Water operation and maintenance data analysis and visualization method and system

Abstract

The invention relates to the technical field of data analysis and visualization, in particular to a water operation and maintenance data analysis and visualization method and system. The method comprises the following steps: acquiring real-time operation and maintenance monitoring data of various parameters, calculating data fluctuation entropy, calculating an interactive adjacent weight of each data object when a front-end visualization request is received, and calculating a cache priority in combination with the data fluctuation entropy, if the cache priority is higher than a preset cache trigger threshold value, triggering an associated data analysis task and storing the task into a data analysis cache, constructing a data visual cache, assigning a rendering parameter according to data fluctuation entropy, calculating cache survival time based on an interactive proximity weight, and obtaining a data analysis result; when the residence duration exceeds the cache survival time or the data fluctuation entropy exceeds a preset fluctuation threshold value, the analysis result and all the associated items are set to be invalid. According to the scheme, the computing resources can be allocated according to needs, so that the high-fluctuation nodes are highlighted on the visual interface, and the timeliness of the cache analysis result is ensured.

Description

Technical Field

[0001] This invention relates to the field of data analysis and visualization technology. More specifically, this invention relates to a method and system for water affairs operation and maintenance data analysis and visualization. Background Technology

[0002] With the deepening of smart water management construction, the scale of water management systems is becoming increasingly large and the number of pipeline monitoring points is dense, generating massive amounts of real-time operation and maintenance data from multiple sources and heterogeneous sources, including pipeline pressure, flow, and water quality.

[0003] Existing water data analysis and visualization systems typically employ a relatively traditional technical architecture. This involves a backend data analysis module periodically calculating all or part of the collected data and storing the results in a database. The frontend visualization interface then queries the database for the required data based on user requests and renders it. However, this architecture has bottlenecks: firstly, comprehensive and continuous analysis of massive amounts of data consumes enormous computing resources, easily leading to high system latency and costs. Furthermore, since data fluctuations at many pipeline nodes are relatively flat for most of the time, frequent periodic analysis of these nodes lacks practical significance, further wasting computing resources. Secondly, when facing complex interactions and large-scale data display, frequent requests for data from the backend by the frontend visualization result in slow response times, severely impacting user experience.

[0004] To address the problems inherent in the traditional technical architecture, researchers employed a caching mechanism for improvement. This caching mechanism involves caching frequently accessed data query results or pre-calculated statistical indicators. This reduces the system's direct access to the backend database when processing requests, thus offloading the database load and improving system response speed and data retrieval efficiency to some extent.

[0005] However, the aforementioned caching mechanisms still have shortcomings. First, existing cache triggering and update mechanisms are usually based on fixed time intervals or simple Least Recently Used (LRU) algorithms, failing to fully consider the time-varying characteristics of water data itself. Specifically, for critical nodes with drastic data fluctuations, a fixed update cycle may lead to a severe disconnect between cached data and the actual situation, while for long-term stable data, frequent cache refreshes will waste computing resources. Second, the management of cached content lacks effective linkage with the visualization interaction needs of front-end users. The system cannot predict the user's interaction intent, which may result in the failure to pre-load the correlation analysis results of areas that users may be interested in into the cache. Finally, in the existing system, data analysis cache and visualization rendering cache are independent of each other, lacking a unified framework for collaborative management. Furthermore, it fails to correlate data characteristics with the presentation of visualization elements, thereby reducing the visualization interface's ability to provide early warnings and insights into key information. Summary of the Invention

[0006] The purpose of this invention is to propose a water affairs operation and maintenance data analysis and visualization method and system to solve the problems of resource waste, lack of linkage between cache management and user interaction needs leading to the inability to predict user intentions and cache in advance, and poor early warning and insight capabilities of the visualization interface in the prior art; to this end, this invention provides solutions in the following two aspects.

[0007] In a first aspect, the present invention provides a water affairs operation and maintenance data analysis and visualization method, comprising: This system acquires real-time operation and maintenance monitoring data of various parameters of pipeline nodes in the water system; for each source data, it calculates data fluctuation entropy based on the joint probability distribution of the change acceleration and change amplitude of the time series data; when a front-end visualization request is received, it parses the request to obtain view parameters; based on the view parameters and the water pipeline network topology, it calculates the interaction proximity weight of each data object; combining the interaction proximity weight with the data fluctuation entropy of the corresponding source data, it calculates the cache priority of each data object; for data objects with a cache priority higher than the preset cache trigger threshold, it triggers the associated data analysis task and stores the analysis results in the data analysis cache; it constructs a data visualization cache, including... It includes a static geometric primitive cache layer, an aggregated data cache layer, and a visualization feature cache layer; the rendering parameters of features in the visualization feature cache layer are assigned values ​​based on the data fluctuation entropy of the corresponding source data in the aggregated data cache layer; when responding to the front-end visualization request, the data from the three cache layers are combined to generate a visualization interface; for any analysis result in the data analysis cache, the cache lifespan is calculated based on the interaction proximity weight of the corresponding data object when the analysis result is generated; when the residence time of the analysis result exceeds the cache lifespan, or the data fluctuation entropy of the dependent source data exceeds a preset fluctuation threshold, the analysis result and all related items in the data visualization cache are set to invalid.

[0008] Preferably, the step of calculating the data fluctuation entropy based on the joint probability distribution of the change acceleration and change amplitude of the time series data for each type of source data includes: dividing the time series data into continuous time windows; calculating the change amplitude and change acceleration for each data point within each time window; dividing the value ranges of change amplitude and change acceleration into 10 discrete intervals to form a 10×10 joint state grid; and counting the number of data point pairs composed of change amplitude and change acceleration falling into each joint interval of the joint state grid within the time window to construct a two-dimensional joint probability distribution matrix. In the formula, For data fluctuation entropy, For data point pairs to fall into the joint interval The probability, It is a logarithmic function with base 2.

[0009] Preferably, the two-dimensional joint probability distribution matrix is ​​constructed as follows: In the formula, For data point pairs to fall into the joint interval The probability, For data point pairs to fall into the joint interval The quantity.

[0010] Preferably, the step of calculating the interaction proximity weights of each data object based on view parameters and the waterworks network topology includes: abstracting the waterworks network as an undirected graph, where network nodes are vertices and pipe segments are edges; determining the central data object specified by the view parameters; and calculating the topological distance from any data object to the central data object using a breadth-first search algorithm in the undirected graph, where the topological distance is the minimum number of pipe segments traversed between them; the formula for calculating the interaction proximity weights is... In the formula, The interaction proximity weight of the data object, The topological distance from the data object to the central data object.

[0011] Preferably, the step of calculating the cache priority of each data object by combining the interaction proximity weight and the data fluctuation entropy of the corresponding source data includes: the calculation formula for the cache priority of any data object is as follows: In the formula, The cache priority of the data object, The interaction proximity weight of the data object, The data fluctuation entropy is the source data corresponding to the data object. This is the theoretical maximum entropy value. These are preset weighting coefficients.

[0012] Preferably, the rendering parameters of the features in the visualization feature cache layer are assigned values ​​based on the data fluctuation entropy of the corresponding source data in the aggregated data cache layer, including: mapping the geometric primitive colors of the pipeline nodes according to the value of the data fluctuation entropy, and mapping the value range of the data fluctuation entropy. Linearly mapped to hue values ​​in the HSV color space from green to red; when the data fluctuation entropy value exceeds the preset warning entropy value, the geometric primitives of the corresponding pipeline node are rendered as a flashing effect twice per second.

[0013] Preferably, the calculation of cache lifetime for any analysis result in the data analysis cache, based on the interaction proximity weight of the corresponding data object when the analysis result was generated, includes: In the formula, For cache lifespan, The interaction proximity weight of the data object, Based on survival time, This is the weighting factor.

[0014] Preferably, when the dwell time of the analysis result exceeds the cache lifespan, or the data fluctuation entropy of the dependent source data exceeds a preset fluctuation threshold, the analysis result and all related items in the data visualization cache are set to invalid, including: the background service polls and checks all analysis results in the data analysis cache every 5 seconds to obtain the dwell time and the current data fluctuation entropy of the dependent source data; when the dwell time exceeds the cache lifespan, or the current data fluctuation entropy exceeds the preset fluctuation threshold, the status field of the analysis result in the data analysis cache and all related items in the data visualization cache are marked as invalid.

[0015] Preferably, the theoretical maximum entropy value The preset weighting coefficient Take 0.6.

[0016] In the second aspect, a water affairs operation and maintenance data analysis and visualization system includes: The system includes a processor and a memory, wherein the memory stores computer program instructions for water affairs operation and maintenance data analysis and visualization, and when the computer program instructions are executed by the processor, the aforementioned water affairs operation and maintenance data analysis and visualization method is implemented.

[0017] The beneficial effects of this invention are as follows: By utilizing data fluctuation entropy to represent the degree of change in data at each monitoring point, and combining data fluctuation entropy with interactive proximity weights to jointly calculate the cache priority of data objects, this invention enables on-demand allocation of computing resources. Simultaneously, it concentrates limited data analysis capabilities on the areas with the most active data changes and the greatest user concern, effectively linking cache management with user interaction needs. Furthermore, by associating data fluctuation entropy with rendering parameters of visualization features, this invention ensures that network nodes with drastically changing data receive prominent visual presentation on the visualization interface, assisting maintenance personnel in quickly identifying status changes and potential anomalies. The cache invalidation mechanism of this invention comprehensively considers interactive proximity weights and data fluctuation entropy, representing the real-time fluctuation of source data, ensuring the timeliness of cache analysis results. Attached Figure Description

[0018] Figure 1 This illustration schematically shows a flowchart of the steps of a water affairs operation and maintenance data analysis and visualization method in this embodiment; Figure 2 The schematic diagram illustrates the structural block diagram of a water affairs operation and maintenance data analysis and visualization system in this embodiment. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0020] like Figure 1 As shown, a water affairs operation and maintenance data analysis and visualization method in this embodiment includes steps S1 to S4: Step S1: Obtain real-time operation and maintenance monitoring data of various parameters of the pipeline nodes in the water system; for each source data, calculate the data fluctuation entropy based on the joint probability distribution of the change acceleration and change amplitude of the time series data.

[0021] The various parameters of the pipeline network nodes include pressure, flow rate, and water quality. Real-time operation and maintenance monitoring data for these parameters are obtained through sensors. Understandably, for a specific parameter of a pipeline network node, such as pressure, a fixed historical time window is selected, which can be 24 hours. All readings from the pressure monitoring sensors within this historical time window are acquired, thus obtaining the time-series data of the pressure parameter, i.e., the source data.

[0022] In one embodiment, calculating the data fluctuation entropy for each type of source data based on the joint probability distribution of the acceleration and magnitude of change in the time series data includes: The time series data is divided into continuous time windows; within each time window, the magnitude and acceleration of change are calculated for each data point; the ranges of the magnitude and acceleration of change are divided into 10 discrete intervals, forming a 10×10 joint state grid; the number of data point pairs composed of magnitude and acceleration of change falling into each joint interval of the joint state grid within the time window is counted, and a two-dimensional joint probability distribution matrix is ​​constructed. In the formula, For data fluctuation entropy, For data point pairs to fall into the joint interval The probability, It is a logarithmic function with base 2.

[0023] Specifically, after obtaining time-series data for parameters such as pressure, flow rate, and water quality, the source data for each parameter, i.e., the time-series data, is divided into continuous time windows of 10 minutes. All pressure data from the pressure monitoring sensor within each time window are acquired. The amplitude and acceleration of change for each data point within each time window are then calculated. The possible range of values ​​for the amplitude is divided into 10 equally wide discrete intervals, and similarly, the possible range of values ​​for the acceleration is divided into 10 equally wide discrete intervals, forming a 10×10 joint state grid. Each grid in the joint state grid represents a joint interval. The amplitude and acceleration of each data point form a data point pair. The number of data point pairs falling into each of the 100 joint intervals is counted, and the probability of a data point pair falling into a joint interval is calculated, constructing a two-dimensional joint probability distribution matrix. The probability of a data point pair falling into a joint interval is then applied to the calculation of the data fluctuation entropy. The formula yields an entropy value that represents the degree of fluctuation in the source data within a time window.

[0024] In one embodiment, the magnitude of the change satisfies: The changing acceleration satisfies: In the formula, For the first The magnitude of change of each data point For the first The acceleration of the change in each data point For the first The magnitude of change of each data point For the first Data points, For the first Data points, For the first Data points, .

[0025] Specifically, the time series data is divided into continuous time windows to obtain multiple time series data segments. Taking the calculation of the change amplitude and acceleration of each data point within a time window as an example, the pressure reading time series data of a pressure monitoring sensor over a 10-minute period is obtained. For example, the data points in the time series data within this time window are 1.5, 1.6, 1.8, 1.7, and 1.5 MPa. For the third data point of 1.8 MPa in the time series data, the change amplitude of the third data point is obtained by subtracting the second (the preceding data point) data point of 1.6 MPa from the third data point of 1.8 MPa, which is 0.2. The change amplitude of the second data point is obtained by subtracting the first (the two preceding data points) data points from the second data point, which is also the change amplitude of the preceding data point of the third data point, which is 0.1. The difference between the change amplitude of the third data point of 0.2 and the change amplitude of the second data point of 0.1 is the acceleration of the third data point, which is 0.1. In this way, the above calculation is repeated for all data points within the entire time window to obtain multiple change amplitudes and accelerations.

[0026] In one embodiment, the two-dimensional joint probability distribution matrix is ​​constructed as follows: ; In the formula, For data point pairs to fall into the joint interval The probability, For data point pairs to fall into the joint interval The quantity.

[0027] Specifically, the magnitude of change for each data point is calculated. and change acceleration Then, multiple data point pairs are formed. Calculate data point pairs Falling into the joint interval The number of items that fall into the first category, i.e., the number of items that fall into the first category. The range of change and the first The number of acceleration intervals, and each joint interval Data point pairs Dividing the number of data points by the total number of data point pairs yields the result of falling into the joint interval. probability That is, the th in the two-dimensional joint probability distribution matrix Line 1 The column values. For example, a total of 1000 data pairs are obtained within the time window, of which 50 data pairs show the magnitude of change. The acceleration varies within the discrete interval 6. In the discrete interval 7, the corresponding joint interval is... count value The value is 50; divide the count value of each joint interval by the total number 1000 to obtain the joint probability, then... Thus, a complete 10×10 two-dimensional joint probability distribution matrix is ​​constructed.

[0028] Step S2: When a front-end visualization request is received, the request is parsed to obtain the view parameters; the interaction proximity weight of each data object is calculated based on the view parameters and the water pipe network topology; the cache priority of each data object is calculated by combining the interaction proximity weight with the data fluctuation entropy of the corresponding source data; for data objects whose cache priority is higher than the preset cache trigger threshold, the associated data analysis task is triggered and the analysis results are stored in the data analysis cache.

[0029] In one embodiment, calculating the interaction proximity weights of each data object based on view parameters and the water supply network topology includes: The water supply network is abstracted as an undirected graph, where the network nodes are the vertices and the pipe segments are the edges. The central data object specified by the view parameters is determined. In the undirected graph, the breadth-first search algorithm is used to calculate the topological distance from any data object to the central data object. The topological distance is the minimum number of pipe segments between the two. The formula for calculating the interaction proximity weight is as follows: In the formula, The interaction proximity weight of the data object, The topological distance from the data object to the central data object.

[0030] Specifically, facilities such as pump stations, valves, and monitoring points in the water supply network are treated as vertices of a graph, and the pipe segments connecting them are considered as edges, constructing a network topology graph, i.e., an undirected graph, as the water supply network topology. This constructed water supply network topology is pre-stored in the system for direct use by the backend algorithm. When a user zooms or pans the map on the front-end interface, the front-end sends the current view center point's geographic coordinates and display level as view parameters to the backend. The backend determines the network node or area of ​​current focus as the central data object based on these view parameters; alternatively, when a user clicks and selects a device named pump station E in the visualization interface, pump station E is identified as the central data object. In the undirected graph, a breadth-first search algorithm is used to traverse the network, calculating the topological distance from any data object to the central data object, i.e., the minimum number of pipe segments between them. For example, the topological distance from pump station E to valve F, which is directly connected to pump station E via a pipe segment, is calculated. The monitoring point G, denoted as 1, is connected to valve F but not directly to pump station E. The distance from pump station E to monitoring point G requires at least two segments of pipeline; this is the topological distance from pump station E to monitoring point G. Let it be 2. Based on topological distance... The proximity of data objects is determined by a decreasing function, such as a Gaussian function or an inverse proportional function, assigning each object an interactive proximity weight between 0 and 1. In this embodiment, an inverse proportional function is used. Calculate the interaction proximity weight. The closer the topological distance, the higher the interaction proximity weight; the farther the topological distance, the lower the interaction proximity weight.

[0031] In one embodiment, the step of calculating the cache priority of each data object by combining the interaction proximity weight and the data fluctuation entropy of the corresponding source data includes: The formula for calculating the cache priority of any data object is as follows: ; In the formula, The cache priority of the data object, The interaction proximity weight of the data object, The data fluctuation entropy is the source data corresponding to the data object. This is the theoretical maximum entropy value. These are preset weighting coefficients.

[0032] The theoretical maximum entropy value The preset weighting coefficient Take 0.6.

[0033] Specifically, preset weighting coefficients This is used to adjust the relative importance of user attention and data change rate; it is preset based on experience and can be set to 0.6. Theoretical maximum entropy value. This refers to the magnitude of change in data points. and change acceleration The entropy value achievable when exhibiting maximum randomness and disorder is derived from the calculation method of data fluctuation entropy, assuming that the probability of each data point pair falling into each joint interval is the same. At that time, the entropy value reaches its maximum. This item is used to calculate the actual data fluctuation entropy. Normalization is performed so that its numerical range is consistent with the interaction neighbor weights, both mapping to between 0 and 1.

[0034] Obtain two metrics for each data object in the pipeline network: interactive proximity weight and data fluctuation entropy. For example, for three data objects, one is valve F, which is close to pump station E, the data object at the center of the user's current view. The interactive proximity weight... The value is 0.5, but the monitoring data is stable, and the data fluctuation entropy is low. Only 1.5; the other is monitoring point G, which is far from the central data object pump station E, with interactive proximity weighting. The value is 0.33, but the data fluctuates wildly, resulting in high data fluctuation entropy. The value is 4.0; another is pump station D, which is directly adjacent to the central data object pump station E, with interactive proximity weights. The value is 0.5, but the data fluctuates greatly, resulting in high data fluctuation entropy. The value is 6.0. The ideal maximum entropy is approximately 6.64, and the buffer priority of valve F is calculated accordingly. The cache priority of monitoring point G is calculated. The buffer priority of pump station D is calculated. .

[0035] The preset cache trigger threshold is set empirically and can be set to 0.6. The cache priority calculated for each data object is compared with the preset cache trigger threshold. For data objects with a cache priority higher than the preset cache trigger threshold, the associated data analysis task is triggered, and the analysis results are stored in the data analysis cache. For data objects with a cache priority lower than the preset cache trigger threshold, no data analysis task is triggered. Of the three data objects whose cache priorities are calculated above, only pump station D has a cache priority higher than the preset cache trigger threshold. Therefore, only the data analysis task associated with pump station D will be triggered, analyzing the source data associated with pump station D and actively caching the analysis results in the data analysis cache.

[0036] Step S3: Construct a data visualization cache, including a static geometric primitive cache layer, an aggregated data cache layer, and a visualization feature cache layer; the rendering parameters of features in the visualization feature cache layer are assigned values ​​based on the data fluctuation entropy of the corresponding source data in the aggregated data cache layer; when responding to the front-end visualization request, the data from the three cache layers are combined to generate a visualization interface.

[0037] The data visualization cache consists of a static geometric primitive cache layer, an aggregated data cache layer, and a visualization feature cache layer. The static geometric primitive cache layer stores the water pipe network topology and geographic information, specifically the start and end coordinates of pipe segments and the geographical locations of nodes. This geographic information typically changes infrequently and can be pre-generated and cached in formats such as vector tiles. The aggregated data cache layer directly retrieves analysis results from the data analysis cache, storing core business data used to drive visualization; for example, the average pressure value of a node or the leakage probability of a pipe segment. The visualization feature cache layer generates specific visual attributes based on the data in the aggregated data cache layer.

[0038] In one embodiment, the rendering parameters of the features in the visualization feature cache layer are assigned values ​​based on the data fluctuation entropy of the corresponding source data in the aggregated data cache layer, including: The colors of the geometric elements of the pipeline nodes are mapped according to the values ​​of data fluctuation entropy, and the range of data fluctuation entropy is defined. Linearly mapped to hue values ​​in the HSV color space from green to red; when the data fluctuation entropy value exceeds the preset warning entropy value, the geometric primitives of the corresponding pipeline node are rendered as a flashing effect twice per second.

[0039] Specifically, data fluctuation entropy The range of values ​​for is from 0 to the theoretical maximum entropy value. Approximately 6.64, linearly mapped to the hue value in the HSV color model. For example, 0 represents a stable green with a hue value of 120; 6.64 represents a warning red with a hue value of 0. When the monitoring data of a pipeline node is very stable, the calculated data fluctuation entropy... When the value is 1.0, it is rendered as a greenish color, such as yellowish-green; when the data of another network node fluctuates drastically, the calculated data fluctuation entropy... When the value is 5.0, it is rendered as a striking orange-red.

[0040] Furthermore, when the data fluctuation entropy of the source data corresponding to any pipeline node... When the data exceeds a preset warning entropy value (an empirical value of 2.8), in addition to rendering the data as a corresponding color, a visual warning effect will also be activated. For example, the data fluctuation entropy of the source data corresponding to a pipeline node is calculated. The value is 3.5, which is rendered as yellow. At the same time, the graphic elements of the pipeline node will be displayed alternately on the visualization interface, flashing twice per second, to highlight the abnormal fluctuation of the data point and remind the operator to pay attention.

[0041] When the backend service receives a visualization request from the frontend, it retrieves the water pipe network topology and geographic information within the corresponding view area from the static geometric primitive cache layer; it also queries the visualization feature cache layer for the corresponding rendering parameters, specifically color and highlight status; and simultaneously obtains specific values ​​from the aggregated data cache layer to display labels or information boxes when the user hovers over the data. The backend combines the geometric information (geographic information), rendering parameters, and numerical values ​​into a unified data package and sends it to the frontend. The frontend's rendering engine, such as a WebGL or SVG library, parses this data package and draws a water pipe operation and maintenance map interface with visual effects, generating the visualization interface.

[0042] Step S4: For any analysis result in the data analysis cache, calculate the cache lifespan based on the interaction proximity weight of the corresponding data object when the analysis result is generated; when the residence time of the analysis result exceeds the cache lifespan, or the data fluctuation entropy dependent on the source data exceeds the preset fluctuation threshold, set the analysis result and all related items in the data visualization cache to invalid.

[0043] In one embodiment, calculating the cache lifetime for any analysis result in the data analysis cache based on the interaction proximity weight of the corresponding data object when the analysis result was generated includes: ; In the formula, For cache lifespan, The interaction proximity weight of the data object, Based on survival time, This is the weighting factor.

[0044] Specifically, the cache lifetime The unit is seconds; basic survival time. Take the empirical value of 60 seconds, and the weighted influence factor. Using an empirical value of 300 seconds, the interaction proximity weight of pump station D is determined when a user is paying attention to it. The cache priority of pump station D is 0.66, which is higher than the preset cache trigger threshold. At this time, the pressure source data associated with pump station D is analyzed and a pressure trend analysis result is generated. The pressure trend analysis result is stored in the data analysis cache, and the cache lifespan of the analysis result is calculated. The duration is 360 seconds, or 6 minutes. This applies when a user is focusing on a data point that is far from the central data object, pump station E, but where the data is stable and the interaction is close to the target location. When monitoring point B has a cache priority of only 0.2 and is higher than the preset cache trigger threshold, the source data associated with monitoring point B is analyzed and an analysis result is generated. The cache lifespan of this analysis result is calculated to be 120 seconds, or 2 minutes. Similarly, the source data of the other parameters associated with pump station D and monitoring point B can be analyzed separately, and the cache lifespan of the corresponding analysis results can be calculated separately. The farther the data object is from the user's current central data object, the shorter the retention time of the analysis result in the cache.

[0045] In one embodiment, when the residence time of the analysis result exceeds the cache lifetime, or the data fluctuation entropy depending on the source data exceeds a preset fluctuation threshold, the analysis result and all related items in the data visualization cache are invalidated, including: The background service polls and checks all analysis results in the data analysis cache every 5 seconds to obtain the dwell time and the current data fluctuation entropy of the dependent source data. When the dwell time exceeds the cache lifespan or the current data fluctuation entropy exceeds the preset fluctuation threshold, the status field of the analysis result in the data analysis cache and all related items in the data visualization cache are marked as invalid.

[0046] Specifically, for each cached entry in the data analysis cache, cache invalidation is managed through two mechanisms. The first is time-based checking. If an analysis result is cached at 10:00 AM, its cache lifetime, calculated based on the interaction proximity weight, is 180 seconds. The background service scans cached entries every 5 seconds. When the background service detects the cached entry at 10:30:05, it finds that the dwell time has reached 185 seconds, exceeding the 180-second cache lifetime. Therefore, the program immediately marks the status field of the cached entry in the data analysis cache as invalid, and simultaneously marks all related items in the data visualization cache as invalid. The second method is data fluctuation detection. Another analysis result was cached at 10:10 AM with a cache lifespan of 300 seconds. Based on the source data corresponding to the analysis result, the data fluctuation entropy at that time was calculated to be 2.5, which is lower than the preset fluctuation threshold. In this embodiment, the preset fluctuation threshold is taken as an empirical value of 3.0. However, at 10:11 AM, an anomaly occurred in the pipeline network, and the real-time data fluctuation entropy of the source data rapidly climbed to 4.2. During the next 5-second polling, the background service detected that the current data fluctuation entropy of 4.2 had exceeded the preset fluctuation threshold of 3.0. It was determined that the cached analysis result could no longer reflect the latest data status. Even if the dwell time was far from expired, the background service would immediately mark all related items in the status field of the data analysis cache and all related items in the data visualization cache as invalid to ensure that the information displayed on the front end is always based on relatively stable or latest data.

[0047] This invention also provides a water affairs operation and maintenance data analysis and visualization system. For example... Figure 2 As shown, the system includes a processor and a memory. The memory stores computer program instructions for water affairs operation and maintenance data analysis and visualization. When the computer program instructions are executed by the processor, a water affairs operation and maintenance data analysis and visualization method according to the present invention is implemented.

[0048] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and therefore will not be described in detail here.

[0049] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented by computer-readable / executable instructions stored or otherwise maintained on such a computer-readable medium.

[0050] In the description of this specification, "multiple" means at least two, such as two, three or more, etc., unless otherwise expressly and specifically defined.

[0051] While various embodiments of the invention have been shown and described in this specification, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention.

Claims

1. A method for analyzing and visualizing water affairs operation and maintenance data, characterized in that, include: To acquire real-time operation and maintenance monitoring data of various parameters of pipeline nodes in the water system; For each source data, the data fluctuation entropy is calculated based on the joint probability distribution of the acceleration and magnitude of change in the time series data; When a front-end visualization request is received, the request is parsed to obtain the view parameters; the interaction proximity weights of each data object are calculated based on the view parameters and the water supply network topology. The cache priority of each data object is calculated by combining the interaction proximity weight and the data fluctuation entropy of the corresponding source data. For data objects whose cache priority is higher than the preset cache trigger threshold, trigger the associated data analysis task and store the analysis results in the data analysis cache. Construct a data visualization cache, including a static geometric primitive cache layer, an aggregated data cache layer, and a visualization feature cache layer; The rendering parameters of features in the visualization feature cache layer are assigned values ​​based on the data fluctuation entropy of the corresponding source data in the aggregated data cache layer; when responding to the front-end visualization request, the data from the three cache layers are combined to generate a visualization interface; For any analysis result in the data analysis cache, the cache lifetime is calculated based on the interaction proximity weight of the corresponding data object when the analysis result was generated. When the dwell time of the analysis results exceeds the cache lifespan, or the data fluctuation entropy that depends on the source data exceeds the preset fluctuation threshold, the analysis results and all related items in the data visualization cache will be invalidated.

2. The water affairs operation and maintenance data analysis and visualization method according to claim 1, characterized in that, For each type of source data, the calculation of data fluctuation entropy based on the joint probability distribution of the acceleration and magnitude of change in the time series data includes: The time series data is divided into continuous time windows; within each time window, the magnitude and acceleration of change are calculated for each data point; the ranges of the magnitude and acceleration of change are divided into 10 discrete intervals, forming a 10×10 joint state grid; the number of data point pairs composed of magnitude and acceleration of change falling into each joint interval of the joint state grid within the time window is counted, and a two-dimensional joint probability distribution matrix is ​​constructed. In the formula, For data fluctuation entropy, For data point pairs to fall into the joint interval The probability, It is a logarithmic function with base 2.

3. The water affairs operation and maintenance data analysis and visualization method according to claim 2, characterized in that, The two-dimensional joint probability distribution matrix is ​​constructed as follows: ； In the formula, For data point pairs to fall into the joint interval The probability, For data point pairs to fall into the joint interval The quantity.

4. The water affairs operation and maintenance data analysis and visualization method according to claim 1, characterized in that, The calculation of the interaction proximity weights of each data object based on view parameters and water supply network topology includes: The water supply network is abstracted as an undirected graph, where the network nodes are the vertices and the pipe segments are the edges. The central data object specified by the view parameters is determined. In the undirected graph, the breadth-first search algorithm is used to calculate the topological distance from any data object to the central data object. The topological distance is the minimum number of pipe segments between the two. The formula for calculating the interaction proximity weight is as follows: In the formula, The interaction proximity weight of the data object, The topological distance from the data object to the central data object.

5. The water affairs operation and maintenance data analysis and visualization method according to claim 1, characterized in that, The calculation of the cache priority of each data object by combining the interaction proximity weight and the data fluctuation entropy of the corresponding source data includes: The formula for calculating the cache priority of any data object is as follows: ; In the formula, The cache priority of the data object, The interaction proximity weight of the data object, The data fluctuation entropy is the source data corresponding to the data object. This is the theoretical maximum entropy value. These are preset weighting coefficients.

6. The water affairs operation and maintenance data analysis and visualization method according to claim 1, characterized in that, The rendering parameters of the features in the visualization feature cache layer are assigned values ​​based on the data fluctuation entropy of the corresponding source data in the aggregated data cache layer, including: The colors of the geometric elements of the pipeline nodes are mapped according to the values ​​of data fluctuation entropy, and the range of data fluctuation entropy is defined. Linearly mapped to hue values ​​in the HSV color space from green to red; when the data fluctuation entropy value exceeds the preset warning entropy value, the geometric primitives of the corresponding pipeline node are rendered as a flashing effect twice per second.

7. The water affairs operation and maintenance data analysis and visualization method according to claim 1, characterized in that, The cache lifetime is calculated for any analysis result in the data analysis cache based on the interaction proximity weight of the corresponding data object when the analysis result was generated, including: ； In the formula, For cache lifespan, The interaction proximity weight of the data object, Based on survival time, This is the weighting factor.

8. The water affairs operation and maintenance data analysis and visualization method according to claim 1, characterized in that, When the residence time of the analysis result exceeds the cache lifespan, or the data fluctuation entropy depending on the source data exceeds a preset fluctuation threshold, the analysis result and all related items in the data visualization cache are invalidated, including: The background service polls and checks all analysis results in the data analysis cache every 5 seconds to obtain the dwell time and the current data fluctuation entropy of the dependent source data. When the dwell time exceeds the cache lifespan or the current data fluctuation entropy exceeds the preset fluctuation threshold, the status field of the analysis result in the data analysis cache and all related items in the data visualization cache are marked as invalid.

9. A water affairs operation and maintenance data analysis and visualization method according to claim 5, characterized in that, The theoretical maximum entropy value The preset weighting coefficient Take 0.

6.

10. A water affairs operation and maintenance data analysis and visualization system, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions for water affairs operation and maintenance data analysis and visualization, and when the computer program instructions are executed by the processor, a water affairs operation and maintenance data analysis and visualization method according to any one of claims 1-9 is implemented.

Images