CN117114446A - Device health assessment method based on data driving - Google Patents
Device health assessment method based on data driving Download PDFInfo
- Publication number
- CN117114446A CN117114446A CN202311104138.1A CN202311104138A CN117114446A CN 117114446 A CN117114446 A CN 117114446A CN 202311104138 A CN202311104138 A CN 202311104138A CN 117114446 A CN117114446 A CN 117114446A
- Authority
- CN
- China
- Prior art keywords
- score
- deviation
- equipment
- parameter
- dimension
- 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
- 230000036541 health Effects 0.000 title claims abstract description 95
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000012423 maintenance Methods 0.000 claims abstract description 87
- 238000013210 evaluation model Methods 0.000 claims abstract description 48
- 238000011156 evaluation Methods 0.000 claims abstract description 32
- 238000005265 energy consumption Methods 0.000 claims abstract description 27
- 238000004364 calculation method Methods 0.000 claims abstract description 17
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 16
- 238000012544 monitoring process Methods 0.000 claims abstract description 16
- 238000010219 correlation analysis Methods 0.000 claims abstract description 7
- 238000007418 data mining Methods 0.000 claims abstract description 7
- 238000007621 cluster analysis Methods 0.000 claims abstract description 6
- 238000005516 engineering process Methods 0.000 claims abstract description 6
- 238000007781 pre-processing Methods 0.000 claims abstract description 6
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 12
- 230000006872 improvement Effects 0.000 description 11
- 230000009471 action Effects 0.000 description 6
- 230000002159 abnormal effect Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 4
- 238000007726 management method Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 230000006399 behavior Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 230000000007 visual effect Effects 0.000 description 3
- 238000012937 correction Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000013439 planning Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000002035 prolonged effect Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000012800 visualization Methods 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 238000005299 abrasion Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000008821 health effect Effects 0.000 description 1
- 230000005802 health problem Effects 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000003064 k means clustering Methods 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000013468 resource allocation Methods 0.000 description 1
- 238000012502 risk assessment Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/27—Regression, e.g. linear or logistic regression
-
- 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
- G06N3/092—Reinforcement 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/067—Enterprise or organisation modelling
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Educational Administration (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- General Business, Economics & Management (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- General Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Probability & Statistics with Applications (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention discloses a device health assessment method based on data driving, which comprises the following steps: step 1: establishing an equipment running condition evaluation model, an equipment fault evaluation model, an equipment maintenance evaluation model and an equipment energy efficiency evaluation model; step 2: collecting equipment running condition records, equipment fault records, equipment maintenance records and equipment energy consumption monitoring records; step 3: the record obtained in the step 2 is subjected to data preprocessing, health parameter correlation analysis, cluster analysis, a data mining regression algorithm and a deep Q network learning technology in sequence; step 4: inputting the deviation degree calculation deviation coefficient obtained in the step 3 into the evaluation model of the step 1), and calculating to obtain an equipment running condition evaluation score; and calculating through the equipment health evaluation model to obtain the equipment health degree. The invention aims to solve the problem that the overall health condition of the equipment cannot be accurately estimated due to single measuring point index variable in the traditional health estimation method.
Description
Technical Field
The invention relates to the field of equipment health assessment, in particular to a data-driven equipment health assessment method.
Background
In the existing technical scheme for evaluating the health of equipment, a real-time monitoring platform is established by collecting the operation data of the equipment, the operation mode of the equipment is identified, and a prediction model is established, so that the health state of the equipment is evaluated. However, due to the complexity of the device, parameters and characteristics of different variables have great influence on the result, the index variables of the measuring points are often evaluated only according to the data of the measuring points, and the information such as fault modes and maintenance modes of the device is not considered, so that the overall health state and future prediction of the device cannot be effectively and comprehensively reflected, and comprehensive consideration of the correlation between different variables is lacking. Even if accurate monitoring and evaluation are performed on a single measurement point index variable, the comprehensive condition of each variable is not analyzed and judged, and the judgment of the health condition cannot be accurately provided.
Disclosure of Invention
The invention discloses a device health assessment method based on data driving, which aims to solve the problem that the overall health condition of a device cannot be accurately assessed due to single measuring point index variable in the traditional health assessment method.
A method for data-driven based device health assessment, comprising the steps of:
step 1: establishing an equipment running condition evaluation model, an equipment fault evaluation model, an equipment maintenance evaluation model and an equipment energy efficiency evaluation model;
step 2: collecting equipment running condition records, equipment fault records, equipment maintenance records and equipment energy consumption monitoring records;
step 3: the method comprises the steps of obtaining a deviation degree calculation deviation coefficient D1-Dn of an operation condition evaluation model, a deviation degree calculation deviation coefficient P1-Pm of an equipment failure evaluation model, a deviation degree calculation deviation coefficient R1-Ri of an equipment maintenance evaluation model and a deviation degree calculation deviation coefficient S1-Sj of an equipment energy efficiency evaluation model through data preprocessing, health parameter correlation analysis, cluster analysis, a data mining regression algorithm and a deep Q network learning technology respectively on the equipment operation condition record, the equipment failure record, the equipment maintenance record and the equipment energy consumption monitoring record obtained in the step 2;
step 4: inputting the deviation degree calculation deviation coefficients D1-Dn of the operation condition evaluation model obtained in the step 3 into the operation condition evaluation model of the equipment in the step 1), and calculating to obtain an operation condition evaluation score of the equipment;
inputting the deviation degree calculation deviation coefficient P1-Pm of the equipment fault evaluation model obtained in the step 3 into the equipment fault evaluation model of the step 1), and calculating to obtain an equipment fault evaluation score;
inputting the deviation degree calculation deviation coefficient R1-Ri of the equipment maintenance evaluation model obtained in the step 3 into the equipment maintenance evaluation model of the step 1), and calculating to obtain an equipment maintenance evaluation score;
inputting the deviation degree calculation deviation coefficient S1-Sj of the equipment energy efficiency evaluation model obtained in the step 3 into the equipment energy efficiency evaluation model, and calculating to obtain an equipment energy efficiency evaluation score;
and calculating through the equipment health evaluation model to obtain the equipment health degree.
In step 1, the establishment of the operation condition evaluation model includes:
the parameter dimension n, the score Hn, the deviation coefficient of the corresponding parameter Dn, for example, is as follows:
1) Parameter 1: a score of H1, calculating a deviation coefficient D1 according to the deviation degree, and grading to be H1 (1-D1);
2) Parameter 2: calculating a deviation coefficient D2 according to the deviation degree, wherein the score is H2 (1-D2);
3) Parameter 3: calculating a deviation coefficient D3 according to the deviation degree, wherein the score is H3 (1-D3);
4) Parameter 4: a score of H4, calculating a deviation coefficient D4 according to the deviation degree, wherein the score is H4 (1-D4);
……
n) parameter n, score Hn, calculating deviation coefficient Dn according to the deviation degree, and scoring Hn (1-Dn);
parameters 1 to n are parameters of the running condition of the equipment, and the dimension value of the running condition Where n is the nth dimension, hn is the score of the nth dimension, and Dn is the deviation value of the nth dimension.
In step 1, the establishment of the equipment fault evaluation model includes:
1) Parameter 1: calculating a deviation coefficient P1 according to the deviation degree, wherein the score is X1 (1-P1);
2) Parameter 2: calculating a deviation coefficient P2 according to the deviation degree, wherein the score is X2 (1-P2);
3) Parameter 3: calculating a deviation coefficient P3 according to the deviation degree, wherein the score is X3 (1-P3);
4) Parameter 4: calculating a deviation coefficient P4 according to the deviation degree, wherein the score is X4 (1-P4);
……
m) parameter m, score Xm, calculating deviation coefficient Pm according to deviation degree, score Xm (1-Pm);
parameters 1 to m are parameters of equipment fault evaluation, and the score of fault dimension Where m is the m dimension, xm is the score of the m dimension, and Pm is the deviation value of the m dimension.
In step 1, the establishment of the equipment maintenance evaluation model includes:
1) Parameter 1: calculating a deviation coefficient R1 according to the deviation degree, wherein the score is C1 (1-R1);
2) Parameter 2: calculating a deviation coefficient R2 according to the deviation degree, wherein the score is C2 (1-R2);
3) Parameter 3: calculating a deviation coefficient R3 according to the deviation degree, wherein the score is C3 (1-R3);
4) Parameter 4: calculating a deviation coefficient R4 according to the deviation degree, wherein the score is C4 (1-R4);
……
i) And calculating a deviation coefficient Ri according to the deviation degree, wherein the score is Ci (1-Ri) according to the parameter i.
Parameters 1 to i are equipment maintenance evaluation parameters, and the maintenance dimension scoreWhere i is the ith dimension, ci is the score of the ith dimension, and Ri is the bias value of the ith dimension.
In step 1, the establishment of the device energy efficiency evaluation model includes:
1) Parameter 1: score F1, the deviation coefficient S1 was calculated from the degree of deviation, and score F1 (1-S1).
2) Parameter 2: score F2, the deviation coefficient S2 was calculated from the degree of deviation, and score F2 x (1-S2).
3) Parameter 3: score F3, the deviation coefficient S3 was calculated from the degree of deviation, and score F3 (1-S3).
4) Parameter 4: score F4, the deviation coefficient S4 was calculated from the degree of deviation, and score F4 (1-S4).
……
j) And calculating a deviation coefficient Sj according to the deviation degree, wherein the score is Fj (1-Sj) according to the parameter j.
Parameters 1 to j are the device energy efficiency evaluation parameters, and the score of the energy efficiency dimension Where j is the j dimension, fj is the j dimension score, and Sj is the j dimension bias value.
In step 4, the device health assessment model includes:
degree of health Z of equipment he ＝H op *(1-K op )+X fm *(1-L fm )+C me *(1-T me )+F ec *(1-Y ec ) Wherein H is op For the score of the run-time dimension, K op The deviation value of the operation condition dimension is customized according to the expert experience knowledge base; x is X fm For the score of the fault dimension, L fm The deviation value of fault dimension is customized according to expert experience knowledge base; c (C) me To maintain the score of maintenance dimension, T me The maintenance dimension deviation value is defined according to an expert experience knowledge base; f (F) ec For the score of energy efficiency dimension, Y ec The deviation value of the energy efficiency dimension is customized according to the expert experience knowledge base.
Compared with the prior art, the invention has the following advantages:
compared with the common methods in the current market, the method has the following advantages that the technical problems of the existing methods can be solved:
(1) overall evaluation: the multivariable health evaluation scheme can comprehensively consider data of multiple aspects such as measurement point indexes, fault records, maintenance records, energy consumption records and the like of the equipment. By comprehensively analyzing the information, the health condition of the equipment can be more comprehensively estimated, and a long-term correction plan can be formulated
(2) And (3) fault data analysis: the fault records are included in the variables of the health assessment, so that potential fault modes and risks can be identified, and corresponding measures are taken through statistics and analysis of equipment fault information, so that loss caused by equipment faults is avoided or reduced.
(3) Maintenance policy monitoring: the maintenance record is included in the health evaluation variable, so that the health condition of equipment maintenance can be identified and judged, and the optimal maintenance period, maintenance method and maintenance content can be determined by analyzing the maintenance record and maintenance information of the equipment, so that the service life of the equipment is prolonged, the reliability is improved, and the maintenance cost is reduced.
(4) Energy consumption efficiency improvement: incorporating the energy consumption record into the health assessment variables can help assess the energy consumption of the device and improve. By analyzing the energy consumption data, the problems of abnormal energy consumption, low energy efficiency, energy waste and the like can be found, and corresponding optimization suggestions are provided so as to reduce energy cost and environmental impact.
(5) Comprehensive decision support: the multi-variable health assessment scheme can provide more comprehensive data support for the management layer, helping them make more accurate and comprehensive decisions. By comprehensively considering various variables of the equipment, the overall condition of the equipment can be better known, and corresponding strategies and plans are formulated so as to improve the operation efficiency and reduce the risk.
In general, the multivariate health assessment scheme can provide a more comprehensive and accurate equipment health assessment result, and makes corrective action more specifically, thereby helping to prevent faults, optimize maintenance strategies, improve energy consumption efficiency, promote equipment health conditions, improve equipment reliability, prolong equipment service life and optimize operation management.
The predictive maintenance method based on equipment health assessment has the following benefits:
(1) the maintenance cost of equipment is reduced: and the health condition of the equipment is accurately analyzed, the equipment faults are predicted, the maintenance plan is reasonably arranged, and the equipment maintenance cost is reduced.
(2) The running efficiency of the equipment is improved: by timely maintaining and optimizing the maintenance strategy, the downtime of the equipment is reduced, and the operation efficiency of the equipment is improved.
(3) Prolonging the service life of the equipment: through predictive maintenance and optimization maintenance strategies, the loss and abrasion of equipment are reduced, and the service life of the equipment is prolonged.
(4) And the reliability of the whole equipment is improved: by continuously optimizing maintenance strategies and updating equipment health curves, the reliability and safety of equipment are improved.
Drawings
Fig. 1 is a flow chart of a method for evaluating health of a device based on data driving according to the present invention.
Detailed Description
The invention has the starting point that the health evaluation is comprehensively and effectively carried out on the equipment, and the intelligent decision-making of equipment management personnel is helped to carry out the measures and actions of preparation on the equipment in reasonable time. And the repeated periodic maintenance tasks are not blindly executed, so that the waste of resources is avoided. The method is characterized in that the method comprises the steps of evaluating the state of equipment in a multi-angle and multi-element manner, comprehensively analyzing the running condition, fault record, maintenance and energy consumption record information, comprehensively evaluating the health condition of the equipment, and quantifying the health degree of the equipment, so that a long-term correction plan covering the aspects of fault prediction, maintenance optimization, energy consumption efficiency improvement and the like is formulated.
As shown in fig. 1, the specific method of the present invention comprises the following steps:
step 1: establishing evaluation criteria
According to the characteristic mode and the behavior rule of the equipment, calculating a health evaluation value of the equipment, wherein the health evaluation value is used for reflecting the state of the equipment and possible problems.
The device health assessment model influencing factor variables are as follows:
(1) device operating condition assessment
The item comprises the running state, running time, running temperature and the like of the equipment, the data is derived from the running condition record of the equipment, and the running time, the temperature and other indexes are analyzed according to the running condition record of the equipment.
The equipment running condition evaluation influence factor weight proportion is determined according to the patent algorithm, the industry expert experience knowledge base and the actual conditions of enterprises. The score evaluation method and criteria are as follows:
the parameter dimension n, the score Hn, the deviation coefficient of the corresponding parameter Dn, for example, is as follows:
1) Parameter 1: the score is H1, the deviation coefficient D1 is calculated according to the deviation degree, and the score is H1 (1-D1).
2) Parameter 2: the score is H2, the deviation coefficient D2 is calculated according to the deviation degree, and the score is H2 (1-D2).
3) Parameter 3: score H3, the deviation factor D3 was calculated from the degree of deviation, and the score H3 was (1-D3).
4) Parameter 4: score H4, a deviation factor D4 was calculated based on the degree of deviation, and scores H4 (1-D4).
……
n) parameter n, score Hn, calculate the deviation factor Dn from the degree of deviation, score Hn (1-Dn).
Score of run Condition dimensionWhere n is the nth dimension, hn is the score for this dimension, and Dn is the deviation value for this dimension.
(2) Device fault assessment
The item comprises the number of faults, the fault type, the fault reason and the like, the data is derived from equipment fault records, and the number of faults, the fault type, the fault reason and the like are analyzed according to the equipment fault records.
The weight proportion of the equipment fault evaluation influencing factors is determined according to the patent algorithm, the industry expert experience knowledge base and the actual conditions of enterprises. The score evaluation method and criteria are as follows:
the parameter dimension m, the score value Xm, the deviation coefficient of the corresponding parameter is Pm, and the example is as follows:
1) Number of failures: the score is X1, the deviation coefficient P1 is calculated according to the deviation degree, and the score is X1 (1-P1).
2) Fault type: score X2, the deviation factor P2 was calculated from the degree of deviation, and score X2X (1-P2).
3) Maintenance time: score X3, the deviation factor P3 was calculated from the degree of deviation, and score X3X (1-P3).
4) Maintenance cost: score X4, the deviation factor P4 was calculated from the degree of deviation, and the score was X4X (1-P4).
……
m) parameter m, score Xm, calculate the deviation coefficient Pm according to the deviation degree, score Xm (1-Pm).
Score of fault dimensionWhere m is the m-th dimension, xm is the score for this dimension, and Pm is the deviation value for this dimension.
(3) Device maintenance assessment
The item includes frequency of maintenance, content of maintenance, quality of maintenance, and the like. The data is derived from the equipment maintenance record in the first step, and the frequency, content, quality and the like of maintenance are analyzed according to the equipment maintenance record.
The weight proportion of the equipment maintenance evaluation influencing factors is determined according to the patent algorithm, the industry expert experience knowledge base and the actual conditions of enterprises. The score evaluation method and criteria are as follows:
the parameter dimension i, the score Ci, the deviation coefficient of the corresponding parameter is Ri, and the example is as follows:
1) Maintaining and recording: score C1, a deviation factor R1 was calculated from the degree of deviation, and score C1 x (1-R1).
2) Maintenance quality: score C2, a deviation factor R2 was calculated from the degree of deviation, and score C2 x (1-R2).
3) Maintenance cost: score C3, a deviation factor R3 was calculated based on the degree of deviation, and score C3 x (1-R3).
4) Maintenance tool and equipment: score C4, a deviation factor R4 was calculated based on the degree of deviation, and scores C4 x (1-R4).
……
i) And calculating a deviation coefficient Ri according to the deviation degree, wherein the score is Ci (1-Ri) according to the parameter i.
Maintenance dimension scoreWhere i is the ith dimension, ci is the score for this dimension, and Ri is the bias value for this dimension.
(4) Device energy efficiency assessment
The method comprises the steps of including energy consumption data, energy efficiency indexes and the like, wherein the data are derived from equipment energy consumption monitoring records in the first step, and the energy consumption conditions and the energy efficiency indexes are analyzed according to the equipment energy consumption monitoring records.
The equipment energy efficiency evaluation influence factor weight proportion is determined according to the patent algorithm, an industry expert experience knowledge base and the actual conditions of enterprises. The score evaluation method and criteria are as follows:
the parameter dimension j, the score value is Fi, the deviation coefficient of the corresponding parameter is Si, and the example is as follows:
1) Energy consumption data: score F1, the deviation coefficient S1 was calculated from the degree of deviation, and score F1 (1-S1).
2) Energy efficiency rating: score F2, the deviation coefficient S2 was calculated from the degree of deviation, and score F2 x (1-S2).
3) Operation mode: score F3, the deviation coefficient S3 was calculated from the degree of deviation, and score F3 (1-S3).
4) Energy utilization rate: score F4, the deviation coefficient S4 was calculated from the degree of deviation, and score F4 (1-S4).
……
j) And calculating a deviation coefficient Sj according to the deviation degree, wherein the score is Fj (1-Sj) according to the parameter j.
Score of energy efficiency dimensionWhere j is the j-th dimension, fj is the score of this dimension, and Sj is the bias of this dimension.
To sum up, the final device health assessment model score
Z he ＝H op *(1-K op )+X fm *(1-L fm )+C me *(1-T me )+F ec *(1-Y ec ) Wherein H is op For running status dimension score, K op Deviation values for the run-state dimension; x is X fm For fault dimension score, L fm Deviation values for fault dimensions; c (C) me To maintain dimension score, T me A deviation value for a maintenance dimension; f (F) ec For energy efficiency dimension score, Y ec Is the deviation value of the energy efficiency dimension.
Step 2: monitoring and data acquisition
(1) Equipment operation condition record: by means of the state monitoring system, sensors are installed on the equipment to monitor various parameters of the equipment, such as temperature, pressure, vibration and the like, and data are collected in real time.
(2) And (3) equipment fault record: and creating a fault report through a fault management system, recording related information including fault description, occurrence time, equipment/system information and the like, and completing the establishment of a solution library and a knowledge base.
(3) And (5) equipment maintenance record: by means of the maintenance system, maintenance activities of equipment and assets are planned, recorded and managed, and detailed information of each maintenance is recorded, including maintenance dates, operation records, replaced parts and the like.
(4) And (3) monitoring and recording equipment energy consumption: and the energy consumption monitoring system is used for monitoring, recording and managing the energy consumption condition of the building or equipment, the energy metering equipment is connected to the building or equipment, acquiring the energy consumption data in real time and recording and storing the historical energy consumption data.
Step 3: data statistics and analysis
And processing and analyzing the acquired data, and identifying the characteristic mode and the behavior rule of the equipment by applying an advanced algorithm and a mode identification technology.
(1) Data preprocessing: preprocessing the collected original data, including the processing steps of data cleaning, abnormal value removal, missing value filling and the like. The quality and the reliability of the data are improved, and a reliable basis is provided for subsequent statistics and analysis.
(2) Health parameter correlation analysis: and carrying out statistics and correlation analysis on the equipment operation data to find out parameters closely related to the health condition of the equipment. By identifying these key parameters, the primary influencing factors affecting the health of the device are determined.
(3) And (3) cluster analysis: data points of similar features are grouped together to help discover different patterns or behaviors in the data. Different working conditions or running states of the equipment are identified, and corresponding processing strategies are adopted for different types of data.
(4) Expert experience knowledge base: the analysis of the health problems of the equipment is carried out by means of expert experience and an existing knowledge base. According to the experience and expertise of the expert, the opinion and advice about the influencing factors of the health effect of the equipment are provided in combination with the equipment operation data and fault records.
(5) Data mining regression algorithm and deep Q network learning technique: and analyzing and modeling the equipment operation data by utilizing a data mining regression algorithm and a deep Q network learning technology to discover potential association factors and influence factors. Factors affecting the health of the device are identified and predictive and decision support is provided.
Step 4: device health assessment
After the equipment health evaluation model is established, a subsequent action plan is output according to the equipment health evaluation model, so that the health condition and performance of the equipment are improved.
(1) Identifying key questions according to the assessment model: accurately identifying problems is the basis for planning actions. And determining key problems and hidden dangers of the equipment according to the result of the equipment health evaluation. Including frequent failures, low performance components or systems, non-standard operation, etc.
(2) Setting priority: priority is set for each question according to the importance and urgency of the question. Attention is focused on the problems with the most influence and urgency on the health condition of the equipment, and reasonable resource allocation is ensured.
(3) And (3) making a specific target: a clear target is formulated for each critical issue. These objectives are quantifiable, specific, and achievable. Such as reduced failure rate, improved energy efficiency, extended equipment life, etc.
(4) Determining improvement measures: for each problem, corresponding improvement measures are formulated. Including repair equipment, replacement of parts, improvement of operating specifications, reinforcement of maintenance planning, improvement of process flows, and the like. It is ensured that each improvement measure is associated with a root cause of the problem to solve the root problem of the problem.
(5) And (5) making a time table: a schedule is formulated for each improvement measure. The time to start and complete each improvement is determined to ensure that the program can be executed on time. Meanwhile, workload and resource requirements are reasonably evaluated, and feasibility of a time table is ensured.
(6) Allocation responsibilities and resources: the responsible person for each improvement measure is determined and the required resources are allocated. This may include human resources, financial resources, equipment and tools, etc. Ensuring that the responsible person has the ability and resources to implement the improvement measure.
Step 5: health visualization display and report
Visual display and report of equipment health assessment are one of the important functions of the technical scheme, and the visual display and report method for the health condition of the equipment is provided, so that a user is helped to comprehensively know the running condition and maintenance requirement of the equipment.
(1) Instrument panel: the system provides an intuitive dashboard showing key indicators and overview information for the device. This may include a summary of information on device status, failure frequency, energy consumption trends, health scores, etc. The user can clearly know the health condition of the whole equipment through the instrument panel.
(2) Equipment distribution map: the system provides a device map showing the location and status of the device in terms of geographic location or building plan. By identifying the health of the device by different symbols or colors, the user can quickly browse and locate devices that may be problematic.
(3) Health score and report: based on the operational data of the devices and the health assessment algorithm, the system generates a health score and report for each device. Health scores are typically based on some key criteria such as energy efficiency, frequency of failures, maintenance records, etc. The report may contain basic information about the device, descriptions of health conditions, problem cues, and recommended maintenance measures.
(4) Alert and exception notification: when the health state of the equipment is abnormal or reaches the early warning condition, the system can send out an alarm notification in time. These alerts may be presented and notified by way of dashboards, mobile applications, or email, etc., so that the user can take action in a timely manner to address the problem.
The user can also customize reports, export data, set customized health assessment indexes and the like according to the needs. Visual presentation and reporting of device health assessment helps users to better understand the health of the device, discover potential problems and opportunities, and take timely action to optimize the device's performance and extend its useful life.
Predictive maintenance methods based on equipment health assessment can play an important role in the field of equipment maintenance. By accurately analyzing the state and fault rule of the equipment, the maintenance strategy is optimized, the maintenance cost is reduced, and the operation efficiency and the overall reliability of the equipment are improved. In future industrial production, the method plays an increasingly important role, and brings more value and innovation for equipment maintenance.
The following describes the embodiment of the present invention in detail by taking a water pump in the water service industry as an example:
(1) and (3) data acquisition: and selecting one water pump, and collecting running condition records, fault records, maintenance records and energy consumption monitoring records of the water pump.
(2) Data preprocessing: the collected water pump data are preprocessed sequentially, the data are smoothed by calculating the average value of the data points in a period of time, the data noise is reduced, and the influence of abnormal values is restrained. And converting the data into standard normal distribution according to the average value and the average difference of the data. Finally, the missing data is extrapolated according to trends and relationships between known data points. The dimension of the data is reduced, unnecessary redundant information is reduced, the format of the data is converted into data content which is easy to analyze, and recorded data which can be suitable for an algorithm model is obtained.
(3) Health parameter correlation analysis and cluster analysis: different algorithms are employed depending on the different data characteristics. In the embodiment of the water pump, a manner of combining the pearson coefficient and the K-means clustering algorithm is adopted, firstly, the distance between each data point and each clustering center is calculated according to the data point, the data point is distributed to the cluster where the closest clustering center is located, the mean value of all the data points in the cluster is calculated, the mean value is used as a new clustering center, the mean value, the mean difference and the covariance of different variables are calculated respectively, the pearson coefficient is calculated, the relation among the variables is expressed, and then different parameters and corresponding deviation coefficients are calculated. A. Parameters and deviation coefficients for equipment operation condition assessment:
parameter 1: run time score H1:83 deviation factor D1 for parameter 1:0.70
Parameter 2: temperature score H2: deviation factor D2 of 80 minutes parameter 2:0.80
Parameter 3: pressure score H3: deviation coefficient D3 of 71 minutes parameter 3:0.85
Parameter 4: flow score H4: deviation factor D4 for parameter 4 of 88: 0.85
Parameter 5: vibration score H5: deviation factor D5 for 95 minutes parameter 5:0.85
Score H of the run Condition dimension op :83 (1-0.70) +80 (1-0.80) +71 (1-0.85) +88 (1-0.85) +95 (1-0.85) =79 minutes
B. Parameters and deviation coefficients for equipment fault assessment
Parameter 1: failure times score X1: deviation coefficient P1 for parameter 1 of 73: 0.80
Parameter 2: maintenance time score X2: deviation coefficient P2 for parameter 2 of 87: 0.80
Parameter 3: fault type score X3: deviation coefficient P3 for 94 minutes parameter 3:0.85
Parameter 4: maintenance cost score X4: deviation coefficient P4 for 84 minutes parameter 4:0.85
Parameter 5: fault cause score X5: deviation coefficient P5 for parameter 5 of 86: 0.85
Parameter 6: treatment measure score X6: deviation coefficient P6 for parameter 6 of 93: 0.85
Score X of failure dimension fm :73 (1-0.80) +87 (1-0.80) +94 (1-0.85) +84 (1-0.85) +86 (1-0.85) +93 (1-0.85) = 85.55 minutes)
C. Parameters and deviation coefficients for equipment maintenance assessment
Parameter 1: maintenance quality score C1:76 deviation from parameter 1 is R1:0.75 parameter 2: maintenance record score C2:94 deviation from parameter 2 is R2:0.80 parameter 3: tool and device score C3: deviation coefficient R3 for parameter 3 of 97: 0.80 parameter 4: personnel skill score C4: deviation coefficient R4 for parameter 4 of 78: 0.80
Parameter 5: maintenance cost score C5: deviation factor R5 of 89 minutes parameter 5: score C of 0.85 maintenance dimension me :76 (1-0.75) +94 (1-0.80) +97 (1-0.80) +78 (1-0.80) +89 (1-0.85) =86.15 minutes
D. Parameters and deviation coefficients for device energy efficiency assessment
Parameter 1: energy consumption data score F1: deviation coefficient S1 for parameter 1 of 70: 0.70
Parameter 2: energy efficiency rating score F2: deviation coefficient S2 of 81 minutes parameter 2:0.75
Parameter 3: energy utilization score F3: deviation coefficient S3 for 96 minutes parameter 3:0.75
Parameter 4: run mode score F4: deviation coefficient S4 for parameter 4 of 71: 0.80
Score F of energy efficiency dimension ec :70 (1-0.70) +81 (1-0.75) +96 (1-0.75) +71 (1-0.80) = 79.45 minutes)
(4) Establishing a device health evaluation model: and formulating deviation values of all dimensions of the equipment according to the expert experience knowledge base, and combining correlation analysis and cluster analysis to obtain scores of all dimensions, so that an overall equipment health evaluation model is established, and the health degree of the water pump is calculated.
Score H of running condition dimension of water pump op 79 minutes, offset value K op 0.70; score X of failure dimension fm 85.55 minutes, offset L fm 0.70; maintenance dimension score C me 86.15 minutes, offset T me 0.80; score F of energy efficiency dimension ec 79.45 minutes, offset Y ec 0.80.
Score of water pump health: 79 (1-0.70) +85.55 (1-0.70) +86.15 (1-0.80) +79.45 (1-0.80) = 82.485 minutes)
(5) Data mining regression algorithm and deep Q network learning technique: after the health model and the current health degree of the water pump are obtained, historical data are input into the model, a data mining regression algorithm is adopted, the model is trained by using a training set, and model parameters are adjusted to minimize errors. And iteratively adjusting model parameters in the training process until convergence conditions are reached, testing the model through historical data, evaluating the performance of the model, and outputting predicted values and errors of model variables. And finally, processing a large-scale data space by using a deep Q network learning technology, obtaining a state function by using a neural network, playing back buffer data empirically, obtaining a predicted target value by using a target network, obtaining a prediction probability by using a strategy network, updating a strategy gradient, further predicting a future equipment health trend, and outputting a possible health value in a future period of time.
(6) Health visualization display and report: according to specific health data, audience demands and use scenes, flexible selection and combination are carried out, and health trend and health risk assessment of the water pump are displayed through the forms of an instrument panel, a graph, a radar chart, a report and the like, so that a user can conveniently and correctly understand and utilize the health degree of the water pump.
Claims (6)
1. A method for data-driven based device health assessment, comprising the steps of:
step 1: establishing an equipment running condition evaluation model, an equipment fault evaluation model, an equipment maintenance evaluation model and an equipment energy efficiency evaluation model;
step 2: collecting equipment running condition records, equipment fault records, equipment maintenance records and equipment energy consumption monitoring records;
step 3: the method comprises the steps of obtaining a deviation degree calculation deviation coefficient D1-Dn of an operation condition evaluation model, a deviation degree calculation deviation coefficient P1-Pm of an equipment failure evaluation model, a deviation degree calculation deviation coefficient R1-Ri of an equipment maintenance evaluation model and a deviation degree calculation deviation coefficient S1-Sj of an equipment energy efficiency evaluation model through data preprocessing, health parameter correlation analysis, cluster analysis, a data mining regression algorithm and a deep Q network learning technology respectively on the equipment operation condition record, the equipment failure record, the equipment maintenance record and the equipment energy consumption monitoring record obtained in the step 2;
step 4: inputting the deviation degree calculation deviation coefficients D1-Dn of the operation condition evaluation model obtained in the step 3 into the operation condition evaluation model of the equipment in the step 1), and calculating to obtain an operation condition evaluation score of the equipment;
inputting the deviation degree calculation deviation coefficient P1-Pm of the equipment fault evaluation model obtained in the step 3 into the equipment fault evaluation model of the step 1), and calculating to obtain an equipment fault evaluation score;
inputting the deviation degree calculation deviation coefficient R1-Ri of the equipment maintenance evaluation model obtained in the step 3 into the equipment maintenance evaluation model of the step 1), and calculating to obtain an equipment maintenance evaluation score;
inputting the deviation degree calculation deviation coefficient S1-Sj of the equipment energy efficiency evaluation model obtained in the step 3 into the equipment energy efficiency evaluation model, and calculating to obtain an equipment energy efficiency evaluation score;
and calculating through the equipment health evaluation model to obtain the equipment health degree.
2. The method for data-driven based device health assessment according to claim 1, wherein in step 1, the establishment of the operation condition assessment model comprises:
the parameter dimension n, the score Hn, the deviation coefficient of the corresponding parameter Dn, for example, is as follows:
1) Parameter 1: a score of H1, calculating a deviation coefficient D1 according to the deviation degree, and grading to be H1 (1-D1);
2) Parameter 2: calculating a deviation coefficient D2 according to the deviation degree, wherein the score is H2 (1-D2);
3) Parameter 3: calculating a deviation coefficient D3 according to the deviation degree, wherein the score is H3 (1-D3);
4) Parameter 4: a score of H4, calculating a deviation coefficient D4 according to the deviation degree, wherein the score is H4 (1-D4);
……
n) parameter n, score Hn, calculating deviation coefficient Dn according to the deviation degree, and scoring Hn (1-Dn);
parameters 1 to n are parameters of the running condition of the equipment, and the dimension value of the running condition Where n is the nth dimension, hn is the score of the nth dimension, and Dn is the deviation value of the nth dimension.
3. The data-driven based device health assessment method according to claim 1, wherein in step 1, the establishment of the device failure assessment model comprises:
1) Parameter 1: calculating a deviation coefficient P1 according to the deviation degree, wherein the score is X1 (1-P1);
2) Parameter 2: calculating a deviation coefficient P2 according to the deviation degree, wherein the score is X2 (1-P2);
3) Parameter 3: calculating a deviation coefficient P3 according to the deviation degree, wherein the score is X3 (1-P3);
4) Parameter 4: calculating a deviation coefficient P4 according to the deviation degree, wherein the score is X4 (1-P4);
……
m) parameter m, score Xm, calculating deviation coefficient Pm according to deviation degree, score Xm (1-Pm);
parameters 1 to m are parameters of equipment fault evaluation, and the score of fault dimension Where m is the m dimension, xm is the score of the m dimension, and Pm is the deviation value of the m dimension.
4. The method for data-driven based device health assessment according to claim 1, wherein in step 1, the establishment of the device maintenance assessment model comprises:
1) Parameter 1: calculating a deviation coefficient R1 according to the deviation degree, wherein the score is C1 (1-R1);
2) Parameter 2: calculating a deviation coefficient R2 according to the deviation degree, wherein the score is C2 (1-R2);
3) Parameter 3: calculating a deviation coefficient R3 according to the deviation degree, wherein the score is C3 (1-R3);
4) Parameter 4: calculating a deviation coefficient R4 according to the deviation degree, wherein the score is C4 (1-R4);
……
i) And calculating a deviation coefficient Ri according to the deviation degree, wherein the score is Ci (1-Ri) according to the parameter i.
Parameters 1 to i are equipment maintenance evaluation parameters, and the maintenance dimension scoreWhere i is the ith dimension, ci is the score of the ith dimension, and Ri is the bias value of the ith dimension.
5. The method for data-driven based device health assessment according to claim 1, wherein in step 1, the establishment of the device energy efficiency assessment model comprises:
1) Parameter 1: score F1, the deviation coefficient S1 was calculated from the degree of deviation, and score F1 (1-S1).
2) Parameter 2: score F2, the deviation coefficient S2 was calculated from the degree of deviation, and score F2 x (1-S2).
3) Parameter 3: score F3, the deviation coefficient S3 was calculated from the degree of deviation, and score F3 (1-S3).
4) Parameter 4: score F4, the deviation coefficient S4 was calculated from the degree of deviation, and score F4 (1-S4).
……
j) And calculating a deviation coefficient Sj according to the deviation degree, wherein the score is Fj (1-Sj) according to the parameter j.
Parameters 1 to j are the device energy efficiency evaluation parameters, and the score of the energy efficiency dimension Where j is the j dimension, fj is the j dimension score, and Sj is the j dimension bias value.
6. The data-driven based device health assessment method of claim 1, wherein in step 4, the device health assessment model comprises:
degree of health Z of equipment he ＝H op *(1-K op )+X fm *(1-L fm )+C me *(1-T me )+F ec *(1-Y ec ) Wherein H is op For the score of the run-time dimension, K op The deviation value of the operation condition dimension is customized according to the expert experience knowledge base; x is X fm For the score of the fault dimension, L fm The deviation value of fault dimension is customized according to expert experience knowledge base; c (C) me To maintain the score of maintenance dimension, T me The maintenance dimension deviation value is defined according to an expert experience knowledge base; f (F) ec For the score of energy efficiency dimension, Y ec The deviation value of the energy efficiency dimension is customized according to the expert experience knowledge base.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311104138.1A CN117114446A (en) | 2023-08-29 | 2023-08-29 | Device health assessment method based on data driving |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311104138.1A CN117114446A (en) | 2023-08-29 | 2023-08-29 | Device health assessment method based on data driving |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117114446A true CN117114446A (en) | 2023-11-24 |
Family
ID=88801790
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311104138.1A Pending CN117114446A (en) | 2023-08-29 | 2023-08-29 | Device health assessment method based on data driving |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117114446A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117668498A (en) * | 2024-01-31 | 2024-03-08 | 和尘自仪(嘉兴)科技有限公司 | Pump health assessment method based on reliability distribution and anomaly detection |
| CN117668498B (en) * | 2024-01-31 | 2024-04-26 | 和尘自仪(嘉兴)科技有限公司 | Pump health assessment method based on reliability distribution and anomaly detection |
-
2023
- 2023-08-29 CN CN202311104138.1A patent/CN117114446A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117668498A (en) * | 2024-01-31 | 2024-03-08 | 和尘自仪(嘉兴)科技有限公司 | Pump health assessment method based on reliability distribution and anomaly detection |
| CN117668498B (en) * | 2024-01-31 | 2024-04-26 | 和尘自仪(嘉兴)科技有限公司 | Pump health assessment method based on reliability distribution and anomaly detection |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20190287005A1 (en) | Diagnosing and predicting electrical pump operation | |
| CN116720752A (en) | Assembled building quality information supervision system based on big data | |
| US20230123527A1 (en) | Distributed client server system for generating predictive machine learning models | |
| CN112446509A (en) | Complex electronic equipment prediction maintenance method | |
| CN117151345A (en) | Enterprise management intelligent decision platform based on AI technology | |
| Sénéchal | Maintenance decision support for sustainable performance: problems and research directions at the crossroads of health management and eco-design | |
| CN117393076B (en) | Intelligent monitoring method and system for heat-resistant epoxy resin production process | |
| CN117196200A (en) | Industrial factory asset management system | |
| CN117114446A (en) | Device health assessment method based on data driving | |
| CN116483042A (en) | Digital lean diagnosis method for lean production control platform | |
| Berrabah et al. | Essential and new maintenance KPIs explained | |
| Sharma et al. | Implementing Big Data Analytics and Machine Learning for Predictive Maintenance in Manufacturing Facilities in South Korea | |
| US20240095853A1 (en) | System and method for supplier risk prediction and interactive risk mitigation in automotive manufacturing | |
| US20220206471A1 (en) | Systems and methods for providing operator variation analysis for transient operation of continuous or batch wise continuous processes | |
| Nwadinobi et al. | Development of Simulation for Condition Monitoring and Evaluation of Manufacturing Systems | |
| JP2023547729A (en) | Systems and methods for scalable automatic maintenance optimization | |
| CN117829554A (en) | Intelligent perception finished product restoration decision support system | |
| Akpan et al. | Risk Management in Nigeria Electric Power Industry | |
| Borger | Maintenance Management Optimization by Asset Categorisation | |
| FARM | 17. WIND FARM DATA COLLECTION AND RELIABILITY ASSESSMENT FOR O&M OPTIMIZATION | |
| Tatarczak et al. | A university-industry collaboration framework and its implementation for the creation of an innovative product | |
| Thelosen | Road to predictive maintenance for capital goods | |
| Al-Najjar et al. | Dynamic and cost-effective maintenance decisions | |
| Hu et al. | Condition assessment model for maintenance of vehicles fleet based on knowledge generation | |
| Akintunde | Process optimization through integration of control and asset management |
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 |