Supply chain management data analysis system and method based on machine learning

Abstract

The invention relates to the technical field of supply chain management data analysis, and discloses a supply chain management data analysis system and method based on machine learning, and the system comprises a supplier object search module, a supplier object service quality analysis module, and a supplier object selection module. The method comprises the following steps: accurately searching a qualified historical supply quality number of a material purchasing target supplier based on target supplier object parameters in combination with an intelligent search algorithm and supplier historical supply behavior information, and accurately analyzing a historical supply quality behavior completion result of the material purchasing target supplier in combination with numerical analysis; supplier service level accurate data analysis is performed by adopting an intelligent search algorithm in combination with the historical supply quantity of a material purchasing target supplier, quality behavior analysis parameters and different supply behavior supplier service level parameters, so that enterprise supply chain supplier service quality dynamic and accurate analysis is realized; and the intelligence and quality of enterprise supply chain management are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of supply chain management data analysis, and specifically to a supply chain management data analysis system and method based on machine learning. Background Art

[0002] Supply chain management refers to optimizing the operation of the supply chain. Supply chain management is to coordinate the internal and external resources of the enterprise to jointly meet the needs of consumers. When we regard the enterprises in each link of the supply chain as a virtual enterprise alliance, and regard any enterprise as a department in this virtual enterprise alliance, the internal management of the alliance is supply chain management. It's just that the composition of the alliance is dynamic and changes at any time according to market needs. Supply chain management needs to divide customer groups according to the service characteristics required by customers; design the enterprise's logistics network according to customer needs and the company's profitability; listen to market demand information and design products that are closer to customers; time delay; strategically determine the source of goods and procurement and establish a win-win cooperation strategy with suppliers; establish an information system in the entire supply chain field; establish performance evaluation criteria for the entire supply chain, etc. The supplier selection in the existing enterprise supply chain management is usually based on human selection, and it is impossible to intelligently and dynamically select supplier objects based on the analysis of the characteristic parameters of the supplier's supply behavior, which reduces the scientificity and security of the enterprise supply chain management.

[0003] A Chinese invention patent with announcement number CN104574024B discloses an ERP remote zero-inventory supply chain management system, which provides lean logistics-related information and scientific and refined enterprise supply chain management for the entire supply chain by setting up a production collection module, a question-and-answer module, a machine operation data management module, a customer tool specification comparison table, a customer data management and setting module, a customer data query module, a report module, and a system management module. However, the above technical solutions cannot realize the selection of the best tool raw material supplier based on intelligent analysis of the supply behavior of the company's tool raw material suppliers. Summary of the invention

[0004] 1. Technical issues to be solved

[0005] In order to solve the problem that the supplier selection in the existing enterprise supply chain management is usually based on human selection, and it is impossible to intelligently and dynamically select supplier objects based on the analysis of supplier supply behavior characteristic parameters, which reduces the scientificity and security of enterprise supply chain management, the above purpose of accurately collecting supply material characteristic information, scientifically searching for target supplier objects, intelligently retrieving supplier historical supply behavior characteristic parameters, scientifically and dynamically evaluating supplier service levels, accurately and dynamically screening the best suppliers, and improving the efficiency and quality of enterprise supply chain management is achieved.

[0006] (II) Technical solution

[0007] The present invention is implemented by the following technical solution: a supply chain management data analysis method, the method comprising the following steps:

[0008] S1. Collecting characteristic text data of supplied materials;

[0009] S2, searching and processing the supplier objects required for the enterprise's supply materials based on the supply material characteristic text data and the supplier supply characteristic text data, and generating target supplier object data;

[0010] S3, based on the target supplier object data and the supplier's historical supply behavior text data, perform a historical supply quantity search process on the target supplier, generate the target supplier's historical supply quantity data, and perform a historical supply quantity behavior analysis process on the target supplier together with the supplier's historical standard supply quantity data, to construct the target supplier's historical supply quantity behavior analysis data;

[0011] S4, performing a historical supply quality search process on the target supplier based on the target supplier object data and the supplier's historical supply behavior text data, generating the target supplier's historical supply quality data, and performing a historical supply quality behavior analysis process on the target supplier with the supplier's historical standard supply quality data, to construct the target supplier's historical supply quality behavior analysis data;

[0012] S5. Perform target supplier service level judgment processing based on the target supplier's historical supply quantity, quality behavior analysis parameters and service level data of suppliers with different supply behaviors, and generate target supplier object service level judgment data;

[0013] S6. Screening the optimal supplier object required for the enterprise to supply materials based on the target supplier object service level judgment data to generate optimal target supplier object data;

[0014] S7. Build the supply chain supplier analysis data required by the enterprise and push feedback to the enterprise supply chain management platform.

[0015] Preferably, the steps of collecting the characteristic text data of the supplied materials are as follows:

[0016] S11. Collect the material feature text information that the enterprise needs to purchase online through the material ERP management system and generate supply material feature text data I, wherein the supply material feature text data includes material name text data, material code text data, material quality standard text data and material specification and size text data.

[0017] The present invention accurately and efficiently collects the characteristic information of purchased materials through the material ERP management system, thereby achieving the effect of providing reliable data support for the scientific search of purchased material supplier objects.

[0018] Preferably, the steps of searching for supplier objects required for the enterprise's supply materials based on the supply material characteristic text data and the supplier supply characteristic text data to generate target supplier object data are as follows:

[0019] S21. Establish supplier supply feature text data set A = (a1,…,a p ,…,a η ), p=1,2,3,…,eta; where a p The supplier supply characteristic text data corresponding to the p-th supplier object is represented, and n represents the maximum number of supplier objects. The supplier supply characteristic text data includes supplier name, supplier address, supplier enterprise code, supplier supply product name text data, supplier supply product code text data, supplier supply product quality standard text data, and supplier supply product specification size text data;

[0020] S22, compare the supply material feature text data I with the supplier supply feature text data a in the supplier supply feature text data set A. p Perform character matching of the supply material feature text, search for the supplier object corresponding to the supply material feature text data I, and generate a target supplier object data set A'=(a' p1 ,…,a' p2 ), 1≤p1≤p≤p2≤η; where a' p1 Indicates the p1th target supplier object data, a' p2 represents the p2th target supplier object data; the specific steps for generating the target supplier object data set A' are as follows:

[0021] S221, initialize each honey source i, that is, the supplier supply feature text data a in the supplier supply feature text data set A p Set as honey source; set the number of honey sources η, the current number of iterations t and the maximum number of iterations T; the initial position of honey source i w i , the formula for generating the nectar source location is as follows: Randomly generated in the search space of the supplier supply feature text data set A, where rand represents a random number in (0,1), and Λ respectively represent the upper bound and lower bound in the search space of the supplier supply feature text data set A;

[0022] S222, is the location of the nectar source w i Assign hired bees. At the beginning of the search, hired bees search the supplier supply feature text data set A in the search space by all the supplier supply feature text data a p The location of the nectar source wi Around, according to the formula w' i =w i +Ξ×(w i -w β ) Search to generate new honey source location w' i , that is, traverse and search for all supplier supply feature text data a in the supplier supply feature text data set A search space p Performing supply material characteristic text character matching with the supply material characteristic text data I, and searching for a supplier object matching the supply material characteristic text data I; wherein means randomly selecting a nectar source not equal to i from n nectar sources, w β represents the position of the randomly selected honey source β in the search space of the supplier supply feature text data set A, and Ξ represents a random number in the interval [-1,1];

[0023] S223, based on the new nectar source location w' i The fitness of the supplier is determined according to the greedy selection method to retain the honey source, that is, to retain the supplier object that matches the supply material feature text data I in the search space of the supplier supply feature text data set A;

[0024] S224, the employed bees share the nectar source information, and the follower bees search for the supplier object nectar source that matches the supply material characteristic text data I in the search space of the supplier supply characteristic text data set A according to the probability value. The follower bees calculate the probability of the nectar source found by the employed bees being followed according to the probability formula based on the nectar source information of the supplier object that matches the supply material characteristic text data I shared by the employed bees. Where Z i Indicates the location of the new nectar source w' i The probability of being followed by an employed bee, Ω i Indicates the location of the new nectar source shared by the employed bees according to the follower bees w' i Adaptability;

[0025] S225, the follower bee searches in the same way as the employed bee, and determines the retained honey source i according to the greedy selection method, that is, searching for a supplier object matching the supply material feature text data I in the search space of the supplier supply feature text data set A;

[0026] S226, during the random search of the supplier supply feature text data set A in the search space, if the honey source location w i After t iterations, the search reaches the iteration threshold No better nectar source was found. The location of the nectar source is w iThe employed bees will be abandoned, and the corresponding employed bees will become scout bees. The calculation formula for the random generation of new nectar source locations in the search space by scout bees is as follows: in It represents the position of the new honey source i in the search space of the supplier supply feature text data set A after t+1 iterations, It represents the position of the new honey source i in the search space of the supplier supply feature text data set A after t iterations;

[0027] S227. When the maximum number of search iterations T is met, the supplier object matching the supply material characteristic text data I is output and a target supplier object data set A' is generated through data identification. The target supplier object data includes the supplier name, supplier address and supplier enterprise code.

[0028] The present invention performs intelligent retrieval of enterprise material procurement supplier objects by scientifically presetting supplier supply characteristic information and combining bee colony optimization algorithm with supply material characteristic information, thereby achieving the effect of accurately searching the enterprise material procurement target supplier objects.

[0029] Preferably, the historical supply quantity search process of the target supplier is performed based on the target supplier object data and the supplier's historical supply behavior text data, the target supplier's historical supply quantity data is generated, and the target supplier's historical supply quantity behavior analysis process is performed with the supplier's historical standard supply quantity data, and the operation steps for constructing the target supplier's historical supply quantity behavior analysis data are as follows:

[0030] S31. Establish a text data set of supplier's historical supply behavior B = (b1,…,b p ,…,b η );where b p The text data of the supplier's historical supply behavior corresponding to the p-th supplier indicates the text data of the supplier's historical supply behavior, which indicates the total quantity, quality-qualified quantity, supply time point, and supply price of materials supplied by the supply chain supplier to the enterprise in a historical time period, and the historical time period includes any one of the last month, the last quarter, and the last year;

[0031] Establish historical standard supply quantity data of suppliers s , the supplier's historical standard supply quantity data represents the standard total quantity data of materials that the enterprise requires the supply chain supplier to supply during a historical period of time;

[0032] S32, using a KD tree nearest neighbor search algorithm to compare the target supplier object data in the target supplier object data set A' with the supplier historical supply behavior text data set B pAccording to the supplier object number and the total quantity of supplied materials, the historical supply quantity of the target supplier is searched, and the target supplier historical supply quantity data set B'=(b' p1 ,…,b' p2 ), where b' p1 Indicates the target supplier object data a' p1 The corresponding target supplier's historical supply quantity data, b' p2 Indicates the target supplier object data a' p2 The corresponding target supplier’s historical supply quantity data;

[0033] S33, compare the target supplier's historical supply quantity data in the target supplier's historical supply quantity data set B' with the supplier's historical standard supply quantity data b s Perform a historical supply quantity numerical comparison, and build a target supplier's historical supply quantity behavior analysis data set B" = (b" p1 ,…,b” p2 ), where b” p1 Indicates the target supplier object data a' p1 The corresponding target supplier’s historical supply quantity behavior analysis data, b” p2 Indicates the target supplier object data a' p2 Corresponding target supplier’s historical supply quantity behavior analysis data;

[0034] When the target supplier's historical supply quantity data is not less than the supplier's historical standard supply quantity data b s , indicating that the historical supply quantity of the supply chain supplier meets the enterprise's supply quantity requirements, then the output target supplier's historical supply quantity behavior analysis data is qualified;

[0035] When the target supplier's historical supply quantity data is less than the supplier's historical standard supply quantity data b s , indicating that the historical supply quantity of the supply chain supplier does not meet the enterprise's supply quantity requirements, then the output target supplier's historical supply quantity behavior analysis data is unqualified.

[0036] The present invention dynamically retrieves the historical supply quantity of the target supplier of material procurement based on the target supplier object parameters in combination with the KD tree nearest neighbor search algorithm and the supplier's historical supply behavior information, and at the same time scientifically determines the completion results of the historical supply quantity behavior of the target supplier of material procurement in combination with numerical analysis, thereby achieving the effect of dynamically and accurately analyzing the completion results of the historical supply quantity behavior of the target supplier of material procurement.

[0037] Preferably, the historical supply quality search process of the target supplier is performed based on the target supplier object data and the supplier's historical supply behavior text data, the target supplier's historical supply quality data is generated, and the target supplier's historical supply quality behavior analysis process is performed with the supplier's historical standard supply quality data, and the operation steps for constructing the target supplier's historical supply quality behavior analysis data are as follows:

[0038] S41. Establish supplier's historical standard supply quality data s , the supplier's historical standard supply quality data represents the quantity data of materials that the enterprise requires the supply chain supplier to supply in accordance with the quality standards in the historical period;

[0039] S42, when the target supplier historical supply quantity behavior analysis data set B' is constructed, an iterative deepening search algorithm is used to compare the target supplier object data in the target supplier object data set A' with the supplier historical supply behavior text data b in the supplier historical supply behavior text data set B. p According to the supplier object number and the key words of the qualified quantity of the supplied material quality, the target supplier's historical supply quality quantity is searched, and the target supplier's historical supply quality data set C'=(c' p1 ,…,c' p2 ), where c' p1 Indicates the target supplier object data a' p1 The corresponding target supplier's historical supply quality data, c' p2 Indicates the target supplier object data a' p2 The corresponding target supplier’s historical supply quality data;

[0040] S43, compare the target supplier's historical supply quality data in the target supplier's historical supply quality data set C' with the supplier's historical standard supply quality data c s Compare the historical supply quality qualified quantity values, and build the target supplier's historical supply quality behavior analysis data set C"=(c" p1 ,…,c” p2 ), where c” p1 Indicates the target supplier object data a' p1 The corresponding target supplier’s historical supply quality behavior analysis data, c” p2 Indicates the target supplier object data a' p2 The corresponding target supplier’s historical supply quality behavior analysis data;

[0041] When the target supplier's historical supply quality data is not less than the supplier's historical standard supply quality data c s, indicating that the number of qualified historical supply quality of the supply chain supplier meets the enterprise's qualified quality requirement, then the output target supplier's historical supply quality behavior analysis data is qualified;

[0042] When the target supplier's historical supply quality data is less than the supplier's historical standard supply quality data c s , indicating that the historical supply quality qualified quantity of the supply chain supplier does not meet the enterprise's supply quality qualified quantity requirements, then the output target supplier's historical supply quality behavior analysis data is unqualified.

[0043] The present invention accurately searches for the qualified quantity of historical supply quality of target suppliers of material procurement based on target supplier object parameters combined with an iterative deepening search algorithm and supplier's historical supply behavior information, and simultaneously accurately analyzes the completion results of the historical supply quality behavior of target suppliers of material procurement in combination with numerical analysis, thereby achieving the effect of dynamically and accurately evaluating the completion results of the historical supply quality behavior of target suppliers of material procurement.

[0044] Preferably, the target supplier service level judgment processing is performed based on the target supplier's historical supply quantity, quality behavior analysis parameters and the service level data of suppliers with different supply behaviors to generate the target supplier object service level judgment data in the following steps:

[0045] S51. Establish a data set of service levels for suppliers with different supply behaviors Where [b" = qualified, c" ​​= qualified] d1 Indicates that the target supplier's historical supply quantity behavior analysis data of the supply chain supplier is qualified and the target supplier's historical supply quality behavior analysis data is qualified, and the corresponding supplier service level data of different supply behaviors is d1, where d1 indicates that the supplier service level is level one;

[0046] Indicates that the target supplier's historical supply quantity behavior analysis data of the supply chain supplier is unqualified and the target supplier's historical supply quality behavior analysis data is qualified, or the target supplier's historical supply quantity behavior analysis data is qualified and the target supplier's historical supply quality behavior analysis data is unqualified. The corresponding different supply behavior supplier service level data is d2, where d2 indicates that the supplier service level is level 2;

[0047] [b” = failed, c” = failed] d3 Indicates that the target supplier's historical supply quantity behavior analysis data of the supply chain supplier is unqualified and the target supplier's historical supply quality behavior analysis data is unqualified, and the corresponding different supply behavior supplier service level data is d3, where d3 indicates that the supplier service level is level three;

[0048] S53, using a breadth-first search algorithm, sequentially perform target supplier service level identification processing on the target supplier historical supply quantity behavior analysis data set B”, the target supplier historical supply quality behavior analysis data set C”, and the different supply behavior supplier service level data set D according to the supplier object number and the target supplier historical supply quantity and quality behavior analysis result keywords, and generate a target supplier object service level judgment data set E = (e p1 ,…,e p2 ), where e p1 Indicates the target supplier object data a' p1 The corresponding target supplier object service level judgment data, e p2 Indicates the target supplier object data a' p2 The corresponding target supplier object service level judgment data; the target supplier object service level judgment data includes any one of the first level, second level and third level.

[0049] The present invention uses an intelligent search algorithm to combine the quantity and quality behavior analysis parameters of the material procurement target supplier's historical supply with the supplier service level parameters of different supply behaviors to perform accurate data analysis of the supplier service level, thereby achieving the effect of dynamic and accurate analysis of the service quality of the enterprise supply chain suppliers.

[0050] Preferably, the optimal supplier object required for the enterprise to supply materials is screened based on the target supplier object service level judgment data, and the operation steps for generating the optimal target supplier object data are as follows:

[0051] S61, using a unified cost search algorithm, searching the target supplier object service level judgment data set E for the supplier object corresponding to the target supplier object service level judgment data with a supplier service level of level 1 according to the service level keyword, and generating the optimal target supplier object data set A" = (a" p3 ,…,a” p4 ), p1≤p3≤p≤p4≤p2; where a” p3 Indicates the p3th optimal target supplier object data, a” p4 Represents the p4th optimal target supplier object data.

[0052] The present invention uses a unified cost search algorithm to accurately screen suppliers with the best service quality for material procurement for enterprises based on target supplier object service level judgment parameters, thereby achieving the effect of providing enterprises with convenient and high-quality supply chain management services.

[0053] Preferably, the steps of constructing the supply chain supplier analysis data required by the enterprise and pushing the data back to the enterprise supply chain management platform are as follows:

[0054] S71, combining the supply material characteristic text data I and the optimal target supplier object data set A” to construct the supply chain supplier analysis data F=(I, A”) required by the enterprise;

[0055] S72. Push the supply chain supplier analysis data F required by the enterprise to the enterprise supply chain management platform in real time through the Internet of Things communication network.

[0056] The present invention combines the supply material characteristic information and the optimal target supplier object information through data processing, realizes the autonomous and efficient construction of the supply chain supplier analysis parameters required by the enterprise and accurately feeds back to the enterprise supply chain management platform, so as to provide scientific data support for the enterprise to select material procurement suppliers.

[0057] A supply chain management data analysis system based on machine learning, used to implement the supply chain management data analysis method, the system includes a supplier object search module, a supplier object service quality analysis module, and a supplier object selection module;

[0058] The supplier object search module includes a supply material characteristic information collection unit, a supplier supply characteristic information storage unit, and a supplier object search unit;

[0059] The supply material characteristic information collection unit collects supply material characteristic text data through the material ERP management system; the supplier supply characteristic information storage unit is used to store supplier supply characteristic text data; the supplier object search unit performs supplier object search processing required for enterprise supply materials based on the supply material characteristic text data and supplier supply characteristic text data based on machine learning to generate target supplier object data;

[0060] The supplier object service quality analysis module includes a supplier historical supply behavior information storage unit, a supplier historical supply quantity identification unit, a supplier historical standard supply quantity storage unit, a supplier historical supply quantity behavior judgment unit, a supplier historical supply quality identification unit, a supplier historical standard supply quality storage unit, a supplier historical supply quality behavior judgment unit, a different supply behavior supplier service level storage unit, and a supplier object service level analysis unit;

[0061] The supplier historical supply behavior information storage unit is used to store the supplier's historical supply behavior text data; the supplier historical supply quantity identification unit performs a target supplier's historical supply quantity search based on the target supplier object data and the supplier's historical supply behavior text data to generate the target supplier's historical supply quantity data; the supplier historical standard supply quantity storage unit is used to store the supplier's historical standard supply quantity data; the supplier historical supply quantity behavior judgment unit performs a target supplier's historical supply quantity behavior analysis on the target supplier's historical supply quantity data and the supplier's historical standard supply quantity data to construct the target supplier's historical supply quantity behavior analysis data; the supplier historical supply quality identification unit performs a target supplier's historical supply quantity behavior analysis based on the target supplier object data and the supplier's historical supply behavior text data The target supplier's historical supply quality search process generates the target supplier's historical supply quality data; the supplier's historical standard supply quality storage unit is used to store the supplier's historical standard supply quality data; the supplier's historical supply quality behavior judgment unit analyzes the target supplier's historical supply quality data and the supplier's historical standard supply quality data to construct the target supplier's historical supply quality behavior analysis data; the different supply behavior supplier service level storage unit is used to store the different supply behavior supplier service level data; the supplier object service level analysis unit judges the target supplier's service level based on the target supplier's historical supply quantity, quality behavior analysis parameters and the different supply behavior supplier service level data to generate the target supplier object service level judgment data;

[0062] The supplier object selection module includes an optimal supplier object screening and analysis unit, an enterprise required supply chain supplier information generation unit, and an enterprise required supply chain supplier information feedback unit;

[0063] The optimal supplier object screening and analysis unit screens the optimal supplier objects required for the enterprise's supply materials based on the target supplier object service level judgment data to generate optimal target supplier object data; the supply chain supplier information generation unit required by the enterprise combines the supply material feature text data and the optimal target supplier object data to construct the supply chain supplier analysis data required by the enterprise; the supply chain supplier information feedback unit required by the enterprise pushes and feeds back the supply chain supplier analysis data required by the enterprise to the enterprise supply chain management platform.

[0064] (III) Beneficial effects

[0065] The present invention provides a supply chain management data analysis system and method based on machine learning. It has the following beneficial effects:

[0066] 1. Accurately and efficiently collect the characteristic information of purchased materials through the material ERP management system, and provide reliable data support for the scientific search of purchased material supplier objects; scientifically preset supplier supply characteristic information combined with machine learning algorithms and supply material characteristic information to conduct intelligent retrieval of enterprise supply material procurement supplier objects, realize the precise search of enterprise material procurement target supplier objects, and improve the efficiency of enterprise supply chain management.

[0067] 2. Dynamically retrieve the historical supply quantity of the target supplier of material procurement based on the target supplier object parameters combined with the intelligent search algorithm and the supplier's historical supply behavior information, and scientifically judge the completion results of the historical supply quantity behavior of the target supplier of material procurement based on numerical analysis; accurately search the historical supply quality quantity of the target supplier of material procurement based on the target supplier object parameters combined with the intelligent search algorithm and the supplier's historical supply behavior information, and accurately analyze the completion results of the historical supply quality behavior of the target supplier of material procurement based on numerical analysis; use intelligent search algorithms combined with the historical supply quantity and quality behavior analysis parameters of the target supplier of material procurement and the service level parameters of suppliers with different supply behaviors to perform precise data analysis of supplier service levels, so as to realize dynamic and precise analysis of the service quality of suppliers in the enterprise supply chain, and improve the intelligence and quality of enterprise supply chain management.

[0068] 3. By adopting intelligent search algorithms to accurately screen suppliers with the best service quality for material procurement based on the target supplier object service level judgment parameters, we can provide enterprises with convenient and high-quality supply chain management services and improve the quality of enterprise material procurement; by using data processing to combine the supply material characteristic information and the optimal target supplier object information, we can independently and efficiently construct the supply chain supplier analysis parameters required by the enterprise and accurately feedback them to the enterprise supply chain management platform, providing scientific data support for enterprises to select material procurement suppliers and reducing the time and economic cost of enterprise supply chain management. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 A module schematic diagram of a supply chain management data analysis system based on machine learning provided by the present invention;

[0070] Figure 2 The present invention provides a flow chart of a supply chain management data analysis method. DETAILED DESCRIPTION

[0071] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0072] The embodiments of the supply chain management data analysis system and method based on machine learning are as follows:

[0073] Embodiment 1:

[0074] See also Figure 1 - Figure 2 , a supply chain management data analysis method, the method comprising the following steps:

[0075] S1. Collecting characteristic text data of supplied materials;

[0076] S2. Search and process the supplier objects required for the enterprise's supply materials based on the supply material feature text data and the supplier supply feature text data, and generate target supplier object data;

[0077] S3. Perform a search process on the target supplier's historical supply quantity based on the target supplier object data and the supplier's historical supply behavior text data, generate the target supplier's historical supply quantity data, and perform a target supplier's historical supply quantity behavior analysis process with the supplier's historical standard supply quantity data, to construct the target supplier's historical supply quantity behavior analysis data;

[0078] S4. Perform a historical supply quality search process on the target supplier based on the target supplier object data and the supplier's historical supply behavior text data, generate the target supplier's historical supply quality data, and perform a historical supply quality behavior analysis process on the target supplier with the supplier's historical standard supply quality data, to construct the target supplier's historical supply quality behavior analysis data;

[0079] S5. Perform target supplier service level judgment processing based on the target supplier's historical supply quantity, quality behavior analysis parameters and service level data of suppliers with different supply behaviors, and generate target supplier object service level judgment data;

[0080] S6. Screening and processing the optimal supplier object required for the enterprise to supply materials based on the target supplier object service level judgment data, and generating optimal target supplier object data;

[0081] S7. Build the supply chain supplier analysis data required by the enterprise and push feedback to the enterprise supply chain management platform.

[0082] For further information, see Figure 1 - Figure 2 The steps for collecting the characteristic text data of the supplied materials are as follows:

[0083] S11. Collect the material feature text information that the enterprise needs to purchase online through the material ERP management system and generate the supply material feature text data I. The supply material feature text data includes material name text data, material code text data, material quality standard text data and material specification and size text data.

[0084] The steps for searching and processing the supplier objects required for the enterprise's supply materials based on the supply material characteristic text data and the supplier supply characteristic text data and generating the target supplier object data are as follows:

[0085] S21. Establish supplier supply feature text data set A = (a1,…,a p ,…,a η ), p=1,2,3,…,eta; where a p The supplier supply characteristic text data corresponding to the p-th supplier object, n represents the maximum number of supplier objects, and the supplier supply characteristic text data includes supplier name, supplier address, supplier enterprise code, supplier supply product name text data, supplier supply product code text data, supplier supply product quality standard text data and supplier supply product specification size text data;

[0086] S22, compare the supply material feature text data I with the supplier supply feature text data a in the supplier supply feature text data set A. p Perform character matching of the supply material feature text, search for the supplier object corresponding to the supply material feature text data I, and generate the target supplier object data set A'=(a' p1 ,…,a' p2 ), 1≤p1≤p≤p2≤η; where a' p1 Indicates the p1th target supplier object data, a' p2 represents the p2th target supplier object data; the specific steps for generating the target supplier object data set A' are as follows:

[0087] S221, initialize each honey source i, that is, the supplier supply feature text data a in the supplier supply feature text data set A p Set as honey source; set the number of honey sources η, the current number of iterations t and the maximum number of iterations T; the initial position of honey source i w i , the formula for generating the nectar source location is as follows: Randomly generated in the search space of the supplier supply feature text data set A, where rand represents a random number in (0,1). and Λ represent the upper and lower bounds of the search space of the supplier supply feature text data set A, respectively;

[0088] S222, is the location of the nectar source w i Assign hired bees. At the beginning of the search, hired bees search the supplier supply feature text data set A in the search space by all the supplier supply feature text data a p The location of the nectar source w i Around, according to the formula w' i =w i +Ξ×(w i -w β ) Search to generate new honey source location w' i , that is, traverse and search for all supplier supply feature text data a in the supplier supply feature text data set A search space p Performing character matching of the supply material characteristic text with the supply material characteristic text data I, and searching for a supplier object matching the supply material characteristic text data I; wherein means randomly selecting a nectar source not equal to i from n nectar sources, w β represents the position of the randomly selected honey source β in the search space of the supplier supply feature text data set A, and Ξ represents a random number in the interval [-1,1];

[0089] S223, based on the new nectar source location w' i The fitness of the honey source is determined according to the greedy selection method, that is, the supplier object matching the supply material feature text data I is retained in the search space of the supplier supply feature text data set A;

[0090] S224, the employed bees share the nectar source information, and the follower bees search for the supplier object nectar source that matches the supply material feature text data I in the supplier supply feature text data set A according to the probability value. The follower bees calculate the probability of the nectar source found by the employed bees being followed according to the probability formula based on the nectar source information of the supplier object that matches the supply material feature text data I shared by the employed bees. Where Z i Indicates the location of the new nectar source w' i The probability of being followed by an employed bee, Ω i Indicates the location of the new nectar source shared by the employed bees according to the follower bees w' i Adaptability;

[0091] S225, the follower bee searches in the same way as the employed bee, and determines the retained honey source i according to the greedy selection method, that is, searching for the supplier object matching the supply material feature text data I in the supplier supply feature text data set A search space;

[0092] S226, during the random search of the scout bee in the search space of the supplier supply feature text data set A, if the honey source location w i After t iterations, the search reaches the iteration threshold No better nectar source was found. The location of the nectar source is w i The employed bees will be abandoned, and the corresponding employed bees will become scout bees. The calculation formula for the random generation of new nectar source locations in the search space by scout bees is as follows: in It represents the position of the new honey source i in the search space of the supplier supply feature text data set A after t+1 iterations. It represents the position of the new honey source i in the search space of the supplier supply feature text data set A after t iterations;

[0093] S227. When the maximum number of search iterations T is met, the supplier object matching the supply material characteristic text data I is output and a target supplier object data set A' is generated through data identification. The target supplier object data includes the supplier name, supplier address and supplier enterprise code.

[0094] Through the supply material characteristic information collection unit, the material ERP management system is used to accurately and efficiently collect the characteristic information of purchased materials, providing reliable data support for the scientific search of purchased material supplier objects; the supplier supply characteristic information storage unit and the supplier object search unit cooperate with each other, and the scientific preset supplier supply characteristic information is combined with the machine learning algorithm and the supply material characteristic information to perform intelligent retrieval of the enterprise's supply material procurement supplier objects, thereby realizing the precise search of the enterprise's material procurement target supplier objects and improving the efficiency of the enterprise's supply chain management.

[0095] For further information, see Figure 1 - Figure 2 Based on the target supplier object data and the supplier's historical supply behavior text data, the target supplier's historical supply quantity search is performed, the target supplier's historical supply quantity data is generated, and the target supplier's historical supply quantity behavior analysis is performed with the supplier's historical standard supply quantity data. The operation steps for constructing the target supplier's historical supply quantity behavior analysis data are as follows:

[0096] S31. Establish a text data set of supplier's historical supply behavior B = (b1,…,b p ,…,b η );where b pThe text data of the supplier's historical supply behavior corresponding to the p-th supplier indicates the text data of the supplier's historical supply behavior, which indicates the total quantity, quality-qualified quantity, supply time point, and supply price of the materials supplied by the supply chain supplier to the enterprise in the historical time period. The historical time period includes any one of the last month, last quarter, and last year.

[0097] Establish historical standard supply quantity data of suppliers s , the supplier's historical standard supply quantity data represents the standard total quantity data of materials that the enterprise requires the supply chain supplier to supply during the historical period;

[0098] S32, using the KD tree nearest neighbor search algorithm to compare the target supplier object data in the target supplier object data set A' with the supplier historical supply behavior text data b in the supplier historical supply behavior text data set B. p According to the supplier object number and the total quantity of supplied materials, the historical supply quantity of the target supplier is searched, and the target supplier historical supply quantity data set B'=(b' p1 ,…,b' p2 ), where b' p1 Indicates the target supplier object data a' p1 The corresponding target supplier's historical supply quantity data, b' p2 Indicates the target supplier object data a' p2 The corresponding target supplier’s historical supply quantity data;

[0099] S33, compare the target supplier's historical supply quantity data in the target supplier's historical supply quantity data set B' with the supplier's historical standard supply quantity data b s Perform a historical supply quantity numerical comparison, and build a target supplier's historical supply quantity behavior analysis data set B" = (b" p1 ,…,b” p2 ), where b” p1 Indicates the target supplier object data a' p1 The corresponding target supplier’s historical supply quantity behavior analysis data, b” p2 Indicates the target supplier object data a' p2 Corresponding target supplier’s historical supply quantity behavior analysis data;

[0100] When the target supplier's historical supply quantity data is not less than the supplier's historical standard supply quantity data b s , indicating that the historical supply quantity of the supply chain supplier meets the enterprise's supply quantity requirements, then the output target supplier's historical supply quantity behavior analysis data is qualified;

[0101] When the target supplier's historical supply quantity data is less than the supplier's historical standard supply quantity data b s , indicating that the historical supply quantity of the supply chain supplier does not meet the enterprise's supply quantity requirements, then the output target supplier's historical supply quantity behavior analysis data is unqualified.

[0102] Based on the target supplier object data and the supplier's historical supply behavior text data, the target supplier's historical supply quality search is performed, the target supplier's historical supply quality data is generated, and the target supplier's historical supply quality behavior analysis is performed with the supplier's historical standard supply quality data. The operation steps for constructing the target supplier's historical supply quality behavior analysis data are as follows:

[0103] S41. Establish supplier's historical standard supply quality data s ,The supplier's historical standard supply quality data represents the quantity data of materials that meet the quality standards that the enterprise requires the supply chain supplier to supply during the historical period;

[0104] S42. When the target supplier historical supply quantity behavior analysis data set B' is constructed, an iterative deepening search algorithm is used to compare the target supplier object data in the target supplier object data set A' with the supplier historical supply behavior text data b in the supplier historical supply behavior text data set B. p According to the supplier object number and the key words of the qualified quantity of the supplied material quality, the target supplier's historical supply quality quantity is searched, and the target supplier's historical supply quality data set C'=(c' p1 ,…,c' p2 ), where c' p1 Indicates the target supplier object data a' p1 The corresponding target supplier's historical supply quality data, c' p2 Indicates the target supplier object data a' p2 The corresponding target supplier’s historical supply quality data;

[0105] S43, compare the target supplier's historical supply quality data in the target supplier's historical supply quality data set C' with the supplier's historical standard supply quality data c s Compare the historical supply quality qualified quantity values, and build the target supplier's historical supply quality behavior analysis data set C"=(c" p1 ,…,c” p2 ), where c” p1 Indicates the target supplier object data a' p1 The corresponding target supplier’s historical supply quality behavior analysis data, c” p2 Indicates the target supplier object data a' p2The corresponding target supplier’s historical supply quality behavior analysis data;

[0106] When the target supplier's historical supply quality data is not less than the supplier's historical standard supply quality data c s , indicating that the number of qualified historical supply quality of the supply chain supplier meets the enterprise's qualified quality requirement, then the output target supplier's historical supply quality behavior analysis data is qualified;

[0107] When the target supplier's historical supply quality data is less than the supplier's historical standard supply quality data c s , indicating that the historical supply quality qualified quantity of the supply chain supplier does not meet the enterprise's supply quality qualified quantity requirements, then the output target supplier's historical supply quality behavior analysis data is unqualified.

[0108] The target supplier's service level is judged based on the target supplier's historical supply quantity, quality behavior analysis parameters and the service level data of suppliers with different supply behaviors. The steps to generate the target supplier's service level judgment data are as follows:

[0109] S51. Establish a data set of service levels for suppliers with different supply behaviors Where [b" = qualified, c" ​​= qualified] d1 It indicates that under the condition that the target supplier's historical supply quantity behavior analysis data of the supply chain supplier is qualified and the target supplier's historical supply quality behavior analysis data is qualified, the corresponding supplier service level data of different supply behaviors is d1, where d1 indicates that the supplier service level is level one;

[0110] The target supplier's historical supply quantity behavior analysis data for the supply chain supplier is unqualified and the target supplier's historical supply quality behavior analysis data is qualified, or the target supplier's historical supply quantity behavior analysis data is qualified and the target supplier's historical supply quality behavior analysis data is unqualified. The corresponding supplier service level data for different supply behaviors is d2, where d2 indicates that the supplier service level is level 2;

[0111] [b” = failed, c” = failed] d3 Indicates that the target supplier of the supply chain supplier's historical supply quantity behavior analysis data is unqualified and the target supplier's historical supply quality behavior analysis data is unqualified. The corresponding supplier service level data of different supply behaviors is d3, where d3 indicates that the supplier service level is level three;

[0112] S53, using a breadth-first search algorithm, sequentially identify the target supplier service level according to the supplier object number and the target supplier's historical supply quantity and quality behavior analysis result keywords, and generate a target supplier object service level judgment data set E = (e p1 ,…,e p2 ), where e p1 Indicates the target supplier object data a' p1 The corresponding target supplier object service level judgment data, e p2 Indicates the target supplier object data a' p2 The corresponding target supplier object service level judgment data; the target supplier object service level judgment data includes any one of the first level, second level and third level.

[0113] Through the cooperation of the supplier historical supply quantity identification unit and the supplier historical supply quantity behavior judgment unit, the historical supply quantity of the target material procurement supplier is dynamically retrieved based on the target supplier object parameters combined with the intelligent search algorithm and the supplier's historical supply behavior information, and at the same time, the completion result of the historical supply quantity behavior of the target material procurement supplier is scientifically judged in combination with numerical analysis; the supplier historical supply quality identification unit and the supplier historical supply quality behavior judgment unit cooperate with each other, and the historical supply quality qualified quantity of the target material procurement supplier is accurately searched based on the target supplier object parameters combined with the intelligent search algorithm and the supplier's historical supply behavior information, and at the same time, the completion result of the historical supply quality behavior of the target material procurement supplier is accurately analyzed in combination with numerical analysis; the supplier object service level analysis unit uses an intelligent search algorithm combined with the material procurement target supplier's historical supply quantity, quality behavior analysis parameters and the service level parameters of suppliers with different supply behaviors to perform accurate data analysis of supplier service levels, thereby realizing dynamic and accurate analysis of the service quality of suppliers in the enterprise supply chain, and improving the intelligence and quality of enterprise supply chain management.

[0114] For further information, see Figure 1 - Figure 2 , based on the target supplier object service level judgment data, the optimal supplier object required for the enterprise to supply materials is screened and processed. The steps for generating the optimal target supplier object data are as follows:

[0115] S61, using a unified cost search algorithm to search for supplier objects corresponding to target supplier object service level judgment data with a supplier service level of level 1 in the target supplier object service level judgment data set E according to the service level keywords, and generate the optimal target supplier object data set A" = (a" p3 ,…,a” p4 ), p1≤p3≤p≤p4≤p2; where a” p3 Indicates the p3th optimal target supplier object data, a” p4 Represents the p4th optimal target supplier object data.

[0116] The steps to construct the supply chain supplier analysis data required by the enterprise and push feedback to the enterprise supply chain management platform are as follows:

[0117] S71, combining the supply material feature text data I and the optimal target supplier object data set A” to construct the supply chain supplier analysis data F=(I, A”) required by the enterprise;

[0118] S72. Push the supply chain supplier analysis data F required by the enterprise to the enterprise supply chain management platform in real time through the Internet of Things communication network.

[0119] Through the optimal supplier object screening and analysis unit, an intelligent search algorithm is used to accurately screen suppliers with the best service quality for material procurement for enterprises based on the target supplier object service level judgment parameters, provide enterprises with convenient and high-quality supply chain management services, and improve the quality of enterprise material procurement; the enterprise's required supply chain supplier information generation unit and the enterprise's required supply chain supplier information feedback unit cooperate with each other, and use data processing to combine the supply material feature information and the optimal target supplier object information, so as to independently and efficiently construct the supply chain supplier analysis parameters required by the enterprise and accurately feedback them to the enterprise supply chain management platform, provide scientific data support for enterprises to select material procurement suppliers, and reduce the time and economic cost of enterprise supply chain management.

[0120] Embodiment 2:

[0121] See also Figure 1 - Figure 2 , a supply chain management data analysis system based on machine learning, used to implement a supply chain management data analysis method, the system includes a supplier object search module, a supplier object service quality analysis module, and a supplier object selection module;

[0122] The supplier object search module includes a supply material characteristic information collection unit, a supplier supply characteristic information storage unit, and a supplier object search unit;

[0123] The supply material characteristic information collection unit collects the supply material characteristic text data through the material ERP management system; the supplier supply characteristic information storage unit is used to store the supplier supply characteristic text data; the supplier object search unit searches for the supplier objects required for the enterprise supply materials based on the supply material characteristic text data and the supplier supply characteristic text data based on machine learning, and generates the target supplier object data;

[0124] The supplier object service quality analysis module includes a supplier historical supply behavior information storage unit, a supplier historical supply quantity identification unit, a supplier historical standard supply quantity storage unit, a supplier historical supply quantity behavior judgment unit, a supplier historical supply quality identification unit, a supplier historical standard supply quality storage unit, a supplier historical supply quality behavior judgment unit, a different supply behavior supplier service level storage unit, and a supplier object service level analysis unit;

[0125] A supplier historical supply behavior information storage unit is used to store the supplier's historical supply behavior text data; a supplier historical supply quantity identification unit searches for the target supplier's historical supply quantity based on the target supplier object data and the supplier's historical supply behavior text data to generate the target supplier's historical supply quantity data; a supplier historical standard supply quantity storage unit is used to store the supplier's historical standard supply quantity data; a supplier historical supply quantity behavior judgment unit analyzes the target supplier's historical supply quantity data and the supplier's historical standard supply quantity data to construct the target supplier's historical supply quantity behavior analysis data; a supplier historical supply quality identification unit performs target supplier quality analysis based on the target supplier object data and the supplier's historical supply behavior text data. The target supplier's historical supply quality search and processing generates the target supplier's historical supply quality data; the supplier's historical standard supply quality storage unit is used to store the supplier's historical standard supply quality data; the supplier's historical supply quality behavior judgment unit analyzes the target supplier's historical supply quality data and the supplier's historical standard supply quality data to construct the target supplier's historical supply quality behavior analysis data; the supplier service level storage unit with different supply behaviors is used to store the supplier service level data with different supply behaviors; the supplier object service level analysis unit judges the target supplier's service level based on the target supplier's historical supply quantity, quality behavior analysis parameters and the supplier service level data with different supply behaviors to generate the target supplier object service level judgment data;

[0126] The supplier object selection module includes an optimal supplier object screening and analysis unit, a supply chain supplier information generation unit required by the enterprise, and a supply chain supplier information feedback unit required by the enterprise;

[0127] The optimal supplier object screening and analysis unit screens the optimal supplier objects required for the enterprise's supply materials based on the target supplier object service level judgment data, and generates the optimal target supplier object data; the supply chain supplier information generation unit required by the enterprise combines the supply material feature text data and the optimal target supplier object data to construct the supply chain supplier analysis data required by the enterprise; the supply chain supplier information feedback unit required by the enterprise pushes the supply chain supplier analysis data required by the enterprise to the enterprise supply chain management platform.

[0128] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A supply chain management data analysis method, characterized in that: The method comprises the following steps: S1. Collecting characteristic text data of supplied materials; S2, searching and processing the supplier objects required for the enterprise's supply materials based on the supply material characteristic text data and the supplier supply characteristic text data, and generating target supplier object data; S3, based on the target supplier object data and the supplier's historical supply behavior text data, perform a historical supply quantity search process on the target supplier, generate the target supplier's historical supply quantity data, and perform a historical supply quantity behavior analysis process on the target supplier together with the supplier's historical standard supply quantity data, to construct the target supplier's historical supply quantity behavior analysis data; S4, performing a historical supply quality search process on the target supplier based on the target supplier object data and the supplier's historical supply behavior text data, generating the target supplier's historical supply quality data, and performing a historical supply quality behavior analysis process on the target supplier with the supplier's historical standard supply quality data, to construct the target supplier's historical supply quality behavior analysis data; S5. Perform target supplier service level judgment processing based on the target supplier's historical supply quantity, quality behavior analysis parameters and service level data of suppliers with different supply behaviors, and generate target supplier object service level judgment data; S6. Screening the optimal supplier object required for the enterprise to supply materials based on the target supplier object service level judgment data to generate optimal target supplier object data; S7. Build the supply chain supplier analysis data required by the enterprise and push feedback to the enterprise supply chain management platform.

2. A supply chain management data analysis method according to claim 1, characterized in that: The S1 comprises the following steps: S11. Collect the material feature text information that the enterprise needs to purchase online through the material ERP management system and generate the supply material feature text data I.

3. A supply chain management data analysis method according to claim 2, characterized in that: The S2 comprises the following steps: S21. Establish supplier supply feature text data set A = (a1,…,a p ,…,a η ), p=1,2,3,…,eta; where a p represents the supplier supply feature text data corresponding to the p-th supplier object, and η represents the maximum number of supplier objects; S22, the I and the a in the A p Perform character matching of the supply material feature text, search for the supplier object corresponding to I, and generate the target supplier object data set A'=(a' p1 ,…,a' p2 ), 1≤p1≤p≤p2≤η; where a' p1 Indicates the p1th target supplier object data, a' p2 represents the p2th target supplier object data; the specific steps for generating the target supplier object data set A' are as follows: S221, initialize each honey source i, that is, the supplier supply feature text data a in the supplier supply feature text data set A p Set as honey source; set the number of honey sources η, the current number of iterations t and the maximum number of iterations T; the initial position of honey source i w i ; S222, is the location of the nectar source w i Assign employed bees. At the beginning of the search, employed bees are composed of all the a p The location of the nectar source w i Around; S223, based on the new nectar source location w' i The fitness of the honey source to be retained is determined according to the greedy selection method, that is, the supplier object matching the I is retained in the search space A; S224, the employed bees share the nectar source information, and the follower bees search for the nectar source of the supplier object matching the I in the A search space according to the probability value, and the follower bees calculate the probability of the nectar source found by the employed bees being followed according to the probability formula based on the nectar source information of the supplier object matching the I shared by the employed bees; S225, the follower bee searches in the same way as the employed bee, and determines the retained nectar source i according to the greedy selection method, that is, searching for a supplier object matching the I in the A search space; S226, during the random search of the scout bee in the A search space, if the nectar source location w i After t iterations, the search reaches the iteration threshold No better nectar source was found. The location of the nectar source is w i The bees will be abandoned, and the corresponding employed bees will become scout bees. The scout bees will randomly generate new nectar source locations in the search space. S227. When the maximum number of search iterations T is met, the supplier object matching the I is output and a target supplier object data set A' is generated through data identification. The target supplier object data includes the supplier name, supplier address and supplier enterprise code.

4. A supply chain management data analysis method according to claim 3, characterized in that: The S3 comprises the following steps: S31. Establish a text data set of supplier's historical supply behavior B = (b1,…,b p ,…,b η );where b p Represents the text data of the supplier's historical supply behavior corresponding to the p-th supplier; Establish historical standard supply quantity data of suppliers s ; S32, using the KD tree nearest neighbor search algorithm to compare the target supplier object data in A' with the target supplier object data in B p According to the supplier object number and the total quantity of supplied materials, the historical supply quantity of the target supplier is searched, and the target supplier historical supply quantity data set B'=(b' p1 ,…,b' p2 ), where b' p1 Indicates the target supplier object data a' p1 The corresponding target supplier's historical supply quantity data, b' p2 Indicates the target supplier object data a' p2 The corresponding target supplier’s historical supply quantity data; S33, compare the historical supply quantity data of the target supplier in B' with the s Perform a historical supply quantity numerical comparison, and build a target supplier's historical supply quantity behavior analysis data set B" = (b" p1 ,…,b” p2 ), where b” p1 Indicates the target supplier object data a' p1 The corresponding target supplier’s historical supply quantity behavior analysis data, b” p2 Indicates the target supplier object data a' p2 Corresponding target supplier’s historical supply quantity behavior analysis data; When the target supplier's historical supply quantity data is not less than b s , then the output target supplier's historical supply quantity behavior analysis data is qualified; When the target supplier's historical supply quantity data is less than b s , then the output target supplier's historical supply quantity behavior analysis data is unqualified.

5. A supply chain management data analysis method according to claim 4, characterized in that: The S4 comprises the following steps: S41. Establish supplier's historical standard supply quality data s ; S42, when the B" is constructed, an iterative deepening search algorithm is used to compare the target supplier object data in A' with the supplier historical supply behavior text data b in B. p According to the supplier object number and the key words of the qualified quantity of the supplied material quality, the target supplier's historical supply quality quantity is searched, and the target supplier's historical supply quality data set C'=(c' p1 ,…,c' p2 ), where c' p1 Indicates the target supplier object data a' p1 The corresponding target supplier's historical supply quality data, c' p2 Indicates the target supplier object data a' p2 The corresponding target supplier’s historical supply quality data; S43, compare the historical supply quality data of the target supplier in C' with the s Compare the historical supply quality qualified quantity values, and build the target supplier's historical supply quality behavior analysis data set C"=(c" p1 ,…,c” p2 ), where c” p1 Indicates the target supplier object data a' p1 The corresponding target supplier’s historical supply quality behavior analysis data, c” p2 Indicates the target supplier object data a' p2 The corresponding target supplier’s historical supply quality behavior analysis data; When the target supplier's historical supply quality data is not less than the supplier's historical standard supply quality data c s , then the output target supplier's historical supply quality behavior analysis data is qualified; When the target supplier's historical supply quality data is less than the supplier's historical standard supply quality data c s , then the output target supplier's historical supply quality behavior analysis data is unqualified.

6. A supply chain management data analysis method according to claim 5, characterized in that: The S5 comprises the following steps: S51, establishing a service level data set D of suppliers with different supply behaviors, wherein the service level data set D of suppliers with different supply behaviors includes service level data of suppliers with different supply behaviors; S53, using a breadth-first search algorithm, sequentially identify the target supplier service level according to the supplier object number and the target supplier's historical supply quantity and quality behavior analysis result keywords, and generate a target supplier object service level judgment data set E = (e p1 ,…,e p2 ), where e p1 Indicates the target supplier object data a' p1 The corresponding target supplier object service level judgment data, e p2 Indicates the target supplier object data a' p2 The corresponding target supplier object service level judgment data.

7. A supply chain management data analysis method according to claim 6, characterized in that: The S6 comprises the following steps: S61, using a unified cost search algorithm to search for the supplier object corresponding to the service level judgment data of the target supplier object with a supplier service level of level 1 in E according to the service level keyword, and generate the optimal target supplier object data set A"=(a" p3 ,…,a” p4 ), p1≤p3≤p≤p4≤p2; where a” p3 Indicates the p3th optimal target supplier object data, a” p4 Represents the p4th optimal target supplier object data.

8. A supply chain management data analysis method according to claim 7, characterized in that: The S7 comprises the following steps: S71, combining the data of I and A” to construct the supply chain supplier analysis data F=(I, A”) required by the enterprise; S72: Push the F to the enterprise supply chain management platform in real time through the Internet of Things communication network.

9. A supply chain management data analysis system based on machine learning, used to implement a supply chain management data analysis method according to any one of claims 1 to 8, characterized in that: The system includes a supplier object search module, a supplier object service quality analysis module, and a supplier object selection module.

Images