Pumped storage power station data fusion intelligent analysis platform

Abstract

The application provides a pumped storage power station data fusion intelligent analysis platform, comprising: a data source comprising a plurality of original data; the data center is configured to collect original data and convert the original data to obtain unified data; the data application comprises an algorithm middle table module, and the algorithm middle table module is configured to construct an operator library, associate corresponding measurement attributes with measurement point identifiers of unified data and form a matching table; obtaining an algorithm construction instruction; according to the algorithm construction instruction, the operator library and the measurement attributes, performing algorithm modeling through a drag-and-drop operation process to obtain a power station data analysis algorithm; instantiating the power station data analysis algorithm to obtain an instantiated algorithm; and processing the corresponding unified data according to the instantiation algorithm and the matching table to generate an analysis result data set. The pumped storage power station data fusion intelligent analysis platform provided by the invention can realize multi-source heterogeneous data integration and analysis, and is high in efficiency, low in cost and good in real-time performance and accuracy.

Description

Technical Field

[0001] This application relates to the field of pumped storage power station technology, and in particular to a data fusion intelligent analysis platform for pumped storage power stations. Background Technology

[0002] Pumped storage hydroelectric power utilizes electricity generated during periods of low electricity demand to pump water into an upper reservoir. During periods of high electricity demand, the water stored in the upper reservoir is released into a lower reservoir to generate electricity. It is also known as energy storage power generation. It can convert excess electricity during periods of low grid load into high-value electricity during periods of high grid load. It is suitable for frequency and phase regulation, can be used to stabilize the frequency and voltage of the power system, is suitable for emergency backup, and can also improve the efficiency of thermal power plants and nuclear power plants in the power system.

[0003] With the rapid development of information technology, the demand for data analysis and processing is increasing across all industries. Traditional operation and maintenance models for pumped storage power stations suffer from problems such as reliance on significant manpower and resources, complex and isolated data sources, and redundant analysis algorithms. Therefore, there is an urgent need for a platform capable of efficiently integrating various types of data for intelligent analysis to support decision-making and equipment management. Summary of the Invention

[0004] In view of this, the purpose of this application is to propose a data fusion intelligent analysis platform for pumped storage power stations to solve the above-mentioned technical problems.

[0005] This application provides a data fusion intelligent analysis platform for pumped storage power stations, comprising: a data source, including various raw data from the pumped storage power station; a data center, configured to collect the raw data, perform time-series alignment and protocol conversion on the raw data to obtain unified data, and store the unified data; and a data application, configured to access the data center to obtain the unified data. The data application includes an algorithm platform module, configured to construct an operator library, associate measurement point identifiers of the unified data with corresponding measurement attributes and form a matching table; obtain algorithm construction instructions; perform algorithm modeling through a drag-and-drop operation process based on the algorithm construction instructions, the operator library, and the measurement attributes to obtain a power station data analysis algorithm; instantiate the power station data analysis algorithm to obtain an instantiated algorithm; and process the corresponding unified data according to the instantiated algorithm and the matching table to generate an analysis result dataset.

[0006] Furthermore, the operator library includes a start operator, an end operator, a time-stamped time operator, a time interval operator, and a threshold alarm operator; the measurement attributes include multiple unit ball valve opening command attributes and multiple unit ball valve fully open result attributes; the algorithm construction instruction is a multiple unit ball valve opening time analysis algorithm construction instruction; the algorithm modeling to obtain the power plant data analysis algorithm through drag-and-drop operation includes: dragging the start operator to add the multiple unit ball valve opening command attributes and the multiple unit ball valve fully open result attributes to the node; dragging the time-stamped time operator to connect with the start operator, selecting the start condition as multiple unit ball valve opening and the end condition as multiple unit ball valve fully open in the node; dragging the time interval operator to connect with the time-stamped time operator, selecting the time range as time-stamped time in the node; dragging the threshold alarm operator to connect with the time interval operator, configuring the alarm threshold in the node; and dragging the end operator to connect with the threshold alarm operator to obtain the multiple unit ball valve opening time analysis algorithm.

[0007] Further, the step of processing the corresponding unified data according to the instantiation algorithm and the matching table to generate the analysis result dataset includes: parsing the instantiation algorithm, obtaining the measurement point identifiers with the measurement attributes of the multiple units' ball valve opening command attribute and the multiple units' ball valve fully open result attribute according to the matching table, selecting the corresponding unified data to form a data set according to the measurement point identifiers; calculating a time segment set using the time-scaled time operator on the data set; processing the time segment set using the time interval operator to obtain the analysis result dataset; and issuing a first alarm message in response to the data value in the analysis result dataset exceeding the alarm threshold.

[0008] Furthermore, configuring alarm thresholds in nodes includes configuring a uniform alarm threshold for all units and / or configuring a corresponding alarm threshold for each unit.

[0009] Further, the instantiation of the power plant data analysis algorithm to obtain an instantiated algorithm includes: instantiating the power plant data analysis algorithm in parallel for each unit of the pumped storage power station to obtain the instantiated algorithm corresponding to each unit; the processing of the corresponding unified data to generate an analysis result dataset based on the instantiated algorithm and the matching table includes: processing the unified data of the corresponding units based on the instantiated algorithm and the matching table to generate the corresponding analysis result dataset; the algorithm platform module is also configured to generate corresponding first trend change curves in the same coordinate system based on the analysis result dataset corresponding to each unit.

[0010] Furthermore, the data application also includes a curve analysis module, which is configured to acquire set parameters and generate a second trend change curve based on the set parameters and the corresponding unified data.

[0011] Furthermore, the data application also includes a health analysis module, which is configured to generate a health analysis report based on a preset period, monitoring indicators, and the unified data.

[0012] Furthermore, the data application also includes a monitoring and alarm module and a system monitoring module. The monitoring and alarm module is configured to comprehensively monitor the unified data and issue a second alarm message in response to data anomalies. The system monitoring module is configured to monitor the pumped storage power station data fusion intelligent analysis platform and issue a third alarm message in response to platform anomalies.

[0013] Furthermore, the pumped storage power station is equipped with multiple cameras, which are used to acquire corresponding unit operation images; the data application also includes a system linkage module, which is configured to associate the work order information of the pumped storage power station's unit equipment with the corresponding unit operation images.

[0014] Furthermore, the data application also includes a special equipment module and a cockpit module. The special equipment module is configured to independently establish equipment files and monitor the special equipment in the pumped storage power station. The cockpit module is configured to display the daily production report of the pumped storage power station.

[0015] As can be seen from the above, this application provides a data fusion intelligent analysis platform for pumped storage power stations, including: a data source, including various raw data from the pumped storage power station; a data center, configured to collect raw data, perform time-series alignment and protocol conversion on the raw data to obtain unified data, and store the unified data; and a data application, configured to access the data center to obtain the unified data. The data application includes an algorithm platform module, which is configured to build an operator library, associate the measurement point identifiers of the unified data with corresponding measurement attributes and form a matching table; obtain algorithm construction instructions; based on the algorithm construction instructions, the operator library, and the measurement attributes, perform algorithm modeling through a drag-and-drop operation process to obtain a power station data analysis algorithm; instantiate the power station data analysis algorithm to obtain an instantiated algorithm; and process the corresponding unified data according to the instantiated algorithm and the matching table to generate an analysis result dataset. By collecting various raw data sources, performing time-series alignment and protocol conversion, standardized unified data is obtained. This achieves the integration and unification of multi-source heterogeneous data, facilitating subsequent data processing and analysis. Through the construction of a unified operator library and drag-and-drop operation workflow for modeling various analysis algorithms, no-code rapid modeling is possible, and a single model can be reused across multiple units, significantly improving development efficiency and resource utilization while reducing maintenance costs. By constructing a matching table that associates measurement point identifiers with measurement attributes, data can be quickly matched for analysis when applying algorithms, improving analysis efficiency and ensuring algorithm reusability. This pumped storage power station data fusion intelligent analysis platform can achieve multi-source heterogeneous data integration and analysis with high efficiency, low cost, and good real-time performance and accuracy. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in this application or related technologies, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the architecture of a data fusion intelligent analysis platform for a pumped storage power station according to an embodiment of this application; Figure 2 This is a business process diagram of the algorithm platform module in the embodiments of this application; Figure 3 This is a flowchart of the curve analysis module in an embodiment of this application; Figure 4 This is a flowchart illustrating the business process for generating monthly analysis reports in the health analysis module of this application embodiment; Figure 5 This is a flowchart of the mid-link monitoring service process of the system monitoring module in this application embodiment; Figure 6 This is a schematic diagram of the first trend change curve in an embodiment of this application; Figure 7 This is a schematic diagram of the second trend change curve in the embodiments of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with specific embodiments and the accompanying drawings.

[0019] It should be noted that, unless otherwise defined, the technical or scientific terms used in the embodiments of this application should have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect.

[0020] Traditional operation and maintenance of pumped-storage power stations relies on manual inspections and record-keeping, which is costly, inefficient, prone to loss, and lacks data integrity. Furthermore, the multi-source heterogeneous data acquired by computer monitoring systems, production management systems, and 500kV monitoring systems is scattered, resulting in severe data silos and fragmented equipment status analysis. There is a lack of an efficient big data analysis platform. One approach is to collect and transform the multi-source heterogeneous data into unified data, enabling centralized analysis and horizontal comparisons. However, the analytical algorithms corresponding to large amounts of data are complex. Editing and storing code for each algorithm separately would consume enormous space and reduce analysis efficiency.

[0021] The following describes specific embodiments in conjunction with... Figures 1 to 7 The technical solution of this application will be described in detail below.

[0022] Some embodiments of this application provide a data fusion intelligent analysis platform for pumped storage power stations, such as... Figure 1As shown, it includes: a data source, including various raw data from pumped storage power stations; a data center, configured to collect the raw data, perform time-series alignment and protocol conversion on the raw data to obtain unified data, and store the unified data; and a data application, configured to access the data center to obtain the unified data. The data application includes an algorithm platform module, which is configured to build an operator library, associate the measurement point identifiers of the unified data with corresponding measurement attributes and form a matching table; obtain algorithm construction instructions; perform algorithm modeling through a drag-and-drop operation process based on the algorithm construction instructions, the operator library, and the measurement attributes to obtain a power station data analysis algorithm; instantiate the power station data analysis algorithm to obtain an instantiated algorithm; and process the corresponding unified data according to the instantiated algorithm and the matching table to generate an analysis result dataset.

[0023] like Figure 1 As shown, pumped storage power stations involve massive amounts of heterogeneous data, such as raw data from computer monitoring, industrial television, plant power temperature measurement, online monitoring, production management, and key raw data during the infrastructure construction phase. The data sources cover computer monitoring systems, production management systems, and 500kV monitoring systems. The data protocols for these various raw data sources may differ, including protocols such as 104 and HTTP.

[0024] The data center comprises a data service module, a data acquisition module, a data processing module, and a data warehouse module. The data acquisition module establishes connections with data sources to collect raw data. The data processing module performs time-series alignment and protocol conversion on the raw data to obtain unified data. Time-series alignment uses algorithms to synchronize timestamps, fill in missing values, and smooth outliers, ensuring consistency across different sources or time periods. Protocol conversion uses algorithms to transform data from different protocols to the same protocol, such as converting all data to HTTP, thus achieving the integration and governance of heterogeneous data across regions. The data warehouse module stores the unified data. The data service module provides a unified data access method, supporting subsequent intelligent analysis applications such as curve analysis, algorithm platforms, health analysis, and monitoring and alerting.

[0025] Data applications can access data centers to obtain unified data for processing and analysis. Data applications can include various processing modules, such as curve analysis modules, algorithm platform modules, health analysis modules, monitoring and alarm modules, system linkage modules, special equipment modules, cockpit modules, system monitoring modules, etc.

[0026] For example, the business process of the algorithm middleware module Figure 2As shown, the system categories of the algorithm are classified, the measurement attributes are divided into the corresponding system categories, and then the unified data is associated with the measurement attributes. The model is modeled using a drag-and-drop algorithm modeling tool. After the model is instantiated, the calculation results are generated through calculation tasks. The alarm calculation determines whether to issue an alarm notification. The algorithm results are compared and analyzed horizontally. At the same time, custom measurement points can be selected to view trend analysis, such as the relationship between ball valve opening time and external temperature, so as to realize the intelligent upgrade of equipment management.

[0027] The algorithm platform module, based on no-code visual orchestration technology, builds an extensible operator library. This library includes various processing operators, such as start operators, end operators, maximum / minimum value operators, time offset calculation operators, time segmentation operators, constant addition / subtraction operators, and exponent calculation operators. Common mathematical algorithms can be encapsulated into operators using the Java language, and each operator provides the foundation for subsequent algorithm construction.

[0028] The algorithm platform module associates the measurement point identifiers of unified data with corresponding measurement attributes and forms a matching table. The measurement point identifier is the identity of the measurement location of the unified data; each piece of unified data has a unique measurement point identifier. For example, if the unified data is the measurement data of the ball valve of Unit 1 being fully open, its corresponding measurement point identifier is "QFQK-1". Measurement attributes are the specific categories corresponding to the unified data and can be pre-stored in the algorithm platform module. One measurement attribute can correspond to multiple measurement point identifiers. For example, if the measurement attribute is "ball valves of Units 1-10 fully open", then the corresponding measurement point identifiers can include "QFQK-1", "QFQK-2", ..., "QFQK-10", without specific limitations. This allows subsequent algorithm construction to directly utilize measurement attributes without needing to select specific measurement data, improving the algorithm's versatility and reducing its computational cost. Furthermore, for cases with numerous measurement attributes, a higher-level system classification can be designed to simplify operations. For example, if the system classification includes a "ball valve system classification", then the corresponding measurement attribute for the "ball valve system classification" can include "multiple units ball valve fully open result attributes".

[0029] The algorithm construction instruction is an instruction for constructing specific power plant data analysis algorithms. Power plant data analysis algorithms include, for example, "multiple unit ball valve opening time analysis algorithm", "generator lift analysis algorithm during power generation start-up", and "water guide swing analysis algorithm during pumping start-up".

[0030] Based on algorithm construction instructions, an operator library, and measurement attributes, a power plant data analysis algorithm is generated through a drag-and-drop workflow. Algorithm construction instructions can include the required operators and measurement attributes, or a pre-built mapping table of algorithm construction instructions and corresponding operators and measurement attributes can be used. Following the algorithm construction instructions, appropriate operators are selected from the operator library using a visual drag-and-drop tool (such as Power BI) and connected. The corresponding measurement attributes are then configured for the operators, thus forming the power plant data analysis algorithm.

[0031] Instantiation of power plant data analysis algorithms results in instantiated algorithms. Instantiation refers to the process of concretizing abstract algorithm concepts into executable programs or data structures, making the algorithm suitable for actual application scenarios. The scenarios support core equipment of pumped storage power plants, such as generators, turbines, ball valves, governors, technical water supply, main transformers, busbar tunnel equipment, excitation, and computer monitoring systems. Instantiation can be achieved through methods such as serialization, deserialization, and the new statement, without any specific limitations.

[0032] By looking up the matching table based on the measurement attributes in the instantiation algorithm, the corresponding measurement point identifier can be determined. Based on the measurement point identifier, the corresponding unified data can be determined. After the unified data is processed by the various operators of the instantiation algorithm, the final analysis result dataset can be obtained, realizing intelligent and efficient data analysis and management.

[0033] By collecting various raw data and performing time-series alignment and protocol conversion, standardized unified data is obtained. This achieves the integration and unification of multi-source heterogeneous data, facilitating subsequent data processing and analysis. By building a unified operator library and drag-and-drop operation process for modeling various analysis algorithms, code-free rapid modeling can be achieved, and multiple units can reuse the model once, greatly improving development efficiency and resource utilization, and reducing maintenance costs. By building a matching table that associates measurement point identifiers with measurement attributes of unified data, data can be quickly matched for data analysis when applying algorithms, improving analysis efficiency.

[0034] This pumped storage power station data fusion intelligent analysis platform can realize the integration and analysis of multi-source heterogeneous data, with high efficiency, low cost, and good real-time performance and accuracy.

[0035] In some embodiments, the operator library includes a start operator, an end operator, a time-stamped operator, a time interval operator, and a threshold alarm operator; the measurement attributes include multiple unit ball valve opening command attributes and multiple unit ball valve fully open result attributes; the algorithm construction instruction is a multiple unit ball valve opening time analysis algorithm construction instruction; the process of obtaining a power plant data analysis algorithm through drag-and-drop operation includes: dragging the start operator to add the multiple unit ball valve opening command attributes and the multiple unit ball valve fully open result attributes to the node; dragging the time-stamped operator to connect with the start operator, selecting the start condition as multiple unit ball valve opening and the end condition as multiple unit ball valve fully open in the node; dragging the time interval operator to connect with the time-stamped operator, selecting the time range as time-stamped time in the node; dragging the threshold alarm operator to connect with the time interval operator, configuring the alarm threshold in the node; and dragging the end operator to connect with the threshold alarm operator to obtain the multiple unit ball valve opening time analysis algorithm.

[0036] Specifically, if the algorithm construction instruction is "Analysis Algorithm for Ball Valve Opening Time of Units 1-10", then the algorithm platform module will create the analysis algorithm for ball valve opening time of Units 1-10. In the system classification, it will create a category called "Ball Valve System". In the measurement attributes under this system category, it needs to create the instruction attribute "Remote opening ball valve (body) of Units 1-10" and the result attribute "Ball valve fully open of Units 1-10". It will also associate all the measurement point identifiers (IDs) of the remote ball valves (body) of Units 1-10 in the unified data with the corresponding measurement attributes, and associate all the measurement point identifiers (IDs) of the switch quantity of fully open ball valves of Units 1-10 in the unified data with the corresponding measurement attributes, thus constructing a "Measurement Point Identifier - Measurement Attribute - System Classification" matching table.

[0037] The specific algorithm modeling process includes: dragging the start operator, adding the instruction attribute "Remotely open ball valves (body) for Units 1-10" and the result attribute "Ball valves fully open for Units 1-10" in the node; dragging the time-stamped operator, connecting the start operator and the time-stamped operator with a line, selecting "Remotely open ball valves (body) for Units 1-10" as the start condition and setting its value to 1, and "Ball valves fully open for Units 1-10" as the end condition and setting its value to 1; dragging the time interval operator, connecting the time-stamped operator and the time interval operator with a line, and selecting the time-stamped time in the node's time range; dragging the threshold alarm operator, connecting the time interval operator and the threshold alarm interval operator with a line, and configuring the alarm threshold in the node; and dragging the end operator, connecting the threshold alarm operator and the end operator with a line, saving the algorithm model, and obtaining the ball valve opening time analysis algorithm for Units 1-10. The construction process is simple and efficient.

[0038] In some embodiments, the step of processing the corresponding unified data according to the instantiation algorithm and the matching table to generate an analysis result dataset includes: parsing the instantiation algorithm; obtaining measurement point identifiers with measurement attributes of the multiple units' ball valve opening command attribute and the multiple units' ball valve fully open result attribute according to the matching table; selecting the corresponding unified data to form a data set according to the measurement point identifiers; calculating a time segment set using the time-scaled time operator on the data set; processing the time segment set using the time interval operator to obtain the analysis result dataset; and issuing a first alarm message in response to a data value in the analysis result dataset exceeding the alarm threshold.

[0039] After algorithm instantiation, task calculations are performed. The task execution cycle and time range can be set. Based on the algorithm instantiation calculations, the final analysis result dataset is generated to support user decision-making. Specifically, the parsed instantiated algorithm determines measurement attributes, and based on these attributes and a matching table, corresponding unified data is determined to form a data set. The data set is processed using a time-stamped time operator to obtain a set of time segments. The time segment set data is processed by a time interval operator to obtain the time data for each time segment interval, ultimately generating a precise analysis result dataset. A threshold alarm operator judges the data values ​​in the analysis result dataset. When the algorithm determines a potential problem, the system will automatically trigger an alarm process and issue the first alarm message, such as "Unit 1 ball valve opening time exceeded." The first alarm message will be simultaneously pushed to the maintenance personnel's intranet office platform and external mobile terminal, ensuring that maintenance personnel can receive and quickly handle problems whether they are in the office or traveling, ensuring stable equipment operation.

[0040] In some embodiments, configuring alarm thresholds in nodes includes configuring uniform alarm thresholds for all units and / or configuring corresponding alarm thresholds for each unit.

[0041] A uniform alarm threshold can be configured for all units, for example, the alarm threshold for the ball valve opening time of units 1-10 can all be 60 seconds. Alternatively, a separate alarm threshold can be configured for each unit, for example, the alarm threshold for the ball valve opening time of unit 1 can be 50 seconds, and the alarm threshold for the ball valve opening time of unit 2 can be 60 seconds, etc. A dual threshold early warning mechanism has been established, which is more in line with practical applications. When an anomaly is triggered, alarm information is automatically pushed to the person responsible for the equipment.

[0042] In some embodiments, instantiating the power plant data analysis algorithm to obtain an instantiated algorithm includes: instantiating the power plant data analysis algorithm in parallel for each unit of the pumped storage power plant to obtain an instantiated algorithm corresponding to each unit; processing the corresponding unified data according to the instantiated algorithm and the matching table to generate an analysis result dataset includes: processing the unified data of the corresponding unit according to the instantiated algorithm and the matching table to generate the corresponding analysis result dataset; the algorithm platform module is further configured to generate corresponding first trend change curves in the same coordinate system according to the analysis result datasets corresponding to each unit.

[0043] In the algorithm model instantiation process, the single-threaded operation mode is broken through, and parallel algorithm instantiation of multiple units is adopted. This allows each unit's algorithm container to run independently with data isolation, achieving the characteristic of one-time algorithm modeling and multi-unit reuse. Correspondingly, the unified data of the corresponding units is processed by each instantiated algorithm to obtain the corresponding analysis result dataset. After obtaining the analysis result datasets, horizontal comparisons between different units can be performed. For example, based on the analysis result dataset of ball valve opening time, a first trend change curve can be constructed, and the evolution trend of equipment status can be traced back through the curve. Figure 6 As shown, by constructing the first trend change curves of units 1-10 in a parallel coordinate system, the results can be compared horizontally. Combined with the temperature change curve, it can be found that the lower the temperature, the longer the ball valve opening time.

[0044] In some embodiments, the data application further includes a curve analysis module, which is configured to acquire set parameters and generate a second trend change curve based on the set parameters and the corresponding unified data.

[0045] The business process of the curve analysis module is as follows: Figure 3 As shown, it provides users with efficient data trend analysis and in-depth analysis capabilities in specific fields, helping users quickly identify key information in the data and optimize the decision-making process. Specifically, for example, it generates a grounding current variation curve over a period of time. Users can configure parameters, such as the time range. The curve analysis module determines the corresponding unified data for the grounding current within that time range and then generates a second trend variation curve for that time range, such as... Figure 7 As shown, one can understand the range of grounding current fluctuations, etc.

[0046] In some embodiments, the data application further includes a health analysis module configured to generate a health analysis report based on a preset period, monitoring indicators, and the unified data.

[0047] The business processes of the health analysis module, for example Figure 4As shown, users can configure preset cycles and monitoring indicators. For example, the preset cycle is one month, and the monitoring indicator is the reservoir water level. By integrating time series analysis algorithms and feature extraction technology, a health analysis report is generated in real time, covering parameter comparison, trend evolution, etc. Multi-level evaluation thresholds can also be set to assess the health level, providing a quantitative assessment basis that is both in-depth and intuitive for operation and maintenance decisions.

[0048] In some embodiments, the data application further includes a monitoring and alarm module and a system monitoring module. The monitoring and alarm module is configured to comprehensively monitor the unified data and issue a second alarm message in response to data anomalies. The system monitoring module is configured to monitor the pumped storage power station data fusion intelligent analysis platform and issue a third alarm message in response to platform anomalies.

[0049] The monitoring and alarm module comprehensively monitors unified data corresponding to measurement attributes, providing flexible alarm settings and efficient alarm information processing. It supports setting multi-level threshold alarms, trend alarms, and correlated alarms, taking different alarm handling measures according to the alarm level. When critical systems or equipment experience data anomalies, the system can trigger alarms in real time, issuing secondary alarm information, such as "reservoir water level too high," providing strong protection for safe and stable operation and effectively improving overall reliability.

[0050] The system monitoring module targets the platform system, enabling real-time monitoring of online user activity, network structure, operational status, and performance metrics. For example... Figure 5 As shown, users can monitor the status of network devices and links in real time by inputting the configuration of the network device and device link relationships. The SNMP protocol is used to monitor link status, dynamically updating the link color indicators in the topology diagram, promptly identifying and resolving network device and link faults, ensuring network stability and efficiency. When the platform malfunctions, it issues a third-party alarm message, such as "data center link failure," allowing users to quickly locate and repair the link.

[0051] The external mobile terminal can perform alarm push notifications and data monitoring. Data monitoring can include monitoring of important data, real-time monitoring, monitoring of unresolved defects, and ongoing work orders. Through the mobile app, users can view the power plant overview, real-time monitoring information, unresolved defect information, ongoing work order information, and receive alarm information in real time, enabling them to perform maintenance proactively and improve management efficiency.

[0052] In some embodiments, the pumped storage power station is equipped with multiple cameras, which are used to acquire corresponding unit operation images; the data application also includes a system linkage module, which is configured to associate the work order information of the pumped storage power station's unit equipment with the corresponding unit operation images.

[0053] The system linkage module achieves real-time linkage between work order information and camera monitoring, as well as defect correlation, through deep collaboration between industrial television and equipment monitoring. Users can intuitively access the unit's operating screen, simultaneously view equipment status, and quickly configure camera positioning, significantly improving the efficiency of operation and maintenance decision-making and the speed of anomaly response, thus helping industrial supervision upgrade towards intelligence and precision.

[0054] In some embodiments, the data application further includes a special equipment module and a cockpit module. The special equipment module is configured to independently establish equipment files and monitor the special equipment in the pumped storage power station. The cockpit module is configured to display the daily production report of the pumped storage power station.

[0055] The special equipment module provides a full lifecycle management solution for special equipment, such as water pumps, turbines, and speed control systems. It establishes a management scheme of "equipment file - inspection record - maintenance decision" and, in conjunction with the inspection window period of the production plan, establishes a configurable early warning trigger mechanism (supporting customizable early warning days, such as 30 days for conventional special equipment and 45 days for precision special equipment). It automatically monitors special equipment data, initiates an early warning before the expiration date, and pushes the warning to the person responsible for the equipment to ensure the safe operation of the equipment.

[0056] The cockpit module can display daily production reports of the pumped storage power station. It adopts a responsive layout design and supports multi-terminal adaptation to PCs and large screens, making it convenient for users to understand key information such as unit status, water level, and power generation, providing strong support for production monitoring and scientific decision-making.

[0057] This pumped-storage power station data fusion and intelligent analysis platform achieves deep integration and intelligent analysis of multi-source heterogeneous data by integrating big data processing and algorithm analysis technologies. The system aggregates massive amounts of operational data from nearly 100,000 monitoring points in real time, develops a no-code algorithm modeling method, supports one-time modeling for reuse across multiple units, and flexibly deploys nearly 500 algorithms. It has constructed eight core functional modules, including curve analysis, algorithm platform, health analysis, and monitoring alarms. The platform employs time-series alignment and protocol conversion technologies to achieve the fusion and governance of heterogeneous data across regions from computer monitoring systems, production management systems, and 500kV monitoring systems. Through a modular architecture, the platform enables efficient collaboration among its functional components. It can not only generate equipment health analysis reports on a regular basis but also accurately predict equipment failures based on horizontal algorithm comparison and trend analysis. By configuring global baseline thresholds and unit-specific thresholds, it automatically pushes alarm information to the responsible personnel when potential problems occur. Practical application shows that the platform can save approximately 50% of operation and maintenance manpower costs and significantly shorten unplanned unit downtime, providing key technical support for the intelligent transformation of pumped-storage power stations and possessing broad application prospects.

[0058] This platform features a highly efficient algorithm reuse mechanism, enabling single-modeling reuse across multiple units through an algorithm middleware platform. It supports rapid deployment, improving development efficiency by over 50% and reducing maintenance costs by 30%. It supports independent operation of cross-unit algorithm containers, with data isolation ensuring operational stability. Employing a parallel computing architecture, the platform overcomes the performance bottlenecks of traditional single-threaded computing, increasing data processing speed by 10 times to meet real-time analysis needs. Algorithms for each unit are independently instantiated, improving resource utilization. The platform employs a dual-threshold alarm system, using a global baseline plus unit-specific threshold monitoring mechanism. It supports both global baseline threshold setting and unit-specific threshold adjustment, automatically pushing tiered alarm information to responsible personnel. The platform supports multi-dimensional visual decision-making, constructing a parallel coordinate system for multi-unit data to achieve horizontal comparison and trend tracking, providing intuitive visual analysis and improving operational decision-making efficiency. The platform can perform automated health analysis. The health analysis module supports dynamic evaluation and multi-dimensional report generation, automatically generating parameter comparison reports and trend curves. Combined with time-series analysis and feature extraction technologies, it increases the accuracy of equipment status assessment and reduces manual intervention. The platform supports multi-system linkage, enabling the association of unit equipment with work orders and defect reports, and supporting the association of work order information with camera video, thereby achieving information sharing and coordinated control, significantly improving the efficiency of operation and maintenance decision-making and the speed of anomaly response.

[0059] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of this application (including the claims) is limited to these examples; within the framework of this application, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of the embodiments of this application as described above, which are not provided in the details for the sake of brevity.

[0060] Furthermore, to simplify the description and discussion, and to avoid obscuring the embodiments of this application, the apparatus may be shown in block diagram form. This is to prevent the embodiments of this application from being difficult to understand, and it also takes into account the fact that the details of the implementation of these block diagram apparatuses are highly dependent on the platform on which the embodiments of this application will be implemented (i.e., these details should be fully within the understanding of those skilled in the art). In setting forth specific details to describe exemplary embodiments of this application, it will be apparent to those skilled in the art that the embodiments of this application may be implemented without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0061] Although this application has been described in conjunction with specific embodiments thereof, many substitutions, modifications and variations of these embodiments will be apparent to those skilled in the art from the foregoing description.

[0062] The embodiments of this application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the embodiments of this application should be included within the protection scope of this application.

Claims

1. A data fusion and intelligent analysis platform for pumped storage power stations, characterized in that, include: Data sources include various raw data from pumped storage power stations; The data center is configured to collect the raw data, perform time-series alignment and protocol conversion on the raw data to obtain unified data, and store the unified data; The data application is configured to access the data center to obtain the unified data. The data application includes an algorithm platform module, which is configured to build an operator library, associate the measurement point identifiers of the unified data with corresponding measurement attributes and form a matching table; obtain algorithm construction instructions; and, based on the algorithm construction instructions, the operator library and the measurement attributes, perform algorithm modeling through a drag-and-drop operation process to obtain a power plant data analysis algorithm. The power plant data analysis algorithm is instantiated to obtain an instantiated algorithm; The corresponding unified data is processed according to the instantiation algorithm and the matching table to generate an analysis result dataset.

2. The intelligent data fusion analysis platform for pumped storage power stations according to claim 1, characterized in that, The operator library includes start operator, end operator, time-stamped time operator, time interval operator, and threshold alarm operator; the measurement attributes include multiple unit ball valve opening command attributes and multiple unit ball valve fully open result attributes; the algorithm construction instruction is a multiple unit ball valve opening time analysis algorithm construction instruction; The algorithm modeling process for obtaining power plant data analysis algorithm through drag-and-drop operation includes: dragging the start operator to add the multiple unit ball valve opening command attribute and the multiple unit ball valve full opening result attribute to the node; Drag and drop the time-stamped operator to connect it to the start operator. In the node, select the start condition as multiple units opening ball valves and the end condition as multiple units fully opening ball valves. Drag and drop the time interval operator to connect it to the time-stamped operator. In the node, select the time range as the time-stamped time. Drag and drop the threshold alarm operator to connect it to the time interval operator. Configure the alarm threshold in the node. Drag and drop the end operator to connect it to the threshold alarm operator to obtain the ball valve opening time analysis algorithm for multiple units.

3. The intelligent data fusion analysis platform for pumped storage power stations according to claim 2, characterized in that, The step of processing the corresponding unified data according to the instantiation algorithm and the matching table to generate the analysis result dataset includes: The instantiation algorithm is parsed, and the measurement point identifiers with the measurement attributes of the multiple units opening ball valve command attribute and the multiple units fully opening ball valve result attribute are obtained according to the matching table. The corresponding unified data is selected according to the measurement point identifier to form a data set. The time segment set is obtained by calculating the data set using the time-scaled time operator; The analysis result dataset is obtained by processing the set of time segments using the time interval operator. If the data value in the analysis result dataset exceeds the alarm threshold, a first alarm message is issued.

4. The intelligent data fusion analysis platform for pumped storage power stations according to claim 2, characterized in that, The configuration of alarm thresholds in nodes includes configuring a uniform alarm threshold for all the units and / or configuring a corresponding alarm threshold for each unit.

5. The intelligent data fusion analysis platform for pumped storage power stations according to claim 1, characterized in that, The instantiation of the power plant data analysis algorithm to obtain the instantiation algorithm includes: instantiating the power plant data analysis algorithm in parallel for each unit of the pumped storage power plant to obtain the instantiation algorithm corresponding to each unit. The step of processing the corresponding unified data according to the instantiation algorithm and the matching table to generate the analysis result dataset includes: processing the unified data of the corresponding unit according to the instantiation algorithm and the matching table to generate the corresponding analysis result dataset; The algorithm platform module is also configured to generate corresponding first trend change curves in the same coordinate system based on the analysis result dataset corresponding to each of the units.

6. The intelligent data fusion analysis platform for pumped storage power stations according to claim 1, characterized in that, The data application also includes a curve analysis module, which is configured to acquire set parameters and generate a second trend change curve based on the set parameters and the corresponding unified data.

7. The intelligent data fusion analysis platform for pumped storage power stations according to claim 1, characterized in that, The data application also includes a health analysis module, which is configured to generate a health analysis report based on a preset period, monitoring indicators, and the unified data.

8. The intelligent data fusion analysis platform for pumped storage power stations according to claim 1, characterized in that, The data application also includes a monitoring and alarm module and a system monitoring module. The monitoring and alarm module is configured to comprehensively monitor the unified data and issue a second alarm message in response to data anomalies. The system monitoring module is configured to monitor the pumped storage power station data fusion intelligent analysis platform and issue a third alarm message in response to platform anomalies.

9. The intelligent data fusion analysis platform for pumped storage power stations according to claim 1, characterized in that, The pumped storage power station is equipped with multiple cameras, which are used to capture the corresponding unit operation images; the data application also includes a system linkage module, which is configured to associate the work order information of the pumped storage power station's unit equipment with the corresponding unit operation images.

10. The intelligent data fusion analysis platform for pumped storage power stations according to claim 1, characterized in that, The data application also includes a special equipment module and a cockpit module. The special equipment module is configured to independently establish equipment files and monitor the special equipment in the pumped storage power station. The cockpit module is configured to display the daily production report of the pumped storage power station.

Images