Distributed data matching method and system based on artificial intelligence
Patent Information
- Application Number
- CN202411118702.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-11-22
AI Technical Summary
[0002]随着信息技术的快速发展,智能制造已经成为各个行业的热门话题,在智能制造中,海量数据是不可避免的,这些数据通过网络和计算技术与传统制造业相结合,实现采购、生产过程的智能化和自动化,提高工业生产的效率和质量,其中,分布式计算匹配数据技术在智能制造中起着重要的作用,可在智能制造中实现大规模数据匹配和实时的决策支持,而现有的分布式数据匹配技术在将生产或采购过程的数据进行匹配找寻时,难以贯通智能制造的上游、中游和下游的数据信息,有时会导致匹配的数据信息并不全面,进而影响数据匹配的准确性,降低数据处理匹配效率,对智能制造的采购或生产会产生延时
[0034] Beneficial effects of the present invention: The present invention uses artificial intelligence enhanced learning technology to connect the upstream, midstream and downstream data information of intelligent manufacturing when performing data matching, making the matched data information more comprehensive and improving the accuracy of data matching. , improve the efficiency of data processing and matching, and avoid delays in procurement or production of smart manufacturing. At the same time, when data matching fails, the procurement and/or production status data of upstream, midstream and downstream manufacturers can be analyzed, and the matching failure data and the procurement and/or production status data of upstream, midstream and downstream manufacturers can be analyzed simultaneously. Correlate features and generate new decision-making strategies through the decision-making path of deep reinforcement learning to rematch the data.
Smart Images

Figure CN119004134A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing technology, and in particular to an artificial intelligence-based distributed data matching method and system. Background Art
[0002] With the rapid development of information technology, intelligent manufacturing has become a hot topic in various industries. In intelligent manufacturing, massive data is inevitable. These data are combined with traditional manufacturing through network and computing technology to realize the intelligence and automation of procurement and production processes, and improve the efficiency and quality of industrial production. Among them, distributed computing matching data technology plays an important role in intelligent manufacturing. It can realize large-scale data matching and real-time decision support in intelligent manufacturing. However, when matching and searching for data in the production or procurement process, the existing distributed data matching technology is difficult to connect the upstream, midstream and downstream data information of intelligent manufacturing, which sometimes leads to incomplete matching data information, thereby affecting the accuracy of data matching, reducing the efficiency of data processing and matching, and causing delays in the procurement or production of intelligent manufacturing. Summary of the invention
[0003] In view of the above problems existing in the technical field of existing data processing, the present invention is proposed.
[0004] Therefore, one of the objects of the present invention is to provide a distributed data matching method and system based on artificial intelligence, which utilizes artificial intelligence enhanced learning technology to connect the upstream, midstream and downstream data information of intelligent manufacturing during data matching, so as to make the matching data information more comprehensive, improve the accuracy of data matching, enhance the efficiency of data processing and matching, and avoid delays in procurement or production of intelligent manufacturing.
[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:
[0006] In one aspect, the present invention provides a distributed data matching method based on artificial intelligence, the method comprising:
[0007] Constructing a data set, collecting historical matching data in the data set, and dividing the historical matching data;
[0008] Acquire feature data in the historical matching data, analyze the association rules of each feature data in the historical matching data, and generate a data model;
[0009] Conduct data tracing in the historical matching data, collect the procurement and / or manufacturing process corresponding to the matching data, upload the data tracing information to the data set, and update the data set;
[0010] The data with the most historical matches is obtained from the data set, and the associated features between the data and the procurement and / or manufacturing process are collected, and the associated features are labeled.
[0011] As a preferred solution of the present invention, the historical matching data is divided according to the business system, time node and / or data capacity, the change characteristics of the matching data among the business system, time node and / or data capacity are analyzed, and the business system range and time node range are preset, and the data with the most matching times within the range is obtained at the same time, and the upstream, midstream and downstream production and procurement information data corresponding to the data are obtained, and the change rules of the matching data and the production and procurement information data within the preset range are calculated. The formula is as follows:
[0012] Among them, Q ij Indicates the jth production and procurement information data collected for the ith time at a certain time node in the preset business system;
[0013] Wherein, m represents the amount of change in production and procurement information data between different time nodes in the preset business system, It represents the different production and procurement information data collected for the gth time by the preset business system at the tth time node in the same period of history, and analyzes the changing rules of the different production and procurement information data collected at the tth time node in the same period of history.
[0014] As a preferred solution of the present invention, the related production and procurement information data among upstream, midstream and downstream manufacturers are obtained, and the production and procurement information data are divided into peak period data and off-peak period data, and the correlation changes between data in different periods are analyzed at the same time, and the Q-learning algorithm is used for enhanced learning to form an enhanced prediction model.
[0015] As a preferred solution of the present invention, the peak period data and the off-peak period data are divided into several evaluation indicators, the evaluation indicator data of the upstream, midstream and downstream manufacturers in the associated business system are counted, the non-numerical data in the evaluation indicator data are converted into numerical data, the null values are filled according to the rules, the data are normalized, and the weight of each evaluation indicator is calculated, which is calculated according to the following formula:
[0016] Among them, w i It represents the maximum evaluation index weight data obtained by upstream, midstream and downstream manufacturers at the i-th time in the same period of history;
[0017] In the formula, k represents the adjustment value in the evaluation index data, δ irepresents the difference between the production and procurement information data collected for the i-th time within the preset business system, and n represents the historical change characteristic value of the data difference between the peak period data and the off-peak period data;
[0018] As a preferred solution of the present invention, the manufacturing time and raw material consumption data of related products of upstream, midstream and downstream manufacturers within the preset business system are collected, the changing rules of the manufacturing time and the raw material consumption data are analyzed, and the manufacturing time is divided into τ1, τ2, ..., τ n , where n represents the nth manufacturing time. At the same time, the change characteristics of the raw material consumption data under different manufacturing times are analyzed, and a safety threshold is preset for the change characteristics of the raw material consumption data. When the raw material consumption of a certain manufacturing time is lower than the safety threshold, it is determined that the manufacturing product corresponding to the manufacturing time is abnormal, and a data matching warning is issued. Otherwise, no determination is made.
[0019] As a preferred solution of the present invention, historical matching data on product manufacturing anomalies within the preset business system is obtained and a database is formed. At the same time, abnormal feature recognition technology and multi-dimensional information fusion data discrimination technology are preset based on the historical matching data in the database, and artificial intelligence assisted processing is performed through the abnormal feature recognition technology and multi-dimensional information data discrimination technology. When the matching data input in the future time period is identified and / or judged by the abnormal feature recognition technology and / or multi-dimensional information data discrimination technology to be consistent with the historical matching data in the database, the input matching data is judged to be abnormal data, otherwise, it is not judged.
[0020] As a preferred solution of the present invention, the historical matching data in the database is divided into Wherein, n represents the nth type of historical matching data, and the abnormal information corresponding to the divided historical matching data is obtained, and the proportion of the divided historical matching data in the database to the overall abnormal information of the upstream, midstream and downstream manufacturers is calculated according to the following formula:
[0021] Among them, C ο It indicates the most abnormal information data in the oth manufacturing process of upstream, midstream and downstream manufacturers;
[0022] Where Y represents the historical matching data with the most input times in the database, P is a given feature, representing the purchasing status features of the upstream, midstream and downstream manufacturers when the user inputs a certain matching data in the future period, and E t Represents the production status characteristics of upstream, midstream and downstream manufacturers at the tth time node, Indicates the classification type of the historical matching data in the database.
[0023] As a preferred solution of the present invention, the matching decision strategies formulated and updated when abnormal matching data occurs in upstream, midstream and downstream manufacturers are obtained, the matching data features after the matching decision strategies are updated are collected, and the decision path of deep reinforcement learning is used to analyze the association between the matching data features and the matching decision strategy updates, and a decision model is generated at the same time.
[0024] As a preferred solution of the present invention, the data capacity of the matching data in the decision model is analyzed, the procurement and / or production status characteristics of the upstream, midstream and downstream manufacturers corresponding to the data capacity are collected, and the data information text in the status characteristics is obtained, the acquired data information text is cleaned, segmented, converted and standardized, the processed data information text is uploaded to the decision model, and the decision model is updated.
[0025] In another aspect, the present invention provides a system for a distributed data matching method based on artificial intelligence, comprising:
[0026] An information acquisition module is used to acquire historical matching data of upstream, midstream and downstream manufacturers, and to acquire production and procurement information data corresponding to the historical matching data, collect feature data in the historical matching data, and construct a data set;
[0027] A data processing module is used to perform correlation analysis on feature data in historical matching data and generate a data model;
[0028] A fusion analysis unit, which includes a traceability module, a state acquisition module, an edge information acquisition module and an analysis module;
[0029] The traceability module is used to trace the historical matching data to obtain the production and procurement information status of upstream, midstream and downstream manufacturers at the time of historical matching data;
[0030] The state acquisition module is used to obtain daily production and procurement data of upstream, midstream and downstream manufacturers, and analyze the changing patterns of production and procurement data at different time points;
[0031] The edge information acquisition module is used to obtain the mutual influencing factors of upstream, midstream and downstream manufacturers;
[0032] The analysis module is used to analyze the historical changes, processing and re-matching data of upstream, midstream and downstream manufacturers in matching data under the affected factors;
[0033] The data decision module obtains the changed matching data features based on the historical changes, processing and re-matching of the matching data by the analysis module, and uses the decision path of deep reinforcement learning to analyze the association between the matching data features and the matching decision strategy update, and generates a decision model at the same time.
[0034] Beneficial effects of the present invention: The present invention utilizes artificial intelligence enhanced learning technology, and when performing data matching, it can connect the data information of the upstream, midstream and downstream of intelligent manufacturing, making the matched data information more comprehensive, improving the accuracy of data matching, improving the efficiency of data processing and matching, and avoiding delays in the procurement or production of intelligent manufacturing. At the same time, when data matching fails, the procurement and / or production status data of upstream, midstream and downstream manufacturers can be analyzed, and the correlation characteristics of the matching failure data and the procurement and / or production status data of upstream, midstream and downstream manufacturers can be analyzed at the same time, and a new decision strategy can be generated through the decision path of deep reinforcement learning to re-match the data. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:
[0036] Figure 1 This is a schematic diagram of a distributed data matching process based on artificial intelligence in an embodiment of the present invention;
[0037] Figure 2 Schematic diagram of the modular structure of a distributed data matching system based on artificial intelligence in an embodiment of the present invention;
[0038] Figure 3 This is a schematic diagram of the data matching state structure in an embodiment of the present invention;
[0039] Numbers in the figure: 110 - information acquisition module; 120 - data processing module; 130 - fusion analysis unit; 1301 - traceability module; 1302 - state acquisition module; 1303 - edge information acquisition module; 1304 - analysis module; 140 - data decision module. DETAILED DESCRIPTION
[0040] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.
[0041] Because the existing distributed data matching technology has difficulty in connecting the upstream, midstream and downstream data information of intelligent manufacturing when matching and searching for data in the production or procurement process, it sometimes leads to incomplete matching data information, which in turn affects the accuracy of data matching, reduces the efficiency of data processing and matching, and causes delays in the procurement or production of intelligent manufacturing.
[0042] Based on this, the present invention proposes a distributed data matching method and system based on artificial intelligence, which utilizes artificial intelligence enhanced learning technology to connect the upstream, midstream and downstream data information of intelligent manufacturing during data matching, making the matching data information more comprehensive, improving the accuracy of data matching, and improving the efficiency of data processing and matching, thereby avoiding delays in procurement or production of intelligent manufacturing.
[0043] The present invention is further described in detail below through embodiments and in conjunction with the accompanying drawings.
[0044] Reference Figures 1 to 3 , is an embodiment of the present invention, which provides a distributed data matching method based on artificial intelligence, the method comprising:
[0045] Build a data set, collect historical matching data in the data set, and divide the historical matching data;
[0046] It should be noted that in this embodiment, the historical matching data is divided according to the business system, time node and / or data capacity, the change characteristics of the matching data among the business system, time node and / or data capacity are analyzed, and the business system range and time node range are preset. At the same time, the data with the most matching times within the range is obtained, and the upstream, midstream and downstream production and procurement information data corresponding to the data are obtained, and the change rules of the matching data and the production and procurement information data within the preset range are calculated. The formula is as follows:
[0047] Among them, Q ij Indicates the jth production and procurement information data collected at a certain time node in the preset business system;
[0048] In the formula, m represents the change in production and procurement information data between different time nodes in the preset business system. Represents the different production and procurement information data collected by the preset business system for the gth time at the tth time node in the same period of history, and analyzes the change rules of the different production and procurement information data collected at the tth time node in the same period of history;
[0049] To further explain, production and procurement information data include raw material consumption, daily goods in and out quantity, product manufacturing time and inventory.
[0050] Obtain feature data from historical matching data, analyze the correlation rules of each feature data in the historical matching data, and generate a data model;
[0051] Conduct data tracing in historical matching data, collect the procurement and / or manufacturing processes corresponding to the matching data, upload the data tracing information to the data set, and update the data set;
[0052] It should be emphasized in this embodiment that the relevant production and procurement information data between upstream, midstream and downstream manufacturers are obtained, and the production and procurement information data are divided into peak period data and off-peak period data, and the correlation changes between data in different periods are analyzed at the same time, and the Q-learning algorithm is used for enhanced learning to form an enhanced prediction model;
[0053] At the same time, based on the above, this embodiment divides the peak period data and the off-peak period data into several evaluation indicators, collects the evaluation indicator data of the upstream, midstream and downstream manufacturers in the associated business system, converts the non-numerical data in the evaluation indicator data into numerical data, fills the null values according to the rules, normalizes the data, and calculates the weight of each evaluation indicator, which is calculated according to the following formula:
[0054] Among them, w i It represents the maximum evaluation index weight data obtained by upstream, midstream and downstream manufacturers at the i-th time in the same period of history;
[0055] In the formula, k represents the adjustment value in the evaluation index data, δ i It represents the difference between the production and procurement information data collected for the i-th time within the preset business system, and n represents the historical change characteristic value of the data difference between the peak period data and the off-peak period data;
[0056] Further, the same historical change characteristics of the peak period data and the off-peak period data are analyzed, and the same state characteristics of the upstream, midstream and downstream manufacturers under the same historical change characteristics are obtained, and the same operation characteristics are uploaded to the prediction model;
[0057] Obtain the data with the most historical matches in the data set, collect the associated features between the data and the procurement and / or manufacturing processes, and annotate the associated features;
[0058] Specifically, in this embodiment, the manufacturing time and raw material consumption data of related products of upstream, midstream and downstream manufacturers within the preset business system are collected, the changing rules of the manufacturing time and raw material consumption data are analyzed, and the manufacturing time is divided into τ1, τ2, ..., τ n , where n represents the nth manufacturing time, and the change characteristics of the raw material consumption data under different manufacturing times are analyzed at the same time, and a safety threshold is preset for the change characteristics of the raw material consumption data. When the raw material consumption of a certain manufacturing time is lower than the safety threshold, it is determined that the manufacturing product corresponding to the manufacturing time is abnormal, and a data matching warning is issued. Otherwise, no determination is made;
[0059] Obtain historical matching data within the preset business system when product manufacturing is abnormal, and form a database. At the same time, preset abnormal feature recognition technology and multi-dimensional information fusion data discrimination technology based on the historical matching data in the database, and perform artificial intelligence-assisted processing through abnormal feature recognition technology and multi-dimensional information data discrimination technology. When the matching data input in the future time period is identified and / or judged to be consistent with the historical matching data in the database by the abnormal feature recognition technology and / or multi-dimensional information data discrimination technology, the input matching data is judged to be abnormal data, otherwise, it is not judged;
[0060] It should be emphasized in this embodiment that, when abnormal data is identified, the system directly displays the identified information to the user through visualization means;
[0061] Based on the above, this embodiment divides the historical matching data in the database into Where n represents the nth type of historical matching data, and the abnormal information corresponding to the divided historical matching data is obtained, and the proportion of the divided historical matching data in the database to the overall abnormal information of the upstream, midstream and downstream manufacturers is calculated according to the following formula:
[0062] Among them, C ο It indicates the most abnormal information data in the oth manufacturing process of upstream, midstream and downstream manufacturers;
[0063] Where Y represents the historical matching data with the most input times in the database, P is a given feature, which represents the purchasing status features of the upstream, midstream and downstream manufacturers when the user inputs a certain matching data in the future period, and E t Represents the production status characteristics of upstream, midstream and downstream manufacturers at the tth time node, Indicates the classification type of historical matching data in the database;
[0064] It should be emphasized in this embodiment that, based on the historical matching data of production status characteristics of upstream, midstream and downstream manufacturers at different time nodes, the Monte Carlo tree updating matching strategy is adopted to generate an optimized data matching decision strategy;
[0065] Abnormal information includes production and procurement information data;
[0066] Obtain matching decision strategies that are updated when abnormal matching data appears in upstream, midstream and downstream manufacturers, collect matching data features after the matching decision strategies are updated, and use deep reinforcement learning decision paths to analyze the correlation between matching data features and matching decision strategy updates, and generate decision models at the same time;
[0067] This embodiment further analyzes the data capacity of the matching data in the decision model, collects the procurement and / or production status characteristics of the upstream, midstream and downstream manufacturers corresponding to the data capacity, obtains the data information text in the status characteristics, cleans, segments, converts and standardizes the obtained data information text, uploads the processed data information text to the decision model, and updates the decision model.
[0068] Based on the above, this application utilizes artificial intelligence enhanced learning technology, which can connect the upstream, midstream and downstream data information of intelligent manufacturing when performing data matching, making the matching data information more comprehensive, improving the accuracy of data matching, and improving data processing and matching efficiency, thereby avoiding delays in procurement or production of intelligent manufacturing.
[0069] In combination with the above-mentioned artificial intelligence-based distributed data matching method, this embodiment also proposes a system applied to the method, as follows:
[0070] The information acquisition module 110 is used to acquire the historical matching data of upstream, midstream and downstream manufacturers, and acquire the production and procurement information data corresponding to the historical matching data, collect the feature data in the historical matching data, and construct a data set;
[0071] The data processing module 120 is used to perform correlation analysis on the feature data in the historical matching data and generate a data model;
[0072] Fusion analysis unit 130, the fusion analysis unit includes a tracing module 1301, a state acquisition module 1302, an edge information acquisition module 1303 and an analysis module 1304;
[0073] The traceability module 1301 is used to trace the historical matching data to obtain the production and procurement information status of the upstream, midstream and downstream manufacturers at the time of the historical matching data;
[0074] The status acquisition module 1302 is used to obtain daily production and procurement data of upstream, midstream and downstream manufacturers, and analyze the changing patterns of production and procurement data at different time points;
[0075] The edge information acquisition module 1303 is used to obtain the mutual influencing factors of upstream, midstream and downstream manufacturers;
[0076] It should be noted that the mutual influencing factors in this embodiment include the manufacturing time difference and the purchasing time difference of the related products;
[0077] The analysis module 1304 is used to analyze the historical changes, processing and re-matching data of the upstream, midstream and downstream manufacturers under the affected factors;
[0078] The data decision module 140 obtains the changed matching data features based on the historical changes, processing and re-matching of the matching data by the analysis module 1304, and uses the decision path of deep reinforcement learning to analyze the association between the matching data features and the matching decision strategy updates, and generates a decision model at the same time.
[0079] In summary, the present invention utilizes artificial intelligence enhanced learning technology, and when performing data matching, it can connect the data information of the upstream, midstream and downstream of intelligent manufacturing, making the matched data information more comprehensive, improving the accuracy of data matching, improving the efficiency of data processing and matching, and avoiding delays in the procurement or production of intelligent manufacturing. At the same time, when data matching fails, the procurement and / or production status data of upstream, midstream and downstream manufacturers can be analyzed, and the correlation characteristics of the matching failure data and the procurement and / or production status data of upstream, midstream and downstream manufacturers can be analyzed at the same time, and a new decision strategy can be generated through the decision path of deep reinforcement learning to re-match the data.
[0080] Some code examples:
[0081] Information acquisition module 110:
[0082] import numpy as np
[0083] import matplotlib.pyplot as plt
[0084] from sklearn.datasets import make_blobs
[0085] from sklearn.cluster import KMeans
[0086] #Sample matching data
[0087] X,y=make_blobs(n_samples=500,centers=4,cluster_std=0.60,random_state=0)
[0088] # Clustering using KMeans
[0089] kmeans=KMeans(n_clusters=4)
[0090] kmeans.fit(X)
[0091] y_kmeans = kmeans.predict(X)
[0092] #Get matching features
[0093] plt.scatter(X[:,0],X[:,1],c=y_kmeans,s=50,cmap='viridis')
[0094] centers=kmeans.cluster_centers_
[0095] plt.scatter(centers[:,0],centers[:,1],c='red',s=200,alpha=0.75)
[0096] plt.title("K-Means Clustering")
[0097] plt.xlabel("Feature 1")
[0098] plt.ylabel("Feature 2")
[0099] plt.show()
[0100] Data processing module 120:
[0101] import numpy as np
[0102] import matplotlib.pyplot as plt
[0103] from scipy.cluster.hierarchy import dendrogram,linkage
[0104] from sklearn.datasets import make_blobs
[0105] #Collect sample matching data
[0106] X,y=make_blobs(n_samples=500, centers=3, random_state=0, cluster_std=0.60)
[0107] # Perform hierarchical clustering
[0108] Z=linkage(X,method='ward')
[0109] # Sample data features
[0110] plt.figure(figsize=(10,7))
[0111] dendrogram(Z,truncate_mode='lastp',p=12,leaf_rotation=90.,leaf_font_size=12.,show_contracted=True)
[0112] plt.title("Hierarchical Clustering Dendrogram")
[0113] plt.xlabel("Sample index or(Cluster size)")
[0114] plt.ylabel("Distance")
[0115] plt.show()
[0116] Fusion Analysis Unit 130:
[0117] import numpy as np
[0118] import matplotlib.pyplot as plt
[0119] from sklearn.datasets import make_moons
[0120] from sklearn.cluster import DBSCAN
[0121] #Get upstream, midstream and downstream manufacturers data information
[0122] X,y=make_moons(n_samples=500,noise=0.1,random_state=0)
[0123] # Clustering using DBSCAN
[0124] dbscan=DBSCAN(eps=0.2,min_samples=5)
[0125] y_dbscan=dbscan.fit_predict(X)
[0126] # Draw the result graph
[0127] plt.scatter(X[:,0],X[:,1],c=y_dbscan,s=50,cmap='viridis')
[0128] #Mark matching data
[0129] core_samples_mask=np.zeros_like(y_dbscan,dtype=bool)
[0130] core_samples_mask[dbscan.core_sample_indices_]=True
[0131] noise_mask=(y_dbscan==-1)
[0132] plt.scatter(X[noise_mask,0],X[noise_mask,1],c='red',s=50,label='Noise')
[0133] plt.title("DBSCAN Clustering")
[0134] plt.xlabel("Feature 1")
[0135] plt.ylabel("Feature 2")
[0136] plt.legend()
[0137] plt.show()
[0138] Data decision module 140:
[0139] import numpy as np
[0140] import matplotlib.pyplot as plt
[0141] from sklearn.datasets import make_blobs
[0142] from sklearn.mixture import GaussianMixture
[0143] #Change matching data
[0144] X,y=make_blobs(n_samples=500, centers=3, random_state=42, cluster_std=0.60)
[0145] # Use GMM for clustering
[0146] gmm=GaussianMixture(n_components=3,random_state=42)
[0147] gmm.fit(X)
[0148] labels = gmm.predict(X)
[0149] #Use Monte Carlo tree search to update the matching strategy
[0150] plt.scatter(X[:,0],X[:,1],c=labels,s=40,cmap='viridis')
[0151] plt.scatter(gmm.means_[:,0],gmm.means_[:,1],c='red',s=200,alpha=0.75,marker='X')#Mark the center point
[0152] plt.title("Gaussian Mixture Model Clustering")
[0153] plt.xlabel("Feature 1")
[0154] plt.ylabel("Feature 2")
[0155] plt.show()
[0156] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function according to the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium.
[0157] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.
[0158] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.
[0159] Any process or method description in the flow chart or otherwise described herein can be understood to represent a module, fragment or portion of a code including one or more executable instructions for implementing the steps of a specific logical function or process. And the scope of the preferred embodiment of the present application includes other implementations, in which the functions may not be performed in the order shown or discussed, including in a substantially simultaneous manner or in a reverse order according to the functions involved.
[0160] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, which can be embodied in any computer-readable medium for use by an instruction execution system, apparatus or device (such as a computer-based system, a system including a processor or other system that can fetch instructions from an instruction execution system, apparatus or device and execute instructions), or used in combination with these instruction execution systems, apparatuses or devices.
[0161] It should be understood that the various parts of the present application can be implemented with hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented with software or firmware stored in a memory and executed by a suitable instruction execution system. All or part of the steps of the above embodiment method can be completed by instructing the relevant hardware through a program, which can be stored in a computer-readable storage medium, and when the program is executed, it includes one of the steps of the method embodiment or a combination thereof.
[0162] In addition, each functional unit in each embodiment of the present application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into one module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of a software functional module. If the above-mentioned integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The storage medium can be a read-only memory, a disk or an optical disk, etc.
[0163] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. A distributed data matching method based on artificial intelligence, characterized in that: The method comprises: Constructing a data set, collecting historical matching data in the data set, and dividing the historical matching data; Acquire feature data in the historical matching data, analyze the association rules of each feature data in the historical matching data, and generate a data model; Conduct data tracing in the historical matching data, collect the procurement and / or manufacturing process corresponding to the matching data, upload the data tracing information to the data set, and update the data set; The data with the most historical matches is obtained from the data set, and the associated features between the data and the procurement and / or manufacturing process are collected, and the associated features are labeled.
2. The distributed data matching method based on artificial intelligence according to claim 1, characterized in that: The historical matching data is divided according to business system, time node and / or data capacity, and the change characteristics of the matching data among the business system, time node and / or data capacity are analyzed. The business system range and time node range are preset, and the data with the most matching times within the range is obtained at the same time, and the upstream, midstream and downstream production and procurement information data corresponding to the data are obtained, and the change rules of the matching data and the production and procurement information data within the preset range are calculated. The formula is as follows: Among them, Q ij Indicates the jth production and procurement information data collected for the ith time at a certain time node in the preset business system; Where m represents the change in production and procurement information data between different time nodes in the preset business system, X t g It represents the different production and procurement information data collected for the gth time by the preset business system at the tth time node in the same period of history, and analyzes the changing rules of the different production and procurement information data collected at the tth time node in the same period of history.
3. The distributed data matching method based on artificial intelligence according to claim 1, characterized in that: The related production and procurement information data between upstream, midstream and downstream manufacturers are obtained, and the production and procurement information data are divided into peak period data and off-peak period data. At the same time, the correlation changes between data in different periods are analyzed, and the Q-learning algorithm is used for enhanced learning to form an enhanced prediction model.
4. The distributed data matching method based on artificial intelligence as claimed in claim 3, characterized in that: The peak period data and the off-peak period data are divided into several evaluation indicators, and the evaluation indicator data of the upstream, midstream and downstream manufacturers in the associated business system are counted. The non-numerical data in the evaluation indicator data are converted into numerical data, and the null values are filled according to the rules. The data is normalized and the weight of each evaluation indicator is calculated according to the following formula: Among them, w i It represents the maximum evaluation index weight data obtained by upstream, midstream and downstream manufacturers at the i-th time in the same period of history; In the formula, k represents the adjustment value in the evaluation index data, δ i It represents the difference between the production and procurement information data collected for the i-th time within the preset business system, and n represents the historical change characteristic value of the data difference between the peak period data and the off-peak period data.
5. The distributed data matching method based on artificial intelligence as claimed in claim 2, characterized in that: Collect the manufacturing time and raw material consumption data of related products of upstream, midstream and downstream manufacturers within the preset business system, analyze the change rules of the manufacturing time and the raw material consumption data, and divide the manufacturing time into τ1, τ2, ..., τ n , where n represents the nth manufacturing time. At the same time, the change characteristics of the raw material consumption data under different manufacturing times are analyzed, and a safety threshold is preset for the change characteristics of the raw material consumption data. When the raw material consumption of a certain manufacturing time is lower than the safety threshold, it is determined that the manufacturing product corresponding to the manufacturing time is abnormal, and a data matching warning is issued. Otherwise, no determination is made.
6. The distributed data matching method based on artificial intelligence as claimed in claim 5, characterized in that: The historical matching data of product manufacturing abnormalities within the preset business system is obtained and a database is formed. At the same time, abnormal feature recognition technology and multi-dimensional information fusion data discrimination technology are preset based on the historical matching data in the database, and artificial intelligence-assisted processing is performed through the abnormal feature recognition technology and multi-dimensional information data discrimination technology. When the matching data input in the future time period is identified and / or judged to be consistent with the historical matching data in the database by the abnormal feature recognition technology and / or the multi-dimensional information data discrimination technology, the input matching data is judged to be abnormal data, otherwise, it is not judged.
7. The distributed data matching method based on artificial intelligence as claimed in claim 6, characterized in that: Divide the historical matching data in the database into Wherein, n represents the nth type of historical matching data, and the abnormal information corresponding to the divided historical matching data is obtained, and the proportion of the divided historical matching data in the database to the overall abnormal information of the upstream, midstream and downstream manufacturers is calculated according to the following formula: Among them, C ο It indicates the most abnormal information data in the oth manufacturing process of upstream, midstream and downstream manufacturers; Where Y represents the historical matching data with the most input times in the database, P is a given feature, representing the purchasing status features of the upstream, midstream and downstream manufacturers when the user inputs a certain matching data in the future period, and E t Represents the production status characteristics of upstream, midstream and downstream manufacturers at the tth time node, Indicates the classification type of the historical matching data in the database.
8. The distributed data matching method based on artificial intelligence as claimed in claim 7, characterized in that: The matching decision strategies that are updated when abnormal matching data occurs in upstream, midstream and downstream manufacturers are obtained, the matching data features after the matching decision strategies are updated are collected, and the decision path of deep reinforcement learning is used to analyze the association between the matching data features and the matching decision strategy updates, and a decision model is generated at the same time.
9. The distributed data matching method based on artificial intelligence as claimed in claim 8, characterized in that: Analyze the data capacity of the matching data in the decision model, collect the procurement and / or production status characteristics of the upstream, midstream and downstream manufacturers corresponding to the data capacity, and obtain the data information text in the status characteristics, clean, segment, convert and standardize the obtained data information text, upload the processed data information text to the decision model, and update the decision model.
10. A system for a distributed data matching method based on artificial intelligence as claimed in claim 1, characterized in that: include: An information acquisition module is used to acquire historical matching data of upstream, midstream and downstream manufacturers, and to acquire production and procurement information data corresponding to the historical matching data, collect feature data in the historical matching data, and construct a data set; A data processing module is used to perform correlation analysis on feature data in historical matching data and generate a data model; A fusion analysis unit, which includes a traceability module, a state acquisition module, an edge information acquisition module and an analysis module; The traceability module is used to trace the historical matching data to obtain the production and procurement information status of upstream, midstream and downstream manufacturers at the time of historical matching data; The state acquisition module is used to obtain daily production and procurement data of upstream, midstream and downstream manufacturers, and analyze the changing patterns of production and procurement data at different time points; The edge information acquisition module is used to obtain the mutual influencing factors of upstream, midstream and downstream manufacturers. The analysis module is used to analyze the historical changes, processing and re-matching data of upstream, midstream and downstream manufacturers in matching data under the affected factors; The data decision module obtains the changed matching data features based on the historical changes, processing and re-matching of the matching data by the analysis module, and uses the decision path of deep reinforcement learning to analyze the association between the matching data features and the matching decision strategy update, and generates a decision model at the same time.
Citation Information
Cited By
Distributed data management system and method based on artificial intelligence
CN120386809A
An artificial intelligence-based distributed data management system and method
CN120386809B
