CN115689443A - Supply chain management method and system based on volatility clustering and block chain - Google Patents
Supply chain management method and system based on volatility clustering and block chain Download PDFInfo
- Publication number
- CN115689443A CN115689443A CN202211298792.6A CN202211298792A CN115689443A CN 115689443 A CN115689443 A CN 115689443A CN 202211298792 A CN202211298792 A CN 202211298792A CN 115689443 A CN115689443 A CN 115689443A
- Authority
- CN
- China
- Prior art keywords
- supply chain
- clustering
- data
- volatility
- link
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Abstract
The invention discloses a supply chain management method and a system based on volatility clustering and block chains, wherein the method comprises the following steps: constructing a block chain according to the relation of each link in the supply chain, wherein each link in the supply chain is respectively used as a block chain node; when the target block chain node acquires real-time supply chain data, inputting the real-time supply chain data and historical supply chain data into a pre-constructed volatility clustering prediction model, and clustering volatility of the input supply chain data by using a neural network model by the volatility clustering prediction model to predict volatility of a supply chain state; and the target block chain node controls to execute the intelligent contract when meeting the corresponding preset fluctuation state condition according to the obtained fluctuation prediction result. The invention can fully integrate the supply chain, improve the transaction processing efficiency, and has the advantages of simple realization method, high automation degree, stability, reliability and the like.
Description
Technical Field
The invention relates to the technical field of supply chain management, in particular to a supply chain management method and system based on volatility clustering and a block chain.
Background
The supply chain is around the core enterprise, starting with the kit, making intermediate products and final products, and finally sending the products to the consumer by the sales network, connecting the supplier, manufacturer, distributor and end user into a whole functional network chain structure. Supply chain management is all that is required to optimize supply chain operation, starting with procurement, to meet the end customer, at minimal cost. Supply chain management is the coordination of resources inside and outside an enterprise to meet consumer demand together. If the enterprises in each link of the supply chain are regarded as a virtual enterprise union, and any enterprise is regarded as a department in the virtual enterprise union, the internal management of the union is the supply chain management, and the composition of the union is dynamic, namely, the composition of the union changes at any time according to the needs.
In the prior art, enterprises of the supply chain are usually managed independently, and lack of integrity can cause imbalance of management and regulation, so that the problems of low transaction processing efficiency in the supply chain and the like are caused. The block chain (Blockchain) is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like, is a chain data structure formed by combining data blocks in a sequential connection mode according to a time sequence, is a distributed account book which is guaranteed to be untrustable and untrustworthy in a cryptographic mode, and has the characteristics of decentralization, information untrusty, autonomy and the like. Some practitioners propose to construct a supply chain management system by using a block chain, that is, a block chain node relationship is established according to a supply chain relationship to construct the block chain, and supply information corresponding to supply chain merchant information is uploaded based on the block chain node relationship, so that supply information analysis management can be performed on each supply chain merchant by using the block chain, and supply adjustment is performed according to a supply relationship analysis result.
However, the above-mentioned manner of implementing supply chain management by using a blockchain is to upload information by using each enterprise as a blockchain node, and does not concern the management control problem of each link of the supply chain, and the targets and requirements of different links are different, the execution of each link may directly affect the transaction efficiency of the whole supply chain, and the supply chain management by using each enterprise as a node still cannot fully integrate the supply chain, and the transaction processing efficiency still needs to be improved, and the stability and reliability of the transaction are also affected.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: aiming at the technical problems in the prior art, the invention provides a supply chain management method and system based on volatility clustering and block chains, which can fully integrate a supply chain, improve transaction processing efficiency, and have the advantages of simple implementation method, high automation degree, stability and reliability.
In order to solve the technical problems, the technical scheme provided by the invention is as follows:
a supply chain management method based on volatility clustering and block chain comprises the following steps:
constructing a block chain according to the relation of each link in the supply chain, wherein each link in the supply chain is respectively used as a block chain node;
when a target block chain node in a supply chain acquires real-time supply chain data, inputting the real-time supply chain data and historical supply chain data into a pre-constructed volatility clustering prediction model, and clustering volatility of the input supply chain data by using a neural network model by the volatility clustering prediction model to predict volatility of a supply chain state;
and the target block chain node controls to execute an intelligent contract when the corresponding preset fluctuation state condition is met according to the obtained fluctuation prediction result.
Furthermore, the volatility clustering prediction model is obtained by clustering input supply chain data by using a density-based noise application space clustering method, and the volatility clustering prediction model is obtained by carrying out unsupervised machine learning training on the supply chain data by using a neural network model in advance and adopting the density-based noise application space clustering method.
Further, the step of clustering the input supply chain data by applying the spatial clustering method to the noise based on the density comprises:
initializing a set of core objects
Initializing cluster number k =0, and initializing inaccessible sample set Γ = D, cluster partitioning
Searching all core objects, and updating a core object set omega;
randomly selecting a core object o from the core object set, initializing a current cluster core object queue Ω cur = { o }, initializing a class sequence number k = k +1, and initializing a current cluster sample set C k = { o }, update unaccessed set of samples Γ = Γ - { o };
if the current cluster core object queue
Then the current cluster C is determined k After generation, update cluster partition C = { C = { C = } 1 ,C 2 ,...,C k H, and update the core object set Ω = Ω -C k Otherwise, in the current cluster core object queue omega cur Taking out a core object o', passing through the neighborhood distanceThe threshold e finds all e neighborhood subsamples set N e (o '), and let Δ = N e (o') # Γ, update the current cluster sample set C k ＝C k U Δ, and updating the set of unvisited samples Γ = Γ - Δ, updating Ω ≧ Δ cur ＝Ω cur U (Δ ≈ Ω) -o' resulting in cluster division C = { C = 1 ,C 2 ,...,C k }。
Further, the step of searching all the core objects includes:
searching a neighborhood subsample set N (xj) of the sample xj in a distance measurement mode, and if the number of samples of the neighborhood subsample set meets the condition that the sample is in the range of the element of the N (x) j ) If | ≧ MinPts, sample x is set j Adding a core object sample set: Ω = Ω & { x + j }。
Further, before inputting the real-time supply chain data and the historical supply chain data into the pre-constructed fluctuating clustering prediction model, the method further comprises the steps of sequentially carrying out data cleaning, data enhancement and data preprocessing on the input supply chain data, wherein the data preprocessing comprises the step of carrying out normalization processing on the supply chain data.
Furthermore, the block chain comprises a purchasing link node, a transportation link node, a storage production link node, a production link node and a sales link node, wherein the purchasing link node and the sales link node predict the volatility of the input supply chain data through the volatility clustering prediction model, and control and execute the intelligent contract according to the volatility prediction result.
Furthermore, the purchasing link node also comprises a step of evaluating the current stocked raw material resources and the product production rate when the raw materials are purchased, inputting the data of the purchasing link into a volatility clustering prediction model for prediction according to the evaluation result so as to control the execution of a purchasing intelligent contract in a low fluctuation interval with the volatility value smaller than a preset threshold value, and a step of evaluating whether the raw material stock is sufficient and whether the product finished product stock is overstocked, wherein if the raw material stock is smaller than the preset stock threshold value, the step of purchasing the raw materials is carried out; and if the backlog quantity of the finished product inventory exceeds a preset backlog threshold, entering a product selling step, evaluating the finished product inventory by the sales product link node before the finished product is sold, and sending sales link data into a prediction model according to an evaluation result so as to control to execute an intelligent sales contract in a low fluctuation interval with a fluctuation value smaller than the preset threshold.
A supply chain management system based on fluctuating clustering and blockchains, comprising:
the block chain is constructed according to the relation of each link in the supply chain, and each link in the supply chain is used as a block chain node;
the volatility prediction module is used for inputting the real-time supply chain data and the historical supply chain data into a volatility clustering prediction model which is constructed in advance when the real-time supply chain data are obtained by a target block chain node in the supply chain, and clustering the volatility of the input supply chain data by using a neural network model by using the volatility clustering prediction model so as to predict the volatility of the state of the supply chain;
and the node control module is used for controlling the target block link points to execute the intelligent contract when the corresponding preset fluctuation state conditions are met according to the obtained fluctuation prediction result.
Further, the volatility prediction module specifically includes:
the supply chain data collection module is used for collecting historical supply chain data and real-time supply chain data required by the prediction model;
the supply chain data processing module is used for cleaning, enhancing and preprocessing the collected supply chain data and then outputting the supply chain data, and extracting and training the characteristics of the supply chain data;
the fluctuation clustering prediction module is used for outputting the cleaned, enhanced and preprocessed real-time supply chain data to a fluctuation clustering prediction model and outputting a fluctuation prediction result;
the supply chain data processing module specifically comprises a data cleaning module, a data enhancing module, a data preprocessing module, a feature engineering module for feature screening and a training module for executing model training, which are connected in sequence.
A computer-readable storage medium storing a computer program which, when executed, implements the method as described above.
Compared with the prior art, the invention has the advantages that:
1. the invention constructs a block chain based on a block chain architecture according to the relation of each link in a supply chain, each link in the supply chain is respectively used as a block chain node to store transaction information in the whole process into the block chain storage, and meanwhile, a fluctuation cluster prediction model constructed based on fluctuation cluster and neural network is combined to carry out fluctuation cluster prediction on the supply chain data of the block chain node points required to be controlled, so that the future fluctuation change of the supply chain state can be predicted timely and accurately, the intelligent contract of an execution node is controlled by using a fluctuation prediction result, the supply chain can be fully integrated, the transaction efficiency of the whole supply chain is greatly improved, and the stable reliability of transaction in the supply chain is improved.
2. The invention further constructs a prediction model by introducing a noise application space clustering method based on density, can improve the convergence speed in the training process, has higher tolerance on abnormal values in the convergence process, is less influenced by the abnormal values, does not need to specify the number of clusters in the initial period, and can realize the convergence on the clusters with any shapes and sizes, thereby ensuring the accuracy and the efficiency of predicting the supply chain state volatility.
3. According to the invention, the purchasing link, the production link and the selling link are abstracted into the state machine, and the intelligent contract is controlled based on the prediction result of the volatility clustering prediction model, so that the automatic execution of the related intelligent contract can be triggered only when a certain volatility condition is met, the volatility prediction result of the data of the supply chain can be effectively utilized to further integrate the supply chain, the transaction processing efficiency is improved, and the digitization and automation degree of the whole supply chain is greatly improved.
4. The invention further realizes the prediction of the trend of a period of time in the future by inputting the historical or real-time supply chain data into the fluctuation clustering prediction model, and arranges the corresponding execution activities to a low fluctuation period by utilizing the control of the prediction result, thereby effectively improving the stability and reliability of the supply chain and reducing the interference of uncertain factors.
Drawings
Fig. 1 is a schematic diagram illustrating a supply chain management method based on fluctuating clustering and block chains according to an embodiment.
Fig. 2 is a schematic diagram illustrating a principle of clustering by applying a spatial clustering algorithm to density-based noise according to this embodiment.
Detailed Description
The invention is further described below with reference to the drawings and specific preferred embodiments of the description, without thereby limiting the scope of protection of the invention.
Clustering is one of the most basic tasks in machine learning and computer vision, and aims to group similar patterns into the same class and different patterns into different classes. The fluctuation clustering is also continuity of fluctuation rate, and the fluctuation of data in each link of the supply chain generally has continuity, for example, if the fluctuation rate lasting for a period of time is high, the probability that the fluctuation rate at a subsequent time is still high is high, correspondingly, if the fluctuation rate lasting for a period of time is low, the fluctuation rate at a following period of time is also low, the future fluctuation can be predicted based on the fluctuation clustering, interference of isolated points and noise can be reduced in the prediction process, and the prediction accuracy is improved. The neural network has the advantages of self-adaptability, self-learning and the like, normal data can be extracted from the neural network-based data analysis through the self-learning, a large amount of data does not need to be accessed, and the neural network has good anti-jamming capability.
According to the invention, on the basis of a block chain architecture, the characteristics of the volatility clustering and the neural network are integrated, a block chain is constructed according to the relation of each link in a supply chain, each link in the supply chain is respectively used as a block chain node, so that transaction information in the whole process is stored in block chain storage, huge supply chain data can be stored safely and stably for a long time, the traceability of all transaction information is ensured, then the influence of a time sequence on data prediction is considered, a volatility clustering prediction model constructed on the basis of the volatility clustering and the neural network is used for carrying out volatility clustering prediction on the supply chain data of the block chain node points to be controlled, the future fluctuation change of the supply chain state can be predicted accurately in time, an intelligent contract of an execution node is controlled by using a volatility prediction result, the supply chain can be fully integrated, the transaction efficiency of the whole supply chain is greatly improved, and the stability and reliability of transaction in the supply chain is improved.
As shown in fig. 1, the steps of the supply chain management method based on volatility clustering and block chain in this embodiment include:
s01, constructing a block chain according to the relation of each link in the supply chain, wherein each link in the supply chain is used as a block chain node;
s02, when a target block chain node in a supply chain acquires real-time supply chain data, inputting the real-time supply chain data and historical supply chain data into a pre-constructed volatility clustering prediction model, and clustering volatility of the input supply chain data by using a neural network model by the volatility clustering prediction model to predict volatility of a supply chain state;
and S03, controlling the target block chain node to execute an intelligent contract when the corresponding preset fluctuation state condition is met according to the obtained fluctuation prediction result.
In this embodiment, the supply chain data may specifically be supply chain resource related data such as price data, transaction data, and storage data, so as to implement a series of supply chain resource transactions such as purchase, transportation, storage, production, and sale of raw materials. The price data may be, for example, a price of a raw material, a transportation cost price, a bitcoin price, a foreign exchange price, a cost sale price, and the like, and may be configured according to actual needs. The supply chain related historical data and the real-time data are acquired in a crawler mode.
In the embodiment, the fluctuation cluster prediction model clusters the input supply chain data by using a density-based noise application spatial clustering method, and the fluctuation cluster prediction model is obtained by carrying out unsupervised machine learning training on the supply chain data by using a neural network model in advance and adopting the density-based noise application spatial clustering method. In the embodiment, a density-based noise application space clustering method is introduced, and a prediction model is built by embedding autoregressive condition variance with a specific time sequence, namely, an unsupervised machine learning clustering algorithm of the density-based noise application space clustering is adopted in the model, so that the convergence speed in the training process can be improved, the tolerance to abnormal values in the convergence process is higher, the influence by the abnormal values is smaller, the number of clusters does not need to be specified initially, the convergence to clusters of any shape and size can be realized, and the accuracy and the efficiency of chain state fluctuation prediction can be considered.
The density-based noise applies a spatial clustering method by dividing an area having a sufficient density into clusters, and finding clusters of an arbitrary shape in a spatial database having noise, wherein a cluster is defined as a maximum set of density-connected points, as shown in fig. 2. Before clustering, the radius of the cluster and the number of points that can be enclosed by the cluster need to be specified. The step of clustering the input supply chain data by using the density-based noise application spatial clustering method in this embodiment specifically includes:
s201, initializing a core object set
Initializing cluster number k =0, and initializing set of unaccessed samples Γ = D, cluster partitioning
S202, searching all core objects, and updating a core object set omega;
s203, in the core object set, a core object o is randomly selected, and a current cluster core object queue omega is initialized cur = o, initializing class index k = k +1 and initializing current cluster sample set C k = { o }, update unvisited sample set Γ = Γ - { o };
s204, if the current cluster core object queue
Judging that the current cluster Ck is generated completely, and updating cluster division C = { C = { (C) } 1 ,C 2 ,.., ck }, and updating the core object set Ω = Ω -Ck, otherwise, proceeding to step S205;
s205, in the current cluster core object queue omega cur Taking out a core object o ', finding out all the belonged neighborhood subsample sets N belonged (o ') through the neighborhood distance threshold value belonged, and updating the current cluster sample set C by enabling delta = N belonged (o ') # k ＝C k U Δ, and updating the set of unvisited samples Γ = Γ - Δ, updating Ω ≧ Δ cur ＝Ω cur U (Δ ≈ Ω) -o' resulting in cluster division C = { C = 1 ,C 2 ,...,Ck}。
The specific step of searching all core objects in step S202 includes:
searching a neighborhood subsample set N e (x) of the sample xj in a distance measurement mode j ) If the number of samples in the neighborhood subsample set satisfies | N ∈ (x) j ) If | ≧ MinPts, add sample xj to the core object sample set: Ω = Ω & { x + j }。
In a specific application embodiment, a noise application space clustering algorithm based on density is constructed, and the following are input: sample set D = (x) 1 ,x 2 ,...,x m ) Neighborhood parameters (e, minPts), and sample distance metric. And (3) outputting: and C, cluster division. The algorithm execution flow comprises the following steps: initializing a set of core objects
Initializing cluster number k =0, initializing unvisited sample set Γ = D, clustering
For j =1,2.. M, all core objects Ω = Ω & { x ] are found as in step S202 above j And then, according to the steps S203 and S204, in the core object set omega, randomly selecting a core object o to perform the current clustering C k Generating, and taking out a core object o 'from the current cluster core object queue Ω cur according to the step S205, finding out all belongings-neighborhood child sample set N belongings (o'),finally obtaining an output result cluster division C = { C = { (C) 1 ,C 2 ,...,C k }。
When the traditional BP neural network is used for trend prediction, the method is very sensitive to weight selection and is very easy to converge to local minimum, and meanwhile, the problems of over-fitting or under-fitting exist in the training process. The traditional FCM clustering algorithm has the following defects: on the one hand, constraints tend to make it sensitive to outliers and noise; on the other hand, the initial clustering center is sensitive and is not easy to converge to the global optimum because the algorithm is an iterative descent algorithm. In the embodiment, the prediction model is built by integrating the neural network model and the clustering algorithm, and the unsupervised machine learning clustering algorithm of noise application space clustering based on density is adopted in the neural network model, so that the characteristic influence caused by a time sequence can be introduced into the machine learning model, the problems of prediction by using the neural network model and the conventional clustering algorithm in the prior art are effectively solved, and the accuracy of supply chain data volatility prediction is improved.
In this embodiment, before inputting the real-time supply chain data and the historical supply chain data into the pre-constructed volatility clustering prediction model, the method further includes the steps of sequentially performing data cleaning, data enhancement and data preprocessing on the input supply chain data, wherein the data cleaning is to perform recheck and check on the data to remove repeated information, errors and other information, the data enhancement is to increase the number and diversity of training samples to improve the robustness of the model by adopting a data enhancement algorithm under the condition that collected supply chain data samples are limited, and the data preprocessing is to perform normalization processing on the supply chain data and the like. The fluctuating clustering prediction model also comprises the steps of feature selection, dimension reduction and the like in the training process so as to screen out required data features, and the effect of model training is further improved. By performing data feature screening and data enhancement in the training process, the over-fitting or under-fitting condition in the model training process can be reduced. And inputting the cleaned and preprocessed data of the supply chain data into a fluctuating clustering prediction model for training, wherein historical data is used as training data for training to obtain a prediction model, and real-time data is used as real-time test data and is input into the prediction model to obtain a real-time prediction result. In the training process, the supply chain data are subjected to unsupervised machine learning by adopting density-based noise application space clustering, and the algorithm is used for dividing the area with sufficient density into clusters, so that the correlation among time sequences and the result of fluctuation clustering are accurately calculated.
Referring to fig. 1, the block chain of this embodiment specifically includes a purchasing link node, a transportation link node, a warehousing production link node, a production link node, and a sales link node, where the purchasing link node, the production link node, and the sales link node are used as target block chain nodes, that is, the purchasing link node and the sales link node predict volatility of input supply chain data through a volatility clustering prediction model, and control the purchasing link node and the sales link node to execute an intelligent contract according to a volatility prediction result.
In this embodiment, the purchasing link node further includes evaluating current stocked raw material resources and a product production rate when raw materials are purchased, inputting data of the purchasing link into a volatility clustering prediction model for prediction according to an evaluation result to control to execute a purchasing intelligent contract in a low fluctuation interval with a volatility value smaller than a preset threshold, and the finished product production link further includes evaluating whether raw material stock is sufficient and whether product finished product stock is overstocked, and if the raw material stock is smaller than the preset stock threshold, the step of raw material purchasing is carried out; and if the backlog quantity of the finished product inventory exceeds a preset backlog threshold, entering a step of selling the products, evaluating the finished product inventory before the finished product is sold by the node of the link of selling the products, and sending the data of the link of selling the products into a prediction model according to an evaluation result so as to control the node of the link of purchasing the intelligent contract for selling to be executed in a low fluctuation interval with a fluctuation value smaller than the preset threshold. The trend of a period of time in the future is predicted by inputting historical or real-time supply chain data into the fluctuation clustering prediction model, and the corresponding execution activities are arranged to a low fluctuation period by utilizing the prediction result control, so that the stability and reliability of the supply chain can be improved, and the interference of uncertain factors can be reduced.
The embodiment abstracts a purchasing link, a production link and a selling link into a state machine, and controls the intelligent contract based on the prediction result of the volatility clustering prediction model, so that the automatic execution of related intelligent contracts can be triggered only when certain volatility conditions are met, the volatility prediction result of the supply chain data can be effectively utilized to further integrate the supply chain, the transaction processing efficiency is improved, the digitization and the automation degree of the whole supply chain are greatly improved, and then the goals of reducing the purchasing cost, improving the sales income and the like are realized.
The supply chain management system based on the volatility clustering and the block chain in the embodiment comprises:
the block chain is constructed according to the relation of each link in the supply chain, and each link in the supply chain is used as a block chain node;
the volatility prediction module is used for inputting the real-time supply chain data and the historical supply chain data into a volatility clustering prediction model which is constructed in advance when the real-time supply chain data are obtained by a target block chain node in the supply chain, and clustering the volatility of the input supply chain data by using a neural network model by using the volatility clustering prediction model so as to predict the volatility of the state of the supply chain;
and the node control module is used for controlling the target block link points to execute the intelligent contract when the corresponding preset fluctuation state conditions are met according to the obtained fluctuation prediction result.
The volatility prediction module of this embodiment specifically includes:
the supply chain data collection module is used for collecting historical supply chain data and real-time supply chain data required by the prediction model;
the supply chain data processing module is used for cleaning, enhancing and preprocessing the collected supply chain data and then outputting the supply chain data, and extracting and training the characteristics of the supply chain data;
and the fluctuation clustering prediction module is used for outputting the cleaned, enhanced and preprocessed real-time supply chain data to a data input fluctuation clustering prediction model and outputting a fluctuation prediction result.
As shown in fig. 1, the supply chain data processing module specifically includes a data cleaning module for cleaning supply chain data, a data enhancement module for enhancing the supply chain data, a data preprocessing module for normalizing the supply chain data, a feature engineering module for performing feature screening, and a training module for performing model training, and the training module specifically performs real-time training by applying a spatial clustering continued learning algorithm to noise based on density, so as to realize the calculation of the correlation between time sequences and the result of the fluctuation clustering. The data cleaning module specifically reviews and verifies the data to remove information such as repetition and errors. Under the condition that the collected supply chain data samples are limited, the data enhancement module increases the number and diversity of training samples by using a data enhancement algorithm so as to improve the robustness of the model.
In this embodiment, the system further includes a purchasing evaluation module, a production evaluation module, and a sales evaluation module, which are respectively used to evaluate a purchasing link node, a production link node, and a sales link node, so as to implement control over execution of the intelligent contract in combination with a volatility prediction result. As mentioned above, the purchasing link node further comprises the steps of evaluating the current stocked raw material resources and the product production rate when the raw materials are purchased, inputting the data of the purchasing link into the volatility clustering prediction model for prediction according to the evaluation result so as to control the execution of the purchasing intelligent contract in the low fluctuation interval with the volatility value smaller than the preset threshold value, the finished product production link further comprises the steps of evaluating whether the raw material stock is sufficient and whether the finished product stock is overstocked, and if the raw material stock is smaller than the preset stock threshold value, the step of purchasing the raw materials is carried out; and if the backlog quantity of the finished product inventory exceeds a preset backlog threshold, entering a step of selling the products, evaluating the finished product inventory before the finished product is sold by the node of the link of selling the products, and sending the data of the link of selling the products into a prediction model according to an evaluation result so as to control the node of the link of purchasing the intelligent contract for selling to be executed in a low fluctuation interval with a fluctuation value smaller than the preset threshold.
In this embodiment, the system further includes an intelligent contract information storage module for storing intelligent contract information and an identity verification module for identity verification, so as to further improve the security and reliability of supply chain transactions.
The supply chain management system based on the volatility clustering and the block chain in this embodiment corresponds to the supply chain management method based on the volatility clustering and the block chain, and is not described in detail herein.
The present invention will be further described below by taking an example of implementing supply chain management based on the above system and the above method of the present invention in a specific application embodiment. The detailed steps of implementing supply chain management based on volatility cluster prediction and block chain in this embodiment are as follows:
(1) Step one, logging in a system and calling an identity authentication module to perform identity authentication.
(2) And secondly, starting a fluctuating clustering prediction module, wherein a supply chain data collection module is used for collecting historical supply chain data and real-time supply chain data required by the prediction model, and the data comprises price data, state data, other data related to transaction and the like. The supply chain data processing module is used for cleaning and preprocessing the supply chain data and then inputting the supply chain data into the fluctuation clustering prediction model, the prediction model adopts a density-based noise application space clustering continuous learning algorithm to carry out real-time training, and the correlation among time sequences and the result of fluctuation clustering are calculated. In the training process, the characteristic engineering module screens out better data characteristics from the original supply chain data through a series of engineering modes such as characteristic selection, dimension reduction and the like so as to improve the effect of model training. The supply chain data analysis model realizes the prediction of the trend of a future period of time by calling out historical or real-time supply chain data from the block chain storage module.
(3) And step three, a raw material purchasing link. Before the raw materials are purchased, a purchasing resource evaluation module is called to evaluate the current inventory raw material resources and the production rate of the product, then the step two is carried out, so that the related historical and real-time data of the purchasing links such as the price of the raw materials and the like are sent into a prediction model, a purchasing intelligent contract is automatically called in a low fluctuation period according to the prediction result to complete the purchasing transaction, and related information is stored in a block chain.
(4) And step four, in the finished product production link, before the production of the product, calling a production evaluation module to evaluate whether the stock of the raw materials is sufficient and whether the stock of the finished product is overstocked. If the raw material inventory is insufficient, turning to the step three to carry out a raw material purchasing step; if the stock of finished products is overstocked, the step five of selling the products is carried out. Otherwise, normal production activities are carried out.
(5) And step five, in the product selling link, calling a sale evaluation module before selling the finished products, evaluating the finished product inventory, sending relevant data of the selling link such as transportation expense price and the like into a prediction model, and controlling to ensure that the selling activity is arranged in a low fluctuation interval with higher price as much as possible according to a prediction result.
The present embodiment also provides a computer-readable storage medium storing a computer program, which when executed implements the method as described above.
The foregoing is considered as illustrative of the preferred embodiments of the invention and is not to be construed as limiting the invention in any way. Although the present invention has been described with reference to the preferred embodiments, it is not intended to be limited thereto. Therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical spirit of the present invention should fall within the protection scope of the technical scheme of the present invention, unless the technical spirit of the present invention departs from the content of the technical scheme of the present invention.
Claims (10)
1. A supply chain management method based on fluctuation clustering and block chains is characterized by comprising the following steps:
constructing a block chain according to the relation of each link in the supply chain, wherein each link in the supply chain is respectively used as a block chain node;
when a target block chain node in a supply chain acquires real-time supply chain data, inputting the real-time supply chain data and historical supply chain data into a pre-constructed volatility clustering prediction model, and clustering volatility of the input supply chain data by using a neural network model by the volatility clustering prediction model to predict volatility of a supply chain state;
and the target block chain node controls to execute an intelligent contract when the corresponding preset fluctuation state condition is met according to the obtained fluctuation prediction result.
2. The supply chain management method based on volatility clustering and block chain according to claim 1, wherein the volatility clustering prediction model uses a density-based noise application space clustering method to cluster the input supply chain data, and the volatility clustering prediction model is obtained by using a neural network model to perform unsupervised machine learning training on the supply chain data in advance by using the density-based noise application space clustering method.
3. The volatility-clustering and blockchain-based supply chain management method of claim 2, wherein the step of clustering input supply chain data using a spatial clustering method for density-based noise comprises:
initializing a core object set
Initializing cluster number k =0, and initializing set of unaccessed samples Γ = D, cluster partitioning
Searching all core objects, and updating a core object set omega;
randomly selecting a core object o in the core object set, and initializing a current cluster core object queue omega cur = o, initializing class index k = k +1 and initializing current cluster sample set C k = { o }, update unvisited sample set Γ = Γ - { o };
if the current cluster core object queue
Then the current cluster C is determined k After generation, update cluster partition C = { C = { C = } 1 ,C 2 ,...,C k And update the kernel pairLike set omega = omega-C k Otherwise, in the current cluster core object queue omega cur Taking out a core object o ', finding out all the belonged neighborhood subsample sets N belonged (o ') through the neighborhood distance threshold value belonged, and updating the current cluster sample set C by enabling delta = N belonged (o ') # k ＝C k U Δ, and updating the set of unvisited samples Γ = Γ - Δ, updating Ω ≧ Δ cur ＝Ω cur U (Δ ≈ Ω) -o' resulting in cluster division C = { C = 1 ,C 2 ,...,C k }。
4. The volatility cluster and blockchain-based supply chain management method of claim 3, wherein said step of searching all core objects comprises:
searching a neighborhood subsample set N (x) of the sample xj in a distance measurement mode j ) If the number of samples in the neighborhood subsample set satisfies | N ∈ (x) j ) If | ≧ MinPts, sample x is set j Adding a core object sample set: Ω = Ω & { x + j }。
5. The volatility clustering and blockchain-based supply chain management method according to claim 1, wherein before inputting the real-time supply chain data and the historical supply chain data into the volatility clustering prediction model constructed in advance, the method further comprises the steps of sequentially performing data cleaning, data enhancement and data preprocessing on the input supply chain data, and the data preprocessing comprises the step of normalizing the supply chain data.
6. The supply chain management method based on volatility clustering and block chains as claimed in any one of claims 1 to 5, wherein the block chains comprise a purchasing link node, a transportation link node, a storage production link node, a production link node and a sales link node, wherein the purchasing link node and the sales link node predict volatility of input supply chain data through the volatility clustering prediction model and control to execute intelligent contracts according to volatility prediction results.
7. The supply chain management method based on fluctuating clustering and block chains as claimed in claim 6, wherein the purchasing link node further comprises evaluating the current stocked raw material resources and the production rate of products when raw materials are purchased, inputting the data of the purchasing link into a fluctuating clustering prediction model for prediction according to the evaluation result so as to control the execution of a purchasing intelligent contract in a low fluctuation interval with a fluctuation value smaller than a preset threshold value, and the finished product production link further comprises evaluating whether the stock of the raw materials is sufficient and the stock of finished products is overstocked, and if the stock of the raw materials is smaller than the preset stock threshold value, the step of raw material purchasing is carried out; and if the backlog quantity of the finished product inventory exceeds a preset backlog threshold, entering a product selling step, evaluating the finished product inventory before the finished product is sold by the product selling link node, and sending sales link data into a prediction model according to an evaluation result so as to control the sales intelligent contract to be executed in a low fluctuation interval with a fluctuation value smaller than the preset threshold.
8. A supply chain management system based on volatility clustering and block chains, comprising:
the block chain is constructed according to the relation of each link in the supply chain, and each link in the supply chain is used as a block chain node;
the volatility prediction module is used for inputting the real-time supply chain data and the historical supply chain data into a volatility clustering prediction model which is constructed in advance when the real-time supply chain data are obtained by a target block chain node in the supply chain, and clustering the volatility of the input supply chain data by using a neural network model by using the volatility clustering prediction model so as to predict the volatility of the state of the supply chain;
and the node control module is used for controlling the target block link points to execute the intelligent contract when the corresponding preset fluctuation state conditions are met according to the obtained fluctuation prediction result.
9. The volatility clustering and block chain based supply chain management system of claim 8, wherein the volatility prediction module specifically comprises:
the supply chain data collection module is used for collecting historical supply chain data and real-time supply chain data required by the prediction model;
the supply chain data processing module is used for cleaning, enhancing and preprocessing the collected supply chain data and then outputting the supply chain data, and extracting and training the characteristics of the supply chain data;
the fluctuation clustering prediction module is used for outputting the cleaned, enhanced and preprocessed real-time supply chain data to a fluctuation clustering prediction model and outputting a fluctuation prediction result;
the supply chain data processing module specifically comprises a data cleaning module, a data enhancing module, a data preprocessing module, a feature engineering module for feature screening and a training module for executing model training, which are connected in sequence.
10. A computer-readable storage medium storing a computer program, characterized in that the computer program realizes the method of any one of claims 1 to 7 when executed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211298792.6A CN115689443A (en) | 2022-10-21 | 2022-10-21 | Supply chain management method and system based on volatility clustering and block chain |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211298792.6A CN115689443A (en) | 2022-10-21 | 2022-10-21 | Supply chain management method and system based on volatility clustering and block chain |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115689443A true CN115689443A (en) | 2023-02-03 |
Family
ID=85067036
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211298792.6A Pending CN115689443A (en) | 2022-10-21 | 2022-10-21 | Supply chain management method and system based on volatility clustering and block chain |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115689443A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116128323A (en) * | 2023-04-07 | 2023-05-16 | 阿里巴巴达摩院(杭州)科技有限公司 | Power transaction decision processing method, storage medium and electronic equipment |
-
2022
- 2022-10-21 CN CN202211298792.6A patent/CN115689443A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116128323A (en) * | 2023-04-07 | 2023-05-16 | 阿里巴巴达摩院(杭州)科技有限公司 | Power transaction decision processing method, storage medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Sharma et al. | Survey of stock market prediction using machine learning approach | |
| Saldivar et al. | Self-organizing tool for smart design with predictive customer needs and wants to realize Industry 4.0 | |
| JP6708847B1 (en) | Machine learning apparatus and method | |
| Cheng et al. | Evolutionary support vector machine inference system for construction management | |
| CN110956273A (en) | Credit scoring method and system integrating multiple machine learning models | |
| CN107545471B (en) | Big data intelligent recommendation method based on Gaussian mixture | |
| CN110188910A (en) | The method and system of on-line prediction service are provided using machine learning model | |
| Wang et al. | A novel framework to evaluate programmable logic controllers: a fuzzy MCDM perspective | |
| Borrajo et al. | Multi-agent neural business control system | |
| EP3948692A1 (en) | Process and system including an optimization engine with evolutionary surrogate-assisted prescriptions | |
| Hao et al. | Manufacturing service supply-demand optimization with dual diversities for industrial internet platforms | |
| Guo et al. | Integration of semi-fuzzy SVDD and CC-Rule method for supplier selection | |
| Saldivar et al. | Identifying smart design attributes for Industry 4.0 customization using a clustering Genetic Algorithm | |
| CN110414627A (en) | A kind of training method and relevant device of model | |
| Ming-Te et al. | Using data mining technique to perform the performance assessment of lean service | |
| CN115689443A (en) | Supply chain management method and system based on volatility clustering and block chain | |
| US11144938B2 (en) | Method and system for predictive modeling of consumer profiles | |
| Eggensperger et al. | Surrogate Benchmarks for Hyperparameter Optimization. | |
| Werth et al. | Simulation-based Optimization of Material Requirements Planning Parameters | |
| Karlsson et al. | Online knowledge extraction and preference guided multi-objective optimization in manufacturing | |
| Makinde et al. | An Improved Customer Relationship Management Model for Business-to-Business E-commerce Using Genetic-Based Data Mining Process | |
| US20200160200A1 (en) | Method and System for Predictive Modeling of Geographic Income Distribution | |
| Li et al. | Real-Time Prediction of Cross-Border e-Commerce Spike Performance Based on Neural Network and Decision Tree | |
| CN111027709B (en) | Information recommendation method and device, server and storage medium | |
| CN108629062A (en) | Methods, devices and systems for optimization of fixing a price |
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 |