Power grid data analysis and dispatching system based on big data

Abstract

The application discloses a power grid data analysis and scheduling system based on big data, relates to the technical field of power grid data scheduling, and aims at solving the problems of inaccurate power grid data acquisition and inaccurate analysis in the prior art. Different evaluation methods and models are adopted for different types of risks, so that the accuracy and pertinence of risk evaluation are improved, different risk level adjustment strategies are formulated according to the comprehensive risk level, the flexible and targeted scheduling strategy can better adapt to the actual demand of power grid operation, the characteristics and change rules of various data can be more accurately captured through special model establishment of different types of power grid data, fine modeling improves the accuracy and reliability of data analysis, the feature data in each model can be extracted and identifier allocated, so that fast data retrieval and analysis can be realized, and the allocation of the identifier makes each feature data have a clear source and meaning.

Description

Technical Field

[0001] This invention relates to the field of power grid data scheduling technology, specifically a power grid data analysis and scheduling system based on big data. Background Technology

[0002] Power grid data dispatch is an important power system management tool. It involves the collection, analysis, processing, and decision-making of various data and information in the power grid, aiming to ensure the safe, stable, and economical operation of the power grid.

[0003] Chinese patent CN103577939A discloses a power grid data management method and system. It primarily integrates data from different professional systems required for power grid operation into a panoramic model, forming a unified whole that is easier to manage, maintain, search, and use, thereby improving the overall management efficiency of the power grid data system. Furthermore, the panoramic model can be expanded online according to different actual situations through extension modules, thus enhancing the scalability of the aforementioned power grid data management system. While this patent solves the problem of power grid data management, the following issues still exist in practical operation:

[0004] 1. The power grid data was not collected and analyzed from multiple perspectives, which made it impossible to analyze and adjust based on the actual power grid data.

[0005] 2. After acquiring the power grid data, the feature data in the data was not extracted more effectively, resulting in insufficient data for subsequent analysis.

[0006] 3. The final acquired power grid data was not subjected to further risk analysis, which resulted in the inability to perform targeted scheduling based on the risk status of each power grid data point. Summary of the Invention

[0007] The purpose of this invention is to provide a power grid data analysis and dispatching system based on big data. By employing different assessment methods and models for different types of risks, the accuracy and relevance of risk assessment are improved. Adjustment strategies for different risk levels are formulated based on the comprehensive risk level. This flexible and targeted dispatching strategy can better adapt to the actual needs of power grid operation. By establishing specialized models for different types of power grid data, the characteristics and changing patterns of various types of data can be captured more accurately. Refined modeling improves the accuracy and reliability of data analysis. By extracting feature data and assigning identifiers to each model, rapid data retrieval and analysis can be easily achieved. The assignment of identifiers ensures that each feature data has a clear source, thus solving problems in existing technologies.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] A power grid data analysis and dispatch system based on big data includes:

[0010] The power grid data acquisition unit is used for:

[0011] Power grid data is collected from the database, and the collected power grid data is distinguished. After the distinction is completed, a unique code is assigned to it, and the target power grid data is obtained after the unique code is assigned.

[0012] The data acquisition and analysis unit is used for:

[0013] The target power grid data is analyzed to obtain target analysis data.

[0014] The data model building unit is used for:

[0015] The target analysis data is used to establish a power grid information model, and the established power grid information model is labeled with feature data. After the feature data is labeled, the power grid model data is obtained.

[0016] The scheduling instruction generation unit is used for:

[0017] The power grid model data is used to generate a scheduling scheme, and the generated scheduling scheme is transmitted to different control centers. The control centers then schedule the power grid data according to the scheduling scheme.

[0018] Preferably, the power grid data acquisition unit is further configured to:

[0019] Retrieve power grid data from the database, which includes basic data, operational data, monitoring data, load data, economic data, management data, and external data;

[0020] The basic data, operational data, monitoring data, load data, economic data, management data, and external data are classified separately, resulting in independent basic data, operational data, monitoring data, load data, economic data, management data, and external data.

[0021] Each type of independent basic data, operational data, monitoring data, load data, economic data, management data, and external data shall be uniquely coded and labeled;

[0022] The unique coding structure includes a data category code, a data type code, a serial number code, and a timestamp. The data category code represents the specific data category, the data type code represents the data type under each data category, the serial number code represents different data records under the same data type, and the timestamp represents the collection time of different data.

[0023] The target power grid data is obtained by uniquely coding and labeling basic data, operational data, monitoring data, load data, economic data, management data, and external data.

[0024] Preferably, the power grid data acquisition unit includes:

[0025] The real-time monitoring module for operating parameters is used to monitor the operating parameters for retrieving power grid data from the database in real time. The operating parameters include data integrity rate, data retrieval response time, and number of concurrent retrieval threads.

[0026] The first operational evaluation coefficient acquisition module is used to acquire a first operational evaluation coefficient using the data integrity rate and data retrieval response time; wherein, the first operational evaluation coefficient is acquired using the following formula:

[0027]

[0028] Among them, F 01 P represents the first performance evaluation coefficient; n represents the total number of data retrievals from the database; P i T represents the data integrity rate corresponding to the i-th data retrieval; i P represents the response time for the i-th data retrieval. b T represents the standard deviation of data integrity rate corresponding to n data retrievals; b T represents the standard deviation of the data retrieval response time for n data retrievals; c This indicates a preset reference value for the data retrieval response time;

[0029] The first coefficient comparison module is used to compare the first operation evaluation coefficient with a preset first operation coefficient threshold and obtain the comparison result.

[0030] The quality assessment module is used to assess the retrieval quality of power grid data from the database based on the comparison result between the first operation evaluation coefficient and the preset first operation coefficient threshold.

[0031] Preferably, the quality determination module includes:

[0032] The first quality evaluation model extraction module is used to extract the comparison results between the first operation evaluation coefficient and the preset first operation coefficient threshold. If the first operation evaluation coefficient is lower than the preset first operation coefficient threshold, then the first quality evaluation model is retrieved.

[0033] The first-level second operation evaluation coefficient acquisition module is used to obtain the second operation evaluation coefficient by combining the first quality evaluation model with the first operation evaluation coefficient and the number of concurrent call threads.

[0034] The structure of the first quality evaluation model is as follows:

[0035]

[0036] Among them, F 02 F represents the second operational evaluation coefficient obtained from the first quality evaluation model. 01 F represents the first performance evaluation coefficient; n represents the total number of data retrievals from the database; F y T represents the preset first operating coefficient threshold; i T represents the response time for the i-th data retrieval; b M represents the standard deviation of the data retrieval response time for n data retrievals; i M represents the number of concurrent data retrieval threads for the i-th data retrieval; b This represents the standard deviation of the number of concurrent data retrieval threads corresponding to n data retrievals.

[0037] The second quality evaluation model extraction module is used to extract the comparison results between the first operation evaluation coefficient and the preset first operation coefficient threshold. If the first operation evaluation coefficient is not lower than the preset first operation coefficient threshold, then the second quality evaluation model is retrieved.

[0038] The Level 2 Second Operation Evaluation Coefficient Acquisition Module is used to obtain the Second Operation Evaluation Coefficient by combining the Second Quality Evaluation Model with the First Operation Evaluation Coefficient and the Number of Concurrent Calling Threads.

[0039] The structure of the second quality evaluation model is as follows:

[0040]

[0041] Among them, F 03 F represents the second operational evaluation coefficient obtained from the second quality evaluation model. 01 F represents the first performance evaluation coefficient; n represents the total number of data retrievals from the database; F y T represents the preset first operating coefficient threshold; n P represents the total latency rate corresponding to the response time of n data retrievals; n M represents the data integrity rate corresponding to n data retrievals; i M represents the number of concurrent data retrieval threads for the i-th data retrieval; i-1 M represents the number of concurrent data retrieval threads for the (i-1)th data retrieval; i-1 T represents the number of concurrent data retrieval threads for the (i-1)th data retrieval; ni T represents the data retrieval response time latency rate corresponding to the i-th data retrieval;ni-1 This represents the data retrieval response time latency rate corresponding to the (i-1)th data retrieval.

[0042] The second coefficient comparison module is used to compare the second operation evaluation coefficient with a preset second operation coefficient threshold.

[0043] The anomaly detection and alarm module is used to determine that the retrieval of power grid data from the database is abnormal when the second operation evaluation coefficient is lower than the preset second operation coefficient threshold, and to issue an early warning of the quality anomaly.

[0044] Preferably, the data collection and analysis unit includes:

[0045] The data acquisition and processing module is used for:

[0046] The dataset is integrated from the basic data, operational data, monitoring data, load data, economic data, management data, and external data in the target power grid data.

[0047] After the dataset is integrated, a unified dataset is obtained, consisting of basic data, operational data, monitoring data, load data, economic data, management data, and external data.

[0048] The acquired integrated dataset will undergo data preprocessing;

[0049] Data preprocessing involves cleaning, transforming, and standardizing the integrated dataset in one step.

[0050] After data preprocessing, the integrated dataset to be analyzed is obtained.

[0051] Preferably, the data collection and analysis unit further includes;

[0052] The data processing and analysis module is used for:

[0053] Perform data analysis on the integrated dataset to be analyzed;

[0054] The basic data, operational data, monitoring data, load data, economic data, management data, and external data in the integrated dataset to be analyzed will be analyzed separately.

[0055] The analysis of basic data involves obtaining the mean, median, and variance through statistical methods; the analysis of operational data involves obtaining the changing trends through Kalman filtering; the analysis of monitoring data involves identifying monitoring signals through Fourier transform; the analysis of load data involves load forecasting through regression analysis; the analysis of economic data involves cost-benefit analysis of the power grid's economic benefits; the analysis of management data involves analyzing management data through fault tree analysis to optimize the allocation and use of power grid resources; and the analysis of external data involves analyzing the impact of weather, economic conditions, and policy changes on the power grid through Pearson correlation coefficient analysis.

[0056] Label the integrated dataset that has been analyzed and is yet to be analyzed as the target analysis data.

[0057] Preferably, the analytical data model building unit includes:

[0058] The model building module is used for:

[0059] Establish a power grid information model for each data point in the target analysis data;

[0060] The model for the basic data is established by using the mean, median, and variance of the basic data as features, and constructing a statistical model of the basic state of the power grid based on these features.

[0061] The model for the operational data is established by using the Kalman filter method to obtain the trend of operational data changes and then constructing a time series model of the operational data.

[0062] The monitoring data model is established by identifying signal features through Fourier transform and then performing signal processing on the monitoring data to build the model.

[0063] The load data model is established using regression analysis to predict the load data, and a load prediction model is established based on the load data.

[0064] The economic data model is established by using economic data obtained through cost-benefit analysis to create an economic evaluation model.

[0065] The management data model is established by applying fault tree analysis to the management data, and a risk assessment model is established for the management data.

[0066] The model for external data is established by using the Pearson correlation coefficient method to analyze the impact of external data, and a correlation analysis model for external data is established.

[0067] After model building, target analysis model data were obtained.

[0068] Preferably, the analytical data model building unit further includes:

[0069] The model feature data labeling module is used for:

[0070] Feature data extraction is performed on the target analysis model data;

[0071] The mean, median, and variance of the statistical model of the power grid's basic state are extracted.

[0072] Extract the trend characteristics of the running data in the time series model;

[0073] Extract the frequency and amplitude features of the signal identified in the signal processing model;

[0074] Extract load forecasting features with high weight values ​​from the load forecasting model;

[0075] Extract economic indicator data from the economic evaluation model;

[0076] Extract management indicator data from the risk assessment model;

[0077] Extract the data of the factors most correlated with power grid operation from the correlation analysis model;

[0078] Assign an identifier to each extracted feature data;

[0079] After the identifiers are assigned, the power grid model data is obtained.

[0080] Preferably, the scheduling instruction generation unit is further configured to:

[0081] Each feature data in the power grid model data is summarized to obtain a feature set;

[0082] The feature set is subject to risk assessment, which includes operational risk assessment, economic risk assessment and management risk assessment.

[0083] Before performing risk assessment on the feature set, each feature in the feature set is weighted.

[0084] After the weight allocation is completed, the operational risk assessment, economic risk assessment, and management risk assessment are carried out in sequence.

[0085] Among them, the operational risk assessment involves using logistic regression to assess the risk of operational data change trends in the time series model and signal frequency and amplitude characteristics in the signal processing model, and then obtaining the risk level of the operational risk assessment.

[0086] Economic risk assessment involves sampling and sensitivity analysis of economic indicator data from an economic assessment model to conduct risk assessment, and then obtaining the risk level of the economic risk assessment.

[0087] Management risk assessment involves using event tree analysis to evaluate the management indicator data in the risk assessment model, and then obtaining the risk level of the management risk assessment.

[0088] Preferably, the scheduling instruction generation unit is further configured to:

[0089] The risk levels in the operational risk assessment, economic risk assessment, and management risk assessment are classified into low risk, medium risk, and high risk.

[0090] After the risk classification is completed, the risk matrix will be constructed by cross-analyzing operational risk, economic risk and management risk, and the comprehensive risk level will be obtained after the cross-analysis.

[0091] Dispatch strategies are formulated based on the comprehensive risk level, and the dispatch strategies include adjustments for low risk, medium risk and high risk.

[0092] The data for low-risk protection is optimized based on the risk threshold; the data for medium-risk protection is adjusted based on the risk threshold; and the data for high-risk protection is used to implement power rationing measures based on the risk threshold.

[0093] The final scheduling scheme is obtained and transmitted to the corresponding control center according to the data type of the scheduling. The control center then schedules the power grid data according to the scheduling scheme.

[0094] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0095] 1. The power grid data analysis and scheduling system based on big data provided by this invention can more accurately predict power grid load and energy demand by comprehensively analyzing and integrating the dataset, thereby optimizing resource allocation. Different analysis methods are adopted for different types of data. Targeted analysis methods can more effectively extract and utilize information from the data, improving the accuracy and efficiency of the analysis.

[0096] 2. The power grid data analysis and scheduling system based on big data provided by this invention can more accurately capture the characteristics and changing patterns of various types of power grid data by establishing specialized models for different types of power grid data. The refined modeling improves the accuracy and reliability of data analysis. By extracting feature data and assigning identifiers to each model, it is possible to easily achieve rapid data retrieval and analysis. The assignment of identifiers ensures that each feature data has a clear source and meaning.

[0097] 3. The power grid data analysis and dispatching system based on big data provided by this invention adopts different assessment methods and models for different types of risks, thereby improving the accuracy and pertinence of risk assessment. According to the comprehensive risk level, adjustment strategies for different risk levels are formulated. This flexible and targeted dispatching strategy can better adapt to the actual needs of power grid operation. Attached Figure Description

[0098] Figure 1 This is a schematic diagram of the power grid data analysis and scheduling unit of the present invention;

[0099] Figure 2 This is a schematic diagram of the power grid data analysis and scheduling process of the present invention. Detailed Implementation

[0100] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0101] To address the problem that existing technologies fail to collect and analyze power grid data from multiple perspectives, thus hindering adjustments based on actual power grid data, please refer to [link to relevant documentation]. Figure 1 and Figure 2 This embodiment provides the following technical solution:

[0102] A power grid data analysis and dispatch system based on big data includes:

[0103] The power grid data acquisition unit is used for:

[0104] Power grid data is collected from the database, and the collected power grid data is distinguished. After the distinction is completed, a unique code is assigned to it, and the target power grid data is obtained after the unique code is assigned.

[0105] The data acquisition and analysis unit is used for:

[0106] The target power grid data is analyzed to obtain target analysis data.

[0107] The data model building unit is used for:

[0108] The target analysis data is used to establish a power grid information model, and the established power grid information model is labeled with feature data. After the feature data is labeled, the power grid model data is obtained.

[0109] The scheduling instruction generation unit is used for:

[0110] The power grid model data is used to generate a scheduling scheme, and the generated scheduling scheme is transmitted to different control centers. The control centers then schedule the power grid data according to the scheduling scheme.

[0111] Specifically, the power grid data acquisition unit can automatically adjust the power grid's operating status through real-time monitoring and analysis of data, thereby achieving optimized allocation and scheduling of power resources. The data acquisition and analysis unit employs different analysis methods for different types of data, enabling more effective extraction and utilization of information from the data and improving the accuracy and efficiency of the analysis. The data model building unit extracts and integrates feature data from multiple models, providing rich information support for power grid scheduling. The scheduling instruction generation unit can derive a comprehensive risk level, providing a more comprehensive and in-depth basis for the formulation of scheduling strategies.

[0112] The power grid data acquisition unit is also used for:

[0113] Retrieve power grid data from the database, which includes basic data, operational data, monitoring data, load data, economic data, management data, and external data;

[0114] The basic data, operational data, monitoring data, load data, economic data, management data, and external data are classified separately, resulting in independent basic data, operational data, monitoring data, load data, economic data, management data, and external data.

[0115] Each type of independent basic data, operational data, monitoring data, load data, economic data, management data, and external data shall be uniquely coded and labeled;

[0116] The unique coding structure includes a data category code, a data type code, a serial number code, and a timestamp. The data category code represents the specific data category, the data type code represents the data type under each data category, the serial number code represents different data records under the same data type, and the timestamp represents the collection time of different data.

[0117] The target power grid data is obtained by uniquely coding and labeling basic data, operational data, monitoring data, load data, economic data, management data, and external data.

[0118] Specifically, power grid data is subdivided into seven major categories: basic data, operational data, monitoring data, load data, economic data, management data, and external data. This makes the management of each type of data clearer and more organized. Each record under each data category is assigned a unique code, including a data category code, data type code, serial number, and timestamp. This greatly improves data traceability and management accuracy. Through unique coding, the system can quickly locate the required data record, improving data retrieval efficiency and supporting faster and more accurate power grid data analysis and scheduling. Unique coding ensures that every data record is traceable, which helps detect and prevent data tampering and misoperation, guaranteeing data authenticity and integrity. Different access permissions can be set according to different user roles and data categories to ensure data security and confidentiality. Through real-time monitoring and analysis of data, the operating status of the power grid can be automatically adjusted, achieving optimized allocation and scheduling of power resources and improving the stability and reliability of the power grid.

[0119] Specifically, the power grid data acquisition unit includes:

[0120] The real-time monitoring module for operating parameters is used to monitor the operating parameters for retrieving power grid data from the database in real time. The operating parameters include data integrity rate, data retrieval response time, and number of concurrent retrieval threads.

[0121] The first operational evaluation coefficient acquisition module is used to acquire a first operational evaluation coefficient using the data integrity rate and data retrieval response time; wherein, the first operational evaluation coefficient is acquired using the following formula:

[0122]

[0123] Among them, F 01 P represents the first performance evaluation coefficient; n represents the total number of data retrievals from the database; P i T represents the data integrity rate corresponding to the i-th data retrieval; i P represents the response time for the i-th data retrieval. b T represents the standard deviation of data integrity rate corresponding to n data retrievals; b T represents the standard deviation of the data retrieval response time for n data retrievals; c This indicates a preset reference value for the data retrieval response time;

[0124] The first coefficient comparison module is used to compare the first operation evaluation coefficient with a preset first operation coefficient threshold and obtain the comparison result.

[0125] The quality assessment module is used to assess the retrieval quality of power grid data from the database based on the comparison result between the first operation evaluation coefficient and the preset first operation coefficient threshold.

[0126] The technical effects of the above solution are as follows: The real-time monitoring module for operating parameters enables real-time monitoring of key operating parameters when the database retrieves power grid data, including data integrity rate, data retrieval response time, and the number of concurrent retrieval threads. These parameters are crucial for evaluating the database's operating status and performance. The first operating evaluation coefficient acquisition module calculates the first operating evaluation coefficient using a specific formula based on the data integrity rate and data retrieval response time. This coefficient considers not only the integrity and response speed of a single data retrieval but also the consistency and stability of multiple retrievals through standard deviations (Pb and Tb), as well as a comparison with a preset response time reference value (Tc), thus providing a comprehensive evaluation index. The first coefficient comparison module compares the calculated first operating evaluation coefficient with a preset first operating coefficient threshold to obtain a comparison result. This step provides a basis for subsequent optimization. If the evaluation coefficient is lower than the threshold, it indicates that there may be a problem with the performance of the database retrieving power grid data, requiring adjustment or optimization. The quality judgment module judges the operating quality of the database retrieving power grid data based on the comparison results. This judgment helps to promptly identify and resolve problems in the data retrieval process, thereby improving data quality and system stability. Overall, this technical solution achieves comprehensive management and optimization of the power grid data retrieval process through real-time monitoring, integrated evaluation, dynamic adjustment, and intelligent judgment. This helps improve the intelligence level of power grid data management and provides strong support for the safe, stable, and efficient operation of the power system.

[0127] In summary, this technical solution provides strong support for the intelligent, efficient, and stable management of power grid data by comprehensively evaluating the performance and quality of power grid data retrieval from the database.

[0128] Specifically, the quality determination module includes:

[0129] The first quality evaluation model extraction module is used to extract the comparison results between the first operation evaluation coefficient and the preset first operation coefficient threshold. If the first operation evaluation coefficient is lower than the preset first operation coefficient threshold, then the first quality evaluation model is retrieved.

[0130] The first-level second operation evaluation coefficient acquisition module is used to obtain the second operation evaluation coefficient by combining the first quality evaluation model with the first operation evaluation coefficient and the number of concurrent call threads.

[0131] The structure of the first quality evaluation model is as follows:

[0132]

[0133] Among them, F 02 F represents the second operational evaluation coefficient obtained from the first quality evaluation model. 01 F represents the first performance evaluation coefficient; n represents the total number of data retrievals from the database; F y T represents the preset first operating coefficient threshold; i T represents the response time for the i-th data retrieval; b M represents the standard deviation of the data retrieval response time for n data retrievals; i M represents the number of concurrent data retrieval threads for the i-th data retrieval; b This represents the standard deviation of the number of concurrent data retrieval threads corresponding to n data retrievals.

[0134] The second quality evaluation model extraction module is used to extract the comparison results between the first operation evaluation coefficient and the preset first operation coefficient threshold. If the first operation evaluation coefficient is not lower than the preset first operation coefficient threshold, then the second quality evaluation model is retrieved.

[0135] The Level 2 Second Operation Evaluation Coefficient Acquisition Module is used to obtain the Second Operation Evaluation Coefficient by combining the Second Quality Evaluation Model with the First Operation Evaluation Coefficient and the Number of Concurrent Calling Threads.

[0136] The structure of the second quality evaluation model is as follows:

[0137]

[0138] Among them, F 03 F represents the second operational evaluation coefficient obtained from the second quality evaluation model. 01 F represents the first performance evaluation coefficient; n represents the total number of data retrievals from the database; F y T represents the preset first operating coefficient threshold; n P represents the total latency rate corresponding to the response time of n data retrievals; n M represents the data integrity rate corresponding to n data retrievals; i M represents the number of concurrent data retrieval threads for the i-th data retrieval; i-1 M represents the number of concurrent data retrieval threads for the (i-1)th data retrieval; i-1 T represents the number of concurrent data retrieval threads for the (i-1)th data retrieval; ni T represents the data retrieval response time latency rate corresponding to the i-th data retrieval; ni-1 This represents the data retrieval response time latency rate corresponding to the (i-1)th data retrieval.

[0139] The second coefficient comparison module is used to compare the second operation evaluation coefficient with a preset second operation coefficient threshold.

[0140] The anomaly detection and alarm module is used to determine that the retrieval of power grid data from the database is abnormal when the second operation evaluation coefficient is lower than the preset second operation coefficient threshold, and to issue an early warning of the quality anomaly.

[0141] The technical effects of the above solution are as follows: By introducing a first quality evaluation model and a second quality evaluation model, this solution achieves refined evaluation of the quality of power grid data retrieval. Based on the comparison between the first operational evaluation coefficient and a preset threshold, a suitable quality evaluation model is selected, and the second operational evaluation coefficient is further calculated by combining parameters such as the number of concurrent retrieval threads. This evaluation method not only considers the efficiency of data retrieval but also takes into account the stability and quality of data retrieval. By comparing the second operational evaluation coefficient with a preset second operational coefficient threshold, potential problems in the process of retrieving power grid data from the database can be detected in a timely manner. When the second operational evaluation coefficient is lower than the threshold, the anomaly detection and alarm module will trigger a quality anomaly warning, prompting management personnel to make timely adjustments and optimizations. This dynamic adjustment mechanism helps ensure the stability and efficiency of the data retrieval process.

[0142] This solution effectively improves data quality and system stability by comprehensively considering multiple key parameters in the data retrieval process, such as data integrity rate, data retrieval response time, and the number of concurrent retrieval threads, and by introducing a quality evaluation model and threshold comparison mechanism. This helps ensure the accuracy and reliability of power grid data, providing strong support for the safe, stable, and efficient operation of the power system. The solution achieves intelligent management of the power grid data retrieval process. Through real-time monitoring, comprehensive evaluation, dynamic adjustment, and intelligent judgment, it enables comprehensive monitoring and optimization of data retrieval quality. This intelligent management approach not only improves management efficiency but also reduces the cost and risk of manual intervention. The solution is designed with a certain degree of scalability and flexibility. As the power system develops and needs change, the quality evaluation model, parameters, and thresholds can be adjusted and optimized according to actual conditions to adapt to new data retrieval requirements and challenges.

[0143] In summary, this technical solution, by introducing a quality evaluation model and a threshold comparison mechanism, achieves refined evaluation and dynamic adjustment of the quality of power grid data retrieval, effectively improving data quality and system stability, and providing strong support for the safe, stable, and efficient operation of the power system. Furthermore, this solution also possesses advantages such as intelligent management, scalability, and flexibility, providing a strong guarantee for the future development of the power system.

[0144] The data collection and analysis unit includes:

[0145] The data acquisition and processing module is used for:

[0146] The dataset is integrated from the basic data, operational data, monitoring data, load data, economic data, management data, and external data in the target power grid data.

[0147] After the dataset is integrated, a unified dataset is obtained, consisting of basic data, operational data, monitoring data, load data, economic data, management data, and external data.

[0148] The acquired integrated dataset will undergo data preprocessing;

[0149] Data preprocessing involves cleaning, transforming, and standardizing the integrated dataset in one step.

[0150] After data preprocessing, the integrated dataset to be analyzed is obtained.

[0151] Specifically, by integrating basic data, operational data, monitoring data, load data, economic data, management data, and external data from the target power grid, the comprehensiveness and integrity of the data are ensured. This comprehensive dataset facilitates a deeper understanding of the power grid's operational status, providing a solid foundation for data analysis. Through data preprocessing steps (including data cleaning, data transformation, and data standardization), errors, redundancy, and inconsistencies in the data can be eliminated, thereby improving the accuracy and reliability of the data. This is crucial for subsequent data analysis and decision-making. The integrated dataset reduces the number of data accesses and complexity during data analysis because all relevant data is centralized in one dataset. This helps improve the efficiency of data analysis, shortens analysis time, and enables a faster response to problems and challenges in power grid operation. By integrating multiple types of data, it provides a foundation for the future expansion of the system. As the power grid expands and technology continues to develop, new data types and sources can be easily added to meet ever-changing analytical needs. By comprehensively analyzing the integrated dataset, power grid load and energy demand can be predicted more accurately, thereby optimizing resource allocation. This helps reduce energy waste and improve the reliability and stability of the power grid.

[0152] The data processing and analysis module is used for:

[0153] Perform data analysis on the integrated dataset to be analyzed;

[0154] The basic data, operational data, monitoring data, load data, economic data, management data, and external data in the integrated dataset to be analyzed will be analyzed separately.

[0155] The analysis of basic data involves obtaining the mean, median, and variance through statistical methods; the analysis of operational data involves obtaining the changing trends through Kalman filtering; the analysis of monitoring data involves identifying monitoring signals through Fourier transform; the analysis of load data involves load forecasting through regression analysis; the analysis of economic data involves cost-benefit analysis of the power grid's economic benefits; the analysis of management data involves analyzing management data through fault tree analysis to optimize the allocation and use of power grid resources; and the analysis of external data involves analyzing the impact of weather, economic conditions, and policy changes on the power grid through Pearson correlation coefficient analysis.

[0156] Label the integrated dataset that has been analyzed and is yet to be analyzed as the target analysis data.

[0157] Specifically, different analytical methods are employed for different types of data. For example, statistical methods are used to analyze basic data, Kalman filtering is used to analyze operational data, and Fourier transform is used to analyze monitoring data. This targeted analytical approach can more effectively extract and utilize information from the data, improving the accuracy and efficiency of the analysis. Power grid data analysis requires multiple dimensions, including basic data, operational data, monitoring data, load data, economic data, management data, and external data. This comprehensive data analysis provides a more complete assessment of the power grid's operating status and performance, helping decision-makers make more accurate judgments. Load forecasting through regression analysis helps power grid dispatchers plan power supply in advance, ensuring the stable operation of the power grid. Simultaneously, cost-benefit economic analysis can assess the economic benefits of the power grid, providing decision support for power grid investment and operation. The application of fault tree analysis in management data can identify potential problems in power grid management, thereby optimizing the allocation and use of power grid resources. This helps reduce resource waste and improve the overall operational efficiency of the power grid. Analyzing the impact of weather, economic conditions, and policy changes on the power grid using the Pearson correlation coefficient method allows for consideration of the potential impact of external factors on power grid operation. This helps power grid dispatchers better respond to external changes, ensuring the stability and security of the power grid.

[0158] To address the problem in existing technologies where the acquisition of power grid data does not result in effective feature extraction, leading to insufficient data for subsequent analysis, please refer to [link to relevant documentation]. Figure 1 and Figure 2 This embodiment provides the following technical solution:

[0159] The data model building unit includes:

[0160] The model building module is used for:

[0161] Establish a power grid information model for each data point in the target analysis data;

[0162] The model for the basic data is established by using the mean, median, and variance of the basic data as features, and constructing a statistical model of the basic state of the power grid based on these features.

[0163] The model for the operational data is established by using the Kalman filter method to obtain the trend of operational data changes and then constructing a time series model of the operational data.

[0164] The monitoring data model is established by identifying signal features through Fourier transform and then performing signal processing on the monitoring data to build the model.

[0165] The load data model is established using regression analysis to predict the load data, and a load prediction model is established based on the load data.

[0166] The economic data model is established by using economic data obtained through cost-benefit analysis to create an economic evaluation model.

[0167] The management data model is established by applying fault tree analysis to the management data, and a risk assessment model is established for the management data.

[0168] The model for external data is established by using the Pearson correlation coefficient method to analyze the impact of external data, and a correlation analysis model for external data is established.

[0169] After model building, target analysis model data were obtained.

[0170] Specifically, by establishing specialized models for different types of power grid data (basic data, operational data, monitoring data, load data, economic data, management data, and external data), the characteristics and changing patterns of various data can be captured more accurately. Refined modeling improves the accuracy and reliability of data analysis, helping the power grid dispatching system make more scientific decisions. Different models, targeting different data types, can quickly extract key information, reducing the complexity and time cost of data processing. Kalman filtering analysis of operational data can accurately predict the changing trends of power grid operating status and promptly identify potential operational risks. Load forecasting models can predict future load demand, providing strong support for power grid dispatching and ensuring the stability and security of power grid operation. Risk assessment models can assess potential risks in power grid operation, providing risk warnings and response suggestions to reduce power grid operational risks. After different types of power grid data are modeled and analyzed using their respective models, data fusion and sharing can be achieved, improving data utilization and value.

[0171] The model feature data labeling module is used for:

[0172] Feature data extraction is performed on the target analysis model data;

[0173] The mean, median, and variance of the statistical model of the power grid's basic state are extracted.

[0174] Extract the trend characteristics of the running data in the time series model;

[0175] Extract the frequency and amplitude features of the signal identified in the signal processing model;

[0176] Extract load forecasting features with high weight values ​​from the load forecasting model;

[0177] Extract economic indicator data from the economic evaluation model;

[0178] Extract management indicator data from the risk assessment model;

[0179] Extract the data of the factors most correlated with power grid operation from the correlation analysis model;

[0180] Assign an identifier to each extracted feature data;

[0181] After the identifiers are assigned, the power grid model data is obtained.

[0182] Specifically, by extracting key feature data from various models, the operating status, development trend, economic efficiency, and risk level of the power grid can be comprehensively reflected. This covers multiple key aspects of the power grid data analysis and dispatch system, including the basic state of the power grid, time-series operating data, signal processing, load forecasting, economic assessment, risk assessment, and correlation analysis. The extracted feature data includes the mean, median, and variance in the statistical model; the changing trends of operating data in the time-series model; and the signal frequency and amplitude in the signal processing model. These are all accurate descriptions of the power grid's operating status. The extraction of high-weight load forecasting features, economic indicators, and management indicators helps to more accurately grasp the future development trend and potential risks of the power grid. By extracting and assigning identifiers to the feature data in each model, rapid data retrieval and analysis can be easily achieved. The assignment of identifiers ensures that each feature data has a clear source and meaning, which helps to enhance the interpretability of the data analysis results. By extracting and integrating feature data from multiple models, rich information support is provided for power grid dispatch.

[0183] To address the issue in existing technologies that fail to perform further risk analysis on the acquired power grid data, thus hindering targeted scheduling based on the risk profile of each power grid data point, please refer to [link to relevant documentation]. Figure 1 and Figure 2 This embodiment provides the following technical solution:

[0184] The scheduling instruction generation unit is also used for:

[0185] Each feature data in the power grid model data is summarized to obtain a feature set;

[0186] The feature set is subject to risk assessment, which includes operational risk assessment, economic risk assessment and management risk assessment.

[0187] Before performing risk assessment on the feature set, each feature in the feature set is weighted.

[0188] After the weight allocation is completed, the operational risk assessment, economic risk assessment, and management risk assessment are carried out in sequence.

[0189] Among them, the operational risk assessment involves using logistic regression to assess the risk of operational data change trends in the time series model and signal frequency and amplitude characteristics in the signal processing model, and then obtaining the risk level of the operational risk assessment.

[0190] Economic risk assessment involves sampling and sensitivity analysis of economic indicator data from an economic assessment model to conduct risk assessment, and then obtaining the risk level of the economic risk assessment.

[0191] Management risk assessment involves using event tree analysis to evaluate the management indicator data in the risk assessment model, and then obtaining the risk level of the management risk assessment.

[0192] The risk levels in the operational risk assessment, economic risk assessment, and management risk assessment are classified into low risk, medium risk, and high risk.

[0193] After the risk classification is completed, the risk matrix will be constructed by cross-analyzing operational risk, economic risk and management risk, and the comprehensive risk level will be obtained after the cross-analysis.

[0194] Dispatch strategies are formulated based on the comprehensive risk level, and the dispatch strategies include adjustments for low risk, medium risk and high risk.

[0195] The data for low-risk protection is optimized based on the risk threshold; the data for medium-risk protection is adjusted based on the risk threshold; and the data for high-risk protection is used to implement power rationing measures based on the risk threshold.

[0196] The final scheduling scheme is obtained and transmitted to the corresponding control center according to the data type of the scheduling. The control center then schedules the power grid data according to the scheduling scheme.

[0197] Specifically, multiple characteristic data points from the power grid model are aggregated into a feature set, which is then used for multi-dimensional assessment of operational, economic, and management risks. This comprehensive and systematic analysis approach more accurately reflects the actual operating status and potential risks of the power grid. For different types of risks (operational, economic, and management risks), the scheme employs different assessment methods and models (such as logistic regression, sensitivity analysis, and event tree analysis), thereby improving the accuracy and relevance of risk assessment. Before conducting the risk assessment, the scheme assigns weights to each feature in the feature set. This weighting method reflects the importance of different features in the risk assessment, thus improving the rationality of the assessment results. Risk levels are divided into low, medium, and high risks. This clear division helps decision-makers better understand the risk situation and adopt corresponding dispatch strategies. By constructing a risk matrix for cross-analysis, the scheme can derive a comprehensive risk level, providing a more comprehensive and in-depth basis for the formulation of dispatch strategies. Based on the comprehensive risk level, the scheme formulates adjustment strategies for different risk levels, including data optimization for low-risk risks, parameter adjustment for medium-risk risks, and power curtailment measures for high-risk risks. This flexible and targeted scheduling strategy can better adapt to the actual needs of power grid operation. Ultimately, the scheduling plan is transmitted to the corresponding control center based on the data type, and the control center then schedules the power grid data according to the plan. This data transmission and control method enables rapid information transmission and efficient execution, improving the real-time performance and accuracy of power grid scheduling.

[0198] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0199] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

Claims

1. A power grid data analysis and dispatch system based on big data, characterized in that, The utility model relates to an electric network data acquisition unit, a data analysis unit, an analysis data model establishment unit and a dispatching instruction generation unit are arranged in the electric network data acquisition unit, and the data analysis unit, the analysis data model establishment unit and the dispatching instruction generation unit are connected with each other. The electric network data acquisition unit comprises: A running parameter real-time monitoring module is arranged in the electric network data acquisition unit, and the running parameter real-time monitoring module is used for monitoring the data integrity rate, the data acquisition response time and the concurrent acquisition thread quantity of the database. A first running evaluation coefficient acquisition module is arranged in the electric network data acquisition unit, and the first running evaluation coefficient acquisition module is used for acquiring the first running evaluation coefficient by using the data integrity rate and the data acquisition response time. A first coefficient comparison module is arranged in the electric network data acquisition unit, and the first coefficient comparison module is used for comparing the first running evaluation coefficient with the preset first running coefficient threshold value to acquire a comparison result. A quality determination module is arranged in the electric network data acquisition unit, and the quality determination module is used for determining the acquisition running quality of the database according to the comparison result between the first running evaluation coefficient and the preset first running coefficient threshold value. The quality determination module comprises: A first quality evaluation model extraction module is arranged in the quality determination module, and the first quality evaluation model extraction module is used for extracting the comparison result between the first running evaluation coefficient and the preset first running coefficient threshold value. A first-level second running evaluation coefficient acquisition module is arranged in the quality determination module, and the first-level second running evaluation coefficient acquisition module is used for acquiring the second running evaluation coefficient by using the first quality evaluation model, the first running evaluation coefficient and the concurrent acquisition thread quantity. The structure of the first quality evaluation model is as follows: A second quality evaluation model extraction module is arranged in the quality determination module, and the second quality evaluation model extraction module is used for extracting the comparison result between the first running evaluation coefficient and the preset first running coefficient threshold value. A second-level second running evaluation coefficient acquisition module is arranged in the quality determination module, and the second-level second running evaluation coefficient acquisition module is used for acquiring the second running evaluation coefficient by using the second quality evaluation model, the first running evaluation coefficient and the concurrent acquisition thread quantity. ； Wherein, F 01 represents the first running evaluation coefficient; n represents the total number of data retrieval from the database; P i represents the data integrity rate corresponding to the i-th data retrieval; T i represents the data retrieval response time corresponding to the i-th data retrieval; P b represents the data integrity rate standard deviation corresponding to n times of data retrieval; T b represents the data retrieval response time standard deviation corresponding to n times of data retrieval; T c represents the preset data retrieval response time reference value; The structure of the second quality evaluation model is as follows: A second coefficient comparison module is arranged in the quality determination module, and the second coefficient comparison module is used for comparing the second running evaluation coefficient with the preset second running coefficient threshold value. An abnormality determination and alarm module is arranged in the quality determination module, and the abnormality determination and alarm module is used for determining that the acquisition running quality of the database is abnormal and performing quality abnormality early warning when the second running evaluation coefficient is lower than the preset second running coefficient threshold value. The electric network data acquisition unit is further used for: ​ ​ ； Wherein, F 02 represents the second operation evaluation coefficient obtained by the first quality evaluation model; F 01 represents the first operation evaluation coefficient; n represents the total number of data retrieval from the database; F y represents the preset first operation coefficient threshold; T i represents the data retrieval response time corresponding to the ith data retrieval; T b represents the data retrieval response time standard deviation corresponding to n times of data retrieval; M i represents the number of concurrent retrieval threads corresponding to the ith data retrieval; M b represents the number of concurrent retrieval threads standard deviation corresponding to n times of data retrieval; ​ ​ ​ ； wherein, F 03 represents a second operation evaluation coefficient obtained by the second quality evaluation model; F 01 represents the first operation evaluation coefficient; n represents the total number of data retrieval from the database; F y represents a preset first operation coefficient threshold; T n represents the total delay rate of the data retrieval response time length corresponding to the n times of data retrieval; P n represents the data integrity rate corresponding to the n times of data retrieval; M i represents the number of concurrent retrieval threads corresponding to the i times of data retrieval; M i-1 represents the number of concurrent retrieval threads corresponding to the i-1 times of data retrieval; M i-1 represents the number of concurrent retrieval threads corresponding to the i-1 times of data retrieval; T ni represents the data retrieval response time length delay rate corresponding to the i times of data retrieval; T ni-1 represents the data retrieval response time length delay rate corresponding to the i-1 times of data retrieval; ​ ​ 2. The big data based power grid data analysis dispatch system of claim 1, wherein, ​ Retrieve power grid data from the database, which includes basic data, operational data, monitoring data, load data, economic data, management data, and external data; The basic data, operational data, monitoring data, load data, economic data, management data, and external data are classified separately, resulting in independent basic data, operational data, monitoring data, load data, economic data, management data, and external data. Each type of independent basic data, operational data, monitoring data, load data, economic data, management data, and external data shall be uniquely coded and labeled; The unique coding structure includes a data category code, a data type code, a serial number code, and a timestamp. The data category code represents the specific data category, the data type code represents the data type under each data category, the serial number code represents different data records under the same data type, and the timestamp represents the collection time of different data. The target power grid data is obtained by uniquely coding and labeling basic data, operational data, monitoring data, load data, economic data, management data, and external data.

3. The big data based power grid data analytics dispatch system of claim 2, wherein, The data collection and analysis unit includes: The data acquisition and processing module is used for: The dataset is integrated from the basic data, operational data, monitoring data, load data, economic data, management data, and external data in the target power grid data. After the dataset is integrated, a unified dataset is obtained, consisting of basic data, operational data, monitoring data, load data, economic data, management data, and external data. The acquired integrated dataset will undergo data preprocessing; Data preprocessing involves cleaning, transforming, and standardizing the integrated dataset in one step. After data preprocessing, the integrated dataset to be analyzed is obtained.

4. The big data based power grid data analytics dispatch system of claim 3, wherein, The data collection and analysis unit also includes; The data processing and analysis module is used for: Perform data analysis on the integrated dataset to be analyzed; The basic data, operational data, monitoring data, load data, economic data, management data, and external data in the integrated dataset to be analyzed will be analyzed separately. The analysis of basic data involves obtaining the mean, median, and variance through statistical methods; the analysis of operational data involves obtaining the changing trends through Kalman filtering; the analysis of monitoring data involves identifying monitoring signals through Fourier transform; the analysis of load data involves load forecasting through regression analysis; the analysis of economic data involves cost-benefit analysis of the power grid's economic benefits; the analysis of management data involves analyzing management data through fault tree analysis to optimize the allocation and use of power grid resources; and the analysis of external data involves analyzing the impact of weather, economic conditions, and policy changes on the power grid through Pearson correlation coefficient analysis. Label the integrated dataset that has been analyzed and is yet to be analyzed as the target analysis data.

5. The big data based power grid data analytics dispatch system of claim 4, wherein, The analytical data model building unit includes: The model building module is used for: Establish a power grid information model for each data point in the target analysis data; The model for the basic data is established by using the mean, median, and variance of the basic data as features, and constructing a statistical model of the basic state of the power grid based on these features. The model establishment of the operation data is a change trend of the operation data obtained by using the Kalman filtering method, and the operation data is subjected to time series model establishment; The model establishment of the monitoring data is a signal feature identified by the Fourier transform method, and the monitoring data is subjected to signal processing model establishment; The model establishment of the load data is load data predicted by using regression analysis, and the load data is subjected to load prediction model establishment; The model establishment of the economic data is economic data obtained by cost-benefit analysis, and the economic model is subjected to economic evaluation model establishment; The model establishment of the management data is management data analyzed by using the fault tree analysis method, and the management data is subjected to risk assessment model establishment; The model establishment of the external data is an influence of the external data analyzed by using the Pearson correlation coefficient method, and the external data is subjected to correlation analysis model establishment; The target analysis model data are obtained after the model establishment.

6. The big data based power grid data analytics dispatch system of claim 5, wherein, The analysis data model establishment unit further comprises: The model feature data marking module is configured to: extract feature data from the target analysis model data; extract the mean value, the median value and the variance in the statistical model of the power grid basic state; extract the operation data change trend feature in the time series model; extract the signal frequency and amplitude features identified in the signal processing model; extract the load prediction feature with a high weight value in the load prediction model; extract the economic index data in the economic evaluation model; extract the management index data in the risk assessment model; extract the factor data with the highest correlation with the power grid operation in the correlation analysis model; assign an identifier to each extracted feature data; the power grid model data are obtained after the identifier assignment is completed.

7. The big data based power grid data analytics dispatch system of claim 6, wherein, The dispatching instruction generation unit is further configured to: aggregate each feature data in the power grid model data, and obtain a feature set after the aggregation; perform risk assessment on the feature set, wherein the risk assessment comprises operation risk assessment, economic risk assessment and management risk assessment; perform weight assignment on each feature in the feature set before the risk assessment on the feature set; perform the operation risk assessment, the economic risk assessment and the management risk assessment in sequence after the weight assignment is completed; the operation risk assessment is a risk assessment on the operation data change trend feature in the time series model and the signal frequency and amplitude features in the signal processing model by using the logistic regression, and a risk level of the operation risk assessment is obtained after the assessment; the economic risk assessment is a risk assessment on the economic index data in the economic evaluation model by using the sampling sensitivity analysis, and a risk level of the economic risk assessment is obtained after the assessment; the management risk assessment is a risk assessment on the management index data in the risk assessment model by using the event tree analysis, and a risk level of the management risk assessment is obtained after the assessment.

8. The big data based power grid data analytics dispatch system of claim 7, wherein, The dispatching instruction generation unit is further configured to: perform level division on the risk levels in the risk level of the operation risk assessment, the risk level of the economic risk assessment and the risk level of the management risk assessment, and the level division is low risk, medium risk and high risk; construct a risk matrix after the level division is completed, and cross analyze the operation risk, the economic risk and the management risk, and obtain a comprehensive risk level after the cross analysis. According to the comprehensive risk level, a scheduling strategy is made, which includes low-risk, medium-risk and high-risk adjustment; The data contained in the low-risk is optimized according to the risk threshold, the data contained in the medium-risk is adjusted according to the risk threshold, and the data protected in the high-risk is implemented by power limiting measures according to the risk threshold; Finally, a completed scheduling scheme is obtained, and the scheduling scheme is transmitted to the corresponding control center according to the type of the scheduled data, and the control center schedules the power grid data according to the scheduling scheme.

Images