CN115952914A - Big data-based electric power metering operation and maintenance work judgment planning method - Google Patents
Big data-based electric power metering operation and maintenance work judgment planning method Download PDFInfo
- Publication number
- CN115952914A CN115952914A CN202310021673.4A CN202310021673A CN115952914A CN 115952914 A CN115952914 A CN 115952914A CN 202310021673 A CN202310021673 A CN 202310021673A CN 115952914 A CN115952914 A CN 115952914A
- Authority
- CN
- China
- Prior art keywords
- maintenance
- maintenance work
- model
- big data
- power metering
- 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
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a big data-based power metering operation and maintenance work judgment planning method, which comprises the following steps of: constructing a multi-dimensional feature library by using feature engineering; acquiring corresponding actual operation and maintenance data based on the multi-dimensional feature library, and acquiring a pre-constructed evaluation score model to obtain each dimension index score; and predicting to obtain the next working plan based on the index scores of the dimensions by using a pre-constructed operation and maintenance planning model, wherein the operation and maintenance planning model is obtained based on a random forest algorithm. Compared with the prior art, the method can provide reliable basis for operation and maintenance work, and has high feasibility.
Description
Technical Field
The invention relates to evaluation of power measurement operation and maintenance work, in particular to a power measurement operation and maintenance work judgment planning method based on big data.
Background
In the trend of new and fiercely developed technology application such as big data, artificial intelligence and the like, the existing resources are fully integrated, and the unified big data application capability is constructed, so that the method is a necessary way for improving lean management and high-quality service level.
The problems of long overload running time of the system, serious equipment abrasion and aging of key components, insufficient timeliness and accuracy of operation and maintenance and the like exist in the power system along with the increase of the service life and the verification requirements of each piece of automation equipment, and the operation and maintenance level of each dimension is difficult to uniformly sum up. The operation and maintenance work level can not be effectively evaluated in the prior art.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a power metering operation and maintenance work judgment planning method based on big data.
The purpose of the invention can be realized by the following technical scheme:
a big data-based power metering operation and maintenance work judgment planning method comprises the following steps:
constructing a multi-dimensional feature library by using feature engineering;
acquiring corresponding actual operation and maintenance data based on the multi-dimensional feature library, and acquiring a pre-constructed evaluation score model to obtain each dimension index score;
and predicting to obtain the next working plan based on the index scores of the dimensions by utilizing a pre-constructed operation and maintenance planning model, wherein the operation and maintenance planning model is obtained based on a random forest algorithm.
Furthermore, the multi-dimensional feature library is a feature library covering four dimensions of equipment operation management, operation and maintenance work management, economic management and personnel management.
Further, the construction of the multi-dimensional feature library comprises:
acquiring all characteristic indexes related to equipment operation management, operation and maintenance work management, economic management and personnel management;
and screening the characteristic indexes with higher correlation to construct the multi-dimensional characteristic library based on the influence correlation of each characteristic index and operation and maintenance work.
Further, the evaluation score model is a continuous model.
Further, the process of obtaining the operation and maintenance planning model based on the random forest algorithm includes:
(a) Reading a characteristic data table and a corresponding comprehensive suggestion list from a database, and associating the two tables to obtain original data, wherein the characteristic data table comprises a plurality of historical dimensional index scores, and the comprehensive suggestion list comprises comprehensive suggestion information corresponding to each dimensional index score;
(b) Randomly extracting n training samples with the same volume as the original data samples from the original data in a release manner by utilizing a resampling method to construct training data;
(c) Extracting K characteristic variables from the K characteristic variables of the training sample by using a random sampling method, wherein K is less than K, and selecting the optimal segmentation characteristic from the K characteristic variables to construct a decision tree classifier;
(d) And (c) repeating the step (b) and the step (c) for multiple times to construct a plurality of decision tree classifiers, wherein all the decision tree classifiers form a random forest, namely the operation and maintenance planning model.
Further, the model parameters of the evaluation score model are adjusted based on the field operation and maintenance condition.
Further, the model parameters include weights and reference values.
Further, the method further comprises: and obtaining a final prediction result by combining a majority voting method according to the prediction result of the operation and maintenance planning model, and obtaining the next work plan based on the final prediction result.
Further, the method further includes updating the operation and maintenance planning model, and specifically includes:
acquiring modified comprehensive suggestion information and corresponding dimensional index scores as a basic sample;
randomly generating a plurality of extension samples for each dimensionality index score in a fluctuation interval, and combining the basic samples and the extension samples to form extension sample data;
and re-performing a random forest algorithm based on the expanded sample data and the original data to obtain an updated operation and maintenance planning model.
Further, the method further comprises: and visualizing the index scores of all the dimensions and the next work plan.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention constructs the multidimensional feature library, can effectively evaluate and judge the operation and maintenance work of the electric power measurement, can predict the next working rule according to the evaluation result, can provide reliable basis for the operation and maintenance work, and has high feasibility.
2. The operation and maintenance planning model is constructed by adopting a random forest algorithm based on big data, and the accuracy of the prediction result is high.
3. The method and the device can obtain the final prediction result by voting on the prediction result of the random forest, and further improve the prediction reliability of the next working plan.
4. The method has the updating function of the evaluation score model and the operation and maintenance planning model, can update related parameters and models according to the actual operation and maintenance condition, and improves the accuracy of judgment and prediction.
5. The operation and maintenance planning model adopts a continuous model, so that the scores of all indexes under different values are ensured to be continuous numerical values, the defects that the value quantity of the segmented model is limited, the segmentation rule has no scientific basis and the like are overcome, and the construction of the evaluation model is closer to the actual operation and maintenance work.
6. By the application of the invention, the existing resources can be fully integrated, the unified big data application capability is constructed, and the deep mining of the analysis scene is realized, so that the management capability is improved, and the marketing high-quality service level is improved.
Drawings
FIG. 1 is a schematic flow diagram of the present invention;
fig. 2 is a schematic diagram of a visualization interface in an embodiment of the invention.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments. The present embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a specific operation process are given, but the scope of the present invention is not limited to the following embodiments.
As shown in fig. 1, the present embodiment provides a method for determining and planning operation and maintenance work of power metering based on big data, which includes the following steps:
s1, constructing a multi-dimensional feature library by using feature engineering;
s2, acquiring corresponding actual operation and maintenance data based on the multi-dimensional feature library, and acquiring a pre-constructed evaluation score model to obtain each dimension index score;
and S3, predicting and obtaining next-step work planning by utilizing a pre-constructed operation and maintenance planning model based on the index scores of the dimensions, wherein the operation and maintenance planning model is obtained based on a random forest algorithm.
The multi-dimensional feature library is a feature library covering four dimensions of equipment operation management, operation and maintenance work management, economic management and personnel management. The construction of the multi-dimensional feature library comprises the following steps:
acquiring all characteristic indexes related to equipment operation management, operation and maintenance work management, economic management and personnel management, wherein the equipment operation management comprises two major categories of a warehousing operation module and a verification line operation module, the warehousing operation module comprises warehousing inventory total quantity, inventory utilization rate and the like, and the verification line operation module comprises verification quantity configuration, verification efficiency configuration and the like; the operation and maintenance work management comprises failure times, failure abnormal defect elimination rate and the like; the economic management comprises the consumption quantity of spare parts, the configuration of operation and maintenance cost and the like; the personnel management comprises daily work attendance condition configuration, work order operation and maintenance configuration and the like;
and screening the characteristic indexes with higher correlation to construct a multi-dimensional characteristic library based on the influence correlation of each characteristic index and operation and maintenance work, wherein the multi-dimensional characteristic library is a characteristic engineering index library suitable for business data and application conditions, and the accuracy and the feasibility of a characteristic engineering index system are improved.
In the method, the corresponding relation between the values and the scores of the evaluation score model representation indexes is a continuous model, the scores of the indexes under different values are ensured to be continuous numerical values, the defects that the value number of the segmented model is limited, the segmentation rule has no scientific basis and the like are overcome, and the construction of the operation and maintenance KPI index system is closer to the actual operation and maintenance work. In this embodiment, three models are designed for the continuous model, which are respectively: a spike distribution model, a partial monotonically increasing model, and a partial monotonically decreasing model.
The evaluation score model calculates recorded data from the system operation and maintenance process according to a data source, and the recorded data comprises a plurality of data related to a multidimensional feature library, including equipment type, equipment quantity, equipment state, warehouse writing time, verification overall conclusion, failure times, single processing work order quantity, inspection plan data, maintenance plan data, spare part consumption quantity, annual operation and maintenance cost overall amount, monthly production cost overall amount, monthly energy consumption overall amount, daily work attendance condition configuration, invested number and the like.
In the method, the process of obtaining the operation and maintenance planning model based on the random forest algorithm comprises the following steps:
(a) Reading a characteristic data table and a corresponding comprehensive suggestion detail table from a database, and associating the two tables to obtain original data, wherein the characteristic data table comprises a plurality of historical index scores of each dimension, the comprehensive suggestion detail table comprises comprehensive suggestion information corresponding to the index score of each dimension, and the comprehensive suggestion information is obtained by an operation and maintenance expert by combining operation and maintenance service experience and field operation and maintenance conditions;
(b) Randomly extracting n training samples with the same volume as the original data samples from the original data in a repeated sampling way to construct training data;
(c) Extracting K characteristic variables from the K characteristic variables of the training sample by using a random sampling method, wherein K is less than K, and selecting the optimal segmentation characteristic from the K characteristic variables to construct a decision tree classifier;
(d) And (c) repeating the step (b) and the step (c) for multiple times to construct a plurality of decision tree classifiers, wherein all the decision tree classifiers form a random forest, namely an operation and maintenance planning model, and the operation and maintenance planning model can obtain the most suitable comprehensive suggestion according to the actual score matching.
In a preferred embodiment, after the prediction result is obtained by the random forest, a final prediction result is obtained by combining a majority voting method, and the next working plan is obtained based on the final prediction result. Assume that a random forest contains T base decision trees { h } 1 ,h 2 ,…,h t Is set as { c) by category label 1 ,c 2 ,…,c n H is I The prediction output at sample x is represented as an N-dimensional vector
Wherein->
Is h I In category label c J An output of (c). The rule of the majority voting method is that the predicted result is the mark with the most votes, if a plurality of marks obtain the highest votes at the same time, one mark is randomly selected from the marks, and the calculation mode is as follows:
in a preferred embodiment, model optimization, including updating and optimizing an evaluation score model and an operation and maintenance planning model, needs to be implemented according to newly added data and suggestions of field use conditions.
(1) Weight management configuration
In general, the model for each index is designed with three parameters: the minimum limit, the optimum value and the maximum limit, and the values of the three parameters determine the distribution of the index scores. Operation and maintenance personnel can properly adjust the value of the model parameter according to the field operation and maintenance condition, and the scoring reliability of the index system is improved.
(2) Automatically updating parameters
In the process of defining the index, the calculation formula of the partial index has the concept of a reference value. The initial setting of the reference value is determined by a service expert, and in the subsequent system operation process, the setting of the reference value can be automatically calculated and updated according to the real situation of on-site operation and maintenance, so that the reference value of each index is closer to the on-site situation of the service.
Taking the calibration efficiency configuration index of the equipment operation management calibration line operation module as an example, the calculation formula of the index is defined as: verification efficiency configuration = (verification number/verification time length)/efficiency benchmark value. The efficiency reference value is set by an expert as an initial value, the model sets the reference value of the timing task to be updated every half year, and the updating rule is as follows: assay duration over total assay/corresponding time from one to two years.
In addition to the above-mentioned reference values defining the indexes being automatically updated, the model design parameters of some of the indexes are also automatically updated.
(3) Integrated proposal optimization
With the use of the operation and maintenance planning model, the field operation and maintenance conditions may change, or better comprehensive suggestions are given to a certain set of index scores, and then the optimized operation and maintenance planning model can be updated.
The optimization process of the model is as follows:
(a) Acquiring data: the system stores the modification opinions provided by the operation and maintenance personnel and the corresponding index scores (namely sample data) into the database, and the model calls the specified data from the database.
(b) Expanding the data volume: only one sample data is needed for modifying the opinions, and the sample size corresponding to the opinions in the algorithm learning process is too small, so that algorithm optimization is not facilitated.
(c) Optimizing the model: and re-running the random forest algorithm based on the expanded data and the past data source. Training data is built again according to the latest data source, and a plurality of decision tree classifiers are built to form a random forest.
(d) And (4) saving a new model: and constructing and storing the optimized model based on the steps. And after the new index score is generated, the optimized latest model prediction comprehensive suggestion is called.
In a preferred embodiment, the dimension index scores and the next work plan may be visualized.
The method can be applied to a single evaluation management scene with four dimensions of equipment operation management, operation and maintenance work management, economic management and personnel management, and can also be applied to a comprehensive evaluation management scene.
(a) Equipment operation management scene: the equipment state and the operation efficiency are taken as research objects, the equipment operation condition is respectively evaluated from the storage module and the calibration line module, the equipment operation management is evaluated by utilizing a plurality of dimensions such as inventory utilization rate, calibration efficiency and the like, calibration tasks and inventory are reasonably arranged, and if the equipment fails, a fault point is timely searched and solved.
(b) Operation and maintenance work management: the operation and maintenance requirements and the operation and maintenance level are used as research objects, operation and maintenance work management is evaluated from multiple dimensions such as failure times, maintenance work completion rate and the like, supervision and management of personnel are enhanced under the condition of non-enthusiasm of operation and maintenance, and corresponding assessment is proposed to improve the operation and maintenance work level.
(c) And (3) economic management: materials, energy and economic cost are taken as research objects, the current economic configuration is evaluated, whether the economic cost is reasonably distributed is analyzed, and if the cost is increased, reasons are found out in time and corrected.
(d) Personnel management: the operation and maintenance personnel are used as research objects, the operation and maintenance enthusiasm and the personnel configuration level of the personnel are evaluated from multiple angles of attendance rate and work order completion rate, and corresponding policies are proposed to strengthen management.
(e) And (4) comprehensive management: the comprehensive score of the index system, the ring ratio (compared with the previous period) and the change value of the same ratio (compared with the same period of the last year) are displayed through the numerical value, so that the operation and maintenance personnel can know the comprehensive current situation and the change trend of the operation and maintenance work, and the operation and maintenance personnel are shown in fig. 2; secondly, the pie chart is used for respectively showing the proportion of each dimension coefficient and the proportion distribution of the comprehensive scores, the bar chart is used for showing the scores of each dimension, and the logical relationship among the three is as follows: composite score = SUM (dimension coefficient dimension score), the origin of the composite score is visually shown by a graph; and finally, according to the scores of the regions and the expression of the text description data, providing operation and maintenance suggestions for the current situation by means of a machine learning algorithm, indicating the next work plan, and providing function points for modifying the suggestions and perfecting the model.
The above method, if implemented in the form of software functional units and sold or used as a stand-alone product, can be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The foregoing detailed description of the preferred embodiments of the invention has been presented. It should be understood that numerous modifications and variations could be devised by those skilled in the art in light of the present teachings without departing from the inventive concepts. Therefore, the technical solutions available to those skilled in the art through logic analysis, reasoning and limited experiments based on the prior art according to the concept of the present invention should be within the scope of protection defined by the claims.
Claims (10)
1. A big data-based power metering operation and maintenance work judgment planning method is characterized by comprising the following steps:
constructing a multi-dimensional feature library by using feature engineering;
acquiring corresponding actual operation and maintenance data based on the multi-dimensional feature library, and acquiring a pre-constructed evaluation score model to obtain each dimension index score;
and predicting to obtain the next working plan based on the index scores of the dimensions by using a pre-constructed operation and maintenance planning model, wherein the operation and maintenance planning model is obtained based on a random forest algorithm.
2. The big data based electric power metering operation and maintenance work judgment planning method according to claim 1, wherein the multi-dimensional feature library is a feature library covering four dimensions of equipment operation management, operation and maintenance work management, economic management and personnel management.
3. The big data based power metering operation and maintenance work judgment planning method according to claim 1, wherein the construction of the multidimensional feature library comprises:
acquiring all characteristic indexes related to equipment operation management, operation and maintenance work management, economic management and personnel management;
and screening the characteristic indexes with higher correlation to construct the multi-dimensional characteristic library based on the influence correlation between each characteristic index and operation and maintenance work.
4. The big data based power metering operation and maintenance work judgment planning method according to claim 1, wherein the evaluation score model is a continuous model.
5. The big data based power metering operation and maintenance work judgment planning method according to claim 1, wherein the process of obtaining the operation and maintenance planning model based on the random forest algorithm comprises:
(a) Reading a characteristic data table and a corresponding comprehensive suggestion list from a database, and associating the two tables to obtain original data, wherein the characteristic data table comprises a plurality of historical dimensional index scores, and the comprehensive suggestion list comprises comprehensive suggestion information corresponding to each dimensional index score;
(b) Randomly extracting n training samples with the same volume as the original data samples from the original data in a release manner by utilizing a resampling method to construct training data;
(c) Extracting K characteristic variables from the K characteristic variables of the training sample by using a random sampling method, wherein K is less than K, and selecting the optimal segmentation characteristic from the K characteristic variables to construct a decision tree classifier;
(d) And (c) repeating the step (b) and the step (c) for multiple times to construct a plurality of decision tree classifiers, wherein all the decision tree classifiers form a random forest, namely the operation and maintenance planning model.
6. The big data based power metering operation and maintenance work judgment planning method according to claim 1, wherein model parameters of the evaluation score model are adjusted based on field operation and maintenance conditions.
7. The big data based power metering operation and maintenance work discriminant planning method of claim 1, wherein the model parameters comprise a weight and a reference value.
8. The big data based power metering operation and maintenance work judgment planning method according to claim 1, characterized by further comprising:
and according to the prediction result of the operation and maintenance planning model, combining a majority voting method to obtain a final prediction result, and obtaining the next working plan based on the final prediction result.
9. The big data-based power metering operation and maintenance work judgment planning method according to claim 5, further comprising updating the operation and maintenance planning model, specifically comprising:
acquiring modified comprehensive suggestion information and corresponding dimensional index scores as a basic sample;
randomly generating a plurality of extension samples for each dimensionality index score in a fluctuation interval, and combining the basic samples and the extension samples to form extension sample data;
and re-performing a random forest algorithm based on the extended sample data and the original data to obtain an updated operation and maintenance planning model.
10. The big data based power metering operation and maintenance work judgment planning method according to claim 1, characterized by further comprising:
and visualizing the index scores of all the dimensions and the next work plan.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310021673.4A CN115952914A (en) | 2023-01-07 | 2023-01-07 | Big data-based electric power metering operation and maintenance work judgment planning method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310021673.4A CN115952914A (en) | 2023-01-07 | 2023-01-07 | Big data-based electric power metering operation and maintenance work judgment planning method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115952914A true CN115952914A (en) | 2023-04-11 |
Family
ID=87290579
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310021673.4A Pending CN115952914A (en) | 2023-01-07 | 2023-01-07 | Big data-based electric power metering operation and maintenance work judgment planning method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115952914A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116611741A (en) * | 2023-07-14 | 2023-08-18 | 湖南省计量检测研究院 | Construction method and system of service quality index system based on wind power equipment |
-
2023
- 2023-01-07 CN CN202310021673.4A patent/CN115952914A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116611741A (en) * | 2023-07-14 | 2023-08-18 | 湖南省计量检测研究院 | Construction method and system of service quality index system based on wind power equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7627454B2 (en) | Method and system for predicting turbomachinery failure events employing genetic algorithm | |
| CN111178624A (en) | Method for predicting new product demand | |
| CN110826237B (en) | Wind power equipment reliability analysis method and device based on Bayesian belief network | |
| CN115564071A (en) | Method and system for generating data labels of power Internet of things equipment | |
| CN117557299B (en) | Marketing planning method and system based on computer assistance | |
| CN115794803B (en) | Engineering audit problem monitoring method and system based on big data AI technology | |
| CN115952914A (en) | Big data-based electric power metering operation and maintenance work judgment planning method | |
| CN116485020A (en) | Supply chain risk identification early warning method, system and medium based on big data | |
| CN113268370A (en) | Root cause alarm analysis method, system, equipment and storage medium | |
| CN114548494B (en) | Visual cost data prediction intelligent analysis system | |
| CN110310012B (en) | Data analysis method, device, equipment and computer readable storage medium | |
| CN116579804A (en) | Holiday commodity sales prediction method, holiday commodity sales prediction device and computer storage medium | |
| CN117787569B (en) | Intelligent auxiliary bid evaluation method and system | |
| CN115238573A (en) | Hydroelectric generating set performance degradation trend prediction method and system considering working condition parameters | |
| CN117291655B (en) | Consumer life cycle operation analysis method based on entity and network collaborative mapping | |
| CN110781206A (en) | Method for predicting whether electric energy meter in operation fails or not by learning meter-dismantling and returning failure characteristic rule | |
| CN117131425B (en) | Numerical control machine tool processing state monitoring method and system based on feedback data | |
| CN117591679A (en) | Intelligent analysis system and method for carbon footprint of building block type product based on knowledge graph | |
| CN115577890A (en) | Intelligent quality management method, electronic device and storage medium | |
| CN117522132A (en) | Vendor risk assessment system and application method | |
| WO2023174431A1 (en) | Kpi curve data processing method | |
| CN106779245A (en) | Civil aviaton's needing forecasting method and device based on event | |
| CN105678430A (en) | Improved user recommendation method based on neighbor project slope one algorithm | |
| CN114881154A (en) | Natural gas station fault detection method and system based on PCA and deep forest | |
| KR20230052010A (en) | Demand forecasting method using ai-based model selector algorithm |
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 |