Disclosure of Invention
The invention aims to provide a power transformation equipment health management method, a power transformation equipment health management system, terminal equipment and a readable storage medium, so as to solve one or more technical problems. The invention can realize real-time evaluation of the power transformation equipment, diagnosis is carried out by combining a machine and an expert algorithm, corresponding auxiliary decisions are given according to an evaluation diagnosis result, and health management can be carried out on the power transformation equipment accurately in real time.
In order to achieve the above purpose, the invention adopts the following technical scheme:
the invention discloses a power transformation equipment health management method, which comprises the following steps:
acquiring real-time data values of various parameters in the operation process of the power transformation equipment;
real-time evaluation is carried out on the real-time data values of the parameters based on a preset evaluation guideline, and an evaluation result is obtained; performing diagnosis based on the evaluation result to obtain an analysis result;
based on the analysis result, auxiliary decision advice is obtained by combining equipment life data, so that health management of the power transformation equipment is realized.
The invention further improves that the step of carrying out real-time evaluation on the real-time data values of the parameters based on a preset evaluation guideline to obtain an evaluation result specifically comprises the following steps:
filtering the real-time data values of the parameters to obtain filtered data; the filtering process is to filter out real-time data values with null data and zero data;
comparing the real-time data value of each parameter after filtering with a preset threshold range of each parameter, and judging whether the real-time data value of each parameter is in the threshold range of the parameter or not; when the device is within the threshold range, the device state is considered to be normal; and when the threshold value range is exceeded, the equipment state is considered to be abnormal, and the equipment state is obtained according to a preset evaluation guideline.
The invention further improves that when the threshold value range is exceeded, the equipment state is considered to be abnormal, and the step of obtaining the equipment state according to the preset evaluation guide comprises the following steps:
when the real-time data value of a certain parameter exceeds the threshold range, the state quantity of a deduction item in a corresponding system 2016 type evaluation guide is searched for in a fuzzy manner through the name of the parameter to be matched, wherein the deduction item comprises a deduction component, a judgment basis and a single deduction value; the state quantity of the deduction item in the corresponding PMS 2008 edition evaluation guide rule is searched for in a fuzzy manner through the name of the monitoring parameter, and the deduction item comprises a deduction component, a judgment basis and a single deduction value;
the system 2016 version evaluation guideline and the PMS 2008 version guideline are matched twice and then summarized; if the same component is repeatedly buckled, taking the evaluation guideline of the system 2016 version as the reference, and removing the buckling items of the component corresponding to the PMS 2008 version; if the deduction is not repeated, the deduction value of the component is normally summarized;
and according to the single deduction range and the total deduction range of the summarized components, referring to a preset general guide rule evaluation standard, obtaining the state of each component, and taking the most serious component state in the evaluation as the final state of the evaluation.
The invention further improves that the step of diagnosing based on the evaluation result and obtaining the analysis result specifically comprises the following steps:
and when the equipment state is abnormal, obtaining an analysis result according to fusion of an expert algorithm and a machine algorithm.
The invention further improves that when the equipment state is abnormal, the step of fusing according to the expert algorithm and the machine algorithm to obtain the analysis result specifically comprises the following steps:
when the equipment state is abnormal, selecting different groups of data after monitoring the operation data;
performing fusion analysis of expert algorithm and machine algorithm on the selected multiple groups of different data; wherein the step of fusion analysis comprises:
according to system case library statistics, carrying out statistics on diagnostic algorithms and diagnostic results of different fault cases, wherein N is the number of fault types, M is the number of algorithms, a weight matrix of N x M is formed, and weight factors of the algorithms are determined;
calculating an algorithm state priority coefficient:in the formula, flag i Representing the algorithm screening tag, ω ij A weight factor representing the algorithm for different diagnostic states; and obtaining a final analysis result based on the obtained algorithm state priority coefficient.
The invention further improves that the step of obtaining the auxiliary decision suggestion by combining the equipment life data based on the analysis result specifically comprises the following steps of:
when the analysis result is in a normal state, no auxiliary decision suggestion exists;
and when the analysis result is in an abnormal state, searching a case library, and outputting auxiliary decision suggestions in the corresponding case library.
The invention further improves that when the analysis result is abnormal, the step of searching the case base and outputting the auxiliary decision advice in the corresponding case base specifically comprises the following steps:
when the analysis result is in an abnormal state, case matching is carried out in a system case library, and a processing method and a rectifying measure of historical cases with data similarity higher than 83% are pushed to realize health management of the power transformation equipment.
The invention relates to a power transformation equipment health management system, which comprises:
the data acquisition module is used for acquiring real-time data values of various parameters in the running process of the power transformation equipment;
the data analysis module is used for carrying out real-time evaluation on the real-time data values of the parameters based on a preset evaluation guideline to obtain an evaluation result; performing diagnosis based on the evaluation result to obtain an analysis result;
and the auxiliary decision making module is used for obtaining auxiliary decision making suggestions by combining the equipment life data according to the analysis result, so as to realize the health management of the power transformation equipment.
An electronic apparatus of the present invention includes: a processor; a memory for storing computer program instructions; when the computer program instructions are loaded and run by the processor, the processor executes the power transformation equipment health management method according to any one of the above aspects of the invention.
A computer readable storage medium of the present invention stores computer program instructions that, when loaded and executed by a processor, perform the power transformation device health management method of any one of the above-described aspects of the present invention.
Compared with the prior art, the invention has the following beneficial effects:
the invention can realize real-time automatic evaluation of the power transformation equipment, equipment with degradation trend can be found in time by means of real-time evaluation and diagnosis, and the diagnosis result can provide investigation of fault reasons and advice of auxiliary decisions in time, so that the power transformation equipment can be managed healthily and accurately in real time.
According to the invention, real-time data is monitored on line by accessing the power transformation equipment, and the real-time and accuracy of evaluation are ensured according to the preset evaluation guideline. The algorithm is good in performance on certain fault types and complements each other, and the invention provides a method for carrying out fusion diagnosis on a machine algorithm and an expert algorithm and pushing out overhaul suggestions with higher accuracy.
Based on the embodiment of the invention, the real-time automatic evaluation is performed by accessing the on-line monitoring data, the reliability of the equipment evaluation is improved by 20% compared with the PMS evaluation result, the accuracy of the diagnosis algorithm is up to 93.6% by fusing expert experience and a machine algorithm, the accuracy of diagnosis is obviously improved, and the auxiliary decision suggestion is more accurate.
Detailed Description
In order to make the purposes, technical effects and technical solutions of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention are clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention; it will be apparent that the described embodiments are some of the embodiments of the present invention. Other embodiments, which may be made by those of ordinary skill in the art based on the disclosed embodiments without undue burden, are within the scope of the present invention.
Referring to fig. 1, the method for managing the health of the power transformation equipment according to the embodiment of the invention is a power transformation equipment health management method embedded in an intelligent analysis management and control system based on power grid operation detection, and specifically comprises the following steps:
step 1, data acquisition: collecting real-time data values of parameters in the running process of equipment; the method comprises the steps of exemplarily, during the running process of the equipment, collecting online monitoring data of the equipment in real time, and recording real-time data values of various parameters;
step 2, data analysis: real-time evaluation and diagnosis are carried out on the real-time data values of all parameters in the running state of the equipment;
step 3, auxiliary decision: and different auxiliary decision suggestions are given out according to different analysis results, so that the health management of the power transformation equipment is realized.
In the method of the embodiment of the invention, the generation of the operation data of the power transformation equipment automatically initiates the evaluation of the equipment state in real time, and the equipment abnormality type and the treatment measure are given by combining the equipment life-span data.
In an embodiment of the present invention, the data analysis includes the steps of:
comparing the acquired real-time data, and filtering the data if the data is empty and zero;
comparing the filtered data with each parameter threshold;
judging whether each item of real-time data is in a corresponding parameter threshold range or not;
when the data value does not exceed the preset parameter threshold, the equipment is in a normal state;
when the data value exceeds the preset parameter threshold, the state of the equipment (attention, abnormality and severity) is obtained according to the evaluation guideline.
In the embodiment of the invention, when the equipment state is abnormal, diagnosis is performed, and analysis results (normal, high-temperature overheat, low-energy discharge, partial discharge, high-energy discharge, low-energy discharge and overheat, medium-temperature overheat, low-temperature overheat and high-energy discharge and overheat) are obtained by fusing an expert algorithm and a machine algorithm.
In an embodiment of the present invention, the auxiliary decision comprises the following steps:
when the final analysis processing result is normal, no auxiliary decision suggestion exists;
and when the final analysis and processing result is in an abnormal state, searching a case library, and outputting auxiliary decision suggestions in the corresponding case library.
In the embodiment of the invention, the power transformation equipment data is collected to the intelligent analysis and control of the power grid operation inspection, the system monitors the threshold value of the collected data by utilizing a real-time monitoring mechanism, and the equipment state evaluation is carried out by utilizing a state evaluation guide rule, and meanwhile, the reasons and treatment measures of abnormality of the defect diagnosis analysis equipment are provided; the case library is used as a training set in the treatment measures, and the cases matched with the real-time data are trained through the cosine similarity algorithm, so that the state of the power transformation equipment can be accurately and timely monitored, and the occurrence of equipment accidents and the possibility of degradation are prevented.
Referring to fig. 2 and 3, for the data analysis process of the monitored device, the following specific steps are given:
when the operation data is monitored, carrying out null judgment and zero judgment on each item of monitoring data, and if the data is null or zero, carrying out filtering treatment on the piece of data which is invalid;
comparing the filtered monitoring data with the threshold values of all parameters;
outputting a normal prompt of the equipment when the monitored real-time data values are all in the threshold range;
when the monitored real-time data exceeds a threshold value, the state quantity of the deduction item in the corresponding system 2016 type evaluation guide is firstly subjected to fuzzy searching through the name of the monitored parameter, the deduction item comprises a deduction component, a judgment basis and a single deduction value. And by analogy, performing matching processing on each piece of online data to obtain a single deduction value of a corresponding component of each piece of data. And then the state quantity of the deduction item in the corresponding PMS 2008 edition evaluation guide rule is subjected to fuzzy searching through the name of the monitoring parameter so as to match, wherein the deduction item comprises a deduction component, a judgment basis and a single deduction value.
And by analogy, carrying out matching processing on each piece of monitoring data to obtain a single deduction value of a corresponding part of each piece of data.
And (5) carrying out twice matching through the system 2016 edition evaluation guideline and the PMS 2008 edition guideline, and then summarizing. If the two matches have repeated deductions for the same component, taking the evaluation guideline of the system 2016 version as the reference, and removing the deduction item of the component corresponding to the PMS 2008 version. If duplicate points do not occur, the point values for the components are normally summarized.
The states (normal, noted, abnormal, severe) of the individual components are obtained by referring to the general guideline evaluation criteria (in the case of transformers, the states corresponding to the different deduction ranges are shown in table 1) based on the single deduction and the total deduction ranges of the summarized components. The most severe part state in this evaluation was taken as the final state (normal, noted, abnormal, severe) for this evaluation.
TABLE 1 states corresponding to different deduction ranges
In the embodiment of the invention, when the equipment state is abnormal, different groups of data are manually selected after the operation data are monitored; performing fusion analysis of expert algorithm and machine algorithm on the selected multiple groups of different data;
according to statistics of a system case library, statistics is carried out on diagnostic algorithms and diagnostic results of different fault cases, N is the number of fault types, M is the number of algorithms, a weight matrix of N x M is formed, and weight factors of the algorithms are determined as shown in table 2:
TABLE 2 weight factors for the respective algorithms
Sequence number
|
Algorithm
|
Low energy discharge
|
Low energy discharge and overheat
|
High energy discharge
|
High-energy discharge and overdischarge
|
High temperature superheating
|
Middle and low temperature superheating
|
1
|
Random forest
|
0.2111
|
0.1795
|
0.2500
|
0.2500
|
0.2704
|
0.2642
|
2
|
GBDT
|
0.2000
|
0.1795
|
0.2500
|
0.5000
|
0.2575
|
0.2642
|
3
|
Decision tree
|
0.2222
|
0.1795
|
0.2500
|
0.2500
|
0
|
0.2830
|
4
|
SVM
|
0.1778
|
0
|
0
|
0
|
0
|
0
|
5
|
Three ratio of
|
0
|
0.2051
|
0.2500
|
0
|
0.2189
|
0.1887
|
6
|
Triangle for sanitation
|
0.1889
|
0.2308
|
0
|
0
|
0
|
0
|
7
|
David pentagon
|
0
|
0.2051
|
0
|
0
|
0.2531
|
0 |
Calculating an algorithm state priority coefficient:in the formula, flag i Representing the algorithm screening tag, ω ij Representing the weighting factors of the algorithm for different diagnostic states.
The monitored operation data is taken as oil chromatographic data for illustration, and 7 diagnostic results corresponding to 7 algorithms are respectively: high temperature superheating, medium and low temperature superheating, and high temperature superheating. There are 4 algorithms to diagnose "high temperature superheat" and 2 algorithms to diagnose "medium and low temperature superheat", then the following conclusions can be drawn for this piece of oil chromatography data:
state priority coefficient of "hyperthermia" =0.2704×1+0.2575×1+0×1+0×1= 0.5279
State priority coefficient of "medium low temperature superheat" =0.2830×1+0.1887×1+0×1= 0.4717
The state priority coefficient of "hyperthermia" was found to be greater than the state priority coefficient of "middle hyperthermia", so the final diagnostic conclusion for this piece of oil chromatography data was "hyperthermia".
And different overhaul suggestions are given according to the diagnosis result of the monitoring data, and if the diagnosis result is normal, no auxiliary decision is made.
If the diagnosis is abnormal (high-temperature overheat, low-energy discharge, partial discharge, high-energy discharge, low-energy discharge and overheat, medium-temperature overheat, low-temperature overheat, high-energy discharge and overheat), performing case matching in a system case library, and pushing a processing method and a rectifying measure of historical cases with data similarity higher than 83%, thereby realizing the health management of the power transformation equipment.
In summary, the traditional intelligent evaluation only accesses the offline and routine data of the equipment, performs equipment evaluation according to the PMS evaluation guideline, does not access the online data to perform real-time automatic evaluation, and lacks the real-time performance and accuracy of the evaluation. The traditional fault diagnosis is only accessed with an expert algorithm, the expert algorithm has not very high accuracy in various fault diagnosis, and the expert algorithm is not supplemented by practical application cases and applied by machine algorithms, so that maintenance suggestions have no powerful data support. The embodiment of the invention provides a power transformation equipment health management method based on embedding of an intelligent analysis management control system for power grid operation detection, and relates to the field of power transformation equipment. The invention accesses the online monitoring real-time data of the power transformation equipment, automatically evaluates according to the new version of evaluation guideline, and ensures the real-time performance and accuracy of evaluation. The algorithm is good in certain fault types and mutually complements, so that fusion diagnosis is carried out on a machine algorithm and an expert algorithm, and overhaul suggestions with high accuracy are provided. Compared with the prior art, the adopted technical scheme has the beneficial effects that: the reliability of equipment evaluation is improved by 20% compared with PMS evaluation by accessing on-line monitoring data, and the accuracy of a diagnosis algorithm is up to 93.6% by fusing an expert algorithm (triangle of David, five sides of David and three ratio) and a machine algorithm (random forest, gradient lifting tree, K nearest neighbor and decision tree), so that the accuracy of diagnosis is obviously improved, and auxiliary decision advice is more accurate.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the above embodiments, one skilled in the art may make modifications and equivalents to the specific embodiments of the present invention, and any modifications and equivalents not departing from the spirit and scope of the present invention are within the scope of the claims of the present invention.