CN119293664A - A device operation evaluation method based on multi-source data fusion - Google Patents
A device operation evaluation method based on multi-source data fusion Download PDFInfo
- Publication number
- CN119293664A CN119293664A CN202411392313.6A CN202411392313A CN119293664A CN 119293664 A CN119293664 A CN 119293664A CN 202411392313 A CN202411392313 A CN 202411392313A CN 119293664 A CN119293664 A CN 119293664A
- Authority
- CN
- China
- Prior art keywords
- data
- equipment
- fusion
- model
- source 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
- 230000004927 fusion Effects 0.000 title claims abstract description 70
- 238000011156 evaluation Methods 0.000 title claims abstract description 41
- 238000012423 maintenance Methods 0.000 claims abstract description 70
- 238000000034 method Methods 0.000 claims abstract description 55
- 238000012544 monitoring process Methods 0.000 claims abstract description 12
- 230000007613 environmental effect Effects 0.000 claims abstract description 11
- 238000013210 evaluation model Methods 0.000 claims description 54
- 238000003745 diagnosis Methods 0.000 claims description 29
- 238000007405 data analysis Methods 0.000 claims description 20
- 238000012545 processing Methods 0.000 claims description 17
- 230000008569 process Effects 0.000 claims description 16
- 239000013598 vector Substances 0.000 claims description 15
- 238000004458 analytical method Methods 0.000 claims description 14
- 238000013527 convolutional neural network Methods 0.000 claims description 14
- 238000005516 engineering process Methods 0.000 claims description 12
- 230000002159 abnormal effect Effects 0.000 claims description 10
- 238000013136 deep learning model Methods 0.000 claims description 10
- 238000007499 fusion processing Methods 0.000 claims description 10
- 238000010801 machine learning Methods 0.000 claims description 10
- 238000010276 construction Methods 0.000 claims description 8
- 238000000605 extraction Methods 0.000 claims description 8
- 230000000694 effects Effects 0.000 claims description 7
- 238000007619 statistical method Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 5
- 238000007418 data mining Methods 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 claims description 3
- 238000009826 distribution Methods 0.000 claims description 3
- 239000000428 dust Substances 0.000 claims description 3
- 239000003344 environmental pollutant Substances 0.000 claims description 3
- 238000011478 gradient descent method Methods 0.000 claims description 3
- 230000036541 health Effects 0.000 claims description 3
- 230000007246 mechanism Effects 0.000 claims description 3
- 231100000719 pollutant Toxicity 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000000611 regression analysis Methods 0.000 claims description 3
- 238000012706 support-vector machine Methods 0.000 claims description 3
- 238000012800 visualization Methods 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000002349 favourable effect Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000002265 prevention Effects 0.000 description 2
- 230000008439 repair process Effects 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000007274 generation of a signal involved in cell-cell signaling Effects 0.000 description 1
- 239000006101 laboratory sample Substances 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 239000000523 sample Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2431—Multiple classes
-
- 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
- G06F18/25—Fusion techniques
- G06F18/251—Fusion techniques of input or preprocessed data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- 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
- G06N3/045—Combinations of 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2123/00—Data types
- G06F2123/02—Data types in the time domain, e.g. time-series data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Operations Research (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention discloses a device operation evaluation method based on multi-source data fusion, which relates to the technical field of device operation evaluation and comprises the following steps of S1, data acquisition, wherein the multi-source data comprises sensor data, operation logs, environment data and maintenance history of a device under the real-time operation state of the device. The equipment operation evaluation method based on the multi-source data fusion can provide accurate equipment state monitoring and fault prediction, can continuously optimize the performance and accuracy according to the actual equipment operation data and environmental conditions of the equipment, thereby realizing efficient and intelligent equipment management and maintenance strategies, clearly defining specific steps and characteristics of the equipment operation evaluation method based on the multi-source data fusion, ensuring the comprehensiveness and the high efficiency of the method, and being suitable for various complex industrial environments to improve the reliability and the maintenance efficiency of the equipment.
Description
Technical Field
The invention relates to the technical field of equipment operation evaluation, in particular to an equipment operation evaluation method based on multi-source data fusion.
Background
With the continuous development of industrial technology, the application of various devices in production and life is wider and wider, the normal operation of the devices is critical to ensuring the production efficiency, improving the product quality and ensuring the personnel safety, however, the traditional device operation evaluation method often depends on a single data source, such as device sensor data or operation logs, and is difficult to comprehensively and accurately reflect the actual operation state of the devices, on one hand, the sensor data alone can not capture some complex situations in the operation of the devices, such as the sensor can not detect potential faults caused by improper operation or environmental factors, and the data of the single sensor can have errors or noise to influence the accuracy of evaluation, on the other hand, the operation logs can record the operation situation of the devices, but the internal operation state of the devices and the disclosure of the potential faults are limited, meanwhile, different types of devices can need different evaluation methods, and the traditional methods lack of generality and flexibility, and complexity are further, along with the intelligence and complexity of the devices, the data quantity generated in the operation process of the devices is larger and more diversified, how to effectively integrate the multiple source data, the data can not be detected, and the valuable information is extracted, so that the current operation state of the devices is to be evaluated accurately.
Disclosure of Invention
In order to solve the technical problems, the invention is realized by the following technical scheme that the equipment operation evaluation method based on multi-source data fusion comprises the following steps:
S1, data acquisition, namely acquiring multi-source data of equipment in a real-time running state, wherein the multi-source data comprises sensor data, operation logs, environment data and maintenance history of the equipment;
S2, data fusion, namely carrying out fusion processing on the multi-source data acquired in the step S1, and integrating the multi-source data by adopting a deep learning model to generate a fusion data set of the running state of the equipment;
S3, constructing a data analysis evaluation model, namely constructing an evaluation model of the running state of the equipment by using a machine learning algorithm based on the fused data set after fusion;
And S4, generating an evaluation result and an alarm, namely outputting the evaluation condition of the equipment operation state according to the evaluation model in the S3, and automatically generating an alarm signal and a maintenance suggestion when the equipment operation state is lower than a preset threshold value or potential faults are predicted so as to assist in equipment maintenance and fault prevention.
Preferably, the data acquisition step S1 further includes:
Sensor data acquisition, namely mounting a sensor on the equipment part for collecting the operation parameters of the equipment part, wherein the sensor comprises a temperature sensor, a pressure sensor and a vibration sensor;
recording operation data of an equipment operator, wherein the operation data comprises operation time, operation property and operation result;
the environmental data acquisition, namely recording environmental data of the running environment of the equipment, wherein the environmental data comprise environmental temperature, humidity, electromagnetic interference, vibration and dust pollutants;
and the maintenance data acquisition is used for recording the maintenance history of the equipment, wherein the maintenance history comprises maintenance time, maintenance type and maintenance effect.
Preferably, the data fusion step S2 further includes:
The data preprocessing comprises the steps of denoising, normalizing and missing value processing on the collected multi-source data, and ensuring that the data quality meets the subsequent processing requirement, wherein the multi-source data consists of information from different data sources, and the data types of the multi-source data comprise the temperature, pressure, vibration, power utilization rate of equipment and the temperature and humidity of the environment;
and data fusion, namely combining information from different data sources by using a convolutional neural network to generate a fusion data set for reflecting the overall state of the equipment.
Preferably, the data analysis and evaluation model construction step S3 further includes:
The characteristic extraction, namely, applying big data analysis technology, including machine learning algorithm and data mining, carrying out characteristic extraction on the fused data set after fusion, and extracting key characteristics;
Training a model, namely training and constructing an evaluation model of the running state of the equipment by using the extracted key features, wherein the obtained evaluation model is used for evaluating and predicting the running state of the equipment in real time, the evaluation model outputs the evaluation state, and the generated evaluation state is used for reflecting the current running state and potential faults of the equipment and comprises the health state, performance indexes and fault prediction of the equipment;
Model optimization, namely adjusting and optimizing an evaluation model according to actual equipment operation data and maintenance feedback, and improving the accuracy and adaptability of the model.
Preferably, the generating an evaluation result and alarming step S4 further includes:
Judging the running state of the equipment by evaluating the evaluation condition of the running state of the equipment output by the evaluation model, wherein the running state comprises normal, warning or fault states;
Generating an alarm, namely generating corresponding alarm signals and fault diagnosis information when the running state of equipment is lower than a preset threshold value or potential faults are predicted, and notifying maintenance personnel to process;
and generating maintenance advice based on the fault diagnosis information and the historical maintenance data generated in the current running state of the equipment, and helping maintenance personnel to effectively maintain and repair the equipment.
Preferably, the alarm signal is generated as follows:
Determining the normal operation parameter range and the preset threshold value of key performance index of the equipment according to the use specification, industry standard and actual operation experience of the equipment, adjusting the preset threshold value according to the historical operation data and the real-time monitoring data of the equipment, and carrying out fusion data set
The method comprises the steps of carrying out monitoring and anomaly detection on a fusion data set through a statistical analysis and a machine learning algorithm to obtain equipment operation state parameters, comparing the equipment operation state parameters with a preset threshold value, triggering an alarm generation mechanism when the equipment operation state parameters are found to exceed the range of the preset threshold value or an anomaly trend and a mode occur, determining the level of an alarm signal according to the anomaly degree of the equipment operation state, dividing the level of the alarm signal into an emergency alarm, an important alarm and a general alarm so that maintenance personnel can take corresponding treatment measures according to the alarm level, generating corresponding alarm signals according to the alarm level, wherein the alarm signals comprise sound, lamplight, short messages and emails so as to timely inform maintenance personnel, and when the alarm signals are generated, the alarm signals also comprise identification information, the alarm level and the anomaly parameters of the equipment, so that the maintenance personnel can quickly locate problem equipment and know the severity of the problem.
Preferably, the fault diagnosis information generation process is as follows:
When the running state of the equipment has abnormal trend, adopting a signal processing technology and a feature extraction algorithm to further analyze the collected running state data of the equipment and extract fault features, wherein the fault features comprise specific values of abnormal parameters, change trend and occurrence time;
According to the extracted fault characteristics, a fault diagnosis algorithm is adopted to classify and position faults, the cause and the influence range of the faults are determined, fault diagnosis results and fault analysis results are obtained, fault diagnosis information is generated according to the fault diagnosis results and the fault analysis results, the fault diagnosis information comprises identification information of fault equipment, fault types, fault causes, influence ranges and suggested treatment measures, and the fault diagnosis information is sent to maintenance personnel in a report mode so that the maintenance personnel can quickly know fault conditions and take corresponding treatment measures.
Preferably, the data fusion process further includes:
performing a matching algorithm on data from different sources by adopting time synchronization and space alignment together to ensure the consistency and the integrity of the data;
multi-dimensional fusion, namely multi-dimensional fusion of multi-source data is carried out by utilizing a multi-source data fusion technology, including multi-mode deep learning and multi-view integrated learning, so as to improve the accuracy and the robustness of an evaluation model;
And (3) dynamic weight adjustment, namely adjusting the weight of the multi-source data in the fusion process according to the reliability and the correlation of different data sources so as to optimize the effect of data fusion.
Preferably, the process of multi-dimensional fusion of multi-source data is as follows:
Firstly, processing image data based on a framework of a convolutional neural network, processing text data based on a long-short-time memory network, constructing a full-connection network aiming at sensor data, combining the convolutional neural network, the long-short-time memory network and the full-connection network to form a multi-modal deep learning model, and completing construction of the multi-modal deep learning model, wherein the multi-source data comprises the image data, the sensor data and the text data;
Extracting image features through a convolutional neural network to obtain a high-dimensional image feature vector, extracting text features through a long-short-term memory network to obtain a text feature vector, extracting sensor features through a fully-connected network to obtain a sensor feature vector, and then splicing different feature vectors together in a splicing mode to form a comprehensive feature vector;
Dividing the preprocessed multi-source data into different views, dividing the image data into a first view, dividing the sensor data into a second view, and dividing the text data into a third view;
The method comprises the steps of respectively constructing an independent learning model for each view, classifying an image view by adopting a convolutional neural network to obtain an image view model, carrying out regression analysis on a sensor view by adopting a support vector machine to obtain a sensor view model, classifying a text view by adopting a long-short-time memory network to obtain a text view model, and obtaining a text view model;
and fusing the image view model, the sensor view model and the text view model by a weighted average method, giving different weights according to the importance of each view model by the weighted average method, and carrying out weighted average on the prediction results of the models to obtain a multi-dimensional fusion result.
Preferably, the data analysis and evaluation model construction process further includes:
Model calibration, namely calibrating the model by comparing the prediction result of the evaluation model with the operation data of the actual equipment so as to reduce deviation and improve the prediction precision;
The method comprises the steps of collecting actual equipment operation data, obtaining an evaluation model prediction result, enabling the actual equipment operation data to correspond to the evaluation model prediction result in time, determining equipment operation state parameters to be compared, enabling the actual equipment operation data to contain the same parameters, enabling the meanings and units of the parameters to be consistent with each other, calculating errors between the evaluation model prediction result and the actual equipment operation data by means of root mean square errors for each parameter to be compared, drawing a chart through a visualization tool, displaying differences and trends between the prediction result and the actual data, carrying out statistical analysis on the errors, knowing distribution conditions and statistical characteristics of the errors, obtaining a comparison analysis result, adjusting parameters of a data analysis evaluation model by means of gradient descent method in an optimization algorithm according to the comparison analysis result, calibrating the data analysis evaluation model, and comparing the data analysis evaluation model with the actual equipment operation data after calibrating the model again, and verifying calibration effects if the errors are descended, describing that the calibration is successful, and continuing the calibration if the errors are not descended, describing that the calibration is unsuccessful;
Specific steps and characteristics of the equipment operation evaluation method based on multi-source data fusion are clearly defined, the comprehensiveness and high efficiency of the method are ensured, and the method is suitable for various complex industrial environments so as to improve the reliability and maintenance efficiency of the equipment.
The invention provides a device operation evaluation method based on multi-source data fusion, which has the following beneficial effects:
1. The equipment operation evaluation method based on multi-source data fusion can comprehensively know the operation state of equipment from multiple angles by collecting the multi-source information such as sensor data, operation logs, environment data and maintenance history of the equipment, is favorable for finding potential problems and risks of the equipment, improves the accuracy and the comprehensiveness of equipment state monitoring, integrates the multi-source data by utilizing a deep learning model and a multi-source data fusion technology, and can fully utilize the advantages of different data sources to generate a fusion data set, overcome the limitation of a single data source and more comprehensively reflect the integral state of the equipment.
2. The equipment operation evaluation method based on the multi-source data fusion can provide accurate equipment state monitoring and fault prediction, can continuously optimize the performance and accuracy according to the actual equipment operation data and environmental conditions of the equipment, thereby realizing efficient and intelligent equipment management and maintenance strategies, clearly defining specific steps and characteristics of the equipment operation evaluation method based on the multi-source data fusion, ensuring the comprehensiveness and the high efficiency of the method, and being suitable for various complex industrial environments to improve the reliability and the maintenance efficiency of the equipment.
3. The equipment operation evaluation method based on multi-source data fusion is characterized in that an evaluation model is built by using a machine learning algorithm based on a fusion data set, the operation condition of the prediction equipment can be evaluated in real time, when the operation state of the equipment is lower than a preset threshold value or potential faults are predicted, an alarm signal and fault diagnosis information are automatically generated, maintenance personnel are notified to process the equipment operation evaluation model, the equipment operation evaluation method is favorable for timely finding out the equipment faults, reducing the equipment downtime, improving the reliability and the availability of the equipment, providing detailed fault information for the maintenance personnel, facilitating the maintenance personnel to quickly and accurately locate and solve the faults, and improving the maintenance efficiency.
Drawings
FIG. 1 is a block flow diagram of a method for evaluating operation of a device based on multi-source data fusion according to the present invention.
Detailed Description
The invention will be described in further detail with reference to the drawings and the detailed description. The embodiments of the invention have been presented for purposes of illustration and description, and are not intended to be exhaustive or limited to the invention in the form disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, and to enable others of ordinary skill in the art to understand the invention for various embodiments with various modifications as are suited to the particular use contemplated.
As shown in FIG. 1, the invention provides a technical scheme that a device operation evaluation method based on multi-source data fusion comprises the following steps:
firstly, data acquisition, namely acquiring multi-source data of equipment in a real-time running state, wherein the multi-source data comprises sensor data, operation logs, environment data and maintenance history of the equipment;
the sensor is arranged on the equipment component and used for collecting operation parameters of the equipment component, the sensor comprises a temperature sensor, a pressure sensor and a vibration sensor, abnormal conditions of the equipment component can be found timely, operation data of an equipment operator are recorded, the operation data comprise operation time, operation property and operation result, the analysis of the influence of the operation on the equipment operation is facilitated, environment data of the equipment operation environment are recorded, the environment data comprise environment temperature, humidity, electromagnetic interference, vibration and dust pollutants, the influence of environment factors on the equipment can be assessed, maintenance history of the equipment is recorded, the maintenance history comprises maintenance time, maintenance type and maintenance effect, and references can be provided for maintenance and management of the equipment.
Carrying out data preprocessing, namely denoising, normalizing and missing value processing on the collected multi-source data to ensure that the data quality meets the subsequent processing requirements, wherein the multi-source data consists of information from different data sources, and the data types of the multi-source data comprise the temperature, the pressure, the vibration, the power utilization rate and the temperature and the humidity of the environment of the equipment;
And then data fusion, namely carrying out fusion processing on the multi-source data acquired in the step S1, integrating the multi-source data by adopting a deep learning model in the fusion processing to generate a fusion data set of the running state of the equipment, and integrating the data of different sources to more comprehensively reflect the whole state of the equipment.
Then constructing a data analysis evaluation model, namely constructing an evaluation model of the running state of the equipment by using a machine learning algorithm based on the fused data set after fusion;
The data analysis evaluation model construction step S3 further comprises the steps of extracting features, applying a big data analysis technology, including a machine learning algorithm and data mining, to conduct feature extraction on the fused data set, extracting key features, providing more valuable information for equipment operation state evaluation, training a model, training and constructing an evaluation model of the equipment operation state by using the extracted key features, enabling the obtained evaluation model to be used for evaluating and predicting the operation state of the equipment in real time, enabling the evaluation model to output the evaluation state, enabling the generated evaluation state to be used for reflecting the current operation state and potential faults of the equipment, enabling the evaluation state to comprise health state, performance index and fault prediction of the equipment, providing decision support for maintenance and management of the equipment, and optimizing the model, and adjusting and optimizing the evaluation model according to actual equipment operation data and maintenance feedback, and improving accuracy and adaptability of the model.
And finally, generating an evaluation result and an alarm, namely outputting the evaluation condition of the equipment operation state according to the evaluation model in the step S3, and automatically generating an alarm signal and a maintenance suggestion when the equipment operation state is lower than a preset threshold value or potential faults are predicted so as to assist in equipment maintenance and fault prevention, thereby being beneficial to timely finding equipment problems and taking measures and improving the reliability and stability of equipment.
The generating evaluation result and alarming step S4 further includes:
judging the running state of the equipment by evaluating the evaluation condition of the running state of the equipment output by the evaluation model, wherein the running state comprises normal, warning or fault states; the running state of the equipment is accurately judged through state judgment, and a basis is provided for maintenance and management of the equipment;
Generating an alarm signal and fault diagnosis information corresponding to the equipment when the running state of the equipment is lower than a preset threshold value or the potential fault is predicted, and notifying maintenance personnel to process the equipment;
And the maintenance suggestion is generated based on the fault diagnosis information and the historical maintenance data generated in the current running state of the equipment, so that maintenance personnel can be helped to effectively maintain and repair the equipment, and the maintenance efficiency is improved.
The alarm signal generation process is as follows:
according to the use specification, industry standard and actual operation experience of the equipment, defining the normal operation parameter range and the preset threshold value of key performance index of the equipment;
The method comprises the steps of monitoring and anomaly detecting a fusion data set through a statistical analysis and a machine learning algorithm to obtain equipment operation state parameters, and comparing the equipment operation state parameters with a preset threshold value;
The method comprises the steps of dividing the level of an alarm signal into an emergency alarm, an important alarm and a general alarm so that maintenance personnel can take corresponding treatment measures according to the alarm level, generating the corresponding alarm signal according to the alarm level, wherein the alarm signal comprises sound, light, a short message and an email so as to inform the maintenance personnel in time, and when the alarm signal is generated, the alarm signal also comprises identification information of equipment, the alarm level and abnormal parameters so that the maintenance personnel can quickly locate the equipment with problems and know the severity of the problems.
The method comprises the steps of determining a normal operation parameter range of equipment and a preset threshold value of key performance indexes in a specific implementation process, adjusting according to historical operation data and real-time monitoring data, improving accuracy and timeliness of alarm signals, detecting abnormal conditions of equipment operation state parameters in time through monitoring and abnormality detection of a fusion data set, triggering an alarm generation mechanism, determining an alarm signal level according to the abnormal degree of the equipment operation state, and generating an alarm signal containing equipment identification information, the alarm level and the abnormal parameters, so that maintenance personnel can quickly locate problem equipment and know the severity of the problem.
The fault diagnosis information generation process comprises the steps of further analyzing the collected equipment operation state data by adopting a signal processing technology and a feature extraction algorithm when an abnormal trend occurs in the equipment operation state, and extracting fault features, wherein the fault features comprise specific numerical values of abnormal parameters, change trend and occurrence time;
According to the extracted fault characteristics, a fault diagnosis algorithm is adopted to classify and position faults, the cause and the influence range of the faults are determined, fault diagnosis results and fault analysis results are obtained, fault diagnosis information is generated according to the fault diagnosis results and the fault analysis results, the fault diagnosis information comprises identification information of fault equipment, fault types, fault causes, influence ranges and suggested processing measures, the fault diagnosis information is sent to maintenance personnel in a report mode so that the maintenance personnel can quickly know fault conditions and take corresponding processing measures, and through the steps, alarm signals and the fault diagnosis information can be timely generated when the running state of the equipment is lower than a preset threshold value or potential faults are predicted, the maintenance personnel are informed of processing, and normal running of the equipment is ensured.
The data fusion processing process further comprises the following steps:
Advanced data matching, namely, performing a matching algorithm on data from different sources by adopting time synchronization and space alignment together, so that the accuracy and usability of multi-source data can be improved, and the consistency and integrity of the data are ensured;
Multi-dimensional fusion, namely multi-dimensional fusion of multi-source data is carried out by utilizing a multi-source data fusion technology, including multi-mode deep learning and multi-view integrated learning, so that the accuracy and the robustness of an evaluation model are improved, and the overall state of equipment can be reflected more comprehensively;
And (3) dynamic weight adjustment, namely adjusting the weight of the multi-source data in the fusion process according to the reliability and the correlation of different data sources so as to optimize the effect of data fusion and improve the accuracy and the reliability of the fused data.
The process of multi-dimensional fusion of multi-source data is as follows:
Firstly, processing image data based on a framework of a convolutional neural network, processing text data based on a long-short-time memory network, constructing a full-connection network aiming at sensor data, combining the convolutional neural network, the long-short-time memory network and the full-connection network to form a multi-modal deep learning model, and completing construction of the multi-modal deep learning model, wherein the multi-source data comprises the image data, the sensor data and the text data;
Extracting image features through a convolutional neural network to obtain a high-dimensional image feature vector, extracting text features through a long-short-term memory network to obtain a text feature vector, extracting sensor features through a fully-connected network to obtain a sensor feature vector, and then splicing different feature vectors together in a splicing mode to form a comprehensive feature vector;
Dividing the preprocessed multi-source data into different views, dividing the image data into a first view, dividing the sensor data into a second view, and dividing the text data into a third view;
The method comprises the steps of respectively constructing an independent learning model for each view, classifying an image view by adopting a convolutional neural network to obtain an image view model, carrying out regression analysis on a sensor view by adopting a support vector machine to obtain a sensor view model, classifying a text view by adopting a long-short-time memory network to obtain a text view model, and obtaining a text view model;
Fusing the image view model, the sensor view model and the text view model by a weighted average method, giving different weights according to the importance of each view model by the weighted average method, and carrying out weighted average on the prediction results of the models to obtain a multi-dimensional fusion result; through the multi-dimensional fusion method, multi-source data in a clinical laboratory sample storage system can be effectively fused, information of different data sources is fully utilized, and accuracy and reliability of sample analysis and storage management are improved.
The data analysis and evaluation model construction process further comprises the steps of model calibration, namely calibrating the model by comparing the prediction result of the evaluation model with actual equipment operation data so as to reduce deviation and improve prediction accuracy;
The method comprises the steps of collecting actual equipment operation data, obtaining an evaluation model prediction result, enabling the actual equipment operation data to correspond to the evaluation model prediction result in time, determining equipment operation state parameters to be compared, enabling the actual equipment operation data to contain the same parameters, enabling the meanings and units of the parameters to be consistent with each other, calculating errors between the evaluation model prediction result and the actual equipment operation data by means of root mean square errors, drawing a chart through a visualization tool, displaying differences and trends between the prediction result and the actual equipment operation data, carrying out statistical analysis on the errors, knowing distribution conditions and statistical characteristics of the errors, obtaining comparison analysis results, adjusting parameters of a data analysis evaluation model by means of gradient descent methods in an optimization algorithm according to the comparison analysis results, calibrating the data analysis evaluation model, and comparing the operation model with the actual equipment operation data after calibrating the model, if the errors descend, describing that the calibration is successful, continuing to calibrate if the errors do not descend, establishing a comprehensive multi-source data fusion technology, an advanced data analysis method and an equipment operation method, and an intelligent equipment operation device failure state monitoring method can not be provided accurately and only can be maintained accurately, and can be maintained accurately and continuously operated according to the conditions of the intelligent equipment operation state monitoring and the intelligent equipment.
The evaluation model is adjusted and optimized according to the actual equipment operation data and maintenance feedback, so that the evaluation model can be continuously adapted to the actual operation condition of the equipment, the accuracy and reliability of the model are improved, and the maintenance management of the equipment is further optimized.
Specific steps and characteristics of the equipment operation evaluation method based on multi-source data fusion are clearly defined, the comprehensiveness and high efficiency of the method are ensured, and the method is suitable for various complex industrial environments so as to improve the reliability and maintenance efficiency of the equipment.
It will be apparent that the described embodiments are only some, but not all, embodiments of the invention. All other embodiments, which can be made by those skilled in the art and which are included in the embodiments of the present invention without the inventive step, are intended to be within the scope of the present invention. Structures, devices and methods of operation not specifically described and illustrated herein, unless otherwise indicated and limited, are implemented according to conventional means in the art.
Claims (10)
1. The equipment operation evaluation method based on multi-source data fusion is characterized by comprising the following steps of:
S1, data acquisition, namely acquiring multi-source data of equipment in a real-time running state, wherein the multi-source data comprises sensor data, operation logs, environment data and maintenance history of the equipment;
S2, data fusion, namely carrying out fusion processing on the multi-source data acquired in the step S1, and integrating the multi-source data by adopting a deep learning model to generate a fusion data set of the running state of the equipment;
S3, constructing a data analysis evaluation model, namely constructing an evaluation model of the running state of the equipment by using a machine learning algorithm based on the fused data set after fusion;
And S4, generating an evaluation result and an alarm, namely outputting the evaluation condition of the equipment operation state according to the evaluation model in S3, and automatically generating an alarm signal and a maintenance suggestion when the equipment operation state is lower than a preset threshold value or potential faults are predicted.
2. The method for evaluating operation of a device based on multi-source data fusion as defined in claim 1, wherein said data acquisition step S1 further comprises:
Sensor data acquisition, namely mounting a sensor on the equipment part for collecting the operation parameters of the equipment part, wherein the sensor comprises a temperature sensor, a pressure sensor and a vibration sensor;
recording operation data of an equipment operator, wherein the operation data comprises operation time, operation property and operation result;
the environmental data acquisition, namely recording environmental data of the running environment of the equipment, wherein the environmental data comprise environmental temperature, humidity, electromagnetic interference, vibration and dust pollutants;
and the maintenance data acquisition is used for recording the maintenance history of the equipment, wherein the maintenance history comprises maintenance time, maintenance type and maintenance effect.
3. The method for evaluating operation of a device based on multi-source data fusion as defined in claim 2, wherein said data fusion step S2 further comprises:
the data preprocessing comprises the steps of denoising, normalizing and missing value processing on the collected multi-source data, wherein the multi-source data consists of information from different data sources, and the data types of the multi-source data comprise the temperature, the pressure, the vibration, the power utilization rate and the temperature and the humidity of the environment of the equipment;
and data fusion, namely combining information from different data sources by using a convolutional neural network to generate a fusion data set for reflecting the overall state of the equipment.
4. The method for evaluating equipment operation based on multi-source data fusion according to claim 3, wherein said data analysis and evaluation model construction step S3 further comprises:
The characteristic extraction, namely, applying big data analysis technology, including machine learning algorithm and data mining, carrying out characteristic extraction on the fused data set after fusion, and extracting key characteristics;
Training a model, namely training and constructing an evaluation model of the running state of the equipment by using the extracted key features, wherein the obtained evaluation model is used for evaluating and predicting the running state of the equipment in real time, the evaluation model outputs the evaluation state, and the generated evaluation state is used for reflecting the current running state and potential faults of the equipment and comprises the health state, performance indexes and fault prediction of the equipment;
Model optimization, namely adjusting and optimizing an evaluation model according to actual equipment operation data and maintenance feedback.
5. The method for evaluating operation of a device based on multi-source data fusion as defined in claim 4, wherein said generating an evaluation result and alerting step S4 further comprises:
Judging the running state of the equipment by evaluating the evaluation condition of the running state of the equipment output by the evaluation model, wherein the running state comprises normal, warning or fault states;
Generating an alarm, namely generating corresponding alarm signals and fault diagnosis information when the running state of equipment is lower than a preset threshold value or potential faults are predicted, and notifying maintenance personnel to process;
And generating maintenance advice based on the fault diagnosis information and the historical maintenance data generated under the current running state of the equipment.
6. The method for evaluating equipment operation based on multi-source data fusion according to claim 5, wherein the alarm signal is generated by the following steps:
Determining the normal operation parameter range and the preset threshold value of key performance index of the equipment according to the use specification, industry standard and actual operation experience of the equipment, adjusting the preset threshold value according to the historical operation data and the real-time monitoring data of the equipment, and carrying out fusion data set
The method comprises the steps of monitoring and anomaly detection on a fusion data set through statistical analysis and a machine learning algorithm to obtain equipment operation state parameters, comparing the equipment operation state parameters with a preset threshold value, triggering an alarm generation mechanism when the equipment operation state parameters are found to exceed the range of the preset threshold value or an anomaly trend and mode occur, determining the level of an alarm signal according to the anomaly degree of the equipment operation state, classifying the level of the alarm signal into emergency alarm, important alarm and general alarm, generating corresponding alarm signals according to the alarm level, wherein the alarm signals comprise sound, lamplight, short messages and emails, and when the alarm signals are generated, the alarm signals also comprise identification information, alarm level and anomaly parameters of equipment.
7. The method for evaluating equipment operation based on multi-source data fusion according to claim 6, wherein the fault diagnosis information generation process is as follows:
when the running state of the equipment has abnormal trend, adopting a signal processing technology and a feature extraction algorithm to further analyze the collected running state data of the equipment and extract fault features;
According to the extracted fault characteristics, a fault diagnosis algorithm is adopted to classify and position the fault, the cause and the influence range of the fault are determined, a fault diagnosis result and a fault analysis result are obtained, fault diagnosis information is generated according to the fault diagnosis result and the fault analysis result, the fault diagnosis information comprises identification information of fault equipment, fault types, the cause of the fault, the influence range and suggested treatment measures, and the fault diagnosis information is sent to maintenance personnel in a report mode.
8. The method for evaluating operation of a device based on multi-source data fusion as defined in claim 7, wherein said data fusion process further comprises:
advanced data matching, namely performing a matching algorithm on data from different sources by adopting time synchronization and space alignment together;
Multidimensional fusion, namely utilizing a multi-source data fusion technology, including multi-mode deep learning and multi-view integrated learning, to carry out multidimensional fusion on multi-source data;
Dynamic weight adjustment, namely adjusting the weight of multi-source data in the fusion process according to the reliability and the correlation of different data sources.
9. The method for evaluating the operation of a device based on multi-source data fusion according to claim 8, wherein the multi-dimensional fusion of the multi-source data is performed as follows:
Firstly, processing image data based on a framework of a convolutional neural network, processing text data based on a long-short-time memory network, constructing a full-connection network aiming at sensor data, combining the convolutional neural network, the long-short-time memory network and the full-connection network to form a multi-modal deep learning model, and completing construction of the multi-modal deep learning model, wherein the multi-source data comprises the image data, the sensor data and the text data;
Extracting image features through a convolutional neural network to obtain a high-dimensional image feature vector, extracting text features through a long-short-term memory network to obtain a text feature vector, extracting sensor features through a fully-connected network to obtain a sensor feature vector, and then splicing different feature vectors together in a splicing mode to form a comprehensive feature vector;
Dividing the preprocessed multi-source data into different views, dividing the image data into a first view, dividing the sensor data into a second view, and dividing the text data into a third view;
The method comprises the steps of respectively constructing an independent learning model for each view, classifying an image view by adopting a convolutional neural network to obtain an image view model, carrying out regression analysis on a sensor view by adopting a support vector machine to obtain a sensor view model, classifying a text view by adopting a long-short-time memory network to obtain a text view model, and obtaining a text view model;
and fusing the image view model, the sensor view model and the text view model by a weighted average method, giving different weights according to the importance of each view model by the weighted average method, and carrying out weighted average on the prediction results of the models to obtain a multi-dimensional fusion result.
10. The method for evaluating equipment operation based on multi-source data fusion according to claim 9, wherein the process for constructing the data analysis and evaluation model further comprises the steps of:
model calibration, namely comparing and evaluating a model prediction result with actual equipment operation data;
The method comprises the steps of collecting actual equipment operation data, obtaining an evaluation model prediction result, enabling the actual equipment operation data to correspond to the evaluation model prediction result in time, determining equipment operation state parameters to be compared, enabling the actual equipment operation data to contain the same parameters, enabling the meanings and units of the parameters to be consistent with each other, calculating errors between the evaluation model prediction result and the actual equipment operation data according to root mean square errors of the parameters to be compared, drawing a chart through a visualization tool, displaying differences and trends between the prediction result and the actual equipment operation data, carrying out statistical analysis on the errors, knowing distribution conditions and statistical characteristics of the errors, obtaining comparison analysis results, adjusting parameters of the data analysis evaluation model according to comparison analysis results, carrying out calibration on the data analysis evaluation model by adopting a gradient descent method in an optimization algorithm, carrying out comparison between the operation model and the actual equipment operation data again after the calibration model, verifying the calibration effect, if the errors are descended, describing that the calibration is successful, and continuing the calibration if the errors are not descended, describing that the calibration is unsuccessful.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411392313.6A CN119293664A (en) | 2024-10-08 | 2024-10-08 | A device operation evaluation method based on multi-source data fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411392313.6A CN119293664A (en) | 2024-10-08 | 2024-10-08 | A device operation evaluation method based on multi-source data fusion |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN119293664A true CN119293664A (en) | 2025-01-10 |
Family
ID=94151396
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411392313.6A Pending CN119293664A (en) | 2024-10-08 | 2024-10-08 | A device operation evaluation method based on multi-source data fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119293664A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119537890A (en) * | 2025-01-22 | 2025-02-28 | 北京新兴科遥信息技术有限公司 | Monitoring and evaluation method for ecological restoration planning implementation based on multi-source data fusion technology |
| CN119716355A (en) * | 2025-02-27 | 2025-03-28 | 闽西职业技术学院 | Electrical equipment fault prediction system and method based on data fusion |
| CN119831112A (en) * | 2025-03-15 | 2025-04-15 | 北京北方永达智能电气有限公司 | Preloaded equipment compartment fault prediction method based on multimode data analysis |
| CN120031549A (en) * | 2025-04-21 | 2025-05-23 | 四川省机场集团有限公司 | 10KV substation equipment multi-source heterogeneous data collaborative fault prediction method and system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118094379A (en) * | 2024-04-07 | 2024-05-28 | 西安众望能源科技有限公司 | Oil well fault diagnosis and prediction method and system based on multi-factor fusion analysis |
| CN118503636A (en) * | 2024-07-22 | 2024-08-16 | 温州职业技术学院 | Power distribution equipment health state assessment method and system |
| CN118503832A (en) * | 2024-07-16 | 2024-08-16 | 杭州祐全科技发展有限公司 | Industrial intelligent detection method and system based on multi-mode large model |
-
2024
- 2024-10-08 CN CN202411392313.6A patent/CN119293664A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118094379A (en) * | 2024-04-07 | 2024-05-28 | 西安众望能源科技有限公司 | Oil well fault diagnosis and prediction method and system based on multi-factor fusion analysis |
| CN118503832A (en) * | 2024-07-16 | 2024-08-16 | 杭州祐全科技发展有限公司 | Industrial intelligent detection method and system based on multi-mode large model |
| CN118503636A (en) * | 2024-07-22 | 2024-08-16 | 温州职业技术学院 | Power distribution equipment health state assessment method and system |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119537890A (en) * | 2025-01-22 | 2025-02-28 | 北京新兴科遥信息技术有限公司 | Monitoring and evaluation method for ecological restoration planning implementation based on multi-source data fusion technology |
| CN119716355A (en) * | 2025-02-27 | 2025-03-28 | 闽西职业技术学院 | Electrical equipment fault prediction system and method based on data fusion |
| CN119831112A (en) * | 2025-03-15 | 2025-04-15 | 北京北方永达智能电气有限公司 | Preloaded equipment compartment fault prediction method based on multimode data analysis |
| CN120031549A (en) * | 2025-04-21 | 2025-05-23 | 四川省机场集团有限公司 | 10KV substation equipment multi-source heterogeneous data collaborative fault prediction method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR102427205B1 (en) | Apparatus and method for generating training data of artificial intelligence model | |
| CN119293664A (en) | A device operation evaluation method based on multi-source data fusion | |
| CN118504794B (en) | Intelligent fault diagnosis system of electric spark forming machine | |
| CN118261584B (en) | Transformer state evaluation method and system based on multi-parameter data | |
| CN117114454A (en) | DC sleeve state evaluation method and system based on Apriori algorithm | |
| CN119312148B (en) | Equipment failure prediction method and related equipment based on industrial Internet of Things | |
| CN118536048A (en) | Switchgear insulation status monitoring and management system | |
| CN118732671B (en) | An intelligent test system for the response time of motor control of new energy vehicles | |
| CN117560300B (en) | Intelligent internet of things flow prediction and optimization system | |
| CN119312205B (en) | Intelligent fire-fighting equipment fault identification method and system | |
| CN109871002B (en) | Concurrent abnormal state identification and positioning system based on tensor label learning | |
| CN112734977B (en) | Equipment risk early warning system and algorithm based on Internet of things | |
| CN118967106A (en) | A special release motor fault detection method | |
| CN117689269A (en) | Fault tracing method for intelligent watch processing stage | |
| KR102602840B1 (en) | Training dataset generating system and training dataset generating method using the same | |
| CN113177646A (en) | Power distribution equipment online monitoring method and system based on self-adaptive edge proxy | |
| CN119249281B (en) | Hydrogen long-distance pipeline sensor optimization method based on artificial intelligence algorithm | |
| CN119618702A (en) | Automatic detection and fault warning system and method for self-service car washing machine | |
| CN119226099A (en) | A computer monitoring system used in water conservancy projects | |
| CN119167240A (en) | A fault cause analysis system based on machine learning | |
| CN119202987A (en) | A bearing fault diagnosis method and device | |
| CN118669281A (en) | Online state evaluation method, device and equipment of wind motor and storage medium | |
| CN118690283A (en) | A method and system for diagnosing power equipment based on pattern recognition algorithm | |
| CN118354287A (en) | Abnormality alarm method and system based on wireless sensor network | |
| CN117806233A (en) | Intelligent monitoring system and method for numerical control machine tool |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |