CN116467592A - Production equipment fault intelligent monitoring method and system based on deep learning - Google Patents
Production equipment fault intelligent monitoring method and system based on deep learning Download PDFInfo
- Publication number
- CN116467592A CN116467592A CN202310349473.1A CN202310349473A CN116467592A CN 116467592 A CN116467592 A CN 116467592A CN 202310349473 A CN202310349473 A CN 202310349473A CN 116467592 A CN116467592 A CN 116467592A
- Authority
- CN
- China
- Prior art keywords
- fault
- production equipment
- feature vector
- parameter feature
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 110
- 238000012544 monitoring process Methods 0.000 title claims abstract description 53
- 238000013135 deep learning Methods 0.000 title claims abstract description 22
- 238000000034 method Methods 0.000 title claims abstract description 17
- 238000012549 training Methods 0.000 claims abstract description 67
- 238000010276 construction Methods 0.000 claims abstract description 3
- 239000013598 vector Substances 0.000 claims description 104
- 230000002159 abnormal effect Effects 0.000 claims description 70
- 238000007781 pre-processing Methods 0.000 claims description 9
- 238000012216 screening Methods 0.000 claims description 5
- 238000002360 preparation method Methods 0.000 abstract description 2
- 238000012545 processing Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012163 sequencing technique Methods 0.000 description 2
- 238000005520 cutting process Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0475—Generative networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/094—Adversarial learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Computational Linguistics (AREA)
- Marketing (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Molecular Biology (AREA)
- Tourism & Hospitality (AREA)
- Strategic Management (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Manufacturing & Machinery (AREA)
- Primary Health Care (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention provides an intelligent monitoring method and system for production equipment faults based on deep learning. Production facility trouble intelligent monitoring system based on deep learning includes: the data acquisition module is used for acquiring historical data of the production equipment and operation data of the production equipment; the model construction module is used for constructing a fault prediction model; the sample preparation module is used for preparing training sample data; the model training module is used for training the fault prediction model; and the fault monitoring module is used for carrying out fault monitoring on the production equipment through the fault prediction model. According to the invention, the first training sample data and the second training sample data are obtained through building the fault prediction model and the production equipment historical data, the fault prediction model is built through training the first training sample data and the second training sample data, the fault prediction model is trained to monitor the fault of the production equipment operation data acquired in real time, and the fault monitoring efficiency is improved.
Description
Technical Field
The invention relates to the technical field of fault monitoring, in particular to an intelligent monitoring method and system for production equipment faults based on deep learning.
Background
In recent years, the production and manufacturing industry of China rapidly develops, more and more enterprises adopt various advanced production for intelligent production, and the production efficiency is remarkably improved. The more complex the device, the higher the requirements for monitoring and maintenance of the device. In the production process, the stability of production equipment is one of the important factors influencing the production efficiency, and once the production equipment fails, the production efficiency of enterprises is directly influenced. When the production equipment is in an operating state, the difficulty of analyzing the operating data of the production equipment by a worker is high, certain requirements are met on the experience of the worker, and the monitoring efficiency is low.
Disclosure of Invention
Aiming at the problems, it is necessary to provide a production equipment fault intelligent monitoring method and system based on deep learning, which are used for analyzing the operation data of the production equipment in real time based on the deep learning technology, so as to improve the monitoring efficiency.
As one aspect of the present application, there is provided a production facility fault intelligent monitoring method based on deep learning, including:
acquiring production equipment historical data, wherein the production equipment historical data comprises operation normal data and operation fault data, the operation normal data comprises normal parameter information, and the operation fault data comprises abnormal parameter information and fault event information;
constructing a fault prediction model, wherein the fault prediction model comprises a fault pre-estimating sub-model and a fault positioning sub-model;
for the historical data of the production equipment, extracting normal parameter information to construct a normal parameter feature vector set, and extracting abnormal parameter information to construct an abnormal parameter feature vector set;
each abnormal parameter feature vector in the abnormal parameter feature vector set is taken as a core, a plurality of groups of fault pre-estimated sample data are constructed and obtained, first training sample data are formed, and the fault pre-estimated sub-model is trained by the first training sample data; respectively taking each fault event as a core, constructing a plurality of groups of fault positioning sample data based on an abnormal parameter feature vector set to form second training sample data, training a fault positioning sub-model by using the constructed plurality of groups of fault pre-estimated sample data, and training to obtain a fault prediction model;
and collecting production equipment operation data, inputting the production equipment operation data into a fault prediction model, and carrying out fault monitoring on the production equipment to obtain a fault monitoring result.
Further, the constructing to obtain a plurality of groups of fault pre-estimated sample data by using each abnormal parameter feature vector in the abnormal parameter feature vector set as a core, and forming the first training sample data includes:
and for any one abnormal parameter feature vector in the abnormal parameter feature vector set, selecting all normal parameter feature vectors with the similarity larger than a preset similarity threshold value from the normal parameter feature vector set, constructing and obtaining a group of fault estimated sample data, traversing the abnormal parameter feature vector set, obtaining fault estimated sample data corresponding to each abnormal parameter feature vector, and forming first training sample data.
Further, each fault event is taken as a core, a plurality of groups of fault location sample data are constructed based on the abnormal parameter feature vector set, and the second training sample data are formed by the steps of:
and extracting fault equipment information from any fault event in the production equipment history data, associating the fault equipment information with abnormal parameter information corresponding to the fault event, constructing a group of fault positioning sample data, traversing all fault events to obtain fault positioning sample data corresponding to each fault event, and forming second training sample data.
Further, the inputting the operation data of the production equipment into the fault prediction model, performing fault monitoring on the production equipment, and obtaining a fault monitoring result includes:
after the fault prediction model receives the production equipment operation data, the production equipment operation data is processed through a fault prediction sub-model to obtain a fault prediction probability, if the fault prediction probability is larger than a preset risk threshold, the production equipment operation data is input into a fault positioning sub-model, the production equipment operation data is processed through the fault positioning sub-model to obtain predicted fault equipment information, and a fault monitoring result is output; otherwise, directly outputting the fault monitoring result.
Further, for any one of the abnormal parameter feature vector sets, selecting all normal parameter feature vectors having a similarity with the abnormal parameter feature vector greater than a preset similarity threshold from the normal parameter feature vector set, and further including:
preprocessing the normal parameter feature vector set, calculating the difference value of each element in the abnormal parameter feature vector and the normal parameter feature vector set, screening out the elements with the difference value larger than a preset difference threshold value in the normal parameter feature vector set, and calculating the similarity of the abnormal parameter feature vector and each element remained in the normal parameter feature vector set.
Further, for the failure prediction model, the method further includes:
the fault prediction model is constructed based on generating a countermeasure network (GAN) as a framework.
As another aspect of the present application, there is provided a deep learning-based production equipment fault intelligent monitoring system, which is applied to the deep learning-based production equipment fault intelligent monitoring method described in any one of the above, including:
the data acquisition module is used for acquiring historical data of the production equipment and acquiring operation data of the production equipment;
the model construction module is used for a fault prediction model, and the fault prediction model comprises a fault pre-estimation sub-model and a fault positioning sub-model;
the sample making module is used for constructing a normal parameter feature vector set and an abnormal parameter feature vector set based on historical data of production equipment, and making first training sample data and second training sample data based on the normal parameter feature vector set and the abnormal parameter feature vector set;
the model training module is used for training the fault prediction model through the first training sample data and the second training sample data;
and the fault monitoring module is used for carrying out fault monitoring on the operation data of the production equipment through the trained fault prediction model.
Further, for the sample preparation module, the method further comprises:
and the preprocessing unit is used for preprocessing the normal parameter feature vector set.
The invention has the following advantages:
1. according to the invention, the first training sample data and the second training sample data are obtained by pre-constructing the fault prediction model and reproducing the historical data of the production equipment, the fault prediction model is constructed based on the training of the first training sample data and the second training sample data, the fault prediction model is used for carrying out fault monitoring on the operation data of the production equipment acquired in real time, and the fault monitoring efficiency is improved.
2. The fault prediction model constructed by the invention comprises a fault pre-estimation sub-model and a fault positioning sub-model, the production equipment operation data acquired in real time can be preprocessed through the fault pre-estimation sub-model, whether the production equipment operation data acquired in real time is further processed through the fault positioning sub-model is determined according to the processing result, and the data processing efficiency is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to the structures shown in these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic structural diagram of an intelligent monitoring system for production equipment faults based on deep learning in an embodiment of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, some embodiments of the present application will be described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application. However, those of ordinary skill in the art will understand that in the various embodiments of the present application, numerous technical details have been set forth in order to provide a better understanding of the present application. However, the technical solutions claimed in the present application can be implemented without these technical details and with various changes and modifications based on the following embodiments.
Example 1
The embodiment 1 of the invention provides a production equipment fault intelligent monitoring method based on deep learning, which comprises the following steps:
acquiring production equipment historical data, wherein the production equipment historical data comprises operation normal data and operation fault data, the operation normal data comprises normal parameter information, and the operation fault data comprises abnormal parameter information and fault event information;
it should be noted that, the production equipment history data may be specific information related to production equipment of a manufacturing enterprise, the number of production equipment may be one or more, in this embodiment, a plurality of production equipment is taken as an example, for example, a certain production line in which automobile door parts are processed is taken as an example, the production equipment history data is specific historical operation data of each equipment on the production line, where the production equipment history data may be specifically divided into normal data and abnormal data, and both include corresponding parameter information, for example, operation power, rotation speed, cutting temperature, working environment parameter and the like of a certain equipment, and the parameter information may be respectively recorded as normal parameter information and abnormal parameter information according to whether a fault occurs, where the fault event information in the abnormal data may be information of the equipment that has the fault.
It is conceivable that if the number of production devices is one, the components corresponding to each process performed by the production devices may be regarded as a "sub-device", and the fault monitoring may also be performed by adopting the scheme provided in the present application.
Constructing a fault prediction model, wherein the fault prediction model comprises a fault pre-estimating sub-model and a fault positioning sub-model;
in this embodiment, the generated countermeasure network (GAN) is used as a frame to construct a fault prediction model, where the fault prediction model specifically includes two sub-models, namely, a fault predictor model and a fault locator sub-model, which are used to generate the countermeasure network (GAN) as the frame, but the training targets are different, the fault predictor model is used to predict the probability of data failure, and the fault locator sub-model is used to locate the device predicted to have failure.
For historical data of production equipment, extracting normal parameter information to construct a normal parameter feature vector set, and extracting abnormal parameter information to construct an abnormal parameter feature vector set;
specifically, the parameter information in the historical data of the production equipment can be sequenced according to a certain mode, so as to construct and obtain feature vectors, for example, conventional initial letter can be adopted for sequencing, in the embodiment, the method is not limited to this, and after the sequencing mode is determined, the parameter information in the historical data of the production equipment can be respectively constructed and obtain a normal parameter feature vector set and an abnormal parameter feature vector set.
Each abnormal parameter feature vector in the abnormal parameter feature vector set is taken as a core, a plurality of groups of fault pre-estimated sample data are constructed and obtained, first training sample data are formed, and a fault pre-estimated sub-model is trained by the first training sample data;
in this embodiment, each abnormal parameter feature vector in the abnormal parameter feature vector set is used as a core, and fault prediction sample data with the same number of groups as the number of abnormal parameter feature vectors in the abnormal parameter feature vector set is constructed and manufactured as first training sample data.
Specifically, first training sample data was prepared by:
for any one abnormal parameter feature vector in the abnormal parameter feature vector set, selecting all normal parameter feature vectors with similarity larger than a preset similarity threshold value from the normal parameter feature vector set, and constructing to obtain a group of fault pre-estimated sample data;
it should be noted that, taking any abnormal parameter feature vector as an example, the similarity of each feature parameter in the abnormal parameter feature vector and the normal parameter feature vector set can be obtained by respectively calculating by using a similarity calculation formula, for example, the similarity is calculated by using a cosine similarity calculation formula, after the similarity is calculated, the normal parameter feature vector is screened by a preset similarity threshold value, and the model training speed can be improved by making sample data with the normal parameter feature vector corresponding to the similarity threshold value larger than the preset similarity threshold value.
Traversing the abnormal parameter feature vector set to obtain fault pre-estimated sample data corresponding to each abnormal parameter feature vector, forming first training sample data, and training the fault pre-estimated sub-model through the first training sample data.
Respectively taking each fault event as a core, constructing a plurality of groups of fault positioning sample data based on an abnormal parameter feature vector set, forming second training sample data, training a fault positioning sub-model by using the constructed plurality of groups of fault pre-estimated sample data, and training to obtain a fault prediction model;
it should be noted that, in this embodiment, with each fault event as a core and the content in the abnormal parameter feature vector set as the associated content, the second training sample data is manufactured, which specifically includes:
taking any fault event in the production equipment history data as an example, extracting to obtain fault equipment information, wherein the fault equipment information is specifically the name of equipment with faults in the fault event, associating the fault equipment information with abnormal parameter information corresponding to the fault event, and constructing to obtain a group of fault positioning sample data;
traversing all fault events in the historical data of the production equipment to respectively obtain fault positioning sample data corresponding to each fault event, forming second training sample data, and training to obtain a fault prediction model.
Collecting production equipment operation data in real time, and inputting the production equipment operation data into a fault prediction model;
after the fault prediction model receives the operation data of the production equipment, the operation data of the production equipment is processed through the fault prediction sub-model to obtain the fault prediction probability;
and processing the result of the fault pre-estimation sub-model through a preset risk threshold value, and judging whether the collected production equipment operation data is required to be input into the fault positioning sub-model for mechanical further processing.
Specifically, if the failure prediction probability is greater than a preset risk threshold, the probability that the production equipment fails in the working state is high, the production equipment operation data can be input into a failure positioning sub-model, the production equipment operation data is processed through the failure positioning sub-model, predicted failure equipment information is obtained, and a failure monitoring result is output; otherwise, directly outputting the fault monitoring result.
The fault monitoring result is specifically a processing result of the fault prediction model on operation data of the production equipment, after the worker obtains the estimated fault equipment information, the worker can further analyze the working state of the corresponding production equipment corresponding to the estimated fault equipment information, adjust the operation parameters, reduce the probability of faults of the corresponding production equipment, and improve the monitoring efficiency.
In a more preferred embodiment, the process of selecting all normal parameter feature vectors with the similarity to the abnormal parameter feature vector greater than the preset similarity threshold from the normal parameter feature vector set further comprises:
preprocessing a normal parameter feature vector set, and calculating the difference value of each element in the abnormal parameter feature vector and the normal parameter feature vector set;
specifically, the difference value calculation may be performed by the following formula:
wherein Q represents a difference value, i represents the number of terms, α i The i-th value, beta, representing the eigenvector of the anomaly parameter I An i-th value representing a normal parameter feature vector, n representing the number of elements in the abnormal parameter feature vector and/or the normal parameter feature vector.
The difference value between the abnormal parameter feature vector and any normal parameter feature vector in the normal parameter feature vector set can be calculated through the formula.
Screening out elements with difference values larger than a preset difference threshold value in the normal parameter feature vector set, and calculating the similarity between the abnormal parameter feature vector and the rest normal parameter feature vectors in the normal parameter feature vector set so as to improve the data screening speed and the data screening efficiency.
Example 2
Referring to fig. 1, on the basis of embodiment 1, embodiment 2 of the present invention further provides a production equipment fault intelligent monitoring system based on deep learning, including:
the data acquisition module is used for acquiring historical data of the production equipment and acquiring operation data of the production equipment;
the model building module is used for building a fault prediction model, wherein the fault prediction model comprises a fault pre-estimation sub-model and a fault positioning sub-model, and in the embodiment, the fault prediction model is built by taking a generated countermeasure network (GAN) as a framework;
the sample making module is used for constructing a normal parameter feature vector set and an abnormal parameter feature vector set based on historical data of production equipment, and making first training sample data and second training sample data based on the normal parameter feature vector set and the abnormal parameter feature vector set;
the model training module is used for training the fault prediction model through the first training sample data and the second training sample data, specifically, training the fault prediction sub-model through the first training sample data and training the fault positioning sub-model through the second training sample data;
and the fault monitoring module is used for carrying out fault monitoring on the operation data of the production equipment through the trained fault prediction model.
In a preferred embodiment, for the sample making module, further comprising:
and the preprocessing unit is used for preprocessing the normal parameter feature vector set.
It will be understood that modifications and variations will be apparent to those skilled in the art from the foregoing description, and it is intended that all such modifications and variations be included within the scope of the following claims. Parts of the specification not described in detail belong to the prior art known to those skilled in the art.
Claims (8)
1. The intelligent monitoring method for the production equipment faults based on deep learning is characterized by comprising the following steps of:
acquiring production equipment historical data, wherein the production equipment historical data comprises operation normal data and operation fault data, the operation normal data comprises normal parameter information, and the operation fault data comprises abnormal parameter information and fault event information;
constructing a fault prediction model, wherein the fault prediction model comprises a fault pre-estimating sub-model and a fault positioning sub-model;
for the historical data of the production equipment, extracting normal parameter information to construct a normal parameter feature vector set, and extracting abnormal parameter information to construct an abnormal parameter feature vector set;
each abnormal parameter feature vector in the abnormal parameter feature vector set is taken as a core, a plurality of groups of fault pre-estimated sample data are constructed and obtained, first training sample data are formed, and the fault pre-estimated sub-model is trained by the first training sample data; respectively taking each fault event as a core, constructing a plurality of groups of fault positioning sample data based on an abnormal parameter feature vector set to form second training sample data, training a fault positioning sub-model by using the constructed plurality of groups of fault pre-estimated sample data, and training to obtain a fault prediction model;
and collecting production equipment operation data, inputting the production equipment operation data into a fault prediction model, and carrying out fault monitoring on the production equipment to obtain a fault monitoring result.
2. The intelligent monitoring method for production equipment faults based on deep learning as claimed in claim 1, wherein the constructing to obtain a plurality of groups of fault pre-estimated sample data by using each abnormal parameter feature vector in the abnormal parameter feature vector set as a core respectively comprises: and for any one abnormal parameter feature vector in the abnormal parameter feature vector set, selecting all normal parameter feature vectors with the similarity larger than a preset similarity threshold value from the normal parameter feature vector set, constructing and obtaining a group of fault estimated sample data, traversing the abnormal parameter feature vector set, obtaining fault estimated sample data corresponding to each abnormal parameter feature vector, and forming first training sample data.
3. The intelligent monitoring method for fault of production equipment based on deep learning as claimed in claim 1, wherein the constructing multiple groups of fault location sample data based on the abnormal parameter feature vector set by using each fault event as a core respectively, and forming the second training sample data comprises:
and extracting fault equipment information from any fault event in the production equipment history data, associating the fault equipment information with abnormal parameter information corresponding to the fault event, constructing a group of fault positioning sample data, traversing all fault events to obtain fault positioning sample data corresponding to each fault event, and forming second training sample data.
4. The intelligent monitoring method for fault of production equipment based on deep learning as claimed in claim 3, wherein the steps of inputting the operation data of the production equipment into a fault prediction model, and performing fault monitoring on the production equipment to obtain a fault monitoring result include:
after the fault prediction model receives the production equipment operation data, the production equipment operation data is processed through a fault prediction sub-model to obtain a fault prediction probability, if the fault prediction probability is larger than a preset risk threshold, the production equipment operation data is input into a fault positioning sub-model, the production equipment operation data is processed through the fault positioning sub-model to obtain predicted fault equipment information, and a fault monitoring result is output; otherwise, directly outputting the fault monitoring result.
5. The intelligent monitoring method for production equipment faults based on deep learning as claimed in claim 4, wherein for any one of the abnormal parameter feature vectors in the abnormal parameter feature vector set, selecting all normal parameter feature vectors with similarity to the abnormal parameter feature vector greater than a preset similarity threshold from the normal parameter feature vector set, further comprising:
preprocessing the normal parameter feature vector set, calculating the difference value of each element in the abnormal parameter feature vector and the normal parameter feature vector set, screening out the elements with the difference value larger than a preset difference threshold value in the normal parameter feature vector set, and calculating the similarity of the abnormal parameter feature vector and each element remained in the normal parameter feature vector set.
6. The intelligent monitoring method for production equipment faults based on deep learning of claim 1, further comprising, for the fault prediction model:
the fault prediction model is constructed based on generating a countermeasure network (GAN) as a framework.
7. A production equipment fault intelligent monitoring system based on deep learning, which is applied to the production equipment fault intelligent monitoring method based on deep learning as set forth in any one of claims 1 to 6, and is characterized by comprising:
the data acquisition module is used for acquiring historical data of the production equipment and acquiring operation data of the production equipment;
the model construction module is used for constructing a fault prediction model, and the fault prediction model comprises a fault pre-estimation sub-model and a fault positioning sub-model;
the sample making module is used for constructing a normal parameter feature vector set and an abnormal parameter feature vector set based on historical data of production equipment, and making first training sample data and second training sample data based on the normal parameter feature vector set and the abnormal parameter feature vector set;
the model training module is used for training the fault prediction model through the first training sample data and the second training sample data;
and the fault monitoring module is used for carrying out fault monitoring on the operation data of the production equipment through the trained fault prediction model.
8. The intelligent monitoring system for production facility failure based on deep learning of claim 7, further comprising, for the sample making module:
and the preprocessing unit is used for preprocessing the normal parameter feature vector set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310349473.1A CN116467592A (en) | 2023-04-04 | 2023-04-04 | Production equipment fault intelligent monitoring method and system based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310349473.1A CN116467592A (en) | 2023-04-04 | 2023-04-04 | Production equipment fault intelligent monitoring method and system based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116467592A true CN116467592A (en) | 2023-07-21 |
Family
ID=87178200
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310349473.1A Pending CN116467592A (en) | 2023-04-04 | 2023-04-04 | Production equipment fault intelligent monitoring method and system based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116467592A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116861218A (en) * | 2023-07-25 | 2023-10-10 | 上海华菱电站成套设备股份有限公司 | Mine winder key equipment state monitoring early warning system |
-
2023
- 2023-04-04 CN CN202310349473.1A patent/CN116467592A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116861218A (en) * | 2023-07-25 | 2023-10-10 | 上海华菱电站成套设备股份有限公司 | Mine winder key equipment state monitoring early warning system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111596604B (en) | Intelligent fault diagnosis and self-healing control system and method for engineering equipment based on digital twinning | |
| CN110221558B (en) | Electro-hydraulic servo valve online fault diagnosis gateway based on edge computing technology | |
| CN111047082B (en) | Early warning method and device of equipment, storage medium and electronic device | |
| CN106888205B (en) | Non-invasive PLC anomaly detection method based on power consumption analysis | |
| CN109829236B (en) | XGboost feature extraction-based compressor fault diagnosis method | |
| CN109146246B (en) | Fault detection method based on automatic encoder and Bayesian network | |
| CN109297689B (en) | Large-scale hydraulic machinery intelligent diagnosis method introducing weight factors | |
| CN111582392B (en) | Multi-working-condition health state online monitoring method for key components of wind turbine generator | |
| CN105607631B (en) | The weak fault model control limit method for building up of batch process and weak fault monitoring method | |
| CN110119787B (en) | Working condition detection method and equipment for rotary mechanical equipment | |
| CN109784668B (en) | Sample feature dimension reduction processing method for detecting abnormal behaviors of power monitoring system | |
| CN116467592A (en) | Production equipment fault intelligent monitoring method and system based on deep learning | |
| CN113496089B (en) | Internet of things equipment abnormal state online monitoring method and system based on artificial intelligence | |
| CN116823227A (en) | Intelligent equipment management system and method based on Internet of things | |
| CN115270491A (en) | Offshore wind power operation and maintenance platform design method based on multivariate information fusion | |
| CN114004512A (en) | Multi-unit operation state outlier analysis method and system based on density clustering | |
| CN113485244A (en) | Numerical control machine tool control system and method based on cutter wear prediction | |
| CN116086537A (en) | Equipment state monitoring method, device, equipment and storage medium | |
| CN117076258A (en) | Remote monitoring method and system based on Internet cloud | |
| CN115313930A (en) | Method for predicting excitation current of large-scale steam turbine generator in real time | |
| CN113627685B (en) | Wind driven generator power prediction method considering wind power internet load limit | |
| CN112819646B (en) | Fault diagnosis system for flexible production line of customized wood furniture | |
| CN111934903B (en) | Docker container fault intelligent prediction method based on time sequence evolution gene | |
| CN113360718A (en) | Linear motor fault detection method and system based on big data analysis | |
| Huang et al. | Static and dynamic joint analysis for operation condition division of industrial process with incremental learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |