Water plant-oriented two-dimensional configuration method and system

Abstract

The invention relates to the technical field of automatic control, in particular to a water plant-oriented two-dimensional configuration method, which comprises the following steps of: inputting working condition data of any equipment of a water plant into a configuration data prediction model to obtain corresponding configuration data; rendering the configuration data on the basis of a Handlebars configuration strategy, and respectively obtaining Web configuration data, bit number point table data and animation description data which are used for supporting two-dimensional configuration of the equipment; the configuration data prediction model is trained based on the following steps: constructing a model training set based on historical operation records of equipment; training a pre-constructed CART decision tree model based on the model training set; and after training is completed, traversing all nodes of the CART decision tree model, and pruning the nodes based on an expert constraint strategy to obtain a configuration data prediction model for predicting equipment configuration data. According to the method, the automation level and the reliability of two-dimensional configuration are remarkably improved, parameter redundancy and overfitting are inhibited, and the configuration efficiency is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automation control, and in particular to a two-dimensional configuration method for water plants and a two-dimensional configuration system for water plants. BACKGROUND

[0002] Two-dimensional configuration is the core carrier for realizing water plant equipment monitoring, process visualization and remote operation and maintenance. When two-dimensional configuration is performed, key tasks such as equipment position number configuration, process animation presentation and multi-system data linkage need to be completed, which directly affects the construction efficiency and operation stability of the water plant automatic control system, and thus is widely used in the automation control scenes of various processes of water plants.

[0003] When the existing water plant configuration system is constructed, an engineer usually needs to first download equipment graph elements from professional configuration software, then manually bind position numbers and real-time data points for each graph element, and subsequently repeatedly check parameters such as position number naming, range, PID initial value. For process animations such as water flow, dosing diffusion and sedimentation, a UI designer needs to manually draw animation effects frame by frame and implement hard coding on the front end. Meanwhile, the naming rules of equipment of different systems such as SCADA, Web configuration and IoT platform are not unified, and manual export of Excel tables is needed to manually import between systems to complete parameter synchronization.

[0004] This manual configuration method is highly dependent on the personal experience of engineers, and the same "dosing-sedimentation" process may have naming conflicts, range deviations, random PID initial values and missing animation parameters in different projects, resulting in repeated rework in subsequent configuration. Moreover, this manual configuration method is inefficient, and the configuration work of a water plant expansion project needs to be completed by multiple people for 2-3 weeks on average. SUMMARY

[0005] (I) Technical problems to be solved

[0006] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present application provides a two-dimensional configuration method for water plants, which solves the technical problems that the manual configuration method is highly dependent on the personal experience of engineers, has a high error rate and is inefficient.

[0007] (II) Technical solutions

[0008] In order to achieve the above-mentioned purposes, the main technical solutions adopted by the present application include:

[0009] In a first aspect, the present application provides a two-dimensional configuration method for water plants, comprising:

[0010] inputting working condition data of any equipment of a water plant into a pre-trained configuration data prediction model to obtain corresponding configuration data;

[0011] render the configuration data based on a pre-deployed Handlebars configuration strategy, to obtain web configuration data, bit number point table data and animation description data for supporting two-dimensional configuration of the device, respectively;

[0012] The configuration data prediction model is trained based on the following steps:

[0013] A model training set is constructed based on the pre-set device historical operation records of each device of the water plant. The device operation records corresponding to any device each include historical working condition data of multiple time periods, and configuration data labels corresponding to the historical working condition data of each time period.

[0014] The pre-constructed CART decision tree model is trained based on the model training set. After the training is completed, all nodes of the CART decision tree model are traversed to perform pruning processing on any node of the CART decision tree model based on a pre-set expert constraint strategy when a split path condition and a node output of the node do not match the expert constraint strategy, to obtain a configuration data prediction model for predicting device configuration data.

[0015] Optionally, in a specific embodiment, the configuration data and the configuration data labels each include device configuration data and animation configuration data.

[0016] The model training set is constructed based on the pre-set device historical operation records of each device of the water plant, including:

[0017] The water plant point table and the process as-built drawing of the water plant are collected, and all historical working condition data of each device of the water plant and device configuration data corresponding to each historical working condition data are obtained according to the water plant point table and the process as-built drawing.

[0018] Based on each historical working condition data and a pre-established mapping relationship table, animation configuration data corresponding to each historical working condition data is obtained.

[0019] The mapping relationship table is a mapping relationship corresponding table between historical working condition data and animation configuration data.

[0020] The historical working condition data, the device configuration data and the animation configuration data corresponding to each device are preprocessed, and a model training set is established based on all preprocessed historical working condition data, device configuration data and animation configuration data. The preprocessing includes data cleaning and field standardization.

[0021] Optionally, in a specific embodiment, training the pre-constructed CART decision tree model based on the model training set includes:

[0022] The CART decision tree model is initialized.

[0023] Feature extraction is performed on the model training set to obtain a first feature matrix corresponding to historical working condition data and a second feature matrix corresponding to configuration data labels;

[0024] The first feature matrix and the second feature matrix are input into a CART decision tree model for recursive splitting;

[0025] During the training process, a node sample set corresponding to each node in the CART decision tree model is obtained; each node sample set is a subset selected from the first feature matrix and the first feature matrix according to a split path condition previously set in a previous node in the CART decision tree model;

[0026] Based on the node sample set corresponding to each node, a first weighted Gini index before splitting of each node is obtained;

[0027] Based on the split path condition of each node and the node sample set, a second weighted Gini index after splitting of each node is obtained, and based on the first weighted Gini index and the second weighted Gini index of each node, an optimization amplitude corresponding to each node is obtained; the optimization amplitude is the difference between the first weighted Gini index and the second weighted Gini index;

[0028] Based on the optimization amplitude corresponding to each node and a previously set business penalty coefficient, a final split score corresponding to each node is obtained;

[0029] Based on the optimization amplitude corresponding to each node and the final split score, model parameters of the CART decision tree model are optimized.

[0030] Optionally, in a specific embodiment, based on the split path condition of each node and the node sample set, the second weighted Gini index after splitting of each node is obtained, including:

[0031] Based on the split path condition of each node, the node sample set of each node is split into a first node sample set and a second node sample set;

[0032] Based on the first node sample set and the second node sample set of each node and a previously set formula one, a third weighted Gini index and a fourth weighted Gini index corresponding to each node are respectively obtained; the formula one is:

[0033] ;

[0034] wherein Gini is a weighted Gini index, K is the total number of variable categories, k is the index of a variable category, w k is a previously set relevance weight of the variable category with the index k, p k is the proportion of all corresponding samples belonging to the variable category with the index k;

[0035] The second weighted Gini index is obtained based on the third weighted Gini index and the fourth weighted Gini index corresponding to each node, and the second weighted Gini index is a sum of a product of a proportion of a sample amount in the first node sample set in a sample amount in the node sample set and the third weighted Gini index and a product of a proportion of a sample amount in the second node sample set in the sample amount in the node sample set and the fourth weighted Gini index.

[0036] Optionally, in a specific embodiment, the final splitting score corresponding to each node is obtained based on the optimization amplitude corresponding to each node and a preset service penalty coefficient, and the method comprises:

[0037] The service conflict judgment strategy is preset.

[0038] If there is no service conflict, the final splitting score corresponding to each node is the optimization amplitude corresponding to the node.

[0039] If there is a service conflict, the final splitting score corresponding to each node is a difference between the optimization amplitude and the service penalty coefficient.

[0040] Optionally, in a specific embodiment, the Handlebars configuration strategy comprises a bit number point table configuration strategy, a Web configuration JSON configuration strategy and an animation description JSON configuration strategy.

[0041] Then, the configuration data is rendered based on the pre-deployed Handlebars configuration strategy, and Web configuration data, bit number point table data and animation description data used for supporting a device to perform two-dimensional configuration are obtained, comprising:

[0042] The configuration parameters are subjected to feature extraction and data integration, and a corresponding JSON array is obtained.

[0043] The JSON array is rendered based on the bit number point table configuration strategy, the Web configuration JSON configuration strategy and the animation description JSON configuration strategy respectively, and Web configuration data, bit number point table data and animation description data used for supporting the device to perform two-dimensional configuration are obtained.

[0044] The Web configuration data, the bit number point table data and the animation description data of the device are written into a preset Web configuration file, a bit number point table file and an animation description file respectively.

[0045] Optionally, in a specific embodiment, all nodes of the CART decision tree model are traversed, and when the splitting path condition and the node output of any node in the CART decision tree model are determined to be unmatched with a preset expert constraint strategy, the node is subjected to pruning processing based on the expert constraint strategy, and the method comprises:

[0046] traversing from the bottommost node of the CART decision tree model upwards layer by layer, and checking the split path condition and node output of each node during the traversal process;

[0047] When the split path condition and node output of any node in the CART decision tree model do not match the pre-set expert constraint strategy, whether the node output of the node is related to the expert constraint strategy is determined, and the node output of the node is replaced / deleted based on the determination.

[0048] Optionally, in a specific embodiment, the Web configuration data, the point table data and the animation description data of the device are respectively written into the pre-set Web configuration file, the point table file and the animation description file, including:

[0049] A unique identification code is generated based on the point table data of the device, and the point table data and the unique identification code of the device are written into the point table file as a piece of data, the Web configuration data and the unique identification code of the device are written into the Web configuration file as a group of data, and the animation description data and the unique identification code of the device are written into the animation description file as a piece of data;

[0050] The Web configuration data, the point table data and the animation description data are mapped and aligned based on the unique identification codes in the Web configuration file, the point table file and the animation description file.

[0051] Optionally, in a specific embodiment, after the Web configuration data, the point table data and the animation description data are mapped and aligned, the method further includes:

[0052] When any Web configuration data, point table data or animation description data in the Web configuration file, the point table file and the animation description file is modified, the contents of the Web configuration data, the point table data or the animation description data in the Web configuration file, the point table file and the animation description file are synchronized.

[0053] In a second aspect, the embodiments of the present application provide a two-dimensional configuration system for a water plant, including a memory, a processor and a computer program stored in the memory, and the processor executes the computer program to implement the two-dimensional configuration method for a water plant.

[0054] (Three) beneficial effects

[0055] The two-dimensional configuration method and system for water plants of the present application can automatically output standardized configuration data based on equipment working condition data through a pre-trained CART decision tree configuration data prediction model, not only reuse the historical operation experience of water plants, but also prune the configuration parameters with the help of expert constraint strategy to ensure that the configuration parameters meet the industry standards and engineering practice, significantly improve the automation level and reliability of the two-dimensional configuration, inhibit parameter redundancy and overfitting, and improve the configuration efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0056] Figure 1 A two-dimensional configuration method flowchart for water plants is provided for the embodiments of the present application.

[0057] Figure 2 A two-dimensional configuration method flowchart for water plants is provided for the embodiments of the present application.

[0058] Figure 3 A configuration data prediction model training flowchart is provided for the embodiments of the present application. DETAILED DESCRIPTION

[0059] In order to better explain the present application and facilitate understanding, the present application is described in detail below through specific embodiments in combination with the drawings.

[0060] Two-dimensional configuration is the core carrier to realize water plant equipment monitoring, process visualization and remote operation and maintenance. In the two-dimensional configuration, key tasks such as device position number configuration, process animation presentation and multi-system data linkage need to be completed, which directly affects the construction efficiency and operation stability of the water plant automatic control system, and is widely used in the automation control scene of each process of the water plant. However, the traditional configuration mode highly depends on manual operation, and engineers need to manually download graphics and bind data points, repeatedly check various process parameters; process animation needs to be hand-drawn frame by frame by UI designers and hard-coded in the front end; and the naming rules of SCADA, Web configuration and other multi-systems are not unified, and manual export table is needed to realize parameter synchronization. This mode not only has low efficiency, but also takes 2-3 weeks for multiple people to configure the water plant expansion project, and due to the dependence on the personal experience of engineers, naming conflicts, range deviations and random PID initial values often occur, which leads to repeated configuration and seriously affects the construction efficiency and operation stability of the automatic control system. The two-dimensional configuration method for water plants proposed in the present application can automatically output standardized configuration data based on device working condition data through a pre-trained CART decision tree configuration data prediction model, which not only fully reuses the historical operation experience of the water plant, but also prunes the model nodes through expert constraint strategies to ensure that the configuration parameters meet the industry specifications and engineering practice. This method not only significantly improves the automation level and parameter reliability of two-dimensional configuration, but also effectively suppresses the problems of model parameter redundancy and overfitting, and solves the problems of experience dependence and rework of manual configuration from the root, greatly improving the overall efficiency of two-dimensional configuration of the water plant.

[0061] In order to better understand the above technical solutions, the exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to enable a clearer, more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0062] The embodiment of the present application provides a two-dimensional configuration method for water plants, as shown in Figure 1 and Figure 3 , comprising:

[0063] S10, inputting the working condition data of any device of the water plant into a pre-trained configuration data prediction model to obtain corresponding configuration data;

[0064] S20, rendering the configuration data based on a pre-deployed Handlebars configuration strategy to obtain Web configuration data, position number point table data and animation description data for supporting two-dimensional configuration of the device, respectively;

[0065] As shown in Figure 2As shown, the configuration data prediction model is trained based on the following steps:

[0066] S01, based on the pre-set device historical operation record of each device of the water plant, a model training set is constructed; the device operation record corresponding to any device includes historical working condition data of multiple time periods, and the configuration data label corresponding to the historical working condition data of each time period;

[0067] S02, based on the model training set, the pre-constructed CART decision tree model is trained; and after the training is completed, all nodes of the CART decision tree model are traversed to determine the split path condition and node output of any node in the CART decision tree model, and when the pre-set expert constraint strategy does not match, the node is pruned based on the expert constraint strategy to obtain a configuration data prediction model for predicting device configuration data.

[0068] Through the pre-trained CART decision tree configuration data prediction model, the embodiment can automatically output standardized configuration data based on device working condition data, which not only fully reuses the historical operation experience of the water plant, but also prunes the model nodes through the expert constraint strategy to ensure that the configuration parameters meet the industry specifications and engineering practice. Not only significantly improves the automation level and parameter reliability of two-dimensional configuration, but also effectively suppresses the problems of model parameter redundancy and overfitting, solves the experience dependence and rework problems of manual configuration from the root, and greatly improves the overall efficiency of two-dimensional configuration of the water plant.

[0069] Optionally, in a specific embodiment, the working condition data includes but is not limited to: process section, device type, flow rate Q, turbidity TU, pump frequency Hz, valve opening degree%; the configuration data includes but is not limited to: bit number naming string, range upper limit, alarm high, PID_P, PID_I, PID_D, animation speed, particle color, diffusion radius. Among them, the common process section includes 7 process sections of water intake, dosing, flocculation, sedimentation, filtration, disinfection and sludge discharge.

[0070] Optionally, in a specific embodiment, the Handlebars configuration strategy includes bit number point table configuration strategy, Web configuration JSON configuration strategy and animation description JSON configuration strategy;

[0071] Then, based on the pre-deployed Handlebars configuration strategy, the configuration data is rendered to obtain Web configuration data, bit number point table data and animation description data for supporting the device to perform two-dimensional configuration, including:

[0072] The configuration parameters are subjected to feature extraction and data integration to obtain corresponding JSON arrays;

[0073] The JSON array is rendered based on the bit number point table configuration strategy, the Web configuration JSON configuration strategy and the animation description JSON configuration strategy respectively to obtain Web configuration data, bit number point table data and animation description data for supporting two-dimensional configuration of the device;

[0074] The Web configuration data, the bit number point table data and the animation description data of the device are written into the pre-set Web configuration file, the bit number point table file and the animation description file respectively.

[0075] Specifically, in order to automatically convert the "candidate table + animation parameters" output by the decision tree into three heterogeneous files which can be directly read by SCADA, Web configuration and animation engine, the system uses a lightweight template engine to complete the synchronous rendering of three sets of configuration files in a single process. The specific steps are as follows:

[0076] Three template files (i.e. Handlebars configuration strategy) are pre-set in the server directory, namely the bit number point table template (bit number point table configuration strategy), the Web configuration JSON template (Web configuration JSON configuration strategy) and the animation description JSON template (animation description JSON configuration strategy). At the same time, an input subdirectory is set up for storing the unified format JSON file generated after the decision tree reasoning, and an output subdirectory is set up for storing the final version of the rendered file.

[0077] The bit number, unique identification code, coordinate, style, interface address, animation speed, particle color, range, alarm value and unit field of each device output by the configuration data prediction model are written into the same JSON array to ensure that the field name corresponds to the template placeholder one by one.

[0078] The template content is illustrated as follows: The bit number point table template: the first line is the Chinese title "bit number, unique identification code, range, unit, alarm high, alarm low, proportional coefficient, integral time, differential time", and the subsequent lines use the placeholder form to write each device record. The Web configuration JSON template: taking "device" as the key, an object array containing bit number, unique identification code, coordinate object, style string and data interface address is generated. The animation description JSON template: taking "animation" as the key, an object array containing bit number, unique identification code, water flow speed, particle color, diffusion radius and sedimentation time is generated.

[0079] The process of rendering the JSON array through the Handlebars configuration strategy is as follows:

[0080] The input JSON array is read to obtain the device array;

[0081] The template engine compilation function is called to inject the device array into the above three templates respectively;

[0082] Generate CSV data (point table data), Web configuration JSON data (Web configuration data) and animation JSON data (animation description data) in parallel within the same process;

[0083] Write three sets of text into the corresponding files in the input directory and calculate the file digest for version checking.

[0084] The rendering result is as follows:

[0085] Point table data: bit number, unique identification code, range, unit, high alarm, low alarm, proportional coefficient, integral time, differential time (PAC03, 3fa2c4b9, 400, L / h, 360, 0, 1.2, 0.3, 0).

[0086] Web configuration JSON data:

[0087] {“device”:[{

[0088] “bit number”:“PAC03”,“unique identification code”:“3fa2c4b9”,“coordinates”:{“x”:120,“y”:80},

[0089] “style”:

[0090] “position:absolute;width:32px;height:32px;background:url(pump.png)”,

[0091] “interface”:“ / api / realtime / 3fa2c4b9”}]}.

[0092] Animation JSON file:

[0093] {“animation”:[{

[0094] “bit number”:“PAC03”,“unique identification code”:“3fa2c4b9”,

[0095] “water flow”:{“speed”:73,“color”:“#00BFFF”},“particle”:{“radius”:84},

[0096] “sediment”:0}]}.

[0097] The preset bit number point table, the Web configuration JSON, and the animation description JSON three sets of templates are combined with the standardized JSON array output by the decision tree, so that the three sets of heterogeneous files of 5000 devices can be synchronously rendered within 300 ms, and the configuration file generation efficiency is greatly improved. Moreover, through the accurate matching of the field name and the template placeholder and the parallel rendering mechanism, the output CSV bit number point table, Web configuration JSON, and animation description JSON can be directly imported into the SCADA, the front-end framework, and the animation engine, so that the "bit number-position-animation" is landed at one time, and the error rate of manual operation is reduced to 0. Meanwhile, the template and the decision tree model are completely decoupled, subsequent field changes only need to update the template, and the model does not need to be adjusted, so that the system maintenance cost is significantly reduced, and the standardized configuration data effectively avoids the synchronization problem caused by the naming and parameter difference between multiple systems, and further guarantees the normativity and stability of the water plant automatic control system.

[0098] Optionally, in a specific embodiment, the Web configuration data, the bit number point table data, and the animation description data of the device are respectively written into the pre-set Web configuration file, the bit number point table file, and the animation description file, including:

[0099] The unique identification code corresponding to the bit number point table data of the device is generated, and the bit number point table data and the unique identification code of the device are written into the bit number point table file as a piece of data, the Web configuration data and the unique identification code of the device are written into the Web configuration file as a group of data, and the animation description data and the unique identification code of the device are written into the animation description file as a piece of data;

[0100] The Web configuration data, the bit number point table data, and the animation description data are mapped and aligned based on the unique identification codes in the Web configuration file, the bit number point table file, and the animation description file.

[0101] Further, after the Web configuration data, the bit number point table data, and the animation description data are mapped and aligned, the method further includes:

[0102] When any Web configuration data, bit number point table data, or animation description data in the Web configuration file, the bit number point table file, and the animation description file is modified, the contents of the Web configuration data, the bit number point table data, or the animation description data in the Web configuration file, the bit number point table file, and the animation description file are synchronized.

[0103] The specific process is as follows:

[0104] Initial synchronization: After rendering, the web configuration module bulk inserts records into the "tag table file", each record generates a UUID; the "tag table file" immediately broadcasts the UUID and neutral fields through the MQTT Tag / Initial topic; the SCADA and IoT platform subscribe to this topic, and after receiving the message, they establish an external UUID→internal tag mapping view, completing the first alignment in <3s.

[0105] Runtime modification (take the SCADA side as an example): The operation and maintenance personnel change the upper limit of the PAC03 range from 400L / h to 450L / h on the SCADA human-machine interface; SCADA calls the RESTful interface PATCH / tag / {UUID} / range, and the request body only contains the field "range:450"; the tag table file checks the existence of UUID→updates the field→writes the operation log→feedback OK; the pig data table immediately pushes a JSON message to the MQTT Tag / Update topic; the web configuration and IoT platform subscribe to the same topic, and after receiving the message: the web configuration updates the VuexStore in real time, and the picture range bar is automatically lengthened, without refreshing the liquid surface, the IoT platform updates the device shadow document, triggering the recalculation of the energy consumption benchmark.

[0106] When two platforms modify the same field at the same time (with a preset time range), the "timestamp late" wins, if the timestamp difference is <1s, then according to the pre-set priority, the arbitration result is broadcast with the MQTT message, and each platform rolls back the local value and prompts the user.

[0107] Hot plug and rollback: Add a device, insert it into the main data table after rendering, automatically trigger "Tag / Add" broadcast, each system dynamically adds mapping, no need to restart; delete the device, mark "deprecated" in the main data table and broadcast, each system interface is displayed in gray, 30 days of cooling-off period, then physically delete, prevent accidental deletion; version rollback, the main data table retains the last 10 version snapshots, administrators can restore to any historical moment through "one-key rollback", the rollback message is broadcast through the "Tag / Rollback" topic, and each system receives it and synchronously covers the local value.

[0108] This embodiment establishes a "tag table file" in the water plant field, using a 32-bit globally unique identifier (UUID) as the primary key, and stores "tag naming, range, unit, alarm upper and lower limit, PID initial value, animation parameters" as neutral fields; any system initiates a write operation, which is pushed to the remaining two systems through the publish-subscribe mechanism within seconds, achieving "one modification, three same changes".

[0109] Further, to overcome the defects of traditional solutions that must rely on professional UI software (InTouch, IFIX, KingView, etc.), the water plant process personnel can complete the whole process of "device placement → data binding → process animation" through a browser without the need to install plugins or hand-draw animations.

[0110] The precondition is that the "bit number master data UUID" initial synchronization has been completed, and the browser end has loaded the "2D Web configuration engine" provided by the system (WebAssembly + JS of about 320KB, automatically cached for the first visit).

[0111] The operation process is: enter the configuration platform URL in the browser address bar, automatically jump to the "project list page"; click "new project", a dialog box pops up to fill in only "project name" and "process section type" (drop-down options: water intake, dosing, flocculation, sedimentation, filtration, disinfection, sludge discharge), The system background immediately calls the decision tree reasoning interface, returns the default device list and animation parameters of this process section, and automatically generates a project UUID.

[0112] Enter the "canvas page", the left side is the "device pool" panel, and the icons in the panel have been pre-filtered by the decision tree according to "process section + device type", only listing the devices that may be used in this project (for example, only PAC metering pump, electromagnetic flowmeter, turbidimeter, and stirrer appear in the dosing room project); Drag the icon to any position on the canvas, and release the mouse: the front end reads the pixel coordinates (x, y) through getBoundingClientRect(), converts it to 0.1% precision relative coordinates, and immediately writes to a temporary JSON; The background sends a "position change" message through WebSocket, and the master data table records the latest coordinates of the device UUID, the whole process is <200ms; If the user wants to place in batches, you can click the "automatic arrangement" button: the system automatically calculates the row and column spacing according to the typical flow direction of the water plant (from top to bottom, return type, concentric type), generates anti-collision grid coordinates, and one-key positioning to avoid overlap.

[0113] Click any device icon that has been placed, and the "property drawer" pops up on the right. The "data binding" area in the drawer has listed all the bindable quantities (flow, frequency, turbidity, valve opening, etc.) corresponding to the device UUID, and the list is pulled in real time from the master data table; The user only needs to select the "real point number of the plant" (the drop-down content is dynamically obtained through the REST interface, consistent with the SCADA point table) in the "binding point" drop-down box, and the system immediately writes the binding field and stores it back to the master data table; After successful binding, the "chain" icon automatically appears in the lower right corner of the icon, and the real-time value can be viewed by hovering the mouse (WebSocket pushes every second), without refreshing the page.

[0114] Process animation configuration: After selecting the device, the drawer switches to the "animation" tab, and the system has pre-filled default values according to the decision tree output: water flow speed, particle color, diffusion radius, and sediment residence time. Users can directly modify the values or click the "visual adjustment" button, and a timeline + slider window will pop up at the bottom of the canvas. When the slider is dragged, the Canvas redraws the particle motion in real time, achieving "what you see is what you get". After confirmation, click "apply", the system writes the latest animation parameters to the ani.json template, synchronizes to the main data table, and broadcasts the MQTT AniUpdate topic to the animation engine. The whole process takes less than 1 s. If you need to turn off the animation, you can switch the "enable / disable" switch without deleting the primitives.

[0115] Save and publish: Click the "save" button at the top, and the front end writes the current canvas JSON (including coordinates, bindings, and animations) to the background through PUT / project / {project UUID}. The background immediately performs three checks: coordinate conflict detection (overlap area > 0 is considered a conflict); binding missing detection (no real-time point binding is highlighted in red); animation overrun detection (speed > 200 px / s or color not in the Web-safe color list is intercepted). After passing the checks, "release version number Vn+1" is automatically generated, and the "ConfigUpdate" event is broadcast to SCADA and IoT platforms. After receiving the event, each platform pulls the latest JSON and completes hot updates without restarting.

[0116] Online preview and rollback: After saving successfully, the "preview" button appears in the upper right corner. Clicking it opens a read-only address in a new tab. The address contains the project UUID, which can be accessed directly by any terminal (PC / tablet / mobile). If you find errors, you can click "rollback to Vn" in the "version history" list, and the system will immediately restore the main data table and canvas JSON to the previous version and broadcast the event again, achieving second-level rollback.

[0117] In one aspect, the global unique identifier (UUID) is used to establish a multi-file data mapping, and the MQTT publish-subscribe mechanism is used to realize "one modification, three same changes". This not only realizes the precise alignment and second-level synchronization of Web configuration, bit number point table and animation description data, but also solves the multi-platform concurrent modification conflict through timestamp priority arbitration. In combination with the hot plug and version rollback mechanism, it not only supports dynamic addition and deletion of devices without system restart, but also realizes one-key rollback through historical snapshots, effectively avoiding the risk of misoperation and ensuring the consistency and security of configuration data. On the other hand, based on the lightweight 2D Web configuration engine, process personnel can complete the whole process configuration only with a browser without the need to install professional software and plug-ins. The device pool can be automatically filtered according to the process section, drag-and-drop device placement and batch automatic arrangement are supported, and the device properties and animation parameters have been pre-filled with default values by decision tree, supporting visual adjustment and real-time preview. The threefold verification when saving can intercept configuration abnormalities, and after publishing, multi-platform hot updates can be realized, which not only reduces the configuration operation threshold and improves the configuration efficiency, but also realizes the "what you see is what you get" process animation configuration, completely eliminating the dependence on professional UI and coding personnel, greatly reducing the configuration error rate, and realizing the lightweight and intelligent of water plant two-dimensional configuration.

[0118] Optionally, in one specific embodiment, the configuration data and the configuration data tag each include: device configuration data and animation configuration data.

[0119] A model training set is constructed based on the pre-set device historical operation records of each device in the water plant, including:

[0120] The water plant point table and the process as-built drawing of the water plant are collected, and all historical working condition data of each device in the water plant and the corresponding device configuration data of each historical working condition data are obtained according to the water plant point table and the process as-built drawing.

[0121] Based on each historical working condition data and the pre-established mapping relationship table, the animation configuration data corresponding to each historical working condition data is obtained.

[0122] The mapping relationship table is a mapping relationship corresponding table between historical working condition data and animation configuration data.

[0123] The historical working condition data, device configuration data and animation configuration data corresponding to each device are preprocessed, and a model training set is established based on all preprocessed historical working condition data, device configuration data and animation configuration data. The preprocessing includes data cleaning and field standardization.

[0124] Specifically, a historical sample library is constructed, and the point table and process as-built drawing of the DCS / SCADA platform of the built water plant in the past 5 years are extracted, covering 7 process sections of water intake, dosing, flocculation, sedimentation, filtration, disinfection and sludge discharge, with a total of 183420 device records.

[0125] Data cleaning, removing sensor fault segments, debugging period abnormal segments, and retaining samples with stable operation for more than 30 days.

[0126] Field unification, mapping heterogeneous names such as "P101", "first-stage pump #1", etc. to "process section_device type_sequence number" format according to existing rules and experience correction "Water Plant Automatic Control Design Unified Coding Guidelines", forming a standard tag.

[0127] Animation labeling: supplemented by UI design, pixel speed, color, and radius of water flow, dosing diffusion, and sediment residence as animation labels.

[0128] The embodiment integrates water plant DCS / SCADA point table, process as-built drawing, and UI animation design experience to construct a high-quality, structured multi-modal training sample library, covering core process sections and containing more than 180,000 stable operation records, unifying equipment naming conventions and supplementing animation configuration labels, effectively solving the problems of heterogeneous historical data, noise, and inconsistent semantics. The model training set formed on this basis enables the configuration data prediction model to simultaneously output engineering-compliant equipment parameters and visually consistent animation parameters, ensuring accurate and reliable control logic such as position number, range, and alarm, and guaranteeing realistic and credible animation effects such as water flow speed, particle color, and diffusion radius, significantly improving configuration efficiency.

[0129] Optionally, in a specific embodiment, the pre-constructed CART decision tree model is trained based on the model training set, including:

[0130] Initializing the CART decision tree model;

[0131] Feature extraction is performed on the model training set to obtain a first feature matrix corresponding to historical working condition data and a second feature matrix corresponding to configuration data labels;

[0132] The first feature matrix and the second feature matrix are input into the CART decision tree model for recursive splitting;

[0133] During the training process, a node sample set corresponding to each node in the CART decision tree model is obtained; each node sample set is a subset selected from the first feature matrix and the first feature matrix according to the split path condition pre-set in the previous node in the CART decision tree model;

[0134] Based on the node sample set corresponding to each node, a first weighted Gini index before splitting of each node is obtained;

[0135] obtaining a second weighted Gini index of each node after splitting based on the splitting path condition and the node sample set of each node, and obtaining an optimization amplitude corresponding to each node based on the first weighted Gini index and the second weighted Gini index of each node; the optimization amplitude is a difference between the first weighted Gini index and the second weighted Gini index;

[0136] obtaining a final splitting score corresponding to each node based on the optimization amplitude corresponding to each node and a preset service penalty coefficient;

[0137] optimizing the model parameters of the CART decision tree model based on the optimization amplitude corresponding to each node and the final splitting score.

[0138] Further, obtaining the second weighted Gini index of each node after splitting based on the splitting path condition and the node sample set of each node, comprises:

[0139] splitting the node sample set of each node into a first node sample set and a second node sample set based on the splitting path condition of each node;

[0140] obtaining a third weighted Gini index and a fourth weighted Gini index corresponding to each node respectively based on the first node sample set and the second node sample set of each node and a preset formula one; the formula one is:

[0141]

[0142] wherein, Gini is a weighted Gini index, K is a total number of variable categories, k is an index of a variable category, w k is a preset relevance weight of the variable category with the index k, p k is a proportion of all corresponding samples belonging to the variable category with the index k;

[0143] obtaining the second weighted Gini index based on the third weighted Gini index and the fourth weighted Gini index corresponding to each node; the second weighted Gini index is a sum of a product of a proportion of a sample amount in the first node sample set in a sample amount in the node sample set and the third weighted Gini index, and a product of a proportion of a sample amount in the second node sample set in the sample amount in the node sample set and the fourth weighted Gini index.

[0144] Further, obtaining the final splitting score corresponding to each node based on the optimization amplitude corresponding to each node and a preset service penalty coefficient, comprises:

[0145] judging whether the splitting path condition corresponding to each node exists conflict based on a preset service conflict judgment strategy;

[0146] if there is no service conflict, the final splitting score corresponding to each node is the optimization amplitude corresponding to the node.​

[0147] If there is a business conflict, the final split score corresponding to each node is the difference between the optimization amplitude and the business penalty coefficient.

[0148] Specifically, the model parameters are initialized, and the basic parameters are set: the tree structure constraint parameters are set, including the maximum depth of the tree is 8, and the minimum sample size of the leaf node is 50; at the same time, the business related parameters are initialized, such as sensor accuracy benchmark (used to determine the range difference threshold), business penalty coefficient λ = 0.05~0.1 (empirical value), mutual information-weight mapping rule (the higher the inter-column mutual information, the smaller the weight w k ).

[0149] The model input is the historical working condition feature, and the output is an 8-dimensional configuration vector (bit number naming, range, alarm value, PID parameter, animation parameter), which ensures that the model completes multi-label joint learning under the same split criterion, rather than 8 times of independent training.

[0150] The historical working condition data of all devices are extracted from the model training set, including process section, device type, flow rate Q, turbidity TU, pump frequency Hz, valve opening %, and other features, which are sorted into a historical working condition feature matrix X (i.e. the first feature matrix). The configuration data label corresponding to each historical working condition data is extracted, and is integrated into an 8-dimensional configuration label matrix Y (i.e. the second feature matrix) according to the fixed dimension (bit number, range upper and lower limit, four-limit alarm value, PID parameter, animation parameter), establishing a one-to-one mapping relationship between "working condition feature X→configuration label Y".

[0151] The complete feature matrix X and label matrix Y are taken as the initial sample set of the root node.

[0152] In the recursive splitting process, the sample set S of the current node is a subset filtered from the sample set of its parent node according to the split path condition of the parent node (such as "process section = dosing and Q>2000m 3 / h").

[0153] For the sample set S of the current node, the first weighted Gini index Gini(S) before splitting is calculated according to formula one; formula one is:

[0154] ;

[0155] Wherein, Gini is the weighted Gini index, K is the total number of variable categories, k is the index of variable category, w k is the correlation weight of the variable category with index k, and p k is the proportion of all samples belonging to the variable category with index k.

[0156] Based on the split path condition of the current node (such as "Q>2000m 3 / h"), the sample set S is split into left child node sample set S left and right child node sample set S right .

[0157] The weighted Gini index of S left and S right is calculated respectively, denoted as the third weighted Gini index Gini(S left ) and the fourth weighted Gini index Gini(S right ), and the calculation logic is consistent with the first weighted Gini index.

[0158] The weighted sum is calculated according to the proportion of sub-node samples to calculate the second weighted Gini index Gini(S split ).

[0159] The improvement range (optimization range) of the purity of the split node △Gini is calculated, and the formula is:

[0160] △Gini=Gini(S)-Gini(S split ), the greater △Gini is, the better the optimization effect of the split scheme on the node purity is.

[0161] The range difference of the corresponding equipment of the left child node sample set S left and the right child node sample set S right is verified, if the difference is greater than 10 times the sensor accuracy (such as the flow sensor accuracy 0.5m 3 / h, then the threshold value is 5m 3 / h), it is determined that there is a business conflict, and the penalty item Penalty=0.05 is triggered; otherwise, Penalty=0.

[0162] The final split score (Score) is calculated by combining the optimization range and the business penalty item, and the formula is:

[0163] Score=△Gini-λ·Penalty;

[0164] If there is no business conflict: Score=ΔGini, the score is determined by the data gain; if there is a business conflict: Score=ΔGini−0.05λ, the priority of the business conflict reduction scheme is reduced.

[0165] Traverse all possible splitting path conditions of the current node, select the scheme with the highest final splitting score to perform node splitting, ensure that the splitting meets both data rules and engineering constraints; split the node according to the selected scheme, if the sample set of the child node meets the condition that "the minimum sample size of the leaf node ≥ 50" or "the maximum depth of the tree = 8", stop the splitting of this branch, mark the child node as a leaf node, and store the 8-dimensional configuration vector joint mode of the sample set of this node (instead of the single majority class of the traditional CART); otherwise, take the sample set of the child node as a new S, and repeat the above recursive splitting process; repeat the above steps until all branches meet the stopping condition, generate a preliminary CART decision tree model, and then prune the model through expert constraint strategies to obtain the final configuration data prediction model.

[0166] The embodiment can accurately quantify the optimization amplitude of the splitting on the node purity by constructing the working condition and configuration double feature matrix, combining the hierarchical screening and weighted Gini index calculation of the node sample set, weakening the interference of redundant columns through mutual information weight distribution, and focusing on the core working condition variables to effectively suppress overfitting; At the same time, the business penalty coefficient and conflict judgment mechanism are integrated to reduce the weight of the splitting scheme with too large range jump, so that the splitting result meets the engineering practice of water plant equipment procurement and operation and maintenance; The multi-label integrated training mode is matched, so that the model can output a complete configuration vector at one time, avoid parameter mismatch, and select the optimal splitting scheme through iterative optimization, and the design of storing the joint mode of the leaf node further improves the consistency and accuracy of the configuration data prediction, lays a high-quality data foundation for subsequent Handlebars rendering and multi-system configuration, and greatly reduces the cost of manual intervention and configuration error rate.

[0167] Optionally, in a specific embodiment, all nodes of the CART decision tree model are traversed to prune a node based on an expert constraint strategy when a splitting path condition and a node output of any node of the CART decision tree model do not match the expert constraint strategy, including:

[0168] Traverse all nodes of the CART decision tree model from the bottom layer of the CART decision tree model to the top layer, and check the splitting path condition and the node output of each node during the traversal process;

[0169] When the splitting path condition and the node output of any node of the CART decision tree model do not match the expert constraint strategy, replace / delete the node output of the node based on whether the node output of the node is related to the expert constraint strategy.

[0170] Specifically, the national standard provisions and enterprise regulations are converted into "hard rules", and the decision tree is pruned after training to ensure that the model is readable and meets the specifications. For example (only a few examples are given):

[0171] Range rule: IF Process Section = Chemical Feed AND Design Flow > 2000 m 3 / h THEN Metering Pump Range = 400 L / h;

[0172] Alarm rule: IF Sedimentation Tank AND Raw Water Turbidity > 20 NTU THEN Effluent Turbidity Alarm High Limit = 18 NTU;

[0173] PID rule: IF Chemical Feed AND Variable Frequency Metering Pump THEN Initial P = 1.2, I = 0.3, D = 0;

[0174] Animation rule: IF Chemical Feed AND PAC Solution THEN Water Flow Color = #00BFFF, Particle Velocity = 60 px / s.

[0175] The above rules are written into the "expert constraint table", and after the CART tree building is completed, the nodes are traversed from bottom to top: branches that conflict with the expert rules are deleted or merged to form a final version of the rule-compliant and interpretable tree.

[0176] The present embodiment can directly identify and process the situation that the split path condition or node output does not match the rules by forced checking of the trained CART decision tree nodes through four types of expert constraint rules of range, alarm, PID and animation, replacing the conflicting output, deleting the illegal branches and other operations, and completely avoiding the prediction results that do not conform to the water plant operation specifications due to the simple fitting of data.

[0177] In addition, the embodiment of the present application provides a two-dimensional configuration system for a water plant, which includes a memory, a processor and a computer program stored in the memory, and the processor executes the computer program to realize the two-dimensional configuration method for a water plant in the above embodiment.

[0178] In the description of the present application, it should be understood that the terms "first", "second" are only for the purpose of description, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "multiple" is two or more, unless otherwise specifically limited.

[0179] In the present application, unless specifically defined otherwise, the terms "mounting", "connected", "connection", "fixed", and the like, should be construed broadly and can be understood as, for example, fixed connection, detachable connection, or integral; mechanical connection, or electrical connection; direct connection, or indirect connection via an intermediate medium; internal communication between two elements, or interaction between two elements. The specific meaning of the above terms in the present application can be understood according to the specific circumstances by those skilled in the art.

[0180] In the present application, unless specifically defined otherwise, the first feature is "on" or "under" the second feature, which can be direct contact between the first and second features, or indirect contact between the first and second features through an intermediate medium. Moreover, the first feature is "above", "over", and "on" the second feature, which can be directly above or obliquely above the second feature, or simply means that the first feature is higher in horizontal height than the second feature. The first feature is "below", "under", and "under" the second feature, which can be directly below or obliquely below the second feature, or simply means that the first feature is lower in horizontal height than the second feature.

[0181] In the description of the present application, the description of the terms "one embodiment", "some embodiments", "embodiment", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present application, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. Furthermore, different embodiments or examples described in the present application and the features of different embodiments or examples can be combined and combined by those skilled in the art without contradiction.

[0182] Although the embodiments of the present application have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limiting the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present application.

Claims

1. A two-dimensional configuration method for water plants, characterized in that, include: Input the operating data of any equipment in the water plant into a pre-trained configuration data prediction model to obtain the corresponding configuration data; The configuration data is rendered based on the pre-deployed Handlebars configuration strategy to obtain Web configuration data, tag point table data, and animation description data for supporting two-dimensional configuration of the device. The configuration data prediction model is trained based on the following steps: The model training set is built based on the pre-set historical operation records of each piece of equipment in the water plant; the equipment operation record corresponding to any piece of equipment includes historical operating data for multiple time periods, as well as the configuration data labels corresponding to the historical operating data for each time period; The pre-built CART decision tree model is trained based on the model training set; After training, all nodes of the CART decision tree model are traversed. If the splitting path conditions and node output of any node in the CART decision tree model do not match the pre-set expert constraint strategy, the node is pruned based on the expert constraint strategy to obtain a configuration data prediction model for predicting device configuration data.

2. The two-dimensional configuration method for water plants according to claim 1, characterized in that, Both the configuration data and the configuration data tags include: device configuration data and animation configuration data; The model training set is constructed based on the pre-set historical operation records of each piece of equipment in the water plant, including: Collect the water plant location table and process as-built drawings of the water plant, and obtain all historical operating condition data of each piece of equipment in the water plant, as well as the equipment configuration data corresponding to each historical operating condition data, based on the water plant location table and process as-built drawings. Based on each historical working condition data and a pre-established mapping table, obtain the animation configuration data corresponding to each historical working condition data. The mapping table is a mapping relationship table between historical working condition data and animation configuration data; For each device, the historical operating condition data, device configuration data, and animation configuration data are preprocessed, and a model training set is established based on all the preprocessed historical operating condition data, device configuration data, and animation configuration data; the preprocessing includes data cleaning and field standardization.

3. The two-dimensional configuration method for water plants according to claim 1, characterized in that, The pre-built CART decision tree model is trained based on the model training set, including: Initialize the CART decision tree model; Feature extraction is performed on the model training set to obtain the first feature matrix corresponding to the historical working condition data and the second feature matrix corresponding to the configuration data labels; The first and second feature matrices are input into the CART decision tree model for recursive splitting; During training, the node sample set corresponding to each node in the CART decision tree model is obtained; each node sample set is a subset selected from the first feature matrix and the first feature matrix according to the splitting path conditions pre-set in the previous node of the CART decision tree model. Based on the node sample set corresponding to each node, obtain the first weighted Gini index of each node before splitting. Based on the splitting path conditions and node sample set of each node, the second weighted Gini index after splitting each node is obtained, and based on the first weighted Gini index and the second weighted Gini index of each node, the optimization magnitude corresponding to each node is obtained; the optimization magnitude is the difference between the first weighted Gini index and the second weighted Gini index. Based on the optimization magnitude corresponding to each node and the pre-set business penalty coefficient, the final split score corresponding to each node is obtained. The model parameters of the CART decision tree model are optimized based on the optimization magnitude and final split score corresponding to each node.

4. The two-dimensional configuration method for water plants according to claim 3, characterized in that, Based on the splitting path conditions and node sample set for each node, the second weighted Gini index after each node split is obtained, including: Based on the splitting path conditions of each node, the node sample set of each node is split into the first node sample set and the second node sample set. Based on the first and second node sample sets for each node, and a pre-set formula (Formula 1), the third and fourth weighted Gini indices for each node are obtained respectively; the first formula is: ； Where Gini is the weighted Gini index, K is the total number of variable categories, k is the index of the variable category, and w k For the variable category with index k, a pre-defined relevance weight is p. k This represents the percentage of all samples belonging to the variable category with index k. Based on the third and fourth weighted Gini indices corresponding to each node, the second weighted Gini index is obtained; the second weighted Gini index is the sum of the product of the proportion of the sample size in the first node sample set to the sample size in the node sample set and the third weighted Gini index, and the product of the proportion of the sample size in the second node sample set to the sample size in the node sample set and the fourth weighted Gini index.

5. The two-dimensional configuration method for water plants according to claim 4, characterized in that, Based on the optimization magnitude for each node and the pre-set business penalty coefficient, the final split score for each node is obtained, including: Based on a pre-set business conflict judgment strategy, determine whether there is a conflict in the split path conditions corresponding to each node. If there is no business conflict, the final split score for each node is the optimization magnitude for that node. If there is a business conflict, the final split score for each node is the difference between the optimization margin and the business penalty coefficient.

6. The two-dimensional configuration method for water plants according to claim 1, characterized in that, Handlebars configuration strategies include tag point table configuration strategy, Web configuration JSON configuration strategy, and animation description JSON configuration strategy; Then, based on the pre-deployed Handlebars configuration strategy, the configuration data is rendered to obtain Web configuration data, tag point table data, and animation description data for supporting two-dimensional device configuration, including: The configuration parameters are subjected to feature extraction and data integration to obtain the corresponding JSON array; The JSON array is rendered based on the tag point table configuration strategy, the Web configuration JSON configuration strategy, and the animation description JSON configuration strategy, respectively, to obtain Web configuration data, tag point table data, and animation description data for supporting two-dimensional configuration of the device; Write the device's Web configuration data, tag point table data, and animation description data into the pre-set Web configuration file, tag point table file, and animation description file, respectively.

7. The two-dimensional configuration method for water plants according to claim 1, characterized in that, Traverse all nodes of the CART decision tree model. When the splitting path conditions and node output of any node in the CART decision tree model do not match the pre-set expert constraint policy, prune the node based on the expert constraint policy, including: The process involves traversing the CART decision tree model layer by layer from the bottom node upwards, and verifying the splitting path conditions and node outputs of each node during the traversal. If the splitting path conditions and node output of any node in the CART decision tree model do not match the pre-set expert constraint strategy, the node output of that node is replaced / deleted based on whether the node output is related to the expert constraint strategy.

8. The two-dimensional configuration method for water plants according to claim 6, characterized in that, Write the device's Web configuration data, tag point table data, and animation description data into the pre-configured Web configuration file, tag point table file, and animation description file, respectively, including: Generate a corresponding unique identification code based on the device's tag number table data, and write the device's tag number table data and unique identification code as a single data entry into the tag number table file. Write the device's Web configuration data and unique identification code as a set of data into the Web configuration file. Write the device's animation description data and unique identification code as a single data entry into the animation description file. Based on the unique identifiers in the Web configuration file, tag point table file, and animation description file, the Web configuration data, tag point table data, and animation description data are mapped and aligned.

9. The two-dimensional configuration method for water plants according to claim 7, characterized in that, After mapping and aligning the Web configuration data, tag point table data, and animation description data, the method further includes: When any Web configuration data, tag point table data, or animation description data in the Web configuration file, tag point table file, and animation description file is modified, the content of the Web configuration data, tag point table data, or animation description data in the Web configuration file, tag point table file, and animation description file is synchronized.

10. A two-dimensional configuration system for water plants, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the two-dimensional configuration method for water plants as described in any one of claims 1 to 9.

Images