CN117114449A - Visual analysis system and method for big electric power data - Google Patents
Visual analysis system and method for big electric power data Download PDFInfo
- Publication number
- CN117114449A CN117114449A CN202311175520.1A CN202311175520A CN117114449A CN 117114449 A CN117114449 A CN 117114449A CN 202311175520 A CN202311175520 A CN 202311175520A CN 117114449 A CN117114449 A CN 117114449A
- Authority
- CN
- China
- Prior art keywords
- data
- module
- power
- sub
- analysis
- 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
- 238000004458 analytical method Methods 0.000 title claims abstract description 101
- 230000000007 visual effect Effects 0.000 title claims abstract description 31
- 238000000034 method Methods 0.000 title claims description 60
- 238000011156 evaluation Methods 0.000 claims abstract description 57
- 238000012545 processing Methods 0.000 claims abstract description 52
- 238000013528 artificial neural network Methods 0.000 claims abstract description 29
- 238000004140 cleaning Methods 0.000 claims abstract description 20
- 230000004927 fusion Effects 0.000 claims abstract description 17
- 238000005516 engineering process Methods 0.000 claims abstract description 9
- 238000000605 extraction Methods 0.000 claims description 36
- 239000013598 vector Substances 0.000 claims description 29
- 238000007726 management method Methods 0.000 claims description 24
- 238000012038 vulnerability analysis Methods 0.000 claims description 23
- 238000010586 diagram Methods 0.000 claims description 22
- 238000007781 pre-processing Methods 0.000 claims description 21
- 238000012549 training Methods 0.000 claims description 21
- 230000036541 health Effects 0.000 claims description 20
- 238000001914 filtration Methods 0.000 claims description 19
- 230000008569 process Effects 0.000 claims description 19
- 238000004422 calculation algorithm Methods 0.000 claims description 18
- 238000009499 grossing Methods 0.000 claims description 15
- 230000004044 response Effects 0.000 claims description 13
- 230000002159 abnormal effect Effects 0.000 claims description 12
- 238000010276 construction Methods 0.000 claims description 10
- 238000012423 maintenance Methods 0.000 claims description 10
- 230000009466 transformation Effects 0.000 claims description 9
- 238000009826 distribution Methods 0.000 claims description 7
- 238000012800 visualization Methods 0.000 claims description 7
- 238000010801 machine learning Methods 0.000 claims description 6
- 238000012544 monitoring process Methods 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 6
- 230000002787 reinforcement Effects 0.000 claims description 5
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000001514 detection method Methods 0.000 claims description 4
- 238000005259 measurement Methods 0.000 claims description 4
- 238000013486 operation strategy Methods 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 5
- 238000000513 principal component analysis Methods 0.000 description 14
- 230000000694 effects Effects 0.000 description 9
- 230000008859 change Effects 0.000 description 7
- 238000005457 optimization Methods 0.000 description 6
- 238000002360 preparation method Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 5
- 238000013461 design Methods 0.000 description 4
- 239000011159 matrix material Substances 0.000 description 3
- 238000012806 monitoring device Methods 0.000 description 3
- 230000005856 abnormality Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000010248 power generation Methods 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 238000012552 review Methods 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 238000012351 Integrated analysis Methods 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000002547 anomalous effect Effects 0.000 description 1
- 238000005311 autocorrelation function Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 238000013524 data verification Methods 0.000 description 1
- 238000013079 data visualisation Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000010183 spectrum analysis Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012731 temporal analysis Methods 0.000 description 1
- 238000000700 time series analysis Methods 0.000 description 1
- 230000001131 transforming effect Effects 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
- G06F18/10—Pre-processing; Data cleansing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/042—Knowledge-based neural networks; Logical representations of neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2123/00—Data types
- G06F2123/02—Data types in the time domain, e.g. time-series data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Data Mining & Analysis (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Educational Administration (AREA)
- Evolutionary Biology (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Bioinformatics & Computational Biology (AREA)
- Game Theory and Decision Science (AREA)
- Molecular Biology (AREA)
- Quality & Reliability (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Development Economics (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Operations Research (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Power Engineering (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a visual analysis system and a visual analysis method for big electric power data, which belong to the technical field of big electric power data. The system improves the data processing precision through data cleaning, signal processing and LSTM model technology; the comprehensive performance of equipment state evaluation and network security analysis is realized by applying multi-index fusion and a graph neural network; by combining visual display and intelligent decision support, a user can manage the power system more accurately and efficiently, so that the stability, efficiency and reliability of the system are improved, the power industry is promoted to advance towards more innovation and more sustainable directions, the beneficial effects brought by the technical innovation are beneficial to the power system to better adapt to increasingly complex energy environments, the overall operation level is improved, and the sustainable development of the industry is promoted.
Description
Technical Field
The invention belongs to the technical field of electric power big data, and particularly relates to an electric power big data visual analysis system and method.
Background
As power systems continue to expand and power demands increase, the size and complexity of large data of power also continues to increase. Traditional power system management and analysis methods often face challenges such as unstable data quality, insufficient analysis precision, insufficient decision support and the like. In order to better monitor, manage and optimize the power system and improve the reliability and efficiency of the power system, an innovative power big data visual analysis system is urgently needed.
Under the background, the visual analysis system based on the power big data is provided, and the data of the power system is comprehensively analyzed and managed through modules such as data acquisition, preprocessing, load prediction, equipment state evaluation, network security analysis, distributed energy management and decision support, so that more accurate, comprehensive and intelligent power system analysis and decision support is realized.
However, the existing system still has some limitations when facing to complicated power big data, such as data noise and abnormal value influence analysis results, the traditional method cannot fully mine data characteristics, load prediction accuracy is not high, equipment state assessment is not comprehensive, network security analysis is not relevant, and the like. Therefore, there is a need for innovative system designs to overcome these limitations and to improve the analysis level and decision support capabilities of power systems.
Under the background, the novel power big data visual analysis system realizes more accurate and reliable data processing and load prediction by introducing advanced technologies such as data cleaning, signal processing, LSTM model and the like. Meanwhile, the multi-index fusion and the application of the graph neural network enable the equipment state evaluation and the network security analysis to be more comprehensive and accurate. Finally, through visual display and intelligent decision support, the analysis result is intuitively presented to the user, so that the user is helped to make more reasonable decisions, and the management level and efficiency of the whole power system are improved.
In summary, the novel power big data visual analysis system brings remarkable beneficial effects on the aspects of power system management and decision support, and provides powerful technical support and innovation solutions for modern transformation of the power industry.
The present invention has been made in view of this.
Disclosure of Invention
In order to solve the technical problems, the invention adopts the basic conception of the technical scheme that:
a power big data visual analysis system, comprising:
a data acquisition module;
the method comprises the steps that on-site monitoring equipment or a sensor is used for collecting current, voltage and load data from a power system in real time, and the data are sampled and filtered to obtain original data;
The data acquisition module comprises a data cleaning submodule and a data processing submodule;
a data preprocessing module;
denoising, interpolating and aligning the original data, processing missing values and abnormal values, performing signal processing and feature extraction by using a wavelet transformation or difference method, and generating time sequence features;
the data preprocessing module comprises a signal processing sub-module and a feature extraction sub-module;
a power load prediction module;
training based on historical power load data by utilizing an LSTM model, constructing a prediction model, inputting time sequence characteristics into the model, carrying out load prediction, calling the result of a data preprocessing module, and obtaining the power load prediction result of a future time period;
the power load prediction module comprises a model training sub-module and a prediction sub-module;
a power device state evaluation module;
combining power engineering knowledge and machine learning technology, performing multi-index fusion by using a PCA method based on multi-source data of electrical parameters, temperature and vibration of equipment, realizing equipment state evaluation, and calling a result of a data preprocessing module;
the power equipment state evaluation module comprises a feature fusion sub-module and a state evaluation sub-module;
The power network safety analysis module;
constructing a topological structure diagram of the power network by using the graph neural network, representing a subsystem by nodes, representing connection by edges, carrying out vulnerability analysis and attack detection based on the graph neural network, evaluating the safety of the power network, and obtaining a power network safety evaluation result;
the power network security analysis module comprises a topology construction sub-module and a vulnerability analysis sub-module;
a distributed energy management module;
optimizing cooperative scheduling of different energy sources in the micro-grid based on market price and user requirements by using a distributed control theory and reinforcement learning, and realizing an energy source distribution and scheduling strategy;
the distributed energy management module comprises a strategy control sub-module and an energy scheduling sub-module.
As a further aspect of the invention: the data cleaning sub-module performs denoising and outlier processing on the original data acquired from the power system, ensures the accuracy of the data, reduces the influence of noise and outlier data, and acquires a clean original data set;
the data processing sub-module integrates and formats clean original data, performs subsequent preprocessing and analysis, and realizes that a unified data source is provided for the subsequent module through an integrated and formatted data set;
A visualization and decision support module;
integrating results from the power load prediction module, the power equipment state evaluation module, the power network safety analysis module and the distributed energy management module, performing visual display on the results including power load prediction, equipment health evaluation, network safety evaluation and energy scheduling strategy information, displaying the running state of a power system in the form of a chart, a map and virtual reality, providing intelligent decision support for users, and recommending reasonable operation strategies according to the system analysis result;
the visualization and decision support module comprises an information display sub-module and a decision recommendation sub-module.
As a further aspect of the invention: the signal processing sub-module performs denoising, smoothing and filtering on the acquired data, eliminates random fluctuation caused by measurement or transmission, processes the processed data, and reduces noise interference in the data;
the feature extraction sub-module performs feature extraction on the data after signal processing, wherein the feature extraction sub-module comprises frequency domain features and time domain features, so that the data feature vectors are analyzed by the subsequent modules, and the extracted data feature vectors contain useful information of signals.
As a further aspect of the invention: the model training sub-module uses the historical power load data and the corresponding characteristic data to train the LSTM model, learn the relation between the load and the characteristic, train the load prediction model, and have the capacity of predicting the future load;
And the prediction submodule inputs characteristic data of a future time period by using the trained LSTM model, and performs load prediction, and a power load value in the predicted future time period.
As a further aspect of the invention: the characteristic fusion submodule fuses the electrical parameters, the temperature and the vibration data from different sensors to generate comprehensive characteristics, and comprehensive display equipment is realized through comprehensive characteristic vectors;
the state evaluation submodule is used for performing equipment state evaluation by using a PCA method based on the comprehensive characteristics, judging the health degree of equipment and obtaining the health state evaluation result of the equipment.
As a further aspect of the invention: the topology construction submodule constructs a topology structure diagram of the power network based on the topology relation of the power system and reflects the connection relation of the power system;
the vulnerability analysis submodule analyzes the vulnerability of the power network by using the graph neural network, predicts the system response under attack, and acquires the vulnerability analysis result of the power network.
As a further aspect of the invention: the information display sub-module visually displays the results of the modules in a chart, a map and other modes, presents the real-time state of the power system and generates a real-time power system state visual interface;
The decision recommendation submodule provides reasonable operation decision suggestions for users by using an intelligent algorithm based on analysis results of the modules, and helps the users to make more optimal decisions through the intelligent decision suggestions.
As a further aspect of the invention: the strategy system submodule formulates a distributed energy scheduling strategy based on market price, user requirements and energy supply conditions, and prescribes allocation and scheduling schemes of energy sources.
As a further aspect of the invention: a visual analysis method for electric power big data comprises the following steps:
s1: data acquisition and pretreatment;
collecting current, voltage and load data from the power system using real-time monitoring devices or sensors;
sampling and filtering the acquired data to remove noise and interference and obtain clean original data;
performing data cleaning to remove abnormal values and form an accurate original data set;
interpolation and alignment are carried out on the original data, missing values are filled, and continuity of the data is ensured;
s2: feature extraction and data preparation;
the signal processing submodule is used for denoising, smoothing and filtering the data, so that the influence of random fluctuation is reduced;
performing signal processing by using a wavelet transformation or difference method, and extracting frequency domain features and time domain features;
The data feature extraction submodule generates a time sequence feature vector which contains useful information of signals;
integrating the extracted time sequence features with historical power load data to prepare for a load prediction module;
s3: predicting the power load;
using a model training sub-module to input historical power load data and characteristic data into an LSTM model for training;
the LSTM model after training learns the relation between the load and the characteristics, and has the capability of predicting future load;
the prediction submodule inputs characteristic data of a future time period by using the trained model to perform load prediction;
the prediction submodule outputs a power load value in a future time period and provides a basis for energy management;
s4: evaluating the state of the power equipment;
the characteristic fusion sub-module fuses different sensor data to generate a comprehensive characteristic vector;
the comprehensive feature vector is input into a state evaluation sub-module, and equipment state evaluation is carried out by using a PCA method;
judging the health degree of the equipment according to the evaluation result, and providing reference for equipment management and maintenance;
s5: safety analysis of the power network;
the topology construction submodule constructs a topology structure diagram of the power network based on the topology relation of the power system;
Constructing nodes and edges of a power network by using a graph neural network, and providing a basis for safety analysis;
the vulnerability analysis submodule utilizes the graph neural network to analyze the vulnerability of the power network, predicts the response of the system and evaluates the safety of the network;
s6: distributed energy management;
the strategy making sub-module makes an energy scheduling strategy based on market price, user demand and energy supply condition;
the energy scheduling submodule executes an energy collaborative scheduling strategy to optimize the allocation and scheduling of various energy sources in the micro-grid;
the scheduling result can be based on the prediction result of the power load prediction module so as to reduce the cost and ensure the supply;
s7: information display and decision support;
the information display sub-module is used for visually displaying analysis results of all the modules in a chart, map and other modes;
the decision recommendation submodule provides operation decision recommendation by using an intelligent algorithm based on the analysis result of the module;
the user makes a more optimal operation decision according to the displayed result and the decision proposal.
Advantageous effects
Firstly, the new system introduces a data cleaning submodule and a data processing submodule, can collect high-quality original data from the power system in real time, and ensures the accuracy of the data through denoising and outlier processing, thereby providing a more reliable data basis for subsequent analysis. Compared with the existing system, the method can reduce the influence of data noise and abnormality, and improves the accuracy and precision of analysis. And secondly, a signal processing sub-module and a feature extraction sub-module of the data preprocessing module enable the system to process the original data more effectively, extract time sequence features and provide richer data features for subsequent model training and analysis. Compared with the existing system, the model can better capture the change rule of the data by extracting the characteristics, and the analysis capability of the system is enhanced.
The model training sub-module and the prediction sub-module of the power load prediction module adopt an LSTM model, so that future power loads can be predicted more accurately. Compared with the traditional method, the model can process the long-term dependency relationship of the sequence data, improves the accuracy of load prediction, and is beneficial to more effectively carrying out load scheduling and planning on the power system. In addition, the characteristic fusion sub-module and the state evaluation sub-module of the power equipment state evaluation module combine power engineering knowledge and machine learning technology, and comprehensive characteristics are generated through fusion of a plurality of data sources, so that the health condition of the equipment can be evaluated more comprehensively. Compared with the existing system, the multi-index fusion method enables evaluation to be more accurate and reliable.
The power network safety analysis module introduces a graph neural network, and realizes safety analysis and vulnerability prediction of the power network from the network topology. Compared with the traditional vulnerability analysis method, the method can further consider the complexity and the relevance of the network, and improves the security assessment capability of the system. Finally, the information display sub-module and the decision recommendation sub-module of the visualization and decision support module can intuitively display the analysis result to the user and provide intelligent decision suggestions for the user. Compared with the existing system, the user-friendly interface and the intelligent suggestion are beneficial to users to make reasonable decisions more quickly and accurately, and the practical application value of the system is improved.
The following describes the embodiments of the present invention in further detail with reference to the accompanying drawings.
Drawings
In the drawings:
FIG. 1 is a schematic diagram of a system flow structure according to the present invention;
FIG. 2 is a block diagram of a system of the present invention;
FIG. 3 is a schematic diagram of a data acquisition process according to the present invention;
FIG. 4 is a schematic diagram of a data preprocessing flow in accordance with the present invention;
FIG. 5 is a schematic diagram of a power load prediction process according to the present invention;
FIG. 6 is a schematic diagram of a power device status evaluation flow structure according to the present invention;
FIG. 7 is a flow chart of the visualized analysis method of the electric power big data.
Description of the embodiments
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions in the embodiments will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and the following embodiments are used to illustrate the present invention.
As shown in fig. 1 to 6, a power big data visual analysis system includes:
a data acquisition module;
the method comprises the steps that on-site monitoring equipment or a sensor is used for collecting current, voltage and load data from a power system in real time, and the data are sampled and filtered to obtain original data;
the data acquisition module comprises a data cleaning submodule and a data processing submodule;
A data preprocessing module;
denoising, interpolating and aligning the original data, processing missing values and abnormal values, performing signal processing and feature extraction by using a wavelet transformation or difference method, and generating time sequence features;
the data preprocessing module comprises a signal processing sub-module and a feature extraction sub-module;
a power load prediction module;
training based on historical power load data by utilizing an LSTM model, constructing a prediction model, inputting time sequence characteristics into the model, carrying out load prediction, calling the result of a data preprocessing module, and obtaining the power load prediction result of a future time period;
the power load prediction module comprises a model training sub-module and a prediction sub-module;
a power device state evaluation module;
combining power engineering knowledge and machine learning technology, performing multi-index fusion by using a PCA method based on multi-source data of electrical parameters, temperature and vibration of equipment, realizing equipment state evaluation, and calling a result of a data preprocessing module;
the power equipment state evaluation module comprises a feature fusion sub-module and a state evaluation sub-module;
the power network safety analysis module;
constructing a topological structure diagram of the power network by using the graph neural network, representing a subsystem by nodes, representing connection by edges, carrying out vulnerability analysis and attack detection based on the graph neural network, evaluating the safety of the power network, and obtaining a power network safety evaluation result;
The power network security analysis module comprises a topology construction sub-module and a vulnerability analysis sub-module;
a distributed energy management module;
optimizing cooperative scheduling of different energy sources in the micro-grid based on market price and user requirements by using a distributed control theory and reinforcement learning, and realizing an energy source distribution and scheduling strategy;
the distributed energy management module comprises a strategy system sub-module and an energy scheduling sub-module;
a visualization and decision support module;
integrating results from the power load prediction module, the power equipment state evaluation module, the power network safety analysis module and the distributed energy management module, performing visual display on the results including power load prediction, equipment health evaluation, network safety evaluation and energy scheduling strategy information, displaying the running state of a power system in the form of a chart, a map and virtual reality, providing intelligent decision support for users, and recommending reasonable operation strategies according to the system analysis result;
the visualization and decision support module comprises an information display sub-module and a decision recommendation sub-module.
The data cleaning sub-module performs denoising and outlier processing on the original data acquired from the power system, ensures the accuracy of the data, reduces the influence of noise and outlier data, and acquires a clean original data set;
Data cleaning is a key step to ensure analysis accuracy, and the following is a method how the data cleaning sub-module processes raw data collected from the power system to denoise, process outliers, ensure data accuracy, and reduce noise and abnormal data effects:
denoising:
filtering technology: a digital filter (e.g., a low pass, high pass, or band pass filter) is applied to attenuate high or low frequency noise, thereby smoothing the data and removing noise.
Moving average: and calculating a moving average value of the data, and reducing the influence of the bursty noise on the data.
Median filtering: the current value is replaced with a median within the data window to suppress the effect of outliers.
Outlier processing:
threshold detection: and setting a proper threshold value, and treating the data exceeding the threshold value as an abnormal value.
Interpolation: outliers are filled in using interpolation methods of neighboring data to reduce interference with the analysis.
And (3) cutting: and data exceeding a certain range is truncated to a reasonable range, so that the influence of extreme values on analysis is prevented.
And (3) guaranteeing data accuracy:
and (3) data verification: comparing the data collected by different data sources, and eliminating possible errors or inconsistent data.
Time synchronization: the time stamp synchronization of different sensor data is ensured, and data errors caused by time differences are avoided.
Noise and anomalous data effects are reduced:
smoothing data: and smoothing the data by adopting methods such as averaging or filtering and the like to reduce random noise in the data.
And (5) repeated sampling: the same data is sampled multiple times and then averaged to reduce the effect of random noise.
Statistical analysis: outliers are identified and processed statistically, such as by replacing outliers with average weights.
Acquiring a clean raw data set:
backing up data: a backup of the original data is maintained so that it can be restarted when a problem occurs during the cleaning process.
And (3) data recording: the cleaning process, including denoising and outlier handling, is recorded in detail to ensure reproducibility of the results.
Through the steps, the data cleaning submodule can effectively remove noise and process abnormal values, ensure the accuracy of data collected from the power system and provide a high-quality clean data set for subsequent analysis.
The data processing sub-module integrates and formats the clean original data, performs subsequent preprocessing and analysis, and realizes that a unified data source is provided for the subsequent module through the integrated and formatted data set.
The data processing sub-module is responsible for integrating and formatting clean original data obtained through the data cleaning sub-module, and carrying out subsequent preprocessing and analysis. The workflow of the data processing sub-module is as follows:
the data from the different sensors, devices or subsystems are integrated together to create a unified data set. And ensures that the time stamps of the data are synchronized so that the data can be properly aligned during analysis. The data is formatted in a standard format, such as a table, matrix, or time series format. And field naming and unit unification of data are ensured, and confusion and errors are avoided. The data is converted into a data type suitable for subsequent analysis, such as numeric, text, etc. The necessary tags, notes or metadata are added to the data describing the source, meaning and importance of the data. Marking abnormal values or special events so as to be capable of being processed in a targeted manner during subsequent analysis;
the data is ordered and indexed according to time stamps or other suitable identification, ensuring that the data is ordered according to time or other rules. This facilitates subsequent time series analysis and cross-time analysis;
The integrated and formatted data is stored in a suitable database or data structure for subsequent data retrieval and processing. Ensuring the safety and reliability of data, and backing up the data to prevent accidental loss;
during the integration and formatting process, a check of the data quality is performed to ensure that there are no missing values, duplicate values, or other anomalies. If a problem is found, appropriate measures can be taken to repair or correct the problem;
by integration and formatting of the data processing sub-modules, clean raw data is converted into an organized, easily-analyzed dataset, providing a good basis for subsequent preprocessing, analysis, and modeling. Such data preparation work can ensure the accuracy and reliability of analysis.
The signal processing sub-module performs denoising, smoothing and filtering on the acquired data, eliminates random fluctuation caused by measurement or transmission, processes the processed data, and reduces noise interference in the data;
the acquisition data acquires raw data from the sensor or the monitoring device, including information such as current, voltage, load and the like. And analyzing the noise characteristics by using a statistical method or a spectrum analysis technology, analyzing the noise characteristics in the data, and determining the type of noise to be removed. Selection of denoising method an appropriate denoising method, such as moving average, median filtering, or wavelet denoising, is selected according to noise characteristics. The smoothing method is selected to select a smoothing method suitable for the characteristics of the data, such as moving average and exponential smoothing. Smoothing technology is applied to the denoised data, so that peaks and fluctuations in the data are reduced. Selection of the filter type the appropriate filter type is selected based on the frequency characteristics of the data, such as low pass filtering, high pass filtering or band pass filtering. The design filter designs parameters of the digital filter, such as cut-off frequency, passband, stopband characteristics and the like, according to the filtering requirement. The application of the filter applies the filter to the data, filtering out unwanted high or low frequency components to eliminate random fluctuations caused by the measurement or transmission. And (5) evaluating the processing effect, comparing the data before and after the processing, and analyzing whether the noise level and the fluctuation situation are obviously improved. And adjusting parameters according to the evaluation result, and adjusting parameters of the denoising, smoothing and filtering method to further optimize the processing effect. The recording process records the processing method, parameters and results of each step in detail for review and reproduction. A report creation process report is generated that includes analysis of raw data, processed data, and noise cancellation effects for reference in subsequent analysis.
The feature extraction sub-module performs feature extraction on the data after signal processing, wherein the feature extraction sub-module comprises frequency domain features and time domain features, so that the data feature vectors are analyzed by the subsequent modules, and the extracted data feature vectors contain useful information of signals.
The feature extraction sub-module performs specific algorithms and methods on the signal-processed data to extract useful features from the data for subsequent analysis, modeling, and prediction. The following specific steps of the feature extraction submodule are as follows:
according to the nature and analysis target of the data, selecting a proper characteristic extraction method. Common methods include time domain features, frequency domain features, wavelet transforms, statistical features, and the like.
Basic time domain statistics such as mean, variance, maximum, minimum, etc. are extracted. First order differences, second order differences, etc. of the time series data are calculated to capture the rate of change and trend of the data.
Fourier transforming the time domain data converts the signal to the frequency domain. Frequency domain features such as energy of frequency components, spectral peaks, etc. are extracted.
Wavelet transformation can decompose a signal into components of different scales, thereby extracting information of different frequencies. The signal is subjected to wavelet decomposition, and wavelet coefficients with different scales are extracted as characteristics.
Statistical features of the signal, such as mean, variance, skewness, kurtosis, etc., are extracted. The statistical features can reflect the distribution and change characteristics of the data.
Relevant features such as lag correlation coefficients, autocorrelation functions, etc. are extracted according to the time sequence of the time series data.
And selecting the extracted features, and eliminating redundant or irrelevant features to reduce the data dimension and improve the efficiency and generalization capability of the model.
Features extracted from different methods are combined into one feature vector as the final data representation. The feature vector may be input for subsequent analysis, modeling, and prediction.
The method, parameters and generated features selected for each step are recorded for review and reproduction. A report of the feature extraction is generated, including the extracted features and their meanings, for reference in subsequent analysis.
Through the feature extraction submodule, the data after signal processing is converted into feature vectors with more information quantity and descriptive capacity, and a valuable data base is provided for subsequent analysis and modeling.
The model training sub-module uses the historical power load data and the corresponding characteristic data to train an LSTM model (long-short-term memory network), learn the relation between the load and the characteristic, train a finished load prediction model, and have the capability of predicting future load;
And the prediction submodule inputs characteristic data of a future time period by using the trained LSTM model, performs load prediction, and predicts a power load value in the future time period.
Training of a specific long-short-term memory network (i.e., LSTM model) is one of the prior arts, and the scheme is not repeated.
The characteristic fusion submodule fuses the electrical parameters, the temperature and the vibration data from different sensors to generate comprehensive characteristics, and comprehensive display equipment is realized through comprehensive characteristic vectors;
the state evaluation submodule performs equipment state evaluation by using a PCA method based on the comprehensive characteristics, judges the health degree of equipment and acquires the health state evaluation result of the equipment.
PCA (Principal Component Analysis) is a common data dimension reduction and feature extraction technique, and the specific principle is not repeated, and is used for extracting main features from high-dimensional data and mapping the data to a new low-dimensional space in the scheme. The goal of PCA is to project the original data onto a new coordinate axis through linear transformation such that the projected data has the greatest variance, specifically:
in the state evaluation sub-module, the process of performing device state evaluation and judging the health degree based on the comprehensive characteristics may be as follows:
And collecting multi-source data such as electrical parameters, temperature, vibration and the like from different sensors to form a multi-dimensional data matrix.
Each feature is normalized to have zero mean and unit variance. This is because PCA is sensitive to the scale and amplitude of the data, and normalization can avoid that certain features have an impact on the result of PCA due to scale differences.
And carrying out PCA calculation on the normalized multidimensional data to obtain a covariance matrix, and corresponding eigenvalues and eigenvectors.
The feature values are ordered according to size, and the principal component, that is, the feature vector corresponding to the maximum feature value is selected. These principal components correspond to the main direction of change in the data.
And projecting the original data onto the selected principal component to obtain the comprehensive feature vector.
Based on the integrated feature vector, a health assessment indicator of the device is calculated. These metrics may be some composite score, distance metric, or other expert-defined health metric.
And comparing the calculated health evaluation index with a predefined threshold value. The state of the device, such as healthy, general, abnormal, etc., is judged according to the setting of the threshold value.
The status judgment result of the output device may be a status label in text form or a status level of a digital representation.
Through the steps, after the comprehensive features are subjected to PCA dimension reduction and feature extraction, the multi-aspect features of the equipment can be better reflected. By utilizing the characteristics and combining with a predefined evaluation index and a predefined threshold value, the state of the equipment can be evaluated and the health degree of the equipment can be judged, so that an important reference is provided for the operation and maintenance of the power equipment.
The topology construction submodule constructs a topology structure diagram of the power network based on the topology relation of the power system and reflects the connection relation of the power system;
and collecting equipment data of the power system, wherein the equipment data comprises information such as positions, connection relations, rated parameters and the like of various equipment. The data may be from engineering drawings, field measured data, etc. of the power system.
According to the collected device data, each power device is used as a node, and the connection relation between the nodes represents the connection between the devices.
And establishing a connecting wire according to the connection relation between the devices to represent the electrical connection between the power devices. The connection lines may be straight lines, arrows or other symbols used to represent physical and electrical relationships between the electrical devices.
And analyzing the topological structure of the power system according to the connection relation between the devices, and determining the relative position and connection mode between the devices. This helps to understand the distribution of the power system and the power transfer path between the devices.
And graphically representing the analyzed topological relation to construct a topological structure diagram. The topology structure diagram can be represented in the forms of a chart, a diagram, a network diagram and the like, so that the connection relation of the power system is clear at a glance.
Critical information such as device name, rating parameters, connection type, etc. is marked in the topology map. These labels help to better understand the structure and operation of the power system.
The topology of the power system may change as equipment changes, upgrades, etc., so that the topology map needs to be updated and maintained periodically to maintain its accuracy and practicality.
The vulnerability analysis submodule analyzes the vulnerability of the power network by using the graph neural network, predicts the system response under attack, and acquires the vulnerability analysis result of the power network.
The vulnerability analysis submodule analyzes the vulnerability of the power network by using the graph neural network, and predicts the system response under attack, the process can be as follows:
the topology structure diagram of the power network is taken as a basis, the nodes represent subsystems of the power system, and the edges represent connection relations. A graph is constructed in which each node represents a power device or subsystem and the edges represent the electrical connections between the devices. Such a diagram is used to represent the topology and connection relationships of the power system;
In order to input a graph into a graph neural network, feature extraction and encoding of nodes and edges are required. Each node and edge has corresponding characteristics, such as the node may be a device parameter, the edge may be a connection strength, etc. These features need to be encoded into a vector form that can be processed by the neural network;
the power network map is modeled using a map neural network (Graph Neural Network, GNN). GNN is a class of neural networks dedicated to processing graph data that is capable of capturing complex relationships between nodes and edges. The GNN can be calculated on each node, and the characteristics of the node and the information of the adjacent nodes are considered, so that a richer representation is generated;
in the graph neural network, the key to vulnerability analysis is how to design the loss function. This loss function should be able to measure the degree of response of the power network when it is under attack. For example, an objective function may be constructed that minimizes a certain performance index of the network (e.g., power transfer capability, node voltage stability, etc.) under attack;
in order to predict the response of the system under attack, different types of attacks, such as node failure, line failure, etc., can be simulated. Adding the simulated attack into the power network graph, updating the characteristics of nodes and edges, and inputting the characteristics into a trained graph neural network for analysis;
And analyzing the simulated power network by using the trained graph neural network, and predicting the system response under the attack. By analyzing changes in node characteristics and degradation in network performance, vulnerability of the power network may be assessed.
The output vulnerability analysis results may be the change of the system response, affected nodes and lines, etc. Such information helps power system operators take corresponding measures to improve the robustness and security of the network.
Through the steps, the vulnerability analysis submodule can conduct vulnerability analysis on the power network by using the graph neural network, forecast system response under attack, help power system operators identify potential risks, and take corresponding precautionary measures.
The information display sub-module visually displays the results of the modules in a chart, a map and other modes, presents the real-time state of the power system and generates a real-time power system state visual interface;
the decision recommendation submodule provides reasonable operation decision suggestions for users by using an intelligent algorithm based on analysis results of the modules, and helps the users to make more optimal decisions through the intelligent decision suggestions.
Based on the analysis results of the modules, the decision recommendation submodule applies an intelligent algorithm to provide reasonable operation decision recommendation for the user, and the process can be as follows:
obtaining analysis results from the previous modules, wherein the results may include information on aspects such as power load prediction, equipment state evaluation, network security evaluation and the like;
and integrating and preprocessing analysis results of different modules so as to facilitate subsequent decision recommendation. This may involve weighting, normalizing, etc. the different results to ensure that the different types of results can be considered together;
based on the integrated analysis results, a series of decision rules are formulated. These rules may be expert knowledge based rules or rules derived based on historical data and machine learning algorithms. The purpose of the rule is to make a reasonable operation decision according to different conditions;
and the formulated decision rule is optimized and adjusted by using intelligent algorithms such as machine learning, reinforcement learning and the like. The intelligent algorithm can continuously adjust the decision rule according to the historical data and the real-time situation so as to provide more accurate decision suggestions;
and generating decision suggestions aiming at different situations based on the optimized decision rule and the result of the intelligent algorithm. These suggestions may relate to operations in terms of power load adjustment, equipment maintenance, network optimization, etc.;
The generated decision advice is presented to the user in an intuitive way, possibly by means of a chart, report, interface, etc. The user can learn about the suggestions through these interfaces in order to make decisions;
the user can make the actual operation according to the decision advice, and the following actual effect can also be recorded. Such feedback information may be used to continually optimize decision rules and algorithms to provide more adaptive suggestions;
the decision recommendation sub-module should be constantly optimized to accommodate changes in the power system and changes in the user's needs. This may be achieved by updating algorithms, optimizing rules, and adopting user feedback, etc.
Through the steps, the decision recommendation sub-module can provide reasonable operation decision recommendation aiming at different conditions for users by using an intelligent algorithm according to the analysis results of the modules so as to optimize the operation and management of the power system.
The strategy system submodule formulates a distributed energy scheduling strategy based on market price, user requirements and energy supply conditions, and specifies the allocation and scheduling scheme of each energy.
The strategy system submodule makes a distributed energy scheduling strategy based on market price, user demand and energy supply condition, and the process of specifying the allocation and scheduling scheme of each energy can be as follows:
Real-time or historical market energy price information is acquired, including prices of various energy sources such as electric power, solar energy, wind energy and the like. Market price fluctuations are analyzed to reasonably consider cost factors in the energy scheduling strategy.
And analyzing the energy demands of users, including prediction of power consumption, peak-valley period demand change and the like. And according to different user demands, formulating an energy scheduling strategy with stronger adaptability so as to meet the actual demands of the users.
And analyzing the supply conditions of the distributed energy sources, including solar energy generating capacity, wind energy generating capacity and the like. Taking the fluctuation of different energy sources into consideration, a scheduling strategy is formulated to utilize renewable energy sources to the greatest extent.
And (3) an optimal energy scheduling strategy is formulated based on market price, user requirements and energy supply conditions by using an optimization algorithm such as linear programming, reinforcement learning and the like. The algorithms may optimize the scheduling policy based on different objective functions, such as cost minimization, reliability maximization, etc.
And (3) formulating a specific energy allocation and scheduling scheme according to the result of the optimization algorithm. For example, the electric load is increased during a low electricity price period, and part of the electric power demand is satisfied by using renewable energy, thereby reducing the cost. And in the high electricity price period, flexible scheduling is carried out according to the user requirements and the energy supply condition so as to ensure the supply stability.
Various constraints of the system, such as energy supply capacity, operating state of the device, priority of user demand, etc., are considered in the preparation of the energy scheduling policy. Ensuring that the formulated strategy is viable in actual operation.
The policy making sub-module should have the ability to adjust and update in real time as market price, user demand and energy supply changes. According to the actual situation, the energy allocation and scheduling scheme is dynamically adjusted to adapt to the changing environment to the greatest extent.
The output formulated energy allocation and scheduling scheme may be information including energy supply schedule, equipment work instruction, etc. Such information may assist the energy management personnel in actual operation and monitoring.
Through the steps, the strategy system submodule can apply an optimization algorithm to formulate a reasonable distributed energy scheduling strategy according to market price, user requirements and energy supply conditions so as to optimize cost, meet requirements and improve energy utilization efficiency.
A visual analysis method for electric power big data comprises the following steps:
s1: data acquisition and pretreatment;
collecting current, voltage and load data from the power system using real-time monitoring devices or sensors;
Sampling and filtering the acquired data to remove noise and interference and obtain clean original data;
performing data cleaning to remove abnormal values and form an accurate original data set;
interpolation and alignment are carried out on the original data, missing values are filled, and continuity of the data is ensured;
s2: feature extraction and data preparation;
the signal processing submodule is used for denoising, smoothing and filtering the data, so that the influence of random fluctuation is reduced;
performing signal processing by using a wavelet transformation or difference method, and extracting frequency domain features and time domain features;
the data feature extraction submodule generates a time sequence feature vector which contains useful information of signals;
integrating the extracted time sequence features with historical power load data to prepare for a load prediction module;
s3: predicting the power load;
using a model training sub-module to input historical power load data and characteristic data into an LSTM model for training;
the LSTM model after training learns the relation between the load and the characteristics, and has the capability of predicting future load;
the prediction submodule inputs characteristic data of a future time period by using the trained model to perform load prediction;
The prediction submodule outputs a power load value in a future time period and provides a basis for energy management;
s4: evaluating the state of the power equipment;
the characteristic fusion sub-module fuses different sensor data to generate a comprehensive characteristic vector;
the comprehensive feature vector is input into a state evaluation sub-module, and equipment state evaluation is carried out by using a PCA method;
judging the health degree of the equipment according to the evaluation result, and providing reference for equipment management and maintenance;
s5: safety analysis of the power network;
the topology construction submodule constructs a topology structure diagram of the power network based on the topology relation of the power system;
constructing nodes and edges of a power network by using a graph neural network, and providing a basis for safety analysis;
the vulnerability analysis submodule utilizes the graph neural network to analyze the vulnerability of the power network, predicts the response of the system and evaluates the safety of the network;
s6: distributed energy management;
the strategy making sub-module makes an energy scheduling strategy based on market price, user demand and energy supply condition;
the energy scheduling submodule executes an energy collaborative scheduling strategy to optimize the allocation and scheduling of various energy sources in the micro-grid;
the scheduling result can be based on the prediction result of the power load prediction module so as to reduce the cost and ensure the supply;
S7: information display and decision support;
the information display sub-module is used for visually displaying analysis results of all the modules in a chart, map and other modes;
the decision recommendation submodule provides operation decision recommendation by using an intelligent algorithm based on the analysis result of the module;
the user makes a more optimal operation decision according to the displayed result and the decision proposal.
Of course, i can provide you with more specific embodiments, incorporating the various functions and implementation details of the power big data visualization analysis system:
example 1: smart micro-grid operation
Data acquisition and pretreatment: the system collects current, voltage and load data of all nodes in the micro-grid in real time through field monitoring equipment, and obtains high-quality original data through sampling and filtering processing.
Feature extraction and data preparation: the signal processing and feature extraction sub-module performs smoothing and feature extraction on the original data to generate a time sequence feature vector, and prepares for load prediction.
Electric load prediction: the model training sub-module trains an LSTM model based on the historical load data and the feature vector to predict future loads of the micro-grid. The prediction submodule inputs the feature vector into the model to obtain a load prediction result in a future time period.
And (3) distributed energy management: and the strategy making sub-module makes distribution and scheduling strategies of different energy sources in the micro-grid according to the future load prediction result and the energy price. The energy scheduling submodule executes a strategy to realize cooperative scheduling of energy so as to ensure supply and reduce cost.
Information presentation and decision support: the information display sub-module displays the load prediction result and the energy scheduling strategy in a chart form, and the decision recommendation sub-module provides reasonable energy scheduling decision suggestion for the micro-grid operator according to the analysis result.
Through the system, a micro-grid operator can know the future load condition in real time, optimize the energy scheduling strategy, ensure the reliability of energy supply and reduce the cost.
Example 2: intelligent maintenance of electrical equipment
A household electrical appliance manufacturer successfully applies the visual analysis system of the power big data to realize intelligent maintenance of the power equipment. The following are implementation details:
data acquisition and pretreatment: and acquiring electrical parameters, temperature and vibration data of the power equipment in real time, and obtaining clean original data through sampling and cleaning.
Feature extraction and data preparation: the signal processing and feature extraction submodule performs smoothing and feature extraction on the data to generate a multidimensional feature vector which contains multiple aspects of the equipment.
Power equipment status assessment: the feature fusion submodule fuses different sensor data to generate comprehensive features, and the state evaluation submodule evaluates the state of the equipment by using a PCA method based on the comprehensive features to judge the health degree of the equipment.
Information presentation and decision support: the information display sub-module displays the health state and trend of the equipment in a chart mode, and the decision recommendation sub-module provides intelligent maintenance suggestions for maintenance personnel based on the state evaluation result.
Through the system, equipment manufacturers can timely find out equipment abnormality, maintain in advance, reduce downtime and maintenance cost, and realize reliable operation of equipment.
Example 3: power network security
The visual analysis system for the big electric data is successfully applied to a national electric power company, so that the safety of an electric power network is improved. The following are implementation details:
data acquisition and pretreatment: the system collects data of all subsystems of the power network in real time, including information such as current and voltage, and accurate original data is obtained through data cleaning and processing.
Building a power network topology: the topology construction submodule constructs a topology structure diagram based on the connection relation of the power network, and shows each subsystem node and the relation.
Power network vulnerability analysis: the vulnerability analysis submodule analyzes the power network topological graph by utilizing the graph neural network, identifies potential vulnerable nodes and attack paths, and predicts system response under attack.
Information presentation and decision support: the information display sub-module visually presents vulnerability analysis results and displays fragile nodes and attack paths in the power network. And the decision recommendation submodule provides a safety precaution suggestion for the power company according to the analysis result and ensures the stable operation of the power network.
Through the system, the electric company can accurately identify potential network weak points, adopts a targeted safety strategy, and improves the resistance of the electric network.
Example 4: real-time energy scheduling optimization
A large energy supplier applies the system to perform real-time energy scheduling optimization. The following are implementation details:
data acquisition and pretreatment: the system collects data generated by different energy sources in real time, such as solar power generation, wind power generation and the like, and clean data is obtained through data processing.
Electric load prediction: the power load prediction module predicts future load changes based on the historical data and the feature extraction results.
And (3) cooperative scheduling of distributed energy sources: and the strategy making submodule makes allocation and scheduling strategies of different energy sources according to the load prediction and the market price, and the energy source scheduling submodule executes the strategies to realize cooperative scheduling of the energy sources and ensure stable supply.
Information presentation and decision support: the information display sub-module displays the generation and scheduling conditions of each energy source in a graph mode, and the decision recommendation sub-module provides real-time scheduling suggestions for suppliers according to the analysis results.
Through the system, an energy supplier can adjust the energy generation and scheduling strategy in real time, so that the actual energy demand is met, the cost is reduced, and the energy utilization efficiency is improved.
The embodiment illustrates the flexibility and adaptability of the power big data visual analysis system in different application scenes, and the system can provide comprehensive decision support for the power industry, optimize operation efficiency and reduce risks. The practical value of the power big data visual analysis system in different application scenes and the profound influence of the power big data visual analysis system on the power industry are highlighted.
The foregoing has shown and described the basic principles, principal features and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, and that the above embodiments and descriptions are merely illustrative of the principles of the present invention, and various changes and modifications may be made without departing from the spirit and scope of the invention, which is defined in the appended claims. The scope of the invention is defined by the appended claims and their equivalents.
Claims (9)
1. A power big data visual analysis system, comprising:
a data acquisition module;
the method comprises the steps that on-site monitoring equipment or a sensor is used for collecting current, voltage and load data from a power system in real time, and the data are sampled and filtered to obtain original data;
the data acquisition module comprises a data cleaning submodule and a data processing submodule;
a data preprocessing module;
denoising, interpolating and aligning the original data, processing missing values and abnormal values, performing signal processing and feature extraction by using a wavelet transformation or difference method, and generating time sequence features;
the data preprocessing module comprises a signal processing sub-module and a feature extraction sub-module;
a power load prediction module;
training based on historical power load data by utilizing an LSTM model, constructing a prediction model, inputting time sequence characteristics into the model, carrying out load prediction, calling the result of a data preprocessing module, and obtaining the power load prediction result of a future time period;
the power load prediction module comprises a model training sub-module and a prediction sub-module;
a power device state evaluation module;
the method comprises the steps of carrying out multi-index fusion by using a PCA method based on multi-source data of electrical parameters, temperature and vibration of equipment, realizing equipment state evaluation and calling a result of a data preprocessing module, wherein the multi-source data comprises power engineering knowledge and machine learning technology;
The power equipment state evaluation module comprises a feature fusion sub-module and a state evaluation sub-module;
the power network safety analysis module;
constructing a topological structure diagram of the power network by using the graph neural network, representing a subsystem by nodes, representing connection by edges, carrying out vulnerability analysis and attack detection based on the graph neural network, evaluating the safety of the power network, and obtaining a power network safety evaluation result;
the power network security analysis module comprises a topology construction sub-module and a vulnerability analysis sub-module;
a distributed energy management module;
optimizing cooperative scheduling of different energy sources in the micro-grid based on market price and user requirements by using a distributed control theory and reinforcement learning, and realizing an energy source distribution and scheduling strategy;
the distributed energy management module comprises a strategy control sub-module and an energy scheduling sub-module;
a visualization and decision support module;
integrating results from the power load prediction module, the power equipment state evaluation module, the power network safety analysis module and the distributed energy management module, performing visual display on the results including power load prediction, equipment health evaluation, network safety evaluation and energy scheduling strategy information, displaying the running state of a power system in the form of a chart, a map and virtual reality, providing intelligent decision support for users, and recommending reasonable operation strategies according to the system analysis result;
The visualization and decision support module comprises an information display sub-module and a decision recommendation sub-module.
2. The visualized analysis system of power big data according to claim 1, wherein the data cleaning sub-module performs denoising and outlier processing on the raw data collected from the power system, reduces noise and outlier influence, and obtains a clean raw data set;
the data processing sub-module integrates and formats clean original data, performs subsequent preprocessing and analysis, and realizes that a unified data source is provided for the subsequent module through the integrated and formatted data set.
3. The visualized analysis system of power big data according to claim 1, wherein the signal processing sub-module performs denoising, smoothing and filtering on the collected data, eliminates random fluctuation caused by measurement or transmission, processes the processed data, and reduces noise interference in the data;
the feature extraction sub-module performs feature extraction on the data after signal processing, wherein the feature extraction sub-module comprises frequency domain features and time domain features, so that the data feature vectors are analyzed by the subsequent modules, and the extracted data feature vectors contain useful information of signals.
4. The visualized analysis system of power big data according to claim 1, wherein the model training sub-module uses historical power load data and characteristic data to train an LSTM model, learn a relationship between load and characteristic, train a completed load prediction model, and have the ability to predict future load;
and the prediction submodule inputs characteristic data of a future time period by using the trained LSTM model, and performs load prediction, and a power load value in the predicted future time period.
5. The visual analysis system of electric power big data according to claim 1, wherein the characteristic fusion submodule fuses electric parameters, temperature and vibration data from different sensors to generate comprehensive characteristics, and comprehensive display equipment is realized through comprehensive characteristic vectors;
the state evaluation submodule is used for performing equipment state evaluation by using a PCA method based on the comprehensive characteristics, judging the health degree of equipment and obtaining the health state evaluation result of the equipment.
6. The visualized analysis system of power big data according to claim 1, wherein the topology construction submodule constructs a topology structure diagram of a power network based on the topology relation of the power system, and reflects the connection relation of the power system;
The vulnerability analysis submodule analyzes the vulnerability of the power network by using the graph neural network, predicts the system response under attack, and acquires the vulnerability analysis result of the power network.
7. The visual analysis system for electric power big data according to claim 1, wherein,
the information display sub-module visually displays the results of the modules in the form of a chart or a map, presents the real-time state of the power system and generates a real-time power system state visual interface;
the decision recommendation submodule provides operation decision suggestions for the user based on the analysis results of the modules, and helps the user to make more optimal decisions through intelligent decision suggestions.
8. The visual analysis system of claim 1, wherein the policy generation sub-module generates a distributed energy scheduling policy based on market price, user demand and energy supply, specifying allocation and scheduling schemes for each energy source.
9. A power big data visual analysis method according to any one of claims 1 to 8, characterized by comprising the steps of:
S1: collecting current, voltage and load data from a power system by using real-time monitoring equipment or a sensor, sampling and filtering the collected data, removing noise and interference to obtain original data, cleaning the data, removing abnormal values to form an accurate original data set, interpolating and aligning the original data, filling missing values, and ensuring the continuity of the data;
s2: the data is denoised, smoothed and filtered by using a signal processing submodule, the influence of random fluctuation is reduced, the signal processing is performed by using a wavelet transformation or difference method, the frequency domain characteristics and the time domain characteristics are extracted, a data characteristic extraction submodule generates a time sequence characteristic vector which contains useful information of signals, and the extracted time sequence characteristics are integrated with historical power load data to prepare for a load prediction module;
s3: the model training submodule is used for inputting historical power load data and characteristic data into an LSTM model for training, the trained LSTM model learns the relation between load and characteristics and has the capability of predicting future load, the prediction submodule utilizes the trained model to input characteristic data of a future time period for load prediction, and the prediction submodule outputs power load values in the future time period to provide basis for energy management;
S4: the feature fusion sub-module fuses different sensor data to generate a comprehensive feature vector, the comprehensive feature vector is input into the state evaluation sub-module, the PCA method is utilized to evaluate the state of the equipment, the health degree of the equipment is judged according to the evaluation result, and a reference is provided for equipment management and maintenance;
s5: the topology construction submodule builds a topology structure diagram of the power network based on the topology relation of the power system, uses the graph neural network to build nodes and edges of the power network, provides a basis for safety analysis, and uses the graph neural network to analyze the vulnerability of the power network, predicts system response and evaluates the safety of the network;
s6: the strategy control submodule formulates an energy scheduling strategy based on market price, user requirements and energy supply conditions, the energy scheduling submodule executes an energy collaborative scheduling strategy, the distribution and scheduling of various energy sources in the micro-grid are optimized, and the scheduling result can be based on the prediction result of the power load prediction module so as to reduce cost and ensure supply;
s7: the information display sub-module is used for visually displaying the analysis results of the modules in a chart or map mode, the decision recommendation sub-module is used for providing operation decision suggestions by using an intelligent algorithm based on the analysis results, and a user makes more optimized operation decisions according to the displayed results and the decision suggestions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311175520.1A CN117114449A (en) | 2023-09-13 | 2023-09-13 | Visual analysis system and method for big electric power data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311175520.1A CN117114449A (en) | 2023-09-13 | 2023-09-13 | Visual analysis system and method for big electric power data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117114449A true CN117114449A (en) | 2023-11-24 |
Family
ID=88796397
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311175520.1A Pending CN117114449A (en) | 2023-09-13 | 2023-09-13 | Visual analysis system and method for big electric power data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117114449A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117410988A (en) * | 2023-12-11 | 2024-01-16 | 广东领卓能源科技有限公司 | Charging control method and device for new energy charging station |
| CN117707027A (en) * | 2024-02-05 | 2024-03-15 | 国网甘肃省电力公司白银供电公司 | Centralized control station one-key sequential control system based on multi-source data fusion |
| CN117788082A (en) * | 2024-02-26 | 2024-03-29 | 南京南自华盾数字技术有限公司 | Power market quotation decision method and system based on electricity price prediction |
| CN117895658A (en) * | 2024-03-13 | 2024-04-16 | 国网山东省电力公司诸城市供电公司 | Intelligent micro-grid data analysis platform |
-
2023
- 2023-09-13 CN CN202311175520.1A patent/CN117114449A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117410988A (en) * | 2023-12-11 | 2024-01-16 | 广东领卓能源科技有限公司 | Charging control method and device for new energy charging station |
| CN117410988B (en) * | 2023-12-11 | 2024-03-29 | 广东领卓能源科技有限公司 | Charging control method and device for new energy charging station |
| CN117707027A (en) * | 2024-02-05 | 2024-03-15 | 国网甘肃省电力公司白银供电公司 | Centralized control station one-key sequential control system based on multi-source data fusion |
| CN117707027B (en) * | 2024-02-05 | 2024-04-16 | 国网甘肃省电力公司白银供电公司 | Centralized control station one-key sequential control system based on multi-source data fusion |
| CN117788082A (en) * | 2024-02-26 | 2024-03-29 | 南京南自华盾数字技术有限公司 | Power market quotation decision method and system based on electricity price prediction |
| CN117895658A (en) * | 2024-03-13 | 2024-04-16 | 国网山东省电力公司诸城市供电公司 | Intelligent micro-grid data analysis platform |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117114449A (en) | Visual analysis system and method for big electric power data | |
| JP5363927B2 (en) | Abnormality detection / diagnosis method, abnormality detection / diagnosis system, and abnormality detection / diagnosis program | |
| CN113239132B (en) | Online out-of-tolerance identification method for voltage transformer | |
| CN110709789B (en) | Method and apparatus for monitoring the condition of subsystems within a renewable power generation device or microgrid | |
| CN117196066A (en) | Intelligent operation and maintenance information analysis model | |
| US20220179393A1 (en) | Machine tool evaluation method, machine tool evaluation system and medium | |
| CN117114454B (en) | DC sleeve state evaluation method and system based on Apriori algorithm | |
| CN116955434A (en) | Full life cycle management and multidimensional energy efficiency analysis system for industrial equipment | |
| CN115372816A (en) | Power distribution switchgear operation fault prediction system and method based on data analysis | |
| CN116992346A (en) | Enterprise production data processing system based on artificial intelligence big data analysis | |
| CN117311295B (en) | Production quality improving method and system based on wireless network equipment | |
| CN114552770A (en) | System and method for managing intelligent alarms | |
| CN117607780A (en) | Abnormality analysis system and analysis method for electric energy metering device | |
| CN117374978A (en) | Grid-connected scheduling management method and system constructed by combining knowledge graph | |
| CN117498432A (en) | New energy power system electric energy quality assessment method | |
| Sim et al. | Missing data imputation algorithm for transmission systems based on multivariate imputation with principal component analysis | |
| Kropatschek et al. | Designing a digital shadow for efficient, low-delay analysis of production quality risk | |
| CN111781467B (en) | Power grid fault intelligent discrimination method based on multidimensional panoramic data | |
| CN118011990B (en) | Industrial data quality monitoring and improving system based on artificial intelligence | |
| MOUHIB et al. | Towards a generic framework of OEE monitoring for driving effectiveness in digitalization era | |
| CN117172139B (en) | Performance test method and system for copper-clad aluminum alloy cable for communication | |
| Mendia et al. | An Intelligent Procedure for the Methodology of Energy Consumption in Industrial Environments | |
| CN116069598A (en) | Intelligent 3D visual IT monitoring management method and system | |
| CN118133068B (en) | Automatic identification and construction method and system for power grid topology | |
| CN118260603A (en) | Equipment state analysis method, medium and equipment based on intelligent large model |
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 |