Modeling analysis method, device and computer equipment for thermal power plant data

Abstract

The application discloses a thermal power plant data modeling analysis method, device and computer equipment. The method comprises the following steps: a graphical interface containing a structured data set identifier and an operator resource pool identifier is displayed; in response to a configuration operation, a target data and a target function operator unit are determined, and a directed data flow graph is constructed according to a dependency relationship; the directed data flow graph is parsed, and a calculation task corresponding to the target data is created; a distributed calculation engine is called, the target function operator unit is scheduled according to an execution order to execute the calculation task, so that a theoretical model reflecting thermal characteristics of the thermal power plant is generated, and a deviation of real-time operation data relative to the theoretical model is calculated; and an execution result including the deviation is displayed. The method of the application realizes full-process visual modeling, reduces the difficulty of thermal power plant data analysis, and improves the processing efficiency of massive data through the distributed calculation engine.

Description

Technical Field

[0001] This application relates to the field of thermal power plant operation analysis technology, and in particular to a modeling and analysis method, apparatus and computer equipment for thermal power plant data. Background Technology

[0002] With the continuous improvement of automation and informatization levels in thermal power plants, massive amounts of historical operating data have been accumulated during the production process, including boiler temperature, pressure, turbine speed, and generator load. This data contains information on equipment operating status, energy efficiency trends, and potential fault characteristics, serving as a crucial foundation for achieving operation optimization, performance evaluation, and predictive maintenance.

[0003] However, current analysis of thermal power plant operation data mainly relies on programming methods (such as Python and MATLAB). This approach suffers from problems such as long development cycles, high maintenance costs, and difficulty for non-professionals to use. Especially when dealing with multi-device, multi-operating-condition, and multi-objective modeling tasks, the lack of a unified framework leads to fragmented processes, difficulty in standardization, and an inability to meet the advanced data analysis needs of thermal power plants. Summary of the Invention

[0004] In view of this, this application provides a modeling and analysis method, apparatus and computer equipment for thermal power plant data, which realizes full-process visualization and automated modeling from data to model to results, reducing technical threshold and improving processing efficiency.

[0005] According to the first aspect of this application, a modeling and analysis method for thermal power plant data is provided, the method comprising: displaying a graphical interface, the graphical interface including an identifier of a structured dataset obtained based on historical operating data of the thermal power plant and an identifier of an operator resource pool registered to a scheduling service; In response to the user's configuration operation on the graphical configuration interface, the target data in the structured dataset and multiple target operators in the operator resource pool are determined, and a directed data flow graph is constructed based on the dependency relationship between the target operators. The directed data flow graph is parsed to create a computational task corresponding to the target data; The distributed computing engine is invoked to schedule the target operators to perform their corresponding target data computing tasks according to the execution order of the computing tasks, so as to generate a theoretical model reflecting the thermal characteristics of the thermal power plant and calculate the deviation of the real-time operating data relative to the theoretical model. The execution results of the computational task are displayed, including the deviation, the theoretical model, and the real-time running data.

[0006] Optionally, the method further includes: The historical operational data is obtained from at least one data source, and the historical operational data includes structured file data and data platform data; The historical operational data is preprocessed to obtain multi-source data that meets the thermodynamic modeling conditions; the multi-source data is then fused to generate the structured dataset. The preprocessing includes at least one of the following: removing outliers whose load change rate exceeds a preset threshold, filling in missing values, aligning timestamps, and unifying units.

[0007] Optionally, constructing a directed data flow graph based on the dependencies between the target operators includes: In response to the call command for the operator resource pool in the configuration operation, the selected target operator is instantiated in the workspace of the graphical configuration interface, and the corresponding operator node is generated. In response to the connection operation of the operator node, the connection relationship between different operator nodes is determined, and the input pins and output pins of each operator node are connected according to the connection relationship to generate the transmission path of the target data between different operator nodes; The input node of the target data is connected to the first operator node called in the transmission path, and the operator node located at the end of the transmission path is connected to the output node to form the directed data flow graph.

[0008] Optionally, parsing the directed data flow graph and creating the computational task corresponding to the target data includes: performing syntax parsing and dependency topology analysis on the directed data flow graph to identify adjacent node pairs, wherein the adjacent node pairs include: The input node of the target data, the first operator node, any two connected operator nodes, the operator node at the end of the transmission path, and the output node; For the adjacent node pair, the output of the previous node is used as the input of the current node, and an independent computation task is generated based on the computation logic of the target operator encapsulated in the current node. The operator nodes are organized into task chains according to the data flow direction in the directed data flow graph to mark the execution order of the computation tasks.

[0009] Optionally, the method further includes: Based on the physical dimension label of the target operator, the semantic legality of the dependency relationship is verified; If the physical quantity dimensions of the target operator to which the adjacent node pair belongs do not match, then the generation of the transmission path for the adjacent node pair is rejected.

[0010] Optionally, displaying the execution result of the computing task includes: exporting the execution result in CSV format or writing it to a data platform; The execution results are displayed through interactive charts, including at least one of trend charts, line charts, scatter plots, heatmaps, and dashboards.

[0011] Optionally, after displaying the execution result of the computation task, the method further includes: Highlight data points in the execution results that exceed the preset alarm threshold; and / or, In response to user input, extract a subset of data from the execution results of the computation task whose timestamp fields fall within the time interval indicated by the input, and update the data source of the interactive chart in real time; and / or, In response to the user's variable comparison operation on the interactive chart, the data sequence of multiple target variables indicated by the variable comparison operation is extracted, and the data sequence of the multiple target variables is mapped to a unified time axis coordinate system for synchronous rendering.

[0012] Optionally, the operator resource pool includes at least one pluggable operator, which includes at least one of the following: data processing operators, model training operators, prediction operators, and evaluation operators; The internal computation logic of the operator is decoupled from the external calling protocol, and it is dynamically connected to the operator resource pool through a unified registration mechanism.

[0013] Optionally, the target operator includes a data segmentation operator, an ideal curve generation operator, a flow rate piecewise linear segmentation operator, a parameter optimization operator, and a deviation analysis operator; scheduling the target operator to execute its corresponding target data calculation task according to the execution order of the executable calculation tasks includes: The data segmentation operator is invoked to segment the target data according to preset segmentation conditions, generating multiple sub-data. The ideal curve generation operator is invoked, the neighborhood range is determined based on the relationship between the key variables of the sub-data, and the ideal relationship curve is generated by fitting through local weighted regression. The flow rate piecewise linear segmentation operator is invoked to perform piecewise linear fitting on the flow rate characteristics of the ideal curve, and the coordinates of the feature points of each segment are extracted. The parameter optimization operator is invoked, and the coordinates of the feature points are used as constraints. The least squares method is used to perform nonlinear fitting on the ideal curve to obtain the model parameters. The theoretical model is then constructed based on the model parameters and the preset thermodynamic mechanism equation. The deviation analysis operator is invoked to compare the output value of the theoretical model with the measured value of the corresponding working point in the real-time operating data, and the deviation is calculated.

[0014] According to a second aspect of this application, a modeling and analysis apparatus for thermal power plant data is provided, the apparatus comprising: The display module is used to display a graphical interface, which includes the identifier of the structured dataset obtained based on the historical operating data of the thermal power plant and the identifier of the operator resource pool that has been connected to the system. The task generation module is used to respond to the user's configuration operation on the graphical configuration interface, determine the target data in the structured dataset and multiple target operators in the operator resource pool, and construct a directed data flow graph based on the dependency relationship between the target operators; and parse the directed data flow graph to create the computation task corresponding to the target data. The task execution module is used to call the distributed computing engine and schedule the target operators to execute the computing tasks corresponding to their target data according to the execution order of the computing tasks, so as to generate a theoretical model reflecting the thermal characteristics of the thermal power plant and calculate the deviation of the real-time operating data relative to the theoretical model. The display module is also used to display the execution results of the computation task, including the deviation, the theoretical model, and the real-time running data.

[0015] Optionally, the device further includes: The data acquisition module is used to acquire the historical operation data from at least one data source, the historical operation data including structured file data and data platform data; The preprocessing module is used to preprocess the historical operating data to obtain multi-source data that meets the thermodynamic modeling conditions; and to fuse the multi-source data to generate the structured dataset. The preprocessing includes at least one of the following: removing outliers whose load change rate exceeds a preset threshold, filling in missing values, aligning timestamps, and unifying units.

[0016] Optionally, the task generation module is specifically used to, in response to the call command for the operator resource pool in the configuration operation, instantiate the selected target operator in the workspace of the graphical configuration interface and generate the corresponding operator node; in response to the connection operation of the operator node, determine the connection relationship between different operator nodes, and connect the input pins and output pins of each operator node according to the connection relationship to generate the transmission path of the target data between different operator nodes; connect the input node of the target data to the first called operator node in the transmission path, and connect the operator node located at the end of the transmission path to the output node to form the directed data flow graph.

[0017] Optionally, the task generation module is specifically used to perform syntax parsing and dependency topology analysis on the directed data flow graph, identify adjacent node pairs, the adjacent node pairs including: the input node of the target data and the first operator node, any two connected operator nodes, the operator node at the end of the transmission path and the output node; for the adjacent node pairs, the output of the previous node is used as the input of the current node, and an independent computation task is generated according to the computation logic of the target operator encapsulated by the current node; the operator nodes are organized into a task chain according to the data flow direction in the directed data flow graph to mark the execution order of the computation tasks.

[0018] Optionally, the device further includes: The verification module is used to perform semantic validity verification on the dependency relationship based on the physical dimension label of the target operator; if the physical quantity dimension of the target operator to which the adjacent node pair belongs does not match, the generation of the transmission path of the adjacent node pair is rejected.

[0019] Optionally, the device further includes: The storage module is used to export the execution results in CSV format or write them to the data platform. The display module is specifically used to display the execution results through interactive charts, including at least one of trend charts, line charts, scatter plots, heatmaps, and dashboards.

[0020] Optionally, the display module is further configured to highlight data points in the execution results that exceed a preset alarm threshold; and / or, in response to user input, extract a subset of data in the execution results of the calculation task whose timestamp fields fall within the time interval indicated by the input operation, and update the data source of the interactive chart in real time; and / or, in response to user variable comparison operations on the interactive chart, extract data sequences of multiple target variables indicated by the variable comparison operation, and map the data sequences of the multiple target variables to a unified time axis coordinate system for synchronous rendering.

[0021] Optionally, the operator resource pool includes at least one pluggable operator, which includes at least one of the following: data processing operators, model training operators, prediction operators, and evaluation operators; The internal computation logic of the operator is decoupled from the external calling protocol, and it is dynamically connected to the operator resource pool through a unified registration mechanism.

[0022] Optionally, the target operator includes a data segmentation operator, an ideal curve generation operator, a flow rate piecewise linear segmentation operator, a parameter optimization operator, and a deviation analysis operator; The task execution module is specifically used to call the data segmentation operator to segment the target data according to preset segmentation conditions, generating multiple sub-data; call the ideal curve generation operator to determine the neighborhood range based on the relationship between key variables of the sub-data, and generate an ideal relationship curve through local weighted regression fitting; call the flow rate piecewise linear segmentation operator to perform piecewise linear fitting on the flow characteristics of the ideal curve, and extract the feature point coordinates of each segment; call the parameter optimization operator to obtain model parameters by nonlinear fitting of the ideal curve using the feature point coordinates as constraints and the least squares method, and construct the theoretical model based on the model parameters and the preset thermodynamic mechanism equation; and call the deviation analysis operator to compare the output value of the theoretical model with the measured value of the corresponding operating point in the real-time operating data, and calculate the deviation.

[0023] According to a third aspect of this application, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the above-described modeling and analysis method for thermal power plant data.

[0024] By employing the aforementioned technical solution, historical data from thermal power plants is decoupled from a standardized operator resource pool within a graphical interface, allowing users to complete visual configuration through drag-and-drop and connection operations. A Directed Data Flow Graph (DAG) is used to transform complex modeling logic into a parallel-scheduling distributed computing task. This enables the automated construction of high-precision thermodynamic characteristic theoretical models from massive amounts of heterogeneous data without the need for coding, significantly reducing modeling difficulty and allowing operation engineers and technicians without programming skills to easily participate in the construction of thermal power plant benchmark models. Simultaneously, this dynamic benchmark enables precise quantification of real-time operational deviations, significantly improving the configuration flexibility of thermal power plant thermal system performance monitoring, the efficiency of large-scale data processing, and the objective accuracy of perceiving equipment degradation trends.

[0025] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0026] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A flowchart illustrating the modeling and analysis method for thermal power plant data provided in an embodiment of this application is shown. Figure 2 A schematic diagram of the graphical interface provided in an embodiment of this application is shown; Figure 3 This illustration shows a schematic diagram of an interactive chart display interface provided in an embodiment of this application; Figure 4 This paper shows a structural block diagram of the modeling and analysis device for thermal power plant data provided in an embodiment of this application; Figure 5 A schematic diagram of the electronic structure of a computer device provided in an embodiment of this application is shown. Detailed Implementation

[0027] The present application will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the embodiments of the present application can be combined with each other.

[0028] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application.

[0029] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this application means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we say an element is “connected” or “attached” to another element, it can be directly connected or attached to the other element, or there may be intermediate elements present. Furthermore, the term “and / or” as used herein includes all or any units and all combinations of one or more associated listed items.

[0030] Exemplary embodiments according to this application will now be described in more detail with reference to the accompanying drawings. However, these exemplary embodiments may be implemented in many different forms and should not be construed as being limited to the embodiments set forth herein. It should be understood that these embodiments are provided so that the disclosure of this application is thorough and complete, and that the concept of these exemplary embodiments is fully conveyed to those skilled in the art.

[0031] This embodiment provides a modeling and analysis method for thermal power plant data, such as Figure 1 As shown, the method includes: Step 101: Display the graphical interface.

[0032] The graphical user interface (GUI) serves as the window through which users interact with the modeling system. The content displayed falls into two main categories. One category consists of identifiers for structured datasets. These datasets originate from historical operating data of thermal power plants (such as boiler temperature, pressure, and turbine speed), and are standardized data resources formed after cleaning and fusion. On the interface, these can be represented by filenames, data table icons, or unique identifiers (IDs). The other category is identifiers for operator resource pools. These pools include at least one operator registered with the scheduling service, such as data processing operators, model training operators, prediction operators, and evaluation operators. On the interface, these are typically presented as functional module icons.

[0033] It's important to note that an operator is an independent functional unit that encapsulates specific computational logic (e.g., Python scripts, C++ dynamic libraries). Operators can be developed, tested, and deployed independently of the main system platform, interacting with it through external calling protocols. When a new analysis algorithm needs to be added, there's no need to recompile the entire platform or perform a system-wide update; only a new operator conforming to the platform specifications needs to be developed and registered in the resource pool. The graphical interface can then automatically scan and display the newly registered operator's identifier, allowing users to use it via drag-and-drop. This pluggable operator reduces system maintenance costs and improves the system's adaptability to different operating scenarios. Through this decoupling, operator developers can focus on optimizing their internal algorithms without worrying about the platform's underlying scheduling logic. The system platform's scheduling service only needs to call according to the external protocol, without needing to understand the operator's internal implementation details. This effectively isolates differences in the technology stack, allowing operators developed in different programming languages ​​to coexist harmoniously in the same resource pool.

[0034] In this embodiment, by displaying data and operators in the form of visual icons, the system provides users with an intuitive visual operation entry point, enabling non-programming professionals to quickly identify and call the required resources, significantly reducing the technical threshold for modeling and analysis.

[0035] In one embodiment, prior to step 101, the modeling and analysis method for thermal power plant data further includes: acquiring historical operating data from at least one data source; preprocessing the historical operating data to obtain multi-source data that meets the thermodynamic modeling conditions; and fusing the multi-source data to generate a structured dataset. Historical operational data includes structured file data and data platform data. Structured file data can be CSV files, Excel spreadsheets, etc. The system can receive files uploaded by users through a file upload interface and parse the data content within them. Data platform data can be PI databases, relational databases, etc. For structured file data, the system can directly read data tables in the platform through pre-configured database connection interfaces (such as JDBC, ODBC, etc.).

[0036] Specifically, preprocessing includes at least one of the following: removing outliers with load variation rates exceeding a preset threshold, imputing missing values, timestamp alignment, and unit unification. The preset threshold can be set reasonably according to the required accuracy of the modeling.

[0037] Understandably, various imputation strategies can be employed to adapt to different scenarios. For short-term missing data, linear interpolation can be used, which involves linearly fitting the values ​​of adjacent data points before and after the missing point to estimate the approximate value of the missing point. For parameters with large fluctuations, a sliding window averaging method can be used, which involves taking the average value of data within a certain time window before and after the missing point as the imputation value.

[0038] In this embodiment, after acquiring historical operating data from multiple sources of thermal power plants, outliers with load fluctuation rates exceeding a preset threshold are removed to eliminate transient interference data. Simultaneously, standardized preprocessing techniques such as missing value imputation, timestamp alignment, and unit unification are employed to filter out high-quality multi-source data that meets the requirements of thermodynamic modeling. Then, a standardized structured dataset is generated through multi-source data fusion. This effectively solves the problems of scattered sources and heterogeneous formats in thermal power plant operating data, improves data accuracy, reduces the complexity and error rate of subsequent modeling and analysis, and ensures the accuracy and reliability of subsequent model training and analysis results.

[0039] Step 102: In response to the user's configuration operation on the graphical configuration interface, determine the target data in the structured dataset and multiple target operators in the operator resource pool, and construct a directed data flow graph based on the dependencies between the target operators.

[0040] In this context, a directed data flow graph is used to represent the complete path of data flowing from a source to a destination. It should be understood that although a directed data flow graph is represented as a two-dimensional planar graph on the interface, internally, it is parsed as a topological data structure.

[0041] For example, configuration operations include, but are not limited to, actions such as clicking, dragging, selecting, and connecting.

[0042] In this embodiment, the user selects target data from the data resource area and the required target operators, their execution order, and data flow direction from the operator resource pool through configuration operations in the graphical configuration interface. Based on these connections, the system automatically constructs a directed data flow graph in the background. This directed data flow graph records the adjacency relationships and dependency weights between operators, providing a logical basis for subsequent task scheduling.

[0043] In practical applications, step 102, which involves constructing a directed data flow graph based on the dependencies between the target operators, specifically includes the following steps: Step 102-1: In response to the call command for the operator resource pool in the configuration operation, the selected target operator is instantiated in the workspace of the graphical configuration interface, and the corresponding operator node is generated.

[0044] Step 102-2: In response to the connection operation of the operator node, determine the connection relationship between different operator nodes, and connect the input pins and output pins of each operator node according to the connection relationship to generate the transmission path of the target data between different operator nodes.

[0045] Step 102-3: Connect the input node of the target data to the first operator node called in the transmission path, and connect the operator node located at the end of the transmission path to the output node to form a directed data flow graph.

[0046] In this embodiment, within the graphical user interface's backend system, the operator resource pool essentially stores class definitions or templates for various operators. When a user selects an operator icon, the system receives a call instruction and creates an instance object of that operator class, i.e., an operator node, in the workspace's memory. This node not only contains the operator's unique identifier ID but also initializes the data structures of its input / output pins. When the system detects a connection operation to the operator node, it establishes a directed edge in the internal data structure from the output pin to the input pin, i.e., a propagation path. This path defines the data flow direction, indicating that the calculation result of the upstream operator will serve as the input parameter for the downstream operator. Finally, by connecting the input node to the first operator and the tail operator to the output node, a closed directed data flow graph is constructed. This transforms the complex thermal power plant thermal data processing logic into an explicit directed data flow graph, achieving deep decoupling and visual orchestration of the modeling process. While lowering the technical threshold for thermal engineers to build complex computational tasks, it ensures the logical rigor of multi-source data in the processing chain and the traceability of task scheduling through defined pin mapping relationships. This helps improve the configuration efficiency and algorithm reuse flexibility of the system for customized modeling of different unit characteristics.

[0047] For example, such as Figure 2 As shown, users can call target operators from the operator resource pool through configuration operations such as dragging / clicking. Target operators include functional operators such as Data_Segmentation, Ideal_Curve, Polyline_Segmentation, and Optimization_Calc.

[0048] The system instantiates corresponding operator nodes in the intermediate workspace: 3 Data_Segmentation nodes, 1 Ideal_Curve node, 2 Polyline_Segmentation nodes, 1 Optimization_Calc node, 4 CSV file output nodes, and the root node 0_alldata_filter (target data input node). Each operator node corresponds to an independent functional unit, undertaking subsequent data processing tasks and completing the instantiation and deployment of the operator.

[0049] Users connect the output pins of input node 0_alldata_filter to the input pins of three Data_Segmentation nodes and one Ideal_Curve node, respectively, to distribute the structured dataset (CSS_X285 / GV1_PO / GV4_PO / y_test) to the parallel processing operators. The output pins of two Data_Segmentation nodes are connected to the input pins of two Polyline_Segmentation nodes, respectively, to pass the segmented data to the flow polyline segmentation operator for fine-tuning of the segmented data. The output pins of the three Data_Segmentation nodes, two Polyline_Segmentation nodes, and one Ideal_Curve node are all connected to the input pins of the Optimization_Calc node, so that intermediate results such as segmented data, ideal curves, and optimized polylines are summarized to the parameter optimization operator. The parameter optimization operator Optimization_Calc at the end of the transmission path connects its output pins to four CSV file output nodes. The resulting complete directed data flow graph, starting from the input node and ending at the output node, clarifies the entire execution sequence of data from input → segmentation → optimization → calculation → export through hierarchical connections of operator nodes.

[0050] It's worth noting that, in addition to defining data types (such as integers and floating-point numbers), each operator unit's input and output pins also have pre-defined physical dimension labels, such as temperature, pressure, flow rate, and valve position. During the generation of transmission paths for target data between different operator nodes, when a user attempts to connect two nodes, the system automatically performs semantic validity checks on the dependencies between target operators by comparing the physical dimension labels of the pins at both ends of the connection line. If the physical dimension of the target operator does not match between adjacent nodes, the system refuses to generate a transmission path for the adjacent node pair. This ensures that every data transmission path conforms to thermodynamic physical laws, guaranteeing the correctness of the modeling logic from the configuration stage, thereby improving the fault tolerance of the configuration operation.

[0051] Step 103: parse the directed data flow graph and create the computation task corresponding to the target data.

[0052] In this embodiment, after the user completes the configuration and triggers the execution command, the system's background parsing engine is activated. The parsing engine traverses the topology of the directed data flow graph, and the system encapsulates each operator node and its required input data into an independent computational task. This automatically transforms the graphically orchestrated business logic into a sequence of specific computational tasks with execution timing constraints and target data mappings. While ensuring that the thermodynamic modeling process strictly follows the physical logical sequence, this improves the automated scheduling efficiency of complex computational processes and the robustness of system operation.

[0053] In practical applications, step 103, which involves parsing the directed data flow graph and creating the computation task corresponding to the target data, specifically includes the following steps: Step 103-1: Perform syntax parsing and dependency topology analysis on the directed data flow graph to identify adjacent node pairs.

[0054] The adjacent node pairs include: the input node of the target data and the first operator node, any two connected operator nodes, and the operator node at the end of the transmission path and the output node.

[0055] Step 103-2: For adjacent node pairs, use the output of the previous node as the input of the current node, and generate independent computation tasks based on the computation logic of the target operator encapsulated in the current node.

[0056] Step 103-3: Organize the operator nodes into task chains according to the data flow direction in the directed data flow graph to mark the execution order of the computation tasks.

[0057] In this embodiment, by performing deep syntax parsing and dependency topology analysis on the directed data flow graph, adjacent node pairs in the data flow are accurately identified. Based on the input-output mapping relationships between nodes, the abstract configuration logic is deconstructed into mutually independent and logically closed-loop computational tasks, and the execution sequence is explicitly defined using task chains. This achieves an automated and highly reliable conversion of complex thermodynamic modeling logic from visual orchestration to a structured task sequence, maximizing the utilization of computing resources and shortening the overall time required for modeling and analysis.

[0058] Step 104: Invoke the distributed computing engine and schedule the target operators to perform their corresponding target data computing tasks according to the execution order of the computing tasks, so as to generate a theoretical model reflecting the thermal characteristics of the thermal power plant and calculate the deviation of the real-time operating data from the theoretical model.

[0059] In this embodiment, a distributed computing engine is used as the underlying execution architecture. The system scheduling service submits tasks sequentially according to the task chain and their dependencies. During execution, the system driver encapsulates target operators of specific algorithms to perform parallel and logical deep computation on multi-source heterogeneous data, constructing a theoretical model of thermodynamic characteristics that can dynamically map the complex operating conditions of thermal power plants. This model serves as a dynamic benchmark for quantifying real-time operational deviations. This not only improves the throughput and response speed of processing massive amounts of historical data through a distributed architecture, but also ensures the rigor of thermodynamic logic through the orderly scheduling of the task chain. It fundamentally solves the technical problems of insufficient accuracy and difficulty in real-time perception of traditional static models under varying operating conditions, assisting users in efficiently understanding modeling results and conducting operational analysis.

[0060] Understandably, during the process execution, the system can automatically record the intermediate calculation results and running logs of each operator node, support process debugging, performance comparison and historical backtracking, which helps to discover abnormal links in the modeling process and improve the maintainability and reliability of the model.

[0061] It is worth mentioning that computational tasks without dependencies can be submitted for execution in parallel to make full use of the cluster's computing resources.

[0062] In practical applications, target operators include data segmentation operators, ideal curve generation operators, flow line segmentation operators, parameter optimization operators, and deviation analysis operators. Step 104 schedules the target operators to perform their corresponding target data calculation tasks according to the execution order of the calculation tasks, specifically including the following steps: Step 104-1: Call the data segmentation operator to segment the target data according to the preset segmentation conditions and generate multiple sub-data.

[0063] Step 104-2: Call the ideal curve generation operator, determine the neighborhood range based on the relationship between the key variables of the sub-data, and generate the ideal relationship curve by fitting through local weighted regression.

[0064] Among them, key variables can be set reasonably according to the application scenario. For example, when performing operation analysis on steam turbine equipment in thermal power plants, the comprehensive valve position and pressure ratio variables can be used as key variables.

[0065] Step 104-3: Call the flow rate piecewise linear operator to perform piecewise linear fitting on the flow characteristics of the ideal curve and extract the coordinates of the feature points of each segment.

[0066] Step 104-4: Call the parameter optimization operator, use the coordinates of the feature points as constraints, and use the least squares method to perform nonlinear fitting on the ideal curve to obtain the model parameters. Then, construct a theoretical model based on the model parameters and the preset thermodynamic mechanism equation.

[0067] Step 104-5: Call the deviation analysis operator to compare the output value of the theoretical model with the measured value of the corresponding working point in the real-time operating data, and calculate the deviation of the real-time operating data relative to the theoretical model.

[0068] In this embodiment, by collaboratively invoking operators such as data segmentation, ideal curve generation, flow piecewise linear segmentation, parameter optimization, and deviation analysis, and utilizing local weighted regression and least squares methods to extract features and invert parameters of complex nonlinear flow characteristics in a multidimensional variable space, a deep coupling modeling of data-driven algorithms and thermodynamic mechanism equations is achieved. This constructs a theoretical benchmark model that can accurately characterize the ideal operating state of thermal power plant equipment under all operating conditions. This effectively solves the problems of insufficient accuracy and poor generalization ability of traditional modeling methods when dealing with strongly nonlinear correlated variables such as valve position and pressure ratio. It can more accurately calculate the difference between actual operation and ideal state, helping to prevent equipment performance anomalies early.

[0069] Step 105: Display the execution results of the computation task.

[0070] The execution results include real-time running data, the theoretical model, and the deviation of the real-time running data from the theoretical model.

[0071] The modeling and analysis method for thermal power plant data provided in this application decouples historical data from a standardized operator resource pool in a graphical interface, allowing users to complete visual configuration through drag-and-drop and connection operations. It utilizes a Directed Data Flow Graph (DAG) to transform complex modeling logic into a parallel-scheduling distributed computing task. This enables automated construction of high-precision thermodynamic characteristic theoretical models from massive heterogeneous data without writing code, significantly reducing modeling difficulty and allowing operation engineers and technicians without programming skills to easily participate in the construction of thermal power plant benchmark models. Simultaneously, based on this dynamic benchmark, accurate quantification of real-time operational deviations can be achieved, significantly improving the configuration flexibility of thermal power plant thermal system performance monitoring, the efficiency of large-scale data processing, and the objective accuracy of perceiving equipment degradation trends.

[0072] In one embodiment, step 105 specifically includes the following steps: exporting the execution result in CSV format or writing it to the data platform; displaying the execution result through interactive charts.

[0073] Interactive charts include at least one of the following: trend charts, line charts, scatter plots, heatmaps, and dashboards.

[0074] In this embodiment, after the calculation task is completed, the generated result data is transmitted back to the front-end interface for display. The display method can be in the form of data tables, trend charts, or dashboards. In particular, for deviation data, the system can display it as a time series curve, intuitively presenting the trend of equipment operating status changes over time. This allows users to clearly perceive complex calculation results, assists them in quickly judging equipment status and formulating optimization strategies, and achieves closed-loop application of data analysis results.

[0075] Specifically, the output results come in two forms. The first form exports the results in CSV (Comma-Separated Values) format. The system serializes the calculated structured data into a CSV file stream in the background and saves it to the user's local terminal via a browser download interface. This method is suitable for scenarios requiring offline analysis, archiving, or secondary processing using third-party tools (such as MATLAB or Excel), meeting the needs of professional researchers for flexible data processing. The second form writes the results to the data platform. The system directly writes the calculation results back to the enterprise's real-time or historical database via a pre-configured database connector (such as a JDBC interface).

[0076] In one embodiment, after step 105, the modeling and analysis method for thermal power plant data further includes: The data points in the execution results that exceed the preset alarm threshold are highlighted.

[0077] In this embodiment, when rendering layers, the system can load preset alarm threshold data in parallel and use the marking function of the front-end graphics library to highlight data points that exceed the threshold range. This visual enhancement quickly attracts the user's attention, helping them rapidly locate abnormal operating conditions within massive amounts of data and significantly improving the response speed of fault warnings.

[0078] In one embodiment, after step 105, the modeling and analysis method for thermal power plant data further includes: In response to user input, a subset of data from the execution results of the computation task that falls within the time interval indicated by the input operation is extracted, and the data source of the interactive chart is updated in real time.

[0079] In this embodiment, after receiving the time interval input by the user, the system extracts a subset of data from the calculation task execution results whose timestamp fields fall within that time interval. This data subset is then used to update the data source of the interactive chart in real time, triggering a re-rendering process for the chart. This eliminates the need to re-execute the entire modeling process; only a partial refresh of the result data is performed, resulting in extremely fast response times. This allows users to smoothly observe the details of device operation over different time periods by zooming in and out on the timeline.

[0080] In one embodiment, after step 105, the modeling and analysis method for thermal power plant data further includes: In response to the user's variable comparison operation on the interactive chart, the data sequence of multiple target variables indicated by the variable comparison operation is extracted, and the data sequence of the multiple target variables is mapped to a unified time axis coordinate system for synchronous rendering.

[0081] In this embodiment, the system maps the data sequences of multiple target variables selected by the user to a unified time axis coordinate system, and employs a dual Y-axis or multi-Y-axis rendering strategy, or performs normalization processing before mapping to the same coordinate system. This enables synchronous rendering of multiple variables in the same chart, allowing for intuitive analysis of the coupling relationships and synergistic trends between different parameters, facilitating operational optimization analysis.

[0082] In one specific embodiment, the optimization modeling of integrated valve position and pressure ratio of a steam turbine is used as an example for illustration.

[0083] (1) Data preparation: Configure a Spark task to read turbine operating data from a data source, including variables such as overall valve position, high-pressure valve 1 opening, high-pressure valve 2 opening, and pressure ratio. After cleaning and preprocessing the data, generate a structured dataset to ensure data integrity and consistency.

[0084] (2) Configuration modeling process construction: like Figure 2 As shown, in the graphical configuration interface, the user first drags the dataset 0_alldata_filter into the canvas as the input source for the modeling process. Then, multiple Data_Segmentation operators are added to segment the data. The first Data_Segmentation operator is configured to segment the data according to the Y-axis segmentation threshold of high-adjustment valve 1, and sets the opening state of the second high-adjustment valve; if it is not fully open, the actual maximum opening value must be manually entered. The second Data_Segmentation operator can select to segment along the X-axis or Y-axis according to different operating conditions, and configure the corresponding segmentation method and threshold parameters.

[0085] After data segmentation, the system adds the Ideal_Curve operator to generate the ideal curve. This operator is configured to select "Comprehensive Valve Position" and "Pressure Ratio" as key variable labels and set the x-neighborhood range to achieve local trend fitting and ideal relationship modeling.

[0086] Based on each data segmentation result, a Polyline_Segmentation operator is added to segment the flow characteristic polyline. Users can configure the high-voltage valve channels of interest and set the positions of segmentation point 1 and segmentation point 2, as well as the number of inflection points in the first, second, and third segments, thereby adapting to the flow variation characteristics of different load ranges.

[0087] Add the Optimization_Calc operator to fit the curve based on the least squares method to find the optimal curve parameters; this operator takes the overall valve position, the opening degree of the first high-pressure valve, the opening degree of the second high-pressure valve, and the pressure ratio as input variables to complete the parameter optimization calculation.

[0088] Finally, users connect the above-mentioned operator modules sequentially according to the data flow direction using connecting lines to form a complete directed data flow graph, constituting an end-to-end visual modeling process, and realizing automated analysis from raw data input to parameter optimization output.

[0089] (3) Process execution: After clicking the "Run" button, the system will automatically execute the modeling process. Specifically, the scheduling engine first parses the user-constructed configuration process structure, identifies each operator module and its connection relationships, and generates the corresponding executable computation task graph. Then, the Spark engine loads the dataset 0_alldata_filter and segments the data according to the configured segmentation conditions. In each segmented subset, the system calls the Ideal_Curve operator to generate an ideal curve based on the relationship between the valve position and pressure ratio. The Polyline_Segmentation operator is then called to perform line segmentation on the flow characteristics of the ideal curve, extracting key feature points from each segment. Finally, the Optimization_Calc operator is called to fit the curve using the least squares method, optimizing the model parameters and outputting the results. Throughout the execution process, the system automatically records the intermediate calculation results and runtime logs of each operator node, supporting subsequent process debugging, performance analysis, and result traceability.

[0090] (4) Results output and display: After modeling is completed, the system provides diverse output options. The calculated analysis results can be exported as a CSV file, which details the optimal curve parameters, the segmented dataset, the ideal curve, and the results of the flow line segmentation, facilitating data archiving and subsequent analysis.

[0091] like Figure 3As shown, to further enhance the readability and intuitiveness of the results, the system also supports various visualization charts, including trend charts, step charts, line charts, bar charts, and scatter plots. These charts can intuitively present key information such as the relationship curve between the integrated valve position and the pressure ratio, and the segmented change trend of the flow characteristics, helping users to efficiently understand the modeling results and conduct operational analysis.

[0092] The modeling and analysis method for thermal power plant data provided in this application can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, etc.; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms.

[0093] It should be noted that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0094] The user information (including but not limited to device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0095] Furthermore, such as Figure 4 As shown, as a specific implementation of the above-mentioned modeling and analysis method for thermal power plant data, this application embodiment provides a modeling and analysis device 400 for thermal power plant data. The modeling and analysis device 400 for thermal power plant data includes: a display module 401, a task generation module 402, and a task execution module 403.

[0096] Among them, the display module 401 is used to display the graphical interface, which includes the identifier of the structured dataset obtained based on the historical operation data of the thermal power plant and the identifier of the operator resource pool that has been connected to the system. The task generation module 402 is used to respond to the user's configuration operation on the graphical configuration interface, determine the target data in the structured dataset and multiple target operators in the operator resource pool, and construct a directed data flow graph based on the dependency relationship between the target operators; and parse the directed data flow graph to create the computation task corresponding to the target data. The task execution module 403 is used to call the distributed computing engine to schedule the target operators to execute the corresponding target data computing tasks according to the execution order of the computing tasks, so as to generate a theoretical model reflecting the thermal characteristics of the thermal power plant and calculate the deviation of the real-time operating data from the theoretical model. The display module 401 is also used to display the execution results of the computation task, including the deviation.

[0097] Furthermore, the modeling and analysis device 400 for thermal power plant data also includes: The data acquisition module (not shown in the figure) is used to acquire historical operational data from at least one data source. The historical operational data includes structured file data and data platform data. The preprocessing module (not shown in the figure) is used to preprocess historical operating data to obtain multi-source data that meets the conditions for thermodynamic modeling; the multi-source data is then fused to generate a structured dataset. The preprocessing includes at least one of the following: removing outliers whose load change rate exceeds a preset threshold, filling in missing values, aligning timestamps, and unifying units.

[0098] Furthermore, the task generation module 402 is specifically used to respond to the call command for the operator resource pool in the configuration operation, instantiate the selected target operator in the workspace of the graphical configuration interface, and generate the corresponding operator node; respond to the connection operation of the operator node, determine the connection relationship between different operator nodes, and connect the input pins and output pins of each operator node according to the connection relationship to generate the transmission path of the target data between different operator nodes; connect the input node of the target data to the first called operator node in the transmission path, and connect the operator node located at the end of the transmission path to the output node to form a directed data flow graph.

[0099] Furthermore, the task generation module 402 is specifically used to perform syntax parsing and dependency topology analysis on the directed data flow graph, identify adjacent node pairs, which include: the input node of the target data and the first operator node, any two connected operator nodes, and the operator node at the end of the transmission path and the output node; for adjacent node pairs, the output of the previous node is used as the input of the current node, and independent computation tasks are generated according to the computation logic of the target operator encapsulated by the current node; the operator nodes are organized into task chains according to the data flow direction in the directed data flow graph to mark the execution order of the computation tasks.

[0100] Furthermore, the modeling and analysis device 400 for thermal power plant data also includes: The verification module (not shown in the figure) is used to perform semantic validity verification on the dependency relationship based on the physical dimension label of the target operator; if the physical quantity dimension of the target operator to which the adjacent node belongs does not match, the generation of the transmission path of the adjacent node pair is rejected.

[0101] Furthermore, the modeling and analysis device 400 for thermal power plant data also includes: The storage module (not shown in the figure) is used to export the execution results in CSV format or write them to the data platform; The display module 401 is specifically used to display the execution results through interactive charts, including at least one of trend charts, line charts, scatter plots, heatmaps, and dashboards.

[0102] Furthermore, the display module 401 is also used to highlight data points in the execution results that exceed the preset alarm threshold; and / or, in response to the user's input operation, to extract a subset of data in the calculation task execution results whose timestamp fields fall within the time interval indicated by the input operation, and to update the data source of the interactive chart in real time; and / or, in response to the user's variable comparison operation on the interactive chart, to extract the data sequence of multiple target variables indicated by the variable comparison operation, and to map the data sequence of multiple target variables to a unified time axis coordinate system for synchronous rendering.

[0103] Furthermore, the operator resource pool includes at least one pluggable operator, which includes at least one of the following: data processing operators, model training operators, prediction operators, and evaluation operators; the internal computation logic of the operator is decoupled from the external calling protocol and dynamically accesses the operator resource pool through a unified registration mechanism.

[0104] Furthermore, the target operators include data segmentation operators, ideal curve generation operators, flow rate piecewise linear segmentation operators, parameter optimization operators, and deviation analysis operators; The task execution module 403 is specifically used to call the data segmentation operator to segment the target data according to preset segmentation conditions, generating multiple sub-data; call the ideal curve generation operator to determine the neighborhood range based on the relationship between key variables of the sub-data, and generate the ideal relationship curve through local weighted regression fitting; call the flow rate piecewise linear segmentation operator to perform piecewise linear fitting on the flow characteristics of the ideal curve and extract the feature point coordinates of each segment; call the parameter optimization operator to obtain the model parameters by nonlinear fitting of the ideal curve using the feature point coordinates as constraints and the least squares method, and construct a theoretical model based on the model parameters and preset thermodynamic mechanism equations; and call the deviation analysis operator to compare the output value of the theoretical model with the measured value of the corresponding operating point in the real-time operation data and calculate the deviation.

[0105] Specific limitations regarding the modeling and analysis device for thermal power plant data can be found in the limitations of the modeling and analysis methods for thermal power plant data described above, and will not be repeated here. Each module in the aforementioned modeling and analysis device for thermal power plant data can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0106] Based on the above, Figure 1 Accordingly, embodiments of this application also provide a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method. Figure 1 The modeling and analysis method for thermal power plant data is shown.

[0107] Based on this understanding, the technical solution of this application can be embodied in the form of a software product. The software product can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, or portable hard drive), and includes several instructions to cause a computer device (such as a personal computer, server, or network device) to execute the methods described in the various implementation scenarios of this application.

[0108] Based on the above, Figure 1 The method shown, and Figure 4 The virtual device embodiment shown is designed to achieve the above objectives, such as... Figure 5 As shown in the illustration, this application also provides a computer device 500, which includes a processor 501 and a memory 502. The memory 502 stores a program or instructions that can run on the processor 501. When the program or instructions are executed by the processor 501, they implement the above-described... Figure 1 The modeling and analysis method for thermal power plant data is shown.

[0109] The memory 502 can be used to store software programs and various data. The memory 502 may primarily include a first storage area for storing programs or instructions and a second storage area for storing data. The first storage area may store the operating system, application programs or instructions required for at least one function (such as sound playback, image playback, etc.). Furthermore, the memory 502 may include volatile memory or non-volatile memory, or both. The non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus RAM (DRRAM). The memory 502 in this embodiment includes, but is not limited to, these and any other suitable types of memory.

[0110] Processor 501 may include one or more processing units; optionally, processor 501 integrates an application processor and a modem processor, wherein the application processor mainly handles operations involving the operating system, user interface, and applications, and the modem processor mainly handles wireless communication signals, such as a baseband processor. It is understood that the aforementioned modem processor may also not be integrated into processor 501.

[0111] Computer equipment can specifically include personal computers, servers, network devices, etc.

[0112] Optionally, the computer device may also include a user interface, a network interface, a camera, radio frequency (RF) circuitry, sensors, audio circuitry, a Wi-Fi module, etc. The user interface may include a display screen, input units such as a keyboard, etc., and optional user interfaces may also include USB ports, card reader ports, etc. The network interface may optionally include standard wired interfaces, wireless interfaces (such as Bluetooth interfaces, Wi-Fi interfaces), etc.

[0113] Those skilled in the art will understand that the computer device structure provided in this embodiment does not constitute a limitation on the computer device, and may include more or fewer components, or combine certain components, or have different component arrangements.

[0114] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platform, or it can be implemented by hardware.

[0115] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of a preferred embodiment, and the modules or processes shown in the drawings are not necessarily essential for implementing this application. Those skilled in the art will understand that the modules in the apparatus of the embodiment can be distributed within the apparatus of the embodiment as described, or can be modified to be located in one or more apparatuses different from this embodiment. The modules of the above-described embodiment can be combined into one module, or further divided into multiple sub-modules.

[0116] The serial numbers in this application are for descriptive purposes only and do not represent the superiority or inferiority of any particular implementation scenario. The above disclosures are merely a few specific implementation scenarios of this application; however, this application is not limited thereto, and any variations conceived by those skilled in the art should fall within the protection scope of this application.

Claims

1. A method for modeling and analyzing thermal power plant data, characterized in that, The method includes: A graphical interface is displayed, which includes the identifier of the structured dataset obtained based on the historical operating data of the thermal power plant and the identifier of the operator resource pool that has been registered to the scheduling service. In response to the user's configuration operation on the graphical configuration interface, the target data in the structured dataset and multiple target operators in the operator resource pool are determined, and a directed data flow graph is constructed based on the dependency relationship between the target operators. The directed data flow graph is parsed to create a computational task corresponding to the target data; The distributed computing engine is invoked to schedule the target operators to perform their corresponding target data computing tasks according to the execution order of the computing tasks, so as to generate a theoretical model reflecting the thermal characteristics of the thermal power plant and calculate the deviation of the real-time operating data relative to the theoretical model. The execution results of the computational task are displayed, including the deviation, the theoretical model, and the real-time running data.

2. The modeling and analysis method for thermal power plant data according to claim 1, characterized in that, The method further includes: The historical operational data is obtained from at least one data source, and the historical operational data includes structured file data and data platform data; The historical operating data is preprocessed to obtain multi-source data that meets the conditions for thermodynamic modeling; The multi-source data is fused to generate the structured dataset; The preprocessing includes at least one of the following: removing outliers whose load change rate exceeds a preset threshold, filling in missing values, aligning timestamps, and unifying units.

3. The modeling and analysis method for thermal power plant data according to claim 1, characterized in that, The construction of the directed data flow graph based on the dependencies between the target operators includes: In response to the call command for the operator resource pool in the configuration operation, the selected target operator is instantiated in the workspace of the graphical configuration interface, and the corresponding operator node is generated. In response to the connection operation of the operator node, the connection relationship between different operator nodes is determined, and the input pins and output pins of each operator node are connected according to the connection relationship to generate the transmission path of the target data between different operator nodes; The input node of the target data is connected to the first operator node called in the transmission path, and the operator node located at the end of the transmission path is connected to the output node to form the directed data flow graph.

4. The modeling and analysis method for thermal power plant data according to claim 3, characterized in that, The step of parsing the directed data flow graph and creating the computation task corresponding to the target data includes: The directed data flow graph is parsed and its dependency topology is analyzed to identify adjacent node pairs, which include: the input node of the target data and the first operator node, any two connected operator nodes, and the operator node at the end of the transmission path and the output node. For the adjacent node pair, the output of the previous node is used as the input of the current node, and an independent computation task is generated based on the computation logic of the target operator encapsulated in the current node. The operator nodes are organized into task chains according to the data flow direction in the directed data flow graph to mark the execution order of the computation tasks.

5. The modeling and analysis method for thermal power plant data according to claim 3, characterized in that, The method further includes: Based on the physical dimension label of the target operator, the semantic legality of the dependency relationship is verified; If the physical quantity dimensions of the target operator to which the adjacent node pair belongs do not match, then the generation of the transmission path for the adjacent node pair is rejected.

6. The modeling and analysis method for thermal power plant data according to claim 1, characterized in that, The display of the execution results of the computation task includes: Export the execution results in CSV format or write them to the data platform; The execution results are displayed through interactive charts, including at least one of trend charts, line charts, scatter plots, heatmaps, and dashboards; After displaying the execution result of the computation task, the method further includes: Highlight data points in the execution results that exceed the preset alarm threshold; and / or, In response to user input, extract a subset of data from the execution results of the computation task whose timestamp fields fall within the time interval indicated by the input, and update the data source of the interactive chart in real time; and / or, In response to the user's variable comparison operation on the interactive chart, the data sequence of multiple target variables indicated by the variable comparison operation is extracted, and the data sequence of the multiple target variables is mapped to a unified time axis coordinate system for synchronous rendering.

7. The modeling and analysis method for thermal power plant data according to claim 1, characterized in that, The operator resource pool includes at least one pluggable operator, which includes at least one of the following: data processing operators, model training operators, prediction operators, and evaluation operators. The internal computation logic of the operator is decoupled from the external calling protocol, and it is dynamically connected to the operator resource pool through a unified registration mechanism.

8. The modeling and analysis method for thermal power plant data according to claim 1, characterized in that, The target operators include data segmentation operators, ideal curve generation operators, flow rate piecewise linear segmentation operators, parameter optimization operators, and deviation analysis operators; the step of scheduling the target operators to execute their corresponding target data calculation tasks according to the execution order of the calculation tasks includes: The data segmentation operator is invoked to segment the target data according to preset segmentation conditions, generating multiple sub-data. The ideal curve generation operator is invoked, the neighborhood range is determined based on the relationship between the key variables of the sub-data, and the ideal relationship curve is generated by fitting through local weighted regression. The flow rate piecewise linear segmentation operator is invoked to perform piecewise linear fitting on the flow rate characteristics of the ideal curve, and the coordinates of the feature points of each segment are extracted. The parameter optimization operator is invoked, and the coordinates of the feature points are used as constraints. The least squares method is used to perform nonlinear fitting on the ideal curve to obtain the model parameters. The theoretical model is then constructed based on the model parameters and the preset thermodynamic mechanism equation. The deviation analysis operator is invoked to compare the output value of the theoretical model with the measured value of the corresponding working point in the real-time operating data, and the deviation is calculated.

9. A modeling and analysis device for thermal power plant data, characterized in that, The device includes: The display module is used to display a graphical interface, which includes the identifier of the structured dataset obtained based on the historical operating data of the thermal power plant and the identifier of the operator resource pool that has been connected to the system. The task generation module, in response to user configuration operations on the graphical configuration interface, determines the target data in the structured dataset and multiple target operators in the operator resource pool, and constructs a directed data flow graph based on the dependencies between the target operators; and, The directed data flow graph is parsed to create a computational task corresponding to the target data; The task execution module is used to call the distributed computing engine and schedule the target operators to execute the computing tasks corresponding to their target data according to the execution order of the computing tasks, so as to generate a theoretical model reflecting the thermal characteristics of the thermal power plant and calculate the deviation of the real-time operating data relative to the theoretical model. The display module is also used to display the execution results of the computation task, including the deviation, the theoretical model, and the real-time running data.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the modeling and analysis method for thermal power plant data as described in any one of claims 1 to 8.

Images