A model calculation analysis visualization automatic layout method and system

Abstract

The present invention discloses a method and system for automatic layout of model calculation and analysis visualization, which relates to the field of graphic visualization, including: dynamically loading data from multiple data sources according to user requests, determining data types and preprocessing; classifying data to form data nodes to draw a preliminary topological map; identifying and optimizing repeated data nodes and redundant edges; analyzing the optimized topological map, and selecting a suitable frame map type for layout adjustment. The present invention uses betweenness centrality, degree centrality and eigenvector centrality to identify and process repeated data nodes, reduce the complexity of graphic layout, and avoid information redundancy; optimize the connection of edges by calculating redundancy, and improve the simplicity and aesthetics of graphic layout. The layout classification and optimization method can select the optimal layout algorithm according to different data structures to ensure the efficiency and intuitiveness of graphic layout. Finally, the graphic layout is further optimized by preventing data node overlap, optimizing cross edges and aligning data nodes.

Description

Technical Field

[0001] The present invention relates to the field of graphic visualization technology, and in particular to a model calculation analysis visualization automatic layout method and system. Background Art

[0002] With the rapid development of big data technology and computing power, data visualization has become an indispensable part of data analysis and mining. By presenting data in a graphical way, data visualization not only improves the efficiency of data analysis, but also enhances the intuitiveness and convenience of data understanding. In the field of data visualization, automatic layout technology is particularly important. Automatic layout technology aims to automatically generate the optimal graphic layout based on the structure and relationship of data, so that users can understand and analyze data more intuitively and efficiently. This technology is widely used in the generation and optimization of various graphic structures such as network diagrams, flow charts, and hierarchical diagrams. However, with the increase in data size and complexity, traditional automatic layout technology faces many challenges in processing massive and complex data.

[0003] At present, the development of automatic layout technology is mainly focused on several classic layout algorithms, such as tree layout, hierarchical layout, grid layout, etc. These algorithms have their own advantages and are widely used in different types of data visualization scenarios. For example, tree layout is suitable for data display with hierarchical structure, ring layout is often used for data display with cyclic or periodic structure, and force-directed layout is widely used in social network analysis and biological network analysis. However, the existing automatic layout technology often shows certain limitations when facing dynamic data, real-time data update and integration of multiple data sources. Specifically, the existing technology has deficiencies in the following aspects: First, the data loading and preprocessing stages lack flexibility and intelligence, and it is difficult to efficiently process multiple data sources and update them in real time. Second, the recognition and processing mechanism of duplicate data nodes and redundant edges is not perfect, which easily leads to the complexity of graph layout and information redundancy. In addition, when dealing with complex network structures, the existing layout algorithms often cannot guarantee the optimization and aesthetics of the layout, which affects users when understanding and analyzing data. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is: how to provide an intelligent automatic layout method in the face of multi-source data and dynamic data updates to ensure the efficiency and aesthetics of the display during data visualization.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a model calculation analysis visualization automatic layout method, which comprises the following steps:

[0007] Dynamically load data from multiple data sources based on user requests, determine the type of data required, and perform preprocessing;

[0008] Apply classification algorithms to classify the preprocessed data, form data nodes corresponding to the data types, and draw a preliminary topological diagram based on the relationship between the data nodes;

[0009] Identify duplicate data nodes and redundant edges, and optimize the preliminary topology graph;

[0010] Analyze the optimized topology diagram, automatically select the appropriate frame diagram type, and make layout adjustments;

[0011] The dynamic loading of data is based on the specific request of the user, and the user request is analyzed by natural language processing technology to determine the major categories of data, format and priority of data sources. In the data loading stage, the intelligent selection algorithm is used to dynamically select the appropriate data source for data capture based on the real-time response of the data source and data quality assessment. After the data is transmitted through the network, it enters preprocessing to remove duplicate and irrelevant data, correct data input errors, format all data, and perform data verification to ensure data integrity and logical consistency;

[0012] The intelligent selection algorithm performs real-time evaluation based on the performance indicators provided by each data source, calculates the comprehensive score of each data source through preset standards, and dynamically determines the data capture task to be assigned to the data source with the highest score based on the obtained score.

[0013] As a preferred solution of the model calculation analysis visualization automatic layout method described in the present invention, wherein: the classification of the preprocessed data is to extract key features from the preprocessed data, and classify the data into multiple specific demand categories according to the features, each specific demand category represents a research direction, and create corresponding data nodes for the data of each specific demand category, and assign a unique identifier to each data node and record its category attributes, and use a graphic algorithm to automatically draw a preliminary topological map according to the predefined relationship rules between the data nodes;

[0014] The expression of the topology graph is:

[0015] ,

[0016] in, is the dataset, including all preprocessed and classified data points, is an edge set, established according to specific relational rules;

[0017] If each pair of data nodes , there is a relationship according to the predefined rules, then and Add an edge between ;

[0018] The expression of edge set is:

[0019] .

[0020] As a preferred solution of the model calculation analysis visualization automatic layout method described in the present invention, wherein: the identification of duplicate data nodes and redundant edges includes identifying and processing duplicate data nodes;

[0021] Identify potential duplicate data nodes through data node attributes, comprehensively evaluate the number of connected edges and the role of each duplicate data node in the graph, and determine whether to retain the current duplicate data nodes based on betweenness centrality, degree centrality, and eigenvector centrality. For data nodes that need to be merged, select representative data nodes, update related edges, and delete redundant data nodes. For data nodes that need to be retained, maintain connection relationships and unique attributes.

[0022] The betweenness centrality, degree centrality and eigenvector centrality are expressed as:

[0023] ,

[0024] ,

[0025] ,

[0026] in, For data nodes The betweenness centrality of It is a data node and The number of all shortest paths between Through the data node The number of shortest paths, For data nodes The degree centrality of It is a data node degree, connected to The number of edges, For data nodes The eigenvector centrality of is the eigenvalue constant, For data nodes The set of neighbor data nodes, Represents a data node and Is there an edge between them? is the eigenvector centrality of the neighbor data node;

[0027] Use betweenness centrality, degree centrality and eigenvector centrality to evaluate the importance of data nodes, set the first comprehensive threshold, if any of the betweenness centrality, degree centrality and eigenvector centrality is greater than the first comprehensive threshold, then the current data node is judged to play a key role in the process, retain the current duplicate data node, otherwise delete the current duplicate data node.

[0028] As a preferred solution of the model calculation analysis visualization automatic layout method described in the present invention, wherein: the identifying of duplicate data nodes and redundant edges also includes identifying redundant edges;

[0029] Use the Dijkstra algorithm to calculate the shortest path between all data node pairs and identify the critical path. If the frequency of all edges in a path is greater than the set second comprehensive threshold, the path is judged to be a critical path. For the edges on the critical path, the current edges are retained; for the edges on the non-critical path, further determine whether they are redundant edges.

[0030] By calculating the redundancy, if the redundancy of an edge is higher than the set first redundancy threshold, the redundant edge is deleted. The expression is:

[0031] ,

[0032] in, For edge The redundancy of is the set of all paths, is the indicator function, if the edge On the path If it appears in ,otherwise, ;

[0033] By deleting duplicate data nodes and duplicate edges, an optimized topology graph is output.

[0034] As a preferred solution of the model calculation analysis visualization automatic layout method described in the present invention, wherein: the topological graph after the analysis optimization automatically selects the appropriate framework graph type by dynamically calculating the data node positions and edges in the graph according to the data node attributes and relationships;

[0035] The types of framework diagrams include tree layout, hierarchical layout, and grid layout;

[0036] If the preliminary topology graph is a non-closed-loop connected graph, there are no isolated subgraphs or independent data nodes, n data nodes correspond to n−1 edges, there is only one connecting edge between any two data nodes, the proportion of high centrality data nodes in the preliminary topology graph is less than the first high centrality threshold, and the clustering coefficient of the preliminary topology graph is less than the first clustering threshold, then it is classified as a tree layout;

[0037] The proportion of high centrality data nodes being less than the first high centrality threshold includes calculating the degree of each data node, the average degree of the preliminary topology graph, and the standard deviation of the average degree, and the expression is:

[0038] ,

[0039] ,

[0040] ,

[0041] in, For data nodes The degree, is the average degree of the preliminary topological graph, is the standard deviation of the mean;

[0042] The statistical degree is higher than The number of data nodes is expressed as:

[0043] ,

[0044] ,

[0045] in, is an exponential function. When the condition is met, the value is 1, otherwise it is 0. is the proportion of high centrality data nodes;

[0046] when If it is less than the first high centrality threshold, we further determine whether the clustering coefficient of the preliminary topology graph is less than the first clustering threshold. The expression is:

[0047] ,

[0048] ,

[0049] in, For data nodes The clustering coefficient, For data nodes The number of edges that actually exist between neighbor data nodes, It is a data node The number of neighbor data nodes, is the average clustering coefficient of the preliminary topological graph;

[0050] If the preliminary topology graph is an undirected non-closed-loop graph or a directed non-closed-loop graph, the in-degree of the root data node is 0, there is only an edge from the upper layer data node to the lower layer data node between any two data nodes, the number of data nodes on each layer is evenly distributed, the variance of the number of data nodes on each layer is less than the first variance threshold, and there is no reverse edge or cross-layer edge, then it is classified as a hierarchical layout;

[0051] The in-degree is 0, which means the number of edges entering the current data node is 0;

[0052] The hierarchy is to determine all root data nodes or source data node sets with an in-degree of 0 in the preliminary topology graph, and use breadth-first search to calculate the hierarchy of each data node, where the hierarchy is the shortest path length from the root data node to the data node;

[0053] The variance of the number of data nodes at each level is less than the first variance threshold, and the expression is:

[0054] ,

[0055] ,

[0056] in, is the variance of the number of data nodes at each level, is the number of data nodes at each level, is the maximum number of levels, is the average number of data nodes per layer;

[0057] If the number of connected edges of any data node in the preliminary topology graph is no more than 4 and there is no isolated data node, it is determined to be a grid layout;

[0058] The first comprehensive threshold, the second comprehensive threshold, the first redundancy threshold, the first high centrality threshold, the first clustering threshold, and the first variance threshold are set according to the historical records of user requests.

[0059] As a preferred solution of the model calculation analysis visualization automatic layout method described in the present invention, the layout adjustment is to prevent data node overlap, optimize edges, and adjust data node alignment according to the framework diagram after classification;

[0060] The method of preventing data nodes from overlapping is to detect the positions of data nodes, calculate the distance between each pair of adjacent data nodes, and if the distance is less than a preset minimum distance, apply a repulsive force to separate the data nodes and adjust their positions, and repeat the detection and adjustment until the distance between all data nodes is greater than or equal to the minimum distance;

[0061] The edge optimization is to use an edge crossing detection algorithm to find the crossing edges, and adjust the data node positions or re-plan the edge paths to reduce the edge crossing points;

[0062] The data node alignment is to determine an alignment rule, calculate the target position of each data node according to the alignment rule, and adjust the data node position to align with the target position;

[0063] The layout adjustment result is output and rendered to the user interface, and the Canvas technology is used for rendering to complete the visual automatic layout.

[0064] Another object of the present invention is to provide a model calculation analysis visualization automatic layout system, which can dynamically load data through an intelligent selection algorithm, comprehensively use betweenness centrality, degree centrality and eigenvector centrality to identify and process duplicate data nodes, calculate redundancy to optimize edge connections, and select the optimal layout according to different data structures, thereby solving the problems of existing automatic layout technology in terms of flexibility, information redundancy and layout aesthetics.

[0065] In order to solve the above technical problems, the present invention provides the following technical solutions: a model calculation analysis visualization automatic layout system, comprising: a data loading module, a data preprocessing module, an optimization module, a layout classification module and a visualization rendering module;

[0066] The data loading module dynamically loads data from multiple data sources according to user requests, parses user requests, determines the data type, format and priority of data sources, performs real-time evaluation based on the performance indicators provided by each data source, and dynamically selects the appropriate data source for data capture;

[0067] The data preprocessing module removes duplication, corrects, formats and verifies the loaded data, extracts key features from the preprocessed data, divides the data into multiple categories according to the features, creates corresponding data nodes, and automatically draws a preliminary topological map using a graphic algorithm according to predefined relationship rules between data nodes;

[0068] The optimization module identifies potential duplicate data nodes through data node attributes, comprehensively evaluates the number of connected edges of each duplicate data node, uses betweenness centrality, degree centrality and eigenvector centrality to determine whether the current duplicate data node should be retained, calculates redundancy, identifies and deletes redundant edges, and reduces or eliminates cross edges by rearranging data node positions or reconnecting edges;

[0069] The layout classification module analyzes the optimized topology graph, automatically selects the appropriate frame graph type, dynamically calculates the data node positions and edges in the graph according to the data node attributes and relationships, and adjusts the layout according to the classification rules;

[0070] The visualization rendering module detects the position of data nodes, calculates the distance between each pair of adjacent data nodes, and if the distance is less than the preset minimum distance, applies repulsive force to separate the data nodes and adjust their positions. It uses an edge crossing detection algorithm to find the crossed edges, and adjusts the position of data nodes or replans the path of the edges to reduce the intersection points of the edges, determines the alignment rules, calculates the target position of each data node according to the alignment rules, adjusts the data node position to align with the target position, outputs the final layout result and renders it to the user interface, uses Canvas technology for rendering, and completes the visualization automatic layout.

[0071] A computer device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the above-mentioned model calculation analysis visualization automatic layout method are implemented.

[0072] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned model calculation analysis visualization automatic layout method.

[0073] Beneficial effects of the present invention: The present invention dynamically loads data through an intelligent selection algorithm to ensure efficient selection and real-time updating of data sources, solving the problem of insufficient flexibility and intelligence in the data loading and preprocessing stages in the prior art. The betweenness centrality, degree centrality and eigenvector centrality are used comprehensively to identify and process duplicate data nodes, effectively reducing the redundancy of the graph layout and avoiding information redundancy. By calculating the redundancy to optimize the connection of edges, the simplicity and aesthetics of the graph layout are greatly improved, ensuring the clarity and readability of the graph. The proposed layout classification and optimization method can select the optimal layout algorithm according to different data structures, ensuring the efficiency and intuitiveness of the graph layout. By preventing data node overlap, optimizing cross edges and data node alignment, the display effect of the graph layout is further optimized, making the final visualization result neater and more standardized. The present invention not only solves the shortcomings of existing automatic layout technology in terms of flexibility, information redundancy and layout aesthetics, but also provides users with an efficient and intelligent automatic layout method, significantly improving the efficiency and user experience of big data analysis and display. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0075] Figure 1An overall flow chart of a model calculation analysis visualization automatic layout method provided by the first embodiment of the present invention;

[0076] Figure 2 A structural diagram of a model calculation analysis visualization automatic layout system provided by a second embodiment of the present invention;

[0077] Figure 3 The figure is a preliminary topological diagram corresponding to a hierarchical layout diagram of a model calculation analysis visualization automatic layout method provided by the third embodiment of the present invention;

[0078] Figure 4 The figure is a hierarchical layout diagram of a model calculation analysis visualization automatic layout method provided by the third embodiment of the present invention;

[0079] Figure 5 The figure is a preliminary topological diagram corresponding to a grid layout diagram of a model calculation analysis visualization automatic layout method provided by the third embodiment of the present invention;

[0080] Figure 6 The figure is a grid layout diagram of a model calculation analysis visualization automatic layout method provided by the third embodiment of the present invention;

[0081] Figure 7 The figure is a preliminary topological diagram corresponding to a tree layout diagram of a model calculation analysis visualization automatic layout method provided by the third embodiment of the present invention;

[0082] Figure 8 The figure is a tree layout diagram of a model calculation analysis visualization automatic layout method provided by the third embodiment of the present invention. DETAILED DESCRIPTION

[0083] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0084] Example 1, reference Figure 1 According to an embodiment of the present invention, a model calculation analysis visualization automatic layout method is provided, comprising:

[0085] Dynamically load data from multiple data sources based on user requests, determine the type of data required, and perform pre-processing.

[0086] The classification algorithm is applied to classify the preprocessed data, forming data nodes of corresponding types of data, and drawing a preliminary topological diagram based on the relationship between data nodes.

[0087] Identify duplicate data nodes and redundant edges, and optimize the preliminary topology graph.

[0088] Analyze the optimized topology diagram, automatically select the appropriate frame diagram type, and make layout adjustments.

[0089] Dynamic data loading is based on the user's specific request. Natural language processing technology is used to parse the user's request to determine the major categories of data, format and priority of data sources. In the data loading stage, intelligent selection algorithms are used to dynamically select appropriate data sources for data capture based on the real-time response of the data source and data quality assessment. After the data is transmitted over the network, it enters preprocessing to remove duplicate and irrelevant data, correct data input errors, format all data, and perform data verification to ensure data integrity and logical consistency.

[0090] The intelligent selection algorithm performs real-time evaluation based on the performance indicators provided by each data source. It calculates the comprehensive score of each data source through preset standards and dynamically determines the data capture task to be assigned to the data source with the highest score based on the obtained score.

[0091] It is further explained that:

[0092] The intelligent selection algorithm evaluates each data source in real time based on predefined criteria to determine the suitability of each data source. These criteria include, but are not limited to, the data source's response time, data integrity, update frequency, accessibility, and historical reliability. The algorithm converts these criteria into quantitative scoring indicators:

[0093] The average time it takes for a data source to respond to a query request. A shorter response time means a higher score. The ratio of complete and correct data provided by a data source means a data source with higher integrity means a higher score.

[0094] The frequency with which the data source updates data; the more frequent the update, the higher the score; the availability ratio of the data source within a given time; the higher the accessibility, the higher the score; based on the evaluation of past performance, data sources with historically stable performance have higher scores.

[0095] These indicators are combined into a comprehensive score through a weighted algorithm, and the weights are pre-set by the system administrator based on business needs and the characteristics of the data source. The intelligent selection algorithm uses this comprehensive score to sort the data sources and dynamically select the data source with the highest score for data crawling. If the highest-scoring data source fails to crawl, the algorithm will automatically select the next higher-scoring data source to try until the data is successfully crawled or all options fail.

[0096] Classification of preprocessed data is to extract key features from the preprocessed data, and divide the data into multiple specific demand categories based on the features. Each specific demand category represents a research direction. Corresponding data nodes are created for the data of each specific demand category, and a unique identifier is assigned to each data node and its category attributes are recorded. According to the predefined relationship rules between data nodes, a preliminary topology map is automatically drawn using a graphics algorithm.

[0097] The expression of the topology graph is:

[0098] ,

[0099] in, is the dataset, including all preprocessed and classified data points, is an edge set, established according to specific relational rules;

[0100] If each pair of data nodes , there is a relationship according to the predefined rules, then and Add an edge between ;

[0101] The expression of edge set is:

[0102] .

[0103] It is further explained that:

[0104] User requests are parsed through natural language processing technology to determine the major categories of data, such as wind farms, hydropower plants, new energy stations, nuclear power plants, etc.

[0105] Taking the new energy station as an example, we can further clarify the types of user needs, such as cost-benefit, environmental impact, and operating efficiency of operating equipment.

[0106] Take cost-benefit as an example to create corresponding data nodes, such as equipment investment, maintenance costs, power generation income, etc.

[0107] Specific data: equipment investment cost, maintenance costs, power generation income, etc.

[0108] Identifying duplicate data nodes and redundant edges includes identifying and processing duplicate data nodes;

[0109] Identify potential duplicate data nodes through data node attributes, comprehensively evaluate the number of connected edges and the role of each duplicate data node in the graph, and determine whether the current duplicate data nodes should be retained based on betweenness centrality, degree centrality, and eigenvector centrality. For data nodes that need to be merged, select representative data nodes, update related edges, and delete redundant data nodes. For data nodes that need to be retained, maintain connection relationships and unique attributes.

[0110] Betweenness centrality, degree centrality, and eigenvector centrality are expressed as:

[0111] ,

[0112] ,

[0113] ,

[0114] in, For data nodes The betweenness centrality of It is a data node and The number of all shortest paths between Through the data node The number of shortest paths, For data nodes The degree centrality of It is a data node degree, connected to The number of edges, For data nodes The eigenvector centrality of is the eigenvalue constant, For data nodes The set of neighbor data nodes, Represents a data node and Is there an edge between them? is the eigenvector centrality of the neighbor data node.

[0115] It should be further explained that A is the adjacency matrix, which represents the connection relationship between the nodes in the graph.

[0116] Assuming the graph has 4 nodes, the adjacency matrix can be expressed as:

[0117] ,

[0118] First row and first column:

[0119] A11=0: There is no edge between node 1 and itself, A12=1: There is an edge between node 1 and node 2, A13=0: There is no edge between node 1 and node 3, A14=0: There is no edge between node 1 and node 4.

[0120] Second row and second column:

[0121] A21=1: There is an edge between node 2 and node 1, A22=0: There is no edge between node 2 and itself, A23=1: There is an edge between node 2 and node 3, A24=1: There is an edge between node 2 and node 4.

[0122] Third row and third column:

[0123] A31=0: There is no edge between node 3 and node 1, A32=1: There is an edge between node 3 and node 2, A33=0: There is no edge between node 3 and itself, A34=1: There is an edge between node 3 and node 4.

[0124] Fourth row and fourth column:

[0125] A41=0: There is no edge between node 4 and node 1, A42=1: There is an edge between node 4 and node 2, A43=1: There is an edge between node 4 and node 3, A44=0: There is no edge between node 4 and itself.

[0126] Use betweenness centrality, degree centrality, and eigenvector centrality to evaluate the importance of data nodes, set the first comprehensive threshold, and if any of the betweenness centrality, degree centrality, and eigenvector centrality is greater than the first comprehensive threshold, then the current data node is judged to have a key role in the process, and the current duplicate data node is retained, otherwise the current duplicate data node is deleted. Dynamically calculate the data node position and edges in the graph based on the data node attributes and relationships;

[0127] Frame diagram types include tree layout, hierarchical layout, and grid layout.

[0128] If the preliminary topology graph is a non-closed-loop connected graph, there are no isolated subgraphs or independent data nodes, n data nodes correspond to n−1 edges, there is only one connecting edge between any two data nodes, the proportion of high centrality data nodes in the preliminary topology graph is less than the first high centrality threshold, and the clustering coefficient of the preliminary topology graph is less than the first clustering threshold, then it is classified as a tree layout.

[0129] The proportion of high centrality data nodes is less than the first high centrality threshold. This includes calculating the degree of each data node, the average degree of the preliminary topology graph, and the standard deviation of the average degree. The expression is:

[0130] ,

[0131] ,

[0132] ,

[0133] in, For data nodes The degree, is the average degree of the preliminary topological graph, is the standard deviation of the mean;

[0134] The statistical degree is higher than The number of data nodes is expressed as:

[0135] ,

[0136] ,

[0137] in, is an exponential function. When the condition is met, the value is 1, otherwise it is 0. is the proportion of high centrality data nodes;

[0138] when If it is less than the first high centrality threshold, we further determine whether the clustering coefficient of the preliminary topology graph is less than the first clustering threshold. The expression is:

[0139] ,

[0140] ,

[0141] in, For data nodes The clustering coefficient, For data nodes The number of edges that actually exist between neighbor data nodes, It is a data node The number of neighbor data nodes, is the average clustering coefficient of the preliminary topological graph.

[0142] If the preliminary topology graph is an undirected non-closed-loop graph or a directed non-closed-loop graph, the in-degree of the root data node is 0, there is only an edge from the upper-layer data node to the lower-layer data node between any two data nodes, the number of data nodes on each level is evenly distributed, the variance of the number of data nodes on each level is less than the first variance threshold, and there are no reverse edges and cross-layer edges, then it is classified as a hierarchical layout.

[0143] An in-degree of 0 means that the number of edges entering the current data node is 0.

[0144] The hierarchy is to determine the set of all root data nodes or source data nodes with in-degree 0 in the preliminary topology graph, and use breadth-first search to calculate the hierarchy of each data node, where the hierarchy is the shortest path length from the root data node to the data node.

[0145] The variance of the number of data nodes at each level is less than the first variance threshold, and the expression is:

[0146] ,

[0147] ,

[0148] in, is the variance of the number of data nodes at each level, is the number of data nodes at each level, is the maximum number of levels, is the average number of data nodes per layer.

[0149] If the number of connected edges of any data node in the preliminary topology graph is no more than 4 and there is no isolated data node, it is determined to be a grid layout.

[0150] The first comprehensive threshold, the second comprehensive threshold, the first redundancy threshold, the first high centrality threshold, the first clustering threshold, and the first variance threshold are set according to the historical records of user requests.

[0151] Layout adjustments include preventing data node overlap, optimizing edges, and aligning data nodes.

[0152] Preventing data nodes from overlapping is to detect the data node positions and calculate the distance between each pair of adjacent data nodes. If the distance is less than the preset minimum distance, a repulsive force is applied to separate the data nodes and adjust their positions. Repeat the detection and adjustment until the distance between all data nodes is greater than or equal to the minimum distance.

[0153] Edge optimization uses an edge crossing detection algorithm to find crossed edges, and adjusts the data node positions or replans the edge paths to reduce edge crossings.

[0154] Data node alignment is to determine the alignment rules, calculate the target position of each data node according to the alignment rules, and adjust the data node position to align with the target position.

[0155] Output the final layout result and render it on the user interface.

[0156] Example 2, reference Figure 2 , as an embodiment of the present invention, provides a system for a model calculation analysis visualization automatic layout method, including: a data loading module, a data preprocessing module, an optimization module, a layout classification module and a visualization rendering module.

[0157] The data loading module dynamically loads data from multiple data sources based on user requests, parses user requests, determines the data type, format, and priority of data sources, performs real-time evaluation based on the performance indicators provided by each data source, and dynamically selects the appropriate data source for data capture.

[0158] The data preprocessing module deduplicates, corrects, formats and verifies the loaded data, extracts key features from the preprocessed data, divides the data into multiple categories based on the features, creates corresponding data nodes, and uses graphics algorithms to automatically draw a preliminary topology map based on predefined relationship rules between data nodes.

[0159] The optimization module identifies potential duplicate data nodes through data node attributes, comprehensively evaluates the number of connected edges of each duplicate data node, uses betweenness centrality, degree centrality and eigenvector centrality to determine whether the current duplicate data nodes should be retained, calculates redundancy, identifies and deletes redundant edges, and reduces or eliminates cross edges by rearranging node positions or reconnecting edges.

[0160] The layout classification module analyzes the optimized topology graph, automatically selects the appropriate frame graph type, dynamically calculates the node positions and edges in the graph based on node attributes and relationships, and adjusts the layout according to classification rules.

[0161] The visualization rendering module detects the node position and calculates the distance between each pair of adjacent nodes. If the distance is less than the preset minimum distance, a repulsive force is applied to separate the nodes and adjust their positions. The edge crossing detection algorithm is used to find the crossed edges, and the node positions are adjusted or the edge paths are replanned to reduce the intersection points of the edges. The alignment rules are determined, and the target position of each node is calculated according to the alignment rules. The node position is adjusted to align with the target position, and the final layout result is output and rendered to the user interface. Canvas technology is used for rendering to complete the visualization automatic layout.

[0162] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0163] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0164] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0165] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0166] Example 3, reference Figure 3~Figure 8 In this embodiment, in order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments. This embodiment experiments the existing traditional method and the method of this embodiment respectively.

[0167] With the changes in the power market, most power plants are currently carrying out coal blending and combustion, wind and smoke system transformation, etc., resulting in insufficient fan speed margin in power plants. Once improper operation or excessive adjustment, the probability of axial flow fans stalling is greatly increased.

[0168] The essence of fan stall is that the fan operating point falls into the stall zone. Due to the irregularity of the stall zone and the deviation of the characteristic curve from the actual situation on site, it is difficult for operators to judge the occurrence of accidents in advance through on-site parameters.

[0169] Generally, corresponding accident handling measures are taken on site based on the parameter changes after the fan stalls, but they are essentially post-event handling, and the only difference is how early or late the accident is discovered.

[0170] Reference Figure 3 It is based on the user's specific request, analyzes the user's request through natural language processing technology, determines the type of data and specific needs, creates corresponding data nodes, and outputs a preliminary topology map.

[0171] Reference Figure 4 The present invention optimizes the preliminary topology diagram and classifies the block diagram to construct a hierarchical layout diagram.

[0172] In a power plant, the heat rate of a steam turbine is one of the key indicators of its performance. The heat rate reflects the ability of a steam turbine to efficiently convert heat energy into electrical energy. A lower heat rate indicates that the conversion process is efficient and there is less energy loss, while a higher heat rate means that the energy conversion efficiency is low and there is a large energy loss. This will not only directly lead to an increase in power generation costs and a decrease in economic benefits, but may also cause additional wear and failure risks for equipment. A sustained high heat rate may indicate problems such as equipment aging, improper maintenance, or unoptimized operating parameters. Therefore, real-time monitoring of the heat rate of the steam turbine and timely warning when it increases abnormally are important measures to ensure the operating efficiency and equipment safety of the power plant. Through an effective monitoring and early warning mechanism, appropriate adjustments and maintenance measures can be taken when problems first occur to prevent small problems from turning into major failures, thereby ensuring the efficient and stable operation of the power plant.

[0173] Reference Figure 5 , Figure 6 According to the user's specific request, natural language processing technology is used to parse the user's request, determine the type of data and specific needs, create corresponding data nodes, output the preliminary topology map and construct the grid layout map.

[0174] It is further explained that:

[0175] Get the real-time value of steam turbine heat rate: Node: Real-time value of steam turbine heat rate.

[0176] Explanation: Obtain the current heat rate value of the turbine in real time from the monitoring system.

[0177] Steam turbine heat rate threshold acquisition: Node: Steam turbine heat rate optimal value search.

[0178] Explanation: The ideal threshold value (benchmark value) of the steam turbine heat rate is obtained from historical data or design parameters.

[0179] Difference calculation: Node: Subtractor.

[0180] Explanation: The difference between the real-time heat rate and the threshold is calculated to obtain the heat rate deviation.

[0181] Deviation judgment: Node: Comparator.

[0182] Explanation: Compare the heat rate deviation with a preset value (e.g. 200) to determine if the deviation exceeds the set range.

[0183] Signal hold: Node: Signal hold.

[0184] Explanation: If the heat rate deviation exceeds the set value, this signal is maintained to ensure that the system can detect this abnormal situation.

[0185] Unit load acquisition: Node: Unit load.

[0186] Explanation: Get the current turbine load data from the monitoring system.

[0187] Load stability judgment: Node: Unit load stability judgment.

[0188] Explanation: Determine whether the current load is stable. If the load fluctuates greatly, it may affect the accuracy of the heat rate, so it is necessary to confirm that the load is stable.

[0189] Conditional merge: Node: AND.

[0190] Explanation: Combine the two conditions of heat rate deviation and load stability. If both conditions are met (i.e., the heat rate deviation is large and the load is stable), the warning is triggered.

[0191] Steam turbine heat rate high warning: Node: Steam turbine heat rate high warning.

[0192] Explanation: When all the above conditions are met, the system generates a high heat rate warning signal to remind the operator to take corresponding measures, such as checking equipment, adjusting operating parameters, etc.

[0193] In the modern power production process, steam turbines and coal mills are important equipment in power plants, which directly affect the operating efficiency and stability of power plants. Steam turbines convert the thermal energy of steam into mechanical energy and then into electrical energy. Their operating status is directly related to the efficiency and safety of power generation. Coal mills are responsible for grinding coal blocks into coal powder. They are one of the core equipment of coal-fired power plants. Their performance and operating status directly affect the combustion efficiency and environmental emissions of boilers. Therefore, real-time monitoring and data analysis of the operating status of steam turbines and coal mills can effectively improve the overall operating efficiency of power plants, reduce equipment failure rates, and reduce maintenance costs.

[0194] Reference Figure 7 , Figure 8 According to the user's specific request, natural language processing technology is used to parse the user's request, determine the type of data and specific needs, create the corresponding data nodes, output the preliminary topology map and construct the corresponding tree layout map.

[0195] Time series data definition: Node: Time series data definition.

[0196] Explanation: Defines the format and structure of time series data collected from various sensors and devices. For example: temperature, pressure, vibration, etc.

[0197] Data collection: Node: turbine data / coal mill data.

[0198] Explanation: Data is collected from actual equipment through sensors such as temperature and pressure.

[0199] Data cleaning: Value range cleaning: Node: Value range cleaning.

[0200] Explanation: Filter out unreasonable values ​​in the data, such as temperature sensor readings outside the normal range.

[0201] Null value cleaning: Node: Null value cleaning.

[0202] Explanation: Handle null values ​​in data, such as filling missing data or deleting records containing null values.

[0203] Data output and visualization: Nodes: data report, rainbow chart, data view, histogram group, data trend chart.

[0204] Interpretation: Generate various types of reports and charts based on the cleaned data to help analysis and decision-making.

[0205] Data Report: Generate detailed statistical reports.

[0206] Rainbow chart: shows the distribution of multidimensional data.

[0207] Data view: displays the processed data in table form.

[0208] Histogram group: displays the frequency distribution of data.

[0209] Data trend chart: shows the changing trend of data over time.

[0210] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A model calculation analysis visualization automatic layout method, characterized in that: include: Dynamically load data from various operating data sources of steam turbines in power plants according to user requests, determine the type of data required, and perform preprocessing; Apply classification algorithms to classify the preprocessed data, form data nodes corresponding to the data types, and draw a preliminary topological diagram based on the relationship between the data nodes; Identify duplicate data nodes and redundant edges, and optimize the preliminary topology graph; Analyze the optimized topology diagram, automatically select the appropriate frame diagram type, and make layout adjustments; Output the layout adjustment results and render them to the user interface, use Canvas technology for rendering, and complete the visual automatic layout; The dynamic loading of data is based on the specific request of the user, and the user request is analyzed by natural language processing technology to determine the major categories of data, format and priority of data sources. In the data loading stage, the intelligent selection algorithm is used to dynamically select the appropriate data source for data capture based on the real-time response of the data source and data quality assessment. After the data is transmitted through the network, it enters preprocessing to remove duplicate and irrelevant data, correct data input errors, format all data, and perform data verification to ensure data integrity and logical consistency; The intelligent selection algorithm is based on real-time evaluation of the performance indicators provided by each data source, and calculates the comprehensive score of each data source through preset standards. According to the obtained score, the data crawling task is dynamically determined to be assigned to the data source with the highest score; The classification of the preprocessed data is to extract key features from the preprocessed data, classify the data into multiple specific demand categories according to the features, each specific demand category represents a research direction, create corresponding data nodes for the data of each specific demand category, assign a unique identifier to each data node and record its category attributes, and automatically draw a preliminary topology map using a graphic algorithm according to predefined relationship rules between data nodes; The expression of the topology graph is: G=(V,E) Among them, V is the data set, including all preprocessed and classified data points, and E is the edge set, which is established according to specific relationship rules; If each pair of data nodes u, v∈V has a relationship according to the predefined rules, then add an edge (u, v)∈E between u and v; The expression of edge set is: E = {(u,v) | u,v∈V and there exists a predefined relation rule connecting u and v}; The identifying of duplicate data nodes and redundant edges includes identifying and processing duplicate data nodes; Identify potential duplicate data nodes through data node attributes, comprehensively evaluate the number of connected edges and the role of each duplicate data node in the graph, and determine whether to retain the current duplicate data nodes based on betweenness centrality, degree centrality, and eigenvector centrality. For data nodes that need to be merged, select representative data nodes, update related edges, and delete redundant data nodes. For data nodes that need to be retained, maintain connection relationships and unique attributes. The betweenness centrality, degree centrality and eigenvector centrality are expressed as: C D (v)=d(v) Among them, C B (v) is the betweenness centrality of data node v, σ st is the number of all shortest paths between data nodes s and t, σ st (v) is the number of shortest paths through data node v, C D (v) is the degree centrality of data node v, d(v) is the degree of data node v, the number of edges connected to v, C E (v) is the eigenvector centrality of data node v, λ is the eigenvalue constant, N(v) is the set of neighboring data nodes of data node v, A uv Indicates whether there is an edge between data nodes u and v, C E is the eigenvector centrality of the neighbor data node; The importance of the data node is evaluated using betweenness centrality, degree centrality and eigenvector centrality, and a first comprehensive threshold is set. If any one of the betweenness centrality, degree centrality and eigenvector centrality is greater than the first comprehensive threshold, the current data node is judged to have a key role in the process, and the current duplicate data node is retained, otherwise the current duplicate data node is deleted; the identifying of duplicate data nodes and redundant edges also includes identifying redundant edges; The Dijkstra algorithm is used to calculate the shortest path between all data node pairs and identify the critical path. If the frequency of occurrence of all edges in a path is greater than the set second comprehensive threshold, the path is judged to be a critical path. For the edges on the critical path, the current edges are retained; for the edges on the non-critical path, further judgment is made as to whether they are redundant edges. By calculating the redundancy, if the redundancy of an edge is higher than the set first redundancy threshold, the redundant edge is deleted. The expression is: Where R(e) is the redundancy of edge e, P is the set of all paths, δ(p,e) is the indicator function, if edge e appears in path p, then δ(p,e) = 1, otherwise, δ(p,e) = 0; Output the optimized topology graph by deleting duplicate data nodes and duplicate edges; The topological graph after analysis and optimization, automatically selecting the appropriate frame graph type is to dynamically calculate the data node positions and edges in the graph according to the data node attributes and relationships; The types of framework diagrams include tree layout, hierarchical layout, and grid layout; If the preliminary topology graph is a non-closed-loop connected graph, has no isolated subgraphs or independent data nodes, has n-1 edges corresponding to n data nodes, has only one connecting edge between any two data nodes, has a high centrality data node ratio less than the first high centrality threshold, and has a clustering coefficient less than the first clustering threshold, then it is classified as a tree layout; The proportion of high centrality data nodes being less than the first high centrality threshold includes calculating the degree of each data node, the average degree of the preliminary topology graph, and the standard deviation of the average degree, and the expression is: Where d(v) is the degree of data node v, μ is the average degree of the preliminary topology graph, and σ is the standard deviation of the average degree; The number of data nodes in the preliminary topological graph with a statistical degree higher than μ+kσ is expressed as: Among them, δ is an exponential function. When the condition is met, the value is 1, otherwise it is 0. h is the proportion of high centrality data nodes; When P h If it is less than the first high centrality threshold, we further determine whether the clustering coefficient of the preliminary topology graph is less than the first clustering threshold. The expression is: Among them, C v is the clustering coefficient of data node v, E v is the number of edges actually existing between neighboring data nodes of data node v, k is the number of neighboring data nodes of data node v, and C is the average clustering coefficient of the preliminary topological graph; If the preliminary topology graph is an undirected non-closed-loop graph or a directed non-closed-loop graph, the in-degree of the root data node is 0, there is only an edge from the upper layer data node to the lower layer data node between any two data nodes, the number of data nodes on each layer is evenly distributed, the variance of the number of data nodes on each layer is less than the first variance threshold, and there is no reverse edge or cross-layer edge, then it is classified as a hierarchical layout; The in-degree is 0, which means the number of edges entering the current data node is 0; The hierarchy is to determine all root data nodes or source data node sets with an in-degree of 0 in the preliminary topology graph, and use breadth-first search to calculate the hierarchy of each data node, where the hierarchy is the shortest path length from the root data node to the data node; The variance of the number of data nodes at each level is less than the first variance threshold, and the expression is: in, is the variance of the number of data nodes at each level, |V L |The number of data nodes at each level, L max is the maximum number of levels, μ L is the average number of data nodes per layer; If the number of connected edges of any data node in the preliminary topology graph is no more than 4 and there is no isolated data node, it is determined to be a grid layout; The first comprehensive threshold, the second comprehensive threshold, the first redundancy threshold, the first high centrality threshold, the first clustering threshold, and the first variance threshold are set according to the historical records of user requests.

2. A model calculation analysis visualization automatic layout method as claimed in claim 1, characterized in that: The layout adjustment is to prevent data node overlap, optimize edges, and adjust data node alignment according to the framework diagram after classification. The method of preventing data nodes from overlapping is to detect the positions of data nodes, calculate the distance between each pair of adjacent data nodes, and if the distance is less than a preset minimum distance, apply a repulsive force to separate the data nodes and adjust their positions, and repeat the detection and adjustment until the distance between all data nodes is greater than or equal to the minimum distance; The edge optimization is to use an edge crossing detection algorithm to find the crossing edges, and adjust the data node positions or re-plan the edge paths to reduce the edge crossing points; The data node alignment is to determine an alignment rule, calculate a target position of each data node according to the alignment rule, and adjust the data node position to align with the target position.

3. A system using a model calculation analysis visualization automatic layout method as claimed in any one of claims 1 to 2, characterized in that: It includes data loading module, data preprocessing module, optimization module, layout classification module and visual rendering module; The data loading module dynamically loads data from various operating data sources of steam turbines in the power plant according to user requests, parses user requests, determines the data type, format and priority of data sources, performs real-time evaluation based on the performance indicators provided by each data source, and dynamically selects appropriate data sources for data capture; The data preprocessing module removes duplication, corrects, formats and verifies the loaded data, extracts key features from the preprocessed data, divides the data into multiple categories according to the features, creates corresponding data nodes, and automatically draws a preliminary topological map using a graphic algorithm according to predefined relationship rules between data nodes; The optimization module identifies potential duplicate data nodes through data node attributes, comprehensively evaluates the number of connected edges of each duplicate data node, uses betweenness centrality, degree centrality and eigenvector centrality to determine whether the current duplicate data node should be retained, calculates redundancy, identifies and deletes redundant edges, and reduces or eliminates cross edges by rearranging data node positions or reconnecting edges; The layout classification module analyzes the optimized topology graph, automatically selects the appropriate frame graph type, dynamically calculates the data node positions and edges in the graph according to the data node attributes and relationships, and adjusts the layout according to the classification rules; The visualization rendering module detects the position of data nodes, calculates the distance between each pair of adjacent data nodes, and if the distance is less than the preset minimum distance, applies repulsive force to separate the data nodes and adjust their positions. It uses an edge crossing detection algorithm to find the crossed edges, and adjusts the position of data nodes or replans the path of the edges to reduce the intersection points of the edges, determines the alignment rules, calculates the target position of each data node according to the alignment rules, adjusts the data node position to align with the target position, outputs the final layout result and renders it to the user interface, uses Canvas technology for rendering, and completes the visualization automatic layout.

4. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the model calculation analysis visualization automatic layout method described in any one of claims 1 to 2 are implemented.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the model calculation analysis visualization automatic layout method described in any one of claims 1 to 2 are implemented.

Images