method of managing of claim 7, wherein said fitting a probability distribution shape of historical sales data for the part comprises:
respectively fitting a plurality of different probability distribution forms according to the historical sales data;
measuring the distance between the actual distribution of the historical sales data and each probability distribution form through KL divergence;
and taking the probability distribution form with the minimum distance as the probability distribution form of the accessory.
9. A method for managing spare parts of engineering machinery, which is characterized by comprising the following steps:
obtaining a predicted demand for the part based on the prediction method of claim 5;
obtaining turnover inventory based on the predicted demand and stock cycle of the parts;
and obtaining the warehouse-dividing spare goods quantity of the accessories in the warehouse dividing based on the turnover inventory and the demand proportional coefficient of the warehouse dividing.
10. The management method according to claim 6 or 9, characterized in that the management method further comprises: monitoring inventory data for the parts, and generating an early warning when one of the following conditions is met:
1) the storage age of the accessories reaches a set storage age threshold value;
2) the demand of the fittings of the sub-bins is larger than the prediction demand of the sub-bins.
11. A work machine accessory stock management system, the system comprising: the device comprises a storage module and a calculation module;
the memory module includes:
the source data warehouse submodule is used for storing historical sales data and current inventory data of the accessories;
the characteristic and result data warehouse submodule is used for storing the processed data characteristics, the intermediate results and the final results of the accessory data;
the model and parameter data warehouse submodule is used for storing a demand prediction model, a probability distribution morphological model, a spectral clustering model, an additive regression model and parameters of the model;
the calculation module comprises:
the demand forecasting submodule is used for obtaining a forecasting demand according to historical sales data of the accessories;
the accessory stocking submodule is used for determining the stocking amount and actual inventory management of accessories;
the accessory early warning submodule is used for monitoring inventory data of accessories and generating early warning when the preset conditions are met;
the total warehouse forecast demand submodule is used for obtaining the stock quantity of the total warehouse;
and the sub-warehouse prediction demand sub-module is used for obtaining the sub-warehouse prediction demand and the sub-warehouse reserve capacity of each sub-warehouse.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010633731.5A CN113962416A (en) | 2020-07-02 | 2020-07-02 | Engineering machinery part stock prediction method, management method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010633731.5A CN113962416A (en) | 2020-07-02 | 2020-07-02 | Engineering machinery part stock prediction method, management method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113962416A true CN113962416A (en) | 2022-01-21 |
Family
ID=79459390
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010633731.5A Pending CN113962416A (en) | 2020-07-02 | 2020-07-02 | Engineering machinery part stock prediction method, management method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113962416A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114118637A (en) * | 2022-01-28 | 2022-03-01 | 树根互联股份有限公司 | Method and device for building prediction model of accessory demand and computer equipment |
| CN117151595A (en) * | 2023-10-31 | 2023-12-01 | 苏州极易科技股份有限公司 | Commodity inventory management method, equipment and storage medium |
-
2020
- 2020-07-02 CN CN202010633731.5A patent/CN113962416A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114118637A (en) * | 2022-01-28 | 2022-03-01 | 树根互联股份有限公司 | Method and device for building prediction model of accessory demand and computer equipment |
| CN117151595A (en) * | 2023-10-31 | 2023-12-01 | 苏州极易科技股份有限公司 | Commodity inventory management method, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108320171B (en) | Hot-sold commodity prediction method, system and device | |
| You et al. | A decision-making framework for precision marketing | |
| Yang et al. | Evaluation of robustness of supply chain information-sharing strategies using a hybrid Taguchi and multiple criteria decision-making method | |
| US7213007B2 (en) | Method for forecasting using a genetic algorithm | |
| US7853462B2 (en) | Optimizing an inventory of a supply chain | |
| CN102184507A (en) | Warehouse replenishing method | |
| Kot et al. | Theory of inventory management based on demand forecasting | |
| EP2273431B1 (en) | Model determination system | |
| JP7105336B2 (en) | smart supply chain system | |
| US20130166350A1 (en) | Cluster based processing for forecasting intermittent demand | |
| CN111415193B (en) | Method for calculating class sales promotion degree based on store similarity of associated commodities | |
| CN113962416A (en) | Engineering machinery part stock prediction method, management method and system | |
| EP3876177A1 (en) | System and method for retail price optimization | |
| CN116051004B (en) | Intelligent management method, system and medium based on big data | |
| Brito et al. | An iterative local search approach applied to the optimal stratification problem | |
| CN112884393A (en) | Inventory replenishment method and device and electronic equipment | |
| Aktunc et al. | Inventory control through ABC/XYZ analysis | |
| US9324026B2 (en) | Hierarchical latent variable model estimation device, hierarchical latent variable model estimation method, supply amount prediction device, supply amount prediction method, and recording medium | |
| Lahrichi et al. | A first MILP model for the parameterization of Demand-Driven MRP | |
| KR101860258B1 (en) | Supply Chain based on Manufacturing System Optimum Design System and Method Using Markov Decision Process and Machine Learning Algorism | |
| CN114186903A (en) | Warehouse product selection method and device, computer equipment and storage medium | |
| Lee et al. | A Predictive Model for Forecasting Spare Parts Demand in Military Logistics | |
| CN113780913A (en) | Method and device for generating safety stock information | |
| CN116342036B (en) | Full-automatic warehouse operation control method and system for WMS | |
| CN113469596A (en) | Intelligent supply chain system and server platform |
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 |