CN117252351B

CN117252351B - Production quality auxiliary decision-making method and system based on AI large model

Info

Publication number: CN117252351B
Application number: CN202311545275.9A
Authority: CN
Inventors: 罗平伟
Original assignee: Shanghai Yiyuan Data Technology Co ltd
Current assignee: Shanghai Yiyuan Data Technology Co ltd
Priority date: 2023-11-20
Filing date: 2023-11-20
Publication date: 2024-02-02
Anticipated expiration: 2043-11-20
Also published as: CN117252351A

Abstract

The application discloses a production quality auxiliary decision-making method and system based on an AI large model, wherein the method comprises the following steps: acquiring a fault task of a large production and manufacturing network model in the production and manufacturing process; determining all production and manufacturing equipment through which each faulty task passes, and determining paths formed by all production and manufacturing equipment in each faulty task as faulty task paths; according to the failure times of each failure task path, positioning final failure equipment in the production and manufacturing network large model under the abnormal working condition; acquiring a target fault description corresponding to the final fault equipment; determining a target fault reason of the final fault equipment in an abnormal working state in the equipment fault knowledge graph based on the target fault description; and carrying out auxiliary decision making on the large production and manufacturing network model based on the target fault cause. The method and the device can rapidly locate the fault equipment and the fault root cause of the fault equipment, and can accurately carry out production quality auxiliary decision in the production and manufacturing process.

Description

Production quality auxiliary decision-making method and system based on AI large model

Technical Field

The application relates to the technical field of artificial intelligence, in particular to a production quality auxiliary decision-making method and system based on an AI large model.

Background

In the production and manufacturing process, production and manufacturing equipment in a large production and manufacturing network model can malfunction, thereby seriously affecting the production quality of products in the production and manufacturing process. In the existing production quality auxiliary decision-making process, the fault equipment is usually positioned by a large model of an unoriented production and manufacture network, and the fault root cause is positioned by an auxiliary decision-making method based on a link calling relation, an auxiliary decision-making method based on traditional rule configuration and an auxiliary decision-making method based on a machine learning algorithm. However, the large model of the non-directional production and manufacturing network cannot determine the traversal time of production and manufacturing equipment, large sample analysis is generally adopted, the sampling time is long, the algorithm is complex, and the rapid positioning of fault equipment is not facilitated. The fault root causes of the fault equipment obtained by adopting the three auxiliary decision methods have larger deviation, so that the fault root causes cannot be accurately analyzed, and the production quality auxiliary decision cannot be accurately carried out.

Disclosure of Invention

The present application aims to solve at least one of the technical problems existing in the related art. Therefore, the embodiment of the application provides the production quality auxiliary decision-making method and the system based on the AI large model, which can rapidly locate the fault equipment and the fault root cause of the fault equipment so as to accurately carry out the production quality auxiliary decision-making in the production and manufacturing process.

In a first aspect, an embodiment of the present application provides a method for assisting in decision making of production quality based on an AI large model, including:

acquiring a fault task of a large production and manufacturing network model in the production and manufacturing process; the production and manufacturing network large model is as follows: taking each production and manufacturing equipment in a target production and manufacturing equipment group of the current production flow as a center, and establishing a network model obtained after branching between each production and manufacturing equipment and adjacent upstream and downstream production and manufacturing equipment; the target production and manufacturing equipment group is the production and manufacturing equipment group with the largest number of production and manufacturing equipment in the current production flow;

determining all production and manufacturing equipment through which each faulty task passes, and determining paths formed by all production and manufacturing equipment in each faulty task as faulty task paths;

determining failure times of each failure task path;

according to the failure times of each failure task path, positioning final failure equipment in the production and manufacturing network large model under the abnormal working condition;

acquiring a target fault description corresponding to the final fault equipment;

determining a target fault reason of the final fault equipment in an abnormal working state in an equipment fault knowledge graph based on the target fault description; the equipment fault knowledge graph is as follows: a plurality of fault descriptions and a plurality of knowledge maps of fault cause association relations;

And carrying out auxiliary decision making on the large production and manufacturing network model based on the target fault cause.

In a second aspect, an embodiment of the present application provides an AI large model-based production quality auxiliary decision-making apparatus, including an acquisition module, a determination module, a positioning module, and an auxiliary decision-making module;

the acquisition module is used for: acquiring a fault task of a large production and manufacturing network model in the production and manufacturing process; the production and manufacturing network large model is as follows: taking each production and manufacturing equipment in a target production and manufacturing equipment group of the current production flow as a center, and establishing a network model obtained after branching between each production and manufacturing equipment and adjacent upstream and downstream production and manufacturing equipment; the target production and manufacturing equipment group is the production and manufacturing equipment group with the largest number of production and manufacturing equipment in the current production flow;

the determining module is used for:

determining failure times of each failure task path;

the positioning module is used for: positioning final fault equipment of the production and manufacturing network large model according to the fault failure times of each fault task path;

The acquisition module is used for: acquiring a target fault description corresponding to the final fault equipment;

the determining module is used for: determining a target fault reason of the final fault equipment in an abnormal working state in an equipment fault knowledge graph based on the target fault description; the equipment fault knowledge graph is as follows: a plurality of fault descriptions and a plurality of knowledge maps of fault cause association relations;

the auxiliary decision module is used for: and making an auxiliary decision on the production manufacturing network large model based on the target fault description.

In a third aspect, embodiments of the present application also provide an electronic device including a memory storing a plurality of computer programs; the processor loads a computer program from the memory to perform any of the AI-large model-based production quality assistance decision making methods provided by embodiments of the present application.

In a fourth aspect, embodiments of the present application further provide a computer readable storage medium storing a plurality of computer programs, where the computer programs are adapted to be loaded by a processor to perform any of the AI-large model-based production quality assistance decision making methods provided by the embodiments of the present application.

In a fifth aspect, embodiments of the present application further provide a computer program product, including a computer program or a computer program, which when executed by a processor implements any of the AI-large model-based production quality assistance decision making methods provided by embodiments of the present application.

According to the method and the device for positioning the fault equipment, the fault equipment in the production and manufacturing process can be rapidly positioned through the large directional production and manufacturing network model, meanwhile, the fault root cause of the fault equipment in the production and manufacturing process can be rapidly positioned through the equipment fault knowledge graph, and accordingly production quality auxiliary decision can be accurately made according to the fault equipment and the fault root cause in the production and manufacturing process.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly introduced below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flow chart of a quality of production aid decision making method based on an AI large model provided in an embodiment of the application;

FIG. 2 is a schematic diagram of a large model of a manufacturing network provided in an embodiment of the present application;

fig. 3 is a schematic structural diagram of an equipment failure knowledge graph according to an embodiment of the present application;

FIG. 4 is a schematic structural diagram of an AI-based mass-model decision making assist apparatus provided in an embodiment of the application;

fig. 5 is a schematic structural diagram of an electronic device provided in an embodiment of the present application.

Detailed Description

The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all, of the embodiments of the present application. All other embodiments, which can be made by those skilled in the art based on the embodiments herein without making any inventive effort, are intended to be within the scope of the present application. Meanwhile, in the description of the embodiments of the present application, the terms "first", "second", and the like are used only to distinguish the description, and are not to be construed as indicating or implying relative importance. Thus, features defining "first", "second" may explicitly or implicitly include one or more features. In the description of the embodiments of the present application, the meaning of "a plurality" is two or more, unless explicitly defined otherwise.

The embodiment of the application provides a production quality auxiliary decision-making method and system based on an AI large model. Specifically, the embodiments of the present application will be described from the point of view of an AI large model-based production quality assistance decision making apparatus, which may be integrated in an electronic device in particular, that is, the AI large model-based production quality assistance decision making method of the embodiments of the present application may be executed by the electronic device. Optionally, the electronic device includes a terminal device. The terminal device may be a mobile phone, a tablet computer, a smart bluetooth device, a notebook computer, a game console, or a personal computer (Personal Computer, PC), etc. Optionally, the electronic device includes a server, which may be a stand alone server, or may be a server network or a server cluster including, but not limited to, a computer, a network host, a single network server, a network server set, or a cloud server formed by servers. Wherein the Cloud server is composed of a large number of computers or web servers based on Cloud Computing (Cloud Computing).

The following description of the embodiments is not intended to limit the preferred embodiments. Although a logical order is depicted in the flowchart, in some cases the steps shown or described may be performed in an order different than depicted in the figures.

The following detailed description is provided with reference to the accompanying drawings, and the embodiments of the present application use the auxiliary decision making device as an execution subject for illustration. Referring to fig. 1, fig. 1 is a flow chart of a method for assisting decision making based on AI large models provided in an embodiment of the present application. The specific flow of the production quality auxiliary decision-making method based on the AI large model provided in the embodiment of the present application may include the following steps 101 to 107:

step 101, obtaining a fault task of a large production and manufacturing network model in the production and manufacturing process;

step 102, determining all production manufacturing equipment through which each faulty task passes, and determining paths formed by all production manufacturing equipment in each faulty task as faulty task paths;

step 103, determining the failure times of each failure task path;

step 104, positioning final fault equipment in the production and manufacturing network large model under the abnormal working state according to the fault failure times of each fault task path;

step 105, obtaining a target fault description corresponding to the final fault equipment;

step 106, determining a target fault reason of the final fault equipment in an abnormal working state in an equipment fault knowledge graph based on the target fault description;

And step 107, making an auxiliary decision on the production and manufacturing network large model based on the target fault cause.

Optionally, the auxiliary decision-making device acquires a failed task generated in the production and manufacturing process of the large production and manufacturing network model, and determines the failed task as a failed task. The production and manufacture network large model is as follows: and taking each production and manufacturing equipment in the target production and manufacturing equipment group of the current production flow as a center, establishing a network model obtained after branching between each production and manufacturing equipment and adjacent upstream and downstream production and manufacturing equipment, wherein the target production and manufacturing equipment group is the production and manufacturing equipment group with the largest number of production and manufacturing equipment in the current production flow.

Further, the auxiliary decision-making device determines all production manufacturing equipment through which each faulty task passes, and determines a path formed by all production manufacturing equipment in each faulty task as a faulty task path.

When the production and manufacturing equipment fails, the failure times of some failure tasks are more than one time, so that the failed production and manufacturing equipment can share the failure times to adjacent branches. Thus, the auxiliary decision means determines the number of failed times of each failed task path, wherein the number of failed times is the number of occurrences of the failed task path.

Further, the auxiliary decision device determines the failure times of each production manufacturing device in each failed task path according to the failure times of each failed task path.

Further, the auxiliary decision-making device compares the failure times of the same manufacturing equipment in all the failure task paths, and determines the same manufacturing equipment with the largest total failure times as the final failure equipment in the manufacturing network large model under the abnormal working condition.

Further, the auxiliary decision device acquires a target fault description corresponding to the final fault device, wherein the target fault description is a phenomenon description corresponding to the final fault device in an abnormal working state. In one embodiment, the target failure of the final malfunctioning device in the abnormal working state of the mechanical failure is described as "excessive wear or fatigue of the device is present, leading to failure of the drive chain or bearings; the equipment is not maintained or maintained on time, so that parts lack lubrication or damage; the motor, gears, pulleys, etc. elements in the apparatus wear or slip.

The target fault of the final fault equipment in the abnormal working state of the electrical fault is described as 'short circuit, overload, open circuit or poor contact of an electric wire or a wiring cabinet'; control circuit faults, such as contactor faults, PLC control card faults, etc.; power failure such as current ripple, voltage undershoot or undershoot.

The target fault of the final fault equipment in the abnormal working state of the hydraulic fault is described as the problem that elements (such as an oil pump, an oil cylinder, a valve and the like) in a hydraulic system are damaged and blocked; oil leakage or insufficient pressure causes the equipment to fail to work normally; the filter in the hydraulic system is not cleaned or damaged, so that impurities and dust are mixed in the oil.

The target fault of the final fault equipment in the abnormal working state of the pneumatic fault is described as the problem that elements (such as a cylinder, a valve and the like) in a pneumatic system are damaged and blocked; abnormal air pressure, such as too low or severe fluctuation; the air leakage in the pneumatic system causes the equipment to fail to work properly.

The target fault of the final fault equipment in the abnormal working state of the automatic control fault is described as a control system fault, such as a program error or improper parameter setting; sensor faults, such as sensor dust accumulation, loose wiring and the like; actuator failure, such as solenoid valve failure or motor drive failure.

Further, under the condition that the equipment fault knowledge graph comprises a plurality of fault descriptions and a plurality of fault reasons, the auxiliary decision device determines the fault reason corresponding to the target fault description as the target fault reason of the final fault equipment in an abnormal working state, wherein the equipment fault knowledge graph is as follows: and a knowledge graph of the association relationship between the plurality of fault descriptions and the plurality of fault reasons. Optionally, the device fault knowledge graph is constructed by a preset tool, for example, apache Spark GraphFrames, and the device fault knowledge graph is constructed by the following first to second steps:

And the first step, carrying out knowledge extraction on the equipment fault information data to obtain entity information and attribute information.

And secondly, filling the entity information and the attribute information into a model layer body of the attribute map to obtain the equipment fault knowledge map.

The device fault knowledge Graph may be a Property Graph (Property Graph). An attribute graph is a graph structure that includes nodes, edges and attributes, the nodes representing entities, the edges connecting a pair of nodes, the edges being relationships between nodes, the nodes and edges having labels and attributes. For example, the attributes of an entity may be fault description, fault name, fault cause, fault solution, solution type, fault possible incident, fault occurrence location and associated fault name, etc. For example, the properties of an edge may be "cause," "location," "description," "schema type," "schema" or "association," etc.

For example, the relationship between the fault description and the fault name is "description", the relationship between the fault occurrence location and the fault name is "location", the relationship between the fault solution and the fault name is "solution", the relationship between the fault solution and the solution type is "solution type", the relationship between the fault possibly induced accident and the fault name is "initiation", the relationship between the fault cause and the fault name is "cause", and the relationship between the associated fault name and the fault name is "association".

The apparatus failure knowledge graph is exemplarily described below with reference to fig. 3. Fig. 3 is a schematic structural diagram of an equipment failure knowledge graph according to an embodiment of the present application. As shown in fig. 3, entities 1 to 8 included in the device failure knowledge graph; wherein, the attribute of entity 1 is fault description, the attribute of entity 2 is fault name, the attribute of entity 3 is fault cause, the attribute of entity 4 is fault solution, the attribute of entity 5 is solution type, the attribute of entity 6 is fault possible accident, the attribute of entity 7 is fault occurrence position, and the attribute of entity 8 is associated fault name.

Further, the auxiliary decision-making device makes an auxiliary decision for the large model of the production and manufacture network according to the target fault cause. In one embodiment, the target fault is "power failure, voltage too low", and the auxiliary decision for producing a large model of the manufacturing network is "turning up the power supply voltage". The objective failure is due to "excessive wear or fatigue of the bearings of the equipment leading to failure of the bearings", then the auxiliary decision for producing a large model of the manufacturing network is "replace bearings of the equipment".

In one embodiment, the step of constructing the manufacturing network large model comprises:

acquiring all processing nodes through which products are produced, classifying each processing node according to the production service type of production and manufacturing equipment in the current production flow, and obtaining a target production equipment class of the current production flow;

acquiring a first-level production equipment class and a second-level production equipment class in the target production equipment classes; the first-level production equipment is the most production equipment in the target production equipment, and the second-level production equipment is the production equipment except the first-level production equipment in the target production equipment;

taking twice the number of the production equipment of the primary production equipment class as the task number, taking each first production and manufacturing equipment in the primary production equipment class as the center, and respectively carrying out test task arrangement on the second production and manufacturing equipment in the secondary production equipment class at two sides to obtain the large production and manufacturing network model; in the large production and manufacturing network model, two branches are respectively arranged at the upstream and downstream of each first production and manufacturing device, and each second production and manufacturing device is connected with at least two first production and manufacturing devices.

Optionally, when the large production and manufacturing network model needs to be built, the auxiliary decision device needs to acquire a processing node through which the production product passes during the current production flow, where the current production flow in the embodiment of the present application is a production and manufacturing flow, and the processing node is a production and manufacturing device, so that the process node can be understood as acquiring the production and manufacturing device through which the production product passes during the current production flow.

Further, the auxiliary decision-making device classifies the processing nodes according to the production service types of the production and manufacturing equipment in the current production flow to obtain the target production equipment type of the current production flow, wherein the production service types can comprise an automobile service type, an electronic product service type, a household appliance manufacturing service type, a communication equipment service type and the like.

Further, the auxiliary decision-making device obtains the number of the production and manufacturing equipment of each production equipment in the target production equipment, compares the number of the production and manufacturing equipment of each production equipment, and determines the production equipment with the largest number of the production and manufacturing equipment.

Optionally, the auxiliary decision-making device determines the production equipment class with the largest number of single production and manufacturing equipment as the first-stage production equipment class in the target production equipment class. Further, the auxiliary decision-making device determines the production equipment class other than the primary production equipment class in the target production equipment class as the secondary production equipment class.

In an embodiment, the target production equipment classes include an a production equipment class, a B production equipment class and a C production equipment class, wherein the number of production and manufacturing equipment of the a production equipment class is i, the number of production and manufacturing equipment of the B production equipment class is j, and the number of production and manufacturing equipment of the C production equipment class is k. The number j is greater than the number i and the number j is greater than the number k, so that the production equipment class B is determined to be a primary production equipment class, and the production equipment class A and the production equipment class C are determined to be secondary production equipment classes.

Further, the auxiliary decision-making device takes twice the number of the production and manufacturing devices of the first-stage production device class as the task number, namely the task number= (the number of the production and manufacturing devices of the first-stage production device class) x 2, and meanwhile, the auxiliary decision-making device takes each first production and manufacturing device of the first-stage production device class as the center, and performs test task arrangement on second production and manufacturing devices of two-stage production device classes on two sides respectively, so that a large production and manufacturing network model of the current production flow is obtained. Thus, in the large production network model, there are two branches upstream and downstream of each first production device, and each second production device is connected to at least two first production devices.

In an embodiment, the target production equipment classes include an a production equipment class, a B production equipment class and a C production equipment class, wherein the number of production and manufacturing equipment of the a production equipment class is i, the number of production and manufacturing equipment of the B production equipment class is j, and the number of production and manufacturing equipment of the C production equipment class is k. The number j is greater than the number i and the number j is greater than the number k, and therefore, the B production equipment class is determined as a primary production equipment class, and the number of tasks to produce the large model of the manufacturing network=max (the number of production manufacturing equipment) ×2=2j.

Taking B1 production and manufacturing equipment, B2 production and manufacturing equipment and B3 production and manufacturing equipment in the B production and manufacturing equipment as centers, respectively carrying out test task arrangement on the production and manufacturing equipment of the A production and manufacturing equipment and the C production and manufacturing equipment on two sides to obtain a large production and manufacturing network model, wherein each production and manufacturing equipment of the A production and manufacturing equipment and the C production and manufacturing equipment is connected with at least two first production and manufacturing equipment due to the fact that the number of j is the largest, so that double-fork is realized, and the concrete analysis is as follows: the B1 production and manufacturing equipment is respectively connected with the A1 production and manufacturing equipment and the A2 production and manufacturing equipment to establish 2 upstream branches, and at the moment, the B1 production and manufacturing equipment is respectively connected with the C1 production and manufacturing equipment and the C2 production and manufacturing equipment to establish 2 downstream branches; the B2 production and manufacturing equipment is connected with the A1 production and manufacturing equipment and the Ai production and manufacturing equipment respectively to establish 2 upstream branches, and at the moment, the B2 production and manufacturing equipment is connected with the C1 production and manufacturing equipment and the Ck production and manufacturing equipment respectively to establish 2 downstream branches; the B3 production and manufacturing equipment is respectively connected with the A2 production and manufacturing equipment and the Ai production and manufacturing equipment to establish 2 upstream branches, and at the moment, the B3 production and manufacturing equipment is respectively connected with the C2 production and manufacturing equipment and the Ck production and manufacturing equipment to establish 2 downstream branches; .. the Bj manufacturing equipment is connected with the A1 manufacturing equipment and the A2 manufacturing equipment respectively to establish 2 upstream branches, and at this time, the Bj manufacturing equipment is connected with the C1 manufacturing equipment and the C2 manufacturing equipment respectively to establish 2 downstream branches, so that a manufacturing network large model is constructed, and referring to FIG. 2, FIG. 2 is a schematic structural diagram of the manufacturing network large model provided in the embodiment of the present application.

According to the method and the device for positioning the fault equipment, the large production and manufacturing network model is built, so that the fault equipment can be positioned through the large production and manufacturing network model, and the production quality auxiliary decision can be accurately made according to the fault equipment in the production and manufacturing process.

In one embodiment, locating the final faulty device of the manufacturing network large model according to the number of faulty failures of each faulty task path includes:

determining the failure times of each production manufacturing device in each failed task path according to the failure times of each failed task path;

adding and summing the failure times of the same production and manufacturing equipment to obtain the total failure times of any same production and manufacturing equipment in all failure task paths;

comparing the total failure times of the same production and manufacturing equipment to obtain a comparison result;

and determining the same production and manufacturing equipment with the largest total failure frequency according to the comparison result, and positioning the same production and manufacturing equipment with the largest total failure frequency as the final failure equipment.

Optionally, the auxiliary decision-making device determines the failure times of each production manufacturing device in each failed task path according to the failure times of each failed task path.

Further, the auxiliary decision device adds and sums the failure times of the same production and manufacturing equipment in all the failure task paths to obtain the total failure times of any same production and manufacturing equipment in all the failure task paths.

Further, the auxiliary decision device compares the total failure times of the same production and manufacturing equipment to obtain a comparison result.

Further, the auxiliary decision-making device determines the same production and manufacturing equipment with the largest total failure frequency according to the comparison result, and positions the same production and manufacturing equipment with the largest total failure frequency as the final fault equipment. In an embodiment, the failed task path and the failure times thereof determined according to the failed task are:

task path 1: A1-B1-C1, failure times are 2 times; task path 2: A1-B1-C2, the failure times are 1 time; task path 3: A1-B2-C1, failure times are 2 times; task path 4: A1-B2-C3, the failure times are 4 times; task path 9: A2-B1-C1, failure times are 2 times; task path 10: A2-B1-C2, failure times are 2 times; task path 11: A2-B2-C1, the failure times are 4 times; task path 12: A2-B2-C3, the failure times are 3 times. Therefore, the number of failure failures of each production manufacturing apparatus in the task path 1 is 2; the failure frequency of each production manufacturing device in the task path 2 is 1 time; the failure times of each production manufacturing device in the task path 3 are 2 times; the failure times of each production manufacturing device in the task path 4 are 4 times; the number of failure times of each production manufacturing apparatus in the task path 9 is 2; the number of failure times of each production manufacturing apparatus in the task path 10 is 2; the number of failure times of each production manufacturing apparatus in the task path 11 is 4; the number of failure failures of each production manufacturing apparatus in the task path 12 was 3.

Therefore, for the production manufacturing apparatus A1, the total number of failures thereof is 2+1+2+4=9 times; for the production manufacturing apparatus A2, the total number of failures thereof is 2+2+4+3=11 times; for the production manufacturing apparatus B1, the total number of failures thereof is 2+1+2+2=7 times; for the production manufacturing apparatus B2, the total number of failures thereof is 2+4+4+3=13 times; for the production manufacturing apparatus C1, the total number of failures thereof is 2+2+2+4=10 times; for the production manufacturing apparatus C2, the total number of failures thereof is 1+2=3 times; for the production manufacturing apparatus C3, the total number of failures thereof was 4+3=7 times.

Since 13 times is the maximum number of times, the production manufacturing apparatus B2 is positioned as the final malfunctioning apparatus.

In the process of positioning faults of the fault equipment through the production and manufacture network large model, each production and manufacture equipment is provided with the adjacent branch, and when the equipment breaks down, the fault failure of the final fault equipment can be shared to the adjacent branch, so that the fault failure times of the final fault equipment are the greatest, and the final fault equipment can be positioned rapidly.

In an embodiment, based on the target fault description, determining, in an equipment fault knowledge graph, a target fault cause of the final faulty equipment in an abnormal working state includes:

Searching a fault name corresponding to the target fault description in the equipment fault knowledge graph;

under the condition that the fault name corresponding to the target fault description is found, determining the fault name as a target fault name;

and determining a target fault reason of the final fault equipment in an abnormal working state in the equipment fault knowledge graph based on the target fault name.

Optionally, the auxiliary decision-making device searches the fault name corresponding to the target fault description in the equipment fault knowledge graph. Optionally, under the condition that the fault name corresponding to the target fault description is found, the auxiliary decision device determines the fault name as the target fault name, and determines a target fault reason of the final fault equipment in an abnormal working state in the equipment fault knowledge graph according to the target fault name.

According to the method and the device for determining the fault root cause of the fault equipment, the fault root cause of the fault equipment in the production and manufacturing process can be rapidly located through the equipment fault knowledge graph, and therefore production quality auxiliary decision can be accurately made according to the fault root cause.

In an embodiment, determining, in the device fault knowledge graph, a target fault cause of the final faulty device in an abnormal working state based on the target fault name includes:

Acquiring a plurality of fault reasons corresponding to the target fault names from the equipment fault knowledge graph;

determining the occurrence times of the fault reasons based on the historical fault information of the final fault equipment aiming at each fault reason in the plurality of fault reasons;

and determining the fault reason with the largest occurrence number as the target fault reason.

Optionally, the auxiliary decision-making device acquires a plurality of fault reasons corresponding to the target fault names from the equipment fault knowledge graph. Optionally, the auxiliary decision device determines, for each of the plurality of fault reasons, the number of occurrences of the fault reason based on historical fault information of the final fault device. Optionally, the auxiliary decision device determines the fault cause with the largest occurrence number as the target fault cause.

According to the method and the device for determining the fault root cause of the fault equipment, the fault root cause of the fault equipment in the production and manufacturing process can be rapidly located through the equipment fault knowledge graph, and therefore production quality auxiliary decision can be accurately made according to the fault root cause in the production and manufacturing process.

Judging whether the target fault corresponding to the target fault name is a hot fault or not;

under the condition that the target fault is a hot fault, processing each fault reason by a first preset model aiming at each fault reason in a plurality of fault reasons corresponding to the target fault name in the equipment fault knowledge graph to obtain prediction scoring of each fault reason; the first preset model is as follows:

；

wherein,indicating a failure of the target and,the cause of the failure is indicated and,indicating the cause of the faultIs provided with a predictive scoring of (a),representing a target faultIs used for the average scoring of (a),a word vector representing the name of the target fault,an associated fault that represents a target fault,representing an associated faultIs a word vector of the associated trouble name,representation ofAndthe degree of similarity of the word vectors between them,a set of all associated fault components representing the target fault,representing the cause of the associated failureIs the cause of the failure of (a)Is provided with a score of (a),representing an associated faultIs used for the average scoring of (a),representing a first preset scoring;

judging whether the fault cause is a hot cause according to each fault cause under the condition that the target fault is a non-hot fault;

under the condition that the fault cause is a hot cause, determining a predictive scoring of the fault cause through a second preset model; the second preset model specifically comprises the following steps:

；

Wherein,representing a second preset scoring;

under the condition that the failure cause is not a hot cause, determining the second preset scoring as the predictive scoring of the failure cause;

and determining the fault reason with the largest predictive scoring among the plurality of fault reasons as the target fault reason.

Optionally, the auxiliary decision device determines the occurrence number of the target fault name based on the historical fault information of the final fault device, and determines that the target fault corresponding to the target fault name is a hot fault when the occurrence number of the target fault name is greater than or equal to a preset number of times, wherein the preset number of times is set according to the actual situation.

Optionally, the auxiliary decision device processes each fault reason through a first preset model aiming at each fault reason in a plurality of fault reasons corresponding to the target fault name in the equipment fault knowledge graph under the condition that the target fault is a hot fault, so as to obtain a prediction score of each fault reason; the first preset model is as follows:

；

wherein,indicating a failure of the target and,the cause of the failure is indicated and,indicating the cause of the faultIs provided with a predictive scoring of (a),representing a target faultIs used for the average scoring of (a),a word vector representing the name of the target fault, An associated fault that represents a target fault,representing an associated faultIs a word vector of the associated trouble name,representation ofAndthe degree of similarity of the word vectors between them,a set of all associated fault components representing the target fault,representing the cause of the associated failureIs the cause of the failure of (a)Is provided with a score of (a),representing an associated faultIs used for the average scoring of (a),a first preset score is indicated.

Optionally, the historical fault information of the final fault device includes a scoring of a plurality of faults and a plurality of fault causes causing the faults, the plurality of faults including associated faultsAnd target failure；The acquisition method of (1) comprises the following steps:

obtaining the target fault from the historical fault informationScoring of a plurality of failure causes of (a);

and determining an average value of scores of a plurality of fault reasons as an average score of the target faults.

Acquisition method of (a)The acquisition method of (2) is similar and will not be described in detail herein.

Indicating a fault at the targetIs the error cause and causes the target failureIn case the scoring of the failure cause i is highest, the first preset score needs to be subtracted.

Optionally, based on the historical fault information of the final fault equipment, statistics is performed on fault reasons in different occurrence times to obtain occurrence times of the fault reasons, and when the occurrence times of the fault reasons are greater than or equal to preset times, the fault reasons are determined to be hot reasons.

Optionally, in the case that the failure cause is a hot cause, the auxiliary decision-making device determines a predictive score of the failure cause through a second preset model; the second preset model is specifically:

；

wherein,representing a second preset scoring;

optionally, in the case that the failure cause is not a hot cause, the auxiliary decision device determines a second preset score as a predictive score for the failure cause.

Optionally, the auxiliary decision device determines a fault cause with the largest predictive score among the plurality of fault causes as the target fault cause.

In an embodiment, the production quality auxiliary decision method based on the AI large model further comprises:

determining, for each of a plurality of fault descriptions, a text similarity of the target fault description and the fault description if the target fault cause is determined to be an error cause;

determining the fault description with the maximum text similarity as an intermediate fault description;

Determining a new fault name based on an intermediate fault name corresponding to the intermediate fault description in the equipment fault knowledge graph;

and determining a new fault reason of the final fault equipment in an abnormal working state in the equipment fault knowledge graph based on the new fault name.

Optionally, the device fault knowledge graph includes a plurality of fault descriptions.

Optionally, in the case that the target fault cause is determined to be an error cause, the auxiliary decision device determines, for each fault description in the plurality of fault descriptions, a text similarity of the target fault description and the fault description.

Optionally, the auxiliary decision-making device determines the fault description with the maximum text similarity as the intermediate fault description. Optionally, the auxiliary decision device determines a new fault name based on an intermediate fault name corresponding to the intermediate fault description in the equipment fault knowledge graph, and optionally, the auxiliary decision device determines a new fault reason of the final fault equipment in an abnormal working state based on the new fault name in the equipment fault knowledge graph.

The following describes the AI large model-based production quality auxiliary decision device provided in the embodiments of the present application, and the AI large model-based production quality auxiliary decision device described below and the AI large model-based production quality auxiliary decision method described above may be referred to correspondingly with each other. Referring to fig. 4, fig. 4 is a schematic structural diagram of an AI large model-based production quality assistance decision making apparatus provided in an embodiment of the present application, and the AI large model-based production quality assistance decision making apparatus may include: an acquisition module 401, a determination module 402, a positioning module 403 and an auxiliary decision module 404;

the acquisition module 401 is configured to: acquiring a fault task of a large production and manufacturing network model in the production and manufacturing process; the production and manufacturing network large model is as follows: taking each production and manufacturing equipment in a target production and manufacturing equipment group of the current production flow as a center, and establishing a network model obtained after branching between each production and manufacturing equipment and adjacent upstream and downstream production and manufacturing equipment; the target production and manufacturing equipment group is the production and manufacturing equipment group with the largest number of production and manufacturing equipment in the current production flow;

the determining module 402 is configured to:

Determining failure times of each failure task path;

the positioning module 403 is configured to: positioning final fault equipment of the production and manufacturing network large model according to the fault failure times of each fault task path;

the acquisition module 401 is configured to: acquiring a target fault description corresponding to the final fault equipment;

the determining module 402 is configured to: determining a target fault reason of the final fault equipment in an abnormal working state in an equipment fault knowledge graph based on the target fault description; the equipment fault knowledge graph is as follows: a plurality of fault descriptions and a plurality of knowledge maps of fault cause association relations;

the auxiliary decision module 404 is configured to: and making an auxiliary decision on the production manufacturing network large model based on the target fault description.

In an alternative example, the AI large model-based production quality assistance decision making apparatus is further configured to:

In an alternative example, the positioning module 403 is further configured to:

In an alternative example, the determination module 402 is further configured to:

；

wherein,indicating a failure of the target and,the cause of the failure is indicated and,indicating the cause of the faultIs provided with a predictive scoring of (a),representing a target faultIs used for the average scoring of (a),a word vector representing the name of the target fault,an associated fault that represents a target fault,representing an associated fault Is a word vector of the associated trouble name,representation ofAndthe degree of similarity of the word vectors between them,all relations representing target faultsA set of linked faults is formed,representing the cause of the associated failureIs the cause of the failure of (a)Is provided with a score of (a),representing an associated faultIs used for the average scoring of (a),representing a first preset scoring;

；

wherein,representing a second preset scoring;

The specific embodiments of the production quality auxiliary decision-making device based on the AI large model provided in the present application are basically the same as each embodiment of the production quality auxiliary decision-making method based on the AI large model, and are not described herein.

Optionally, as shown in fig. 5, fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present application. The electronic device may include: processor 510, communication interface (Communication Interface) 520, memory 530, and communication bus 540, wherein processor 510, communication interface 520, memory 530 complete communication with each other through communication bus 540. Processor 510 may invoke a computer program in memory 530 to perform the quality of production aid decision making method based on AI large models, including, for example:

determining failure times of each failure task path;

The embodiment of the electronic device in the present application is substantially the same as the embodiment of the above-mentioned production quality auxiliary decision method based on AI big model, and will not be described herein.

Furthermore, the logic computer program in the memory 530 may be implemented in the form of a software functional unit and may be stored in a computer readable storage medium when sold or used as a separate product. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising a number of computer programs for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.

In another aspect, embodiments of the present application further provide a non-transitory computer readable storage medium, where the non-transitory computer readable storage medium includes a computer program, where the computer program is stored on the non-transitory computer readable storage medium, and when the computer program is executed by a processor, the computer program is capable of executing the steps of the AI-based large model production quality auxiliary decision method provided by the foregoing embodiments, for example, including:

determining failure times of each failure task path;

The specific embodiments of the readable storage medium of the present application are substantially the same as the specific embodiments of the above-mentioned production quality auxiliary decision method based on AI big model, and will not be repeated here.

In yet another aspect, embodiments of the present application further provide a computer product, where the computer product includes a computer program, where the computer program may be stored on the computer product, and when the computer program is executed by a processor, the computer is capable of executing the steps of the AI large model-based production quality auxiliary decision method provided in the foregoing embodiments, for example, including:

determining failure times of each failure task path;

The embodiments of the computer product of the present application are substantially the same as the embodiments of the above-mentioned quality-of-product decision-making method based on AI big model, and will not be described herein.

The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.

From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the above technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., comprising a number of computer programs for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or parts of the embodiments.

Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present application, and are not limiting thereof; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the corresponding technical solutions.

Claims

1. The production quality auxiliary decision-making method based on the AI large model is characterized by comprising the following steps of:

Determining failure times of each failure task path;

performing auxiliary decision making on the large production and manufacturing network model based on the target fault cause;

the determining, based on the target fault description, a target fault cause of the final fault device in an abnormal working state in a device fault knowledge graph includes:

determining a target fault reason of the final fault equipment in an abnormal working state in the equipment fault knowledge graph based on the target fault name;

The determining, based on the target fault name, a target fault cause of the final faulty device in an abnormal working state in the device fault knowledge graph includes:

where u denotes the target fault, i denotes the cause of the fault, P (u, i) denotes the predictive scoring of the cause of the fault i,mean scoring representing target fault u, P _u Word vector representing target fault name, v representing associated fault of target fault, P _v Word vector representing associated fault name of associated fault v, sim (P _u ,P _v ) Representing P _u And P _v Word vector similarity between the target faults, V represents the set of all associated fault components of the target faults, and r _vi Scoring of the cause of failure i, which leads to an associated failure v>Mean score, r, representing an associated fault v _ui Representing a first preset scoring; the historical fault information of the final fault equipment comprises a plurality of faults and scores of a plurality of fault reasons causing the faults, wherein the plurality of faults comprise associated faults v and target faults u; obtaining scores of a plurality of fault reasons causing the target fault u from the historical fault information, and determining an average value of the scores of the plurality of fault reasons as an average score +. >Obtaining scores of a plurality of fault reasons causing the associated fault v from the historical fault information, anddetermining an average value of scores of a plurality of fault causes as an average score +.>First preset scoring r _ui Is a preset fixed score;

wherein r is _b Representing a second preset score, r _b Is a preset fixed score;

2. The AI-large-model-based production quality assistance decision making method of claim 1, wherein the step of constructing the production manufacturing network large model comprises:

taking twice the number of the production equipment of the primary production equipment class as the task number, taking each first production and manufacturing equipment in the primary production equipment class as the center, and respectively carrying out test task arrangement on the second production and manufacturing equipment in the secondary production equipment class at two sides to obtain the large production and manufacturing network model; in the large production and manufacturing network model, two branches are respectively arranged at the upstream and downstream of each first production and manufacturing device, and each second production and manufacturing device is connected with at least two first production and manufacturing devices; the first production manufacturing equipment is any one of the primary production equipment classes, and the second production manufacturing equipment is any one of the secondary production equipment classes.

3. The AI-large-model-based production quality assistance decision making method according to claim 1, wherein said locating the final malfunctioning device of the production manufacturing network large model according to the number of failure times of each of the malfunctioning task paths comprises:

4. The AI-large-model-based production quality auxiliary decision method according to claim 1, wherein the determining, in the equipment fault knowledge graph, a target fault cause of the final faulty equipment in an abnormal operation state based on the target fault name includes:

5. The AI-large-model-based production quality assistance decision making method according to any one of claims 1 to 4, wherein a plurality of fault descriptions are included in the equipment fault knowledge graph; the method further comprises the steps of:

6. An AI large model-based production quality auxiliary decision-making device, comprising: the system comprises an acquisition module, a determination module, a positioning module and an auxiliary decision-making module;

The determining module is used for:

determining failure times of each failure task path;

the auxiliary decision module is used for: making an auxiliary decision on the production manufacturing network large model based on the target fault description;

where u denotes the target fault, i denotes the cause of the fault, P (u, i) denotes the predictive scoring of the cause of the fault i,mean scoring representing target fault u, P _u Word vector representing target fault name, v representing associated fault of target fault, P _v Word vector representing associated fault name of associated fault v, sim (P _u ,P _v ) Representing P _u And P _v Word vector similarity between the target faults, V represents the set of all associated fault components of the target faults, and r _vi Scoring of the cause of failure i, which leads to an associated failure v>Mean score, r, representing an associated fault v _ui Representing a first preset scoring; the historical fault information of the final fault equipment comprises a plurality of faults and scores of a plurality of fault reasons causing the faults, wherein the plurality of faults comprise associated faults v and target faults u; obtaining scores of a plurality of fault reasons causing the target fault u from the historical fault information, and determining an average value of the scores of the plurality of fault reasons as an average score +.>Obtaining scores of a plurality of fault reasons causing the associated fault v from the historical fault information, and determining an average value of the scores of the plurality of fault reasons as an average score +.>First preset scoring r _ui Is a preset fixed score;

7. An electronic device comprising a processor and a memory, the memory storing a plurality of computer programs; the processor loads a computer program from the memory to perform the AI-large model-based production quality assistance decision making method of any one of claims 1-5.

8. A computer-readable storage medium, characterized in that it stores a plurality of computer programs adapted to be loaded by a processor for executing the AI big model-based production quality assistance decision making method according to any one of claims 1 to 5.