Water supply pipe network equipment management system based on digital twinning

Abstract

The invention relates to the technical field of intelligent management of water supply pipe networks, in particular to a water supply pipe network equipment management system based on digital twinning. Comprising a data acquisition module which is used for periodically acquiring and preprocessing pressure and flow data of each monitoring point of the water supply network; a reliability evaluation module; the reliability evaluation module is used for calculating the timeliness rate and the integrity rate of pressure and flow data of each monitoring point position and carrying out reliability grading based on the calculated timeliness rate and the integrity rate; the method has the beneficial effects that 1, a fault point location can be positioned through reliability evaluation, and the manual inspection cost is reduced; 2, a daily score drives the model to continuously iterate, and the hydraulic simulation precision is improved; and 3, data support is provided for pipe network pressure regulation and control and leakage detection, the pipe explosion accident rate is reduced, and the intelligent level of water operation is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent management of water supply pipe networks, and in particular to a water supply pipe network equipment management system based on digital twins. Background Art

[0002] The traditional water supply network management model relies on manual experience and decentralized systems. Equipment ledgers, inspection records, and operation data are scattered in Excel, paper documents, and SCADA systems, and cannot be analyzed in a linked manner. Traditional data only records "point status" such as pressure values at a certain moment, lacking spatial topological relationships such as pipeline connection relationships and time evolution laws, making it difficult to support dynamic decision-making. At the same time, operation and maintenance rely on manual labor, which is inefficient and costly. Manual inspections have problems such as "missed inspections, misjudgments, and lags." For example, the corrosion of buried pipelines requires excavation and inspection, which is time-consuming and laborious; nighttime pipe bursts often result in a large waste of water resources due to long inspection intervals. In addition, traditional models are not accurate enough and have weak predictive capabilities. Traditional hydraulic models (such as EPANET) are built offline based on historical data and cannot update the network status (such as valve opening changes, new users) in real time, resulting in large deviations between simulation results and actual results. Therefore, it is necessary to design a new type of water supply network equipment management system to better realize the operation monitoring and management of the water supply network. Summary of the invention

[0003] In view of the above problems, the present invention proposes a water supply network equipment management system based on digital twin, comprising:

[0004] Data acquisition module: The data acquisition module is used to periodically collect and pre-process the pressure and flow data of each monitoring point in the water supply network;

[0005] Reliability evaluation module; the reliability evaluation module includes calculating the timeliness and completeness of the pressure and flow data at each monitoring point, and performing reliability grading based on the calculated timeliness and completeness;

[0006] Data analysis module: The data analysis module performs comprehensive scoring on monitoring points with high reliability, and obtains the comprehensive score of each monitoring point and the overall score of the hydraulic model system;

[0007] Digital twin module: The digital twin module includes a virtual model, a data center and a visualization module;

[0008] The virtual model includes a water supply network hydraulic model system built based on a GIS map, and in the hydraulic model system, each virtual node matches the predicted value of the corresponding monitoring point, and synchronizes the monitoring data of each monitoring point in real time;

[0009] The data middle platform includes a time series database for storing pressure and flow time series data and supporting high-concurrency writing and second-level query; and a metadata module for recording metadata such as acquisition frequency, transmission frequency, and device type, and for subsequent statistical rule matching.

[0010] The visualization module displays the timeliness rate and integrity rate distribution of each monitoring point in real time, and marks the reliability status with traffic lights; at the same time, it displays the curves of the actual measurement values and model prediction values of pressure and flow, and displays the calculation errors of pressure and flow.

[0011] Abnormal point warning module: The abnormal point warning module sets a first-level warning according to the monitoring points with low reliability; and sets a second-level warning according to the model score.

[0012] Preferably, the calculation method of the timeliness rate of monitoring data includes:

[0013]

[0014] Wherein, R represents the daily actual received data volume, and T represents the theoretical data volume, specifically 288 pieces per day.

[0015] The calculation method of the integrity rate of monitoring data includes:

[0016]

[0017] Preferably, the reliability classification includes:

[0018] If TR≥X and CR≥X, then this monitoring point is marked as having high reliability;

[0019] If TR≤Y and CR≤Y, then this monitoring point is marked as having low reliability;

[0020] Wherein, X and Y are set thresholds.

[0021] Preferably, the comprehensive score of the pressure value of the monitoring point and the comprehensive score of the flow value of the monitoring point, wherein the comprehensive score of the pressure value of the monitoring point includes:

[0022] Calculate the mean absolute error of the pressure value of the monitoring points with high reliability:

[0023]

[0024] Wherein, represents the predicted value of the i-th sample output by the digital twin module, represents the actual pressure measurement value of the i-th sample, and n represents the number of samples;

[0025] Calculate the mean Nash coefficient of the pressure value of the monitoring points with high reliability:

[0026]

[0027] Among them, is expressed as the average value of all measured pressure data;

[0028] Calculate the error fluctuation of the pressure value at the monitoring points with high reliability:

[0029] , and

[0030] Among them, is expressed as the error between the actual pressure measurement value and the pressure prediction value of the i-th sample, is expressed as the average value of all pressure prediction errors from 1 to n;

[0031] Comprehensive scoring system for the pressure value of the monitoring points:

[0032] Single-index score, with a full score of 100 points:

[0033] MAE score: If ≤ threshold A, get 100 points; for every 0.1 exceeding threshold A, deduct 5 points until 0 points;

[0034] NSE score: If ≥ threshold B, get 100 points; for every 0.1 lower than threshold B, deduct 5 points until 0 points;

[0035] Error fluctuation score: If the standard deviation of the pressure value ≤ threshold C, get 100 points; for every 0.1 exceeding threshold C, deduct 2 points until 0 points;

[0036] Comprehensive score of the pressure value of the monitoring points:

[0037] Comprehensive score of the pressure value of the monitoring point = 0.4 × score + 0.4 × score + 0.2 × error fluctuation score of the pressure value

[0038] Overall score of the pressure value of the hydraulic model system:

[0039]

[0040] Among them, is the comprehensive score of the pressure value of the k-th monitoring point, and M is the number of high-reliability points.

[0041] Preferably, the comprehensive score of the flow value of the monitoring points includes:

[0042] Calculate the mean absolute error of the flow value at the monitoring points with high reliability:

[0043]

[0044] Among them, represents the predicted value of the i-th sample output by the digital twin module, represents the actual flow measurement value of the i-th sample, and n represents the number of samples;

[0045] Calculate the average Nash coefficient of the flow values at monitoring points with high reliability:

[0046]

[0047] Among them, represents the average value of all measured flow data from 1 to n;

[0048] Calculate the error fluctuation of the flow values at monitoring points with high reliability:

[0049] , and

[0050] Among them, represents the error between the actual flow measurement value and the flow prediction value of the i-th sample, represents the average value of all flow prediction errors from 1 to n;

[0051] Comprehensive scoring system for flow values at monitoring points:

[0052] MAE score: If ≤ threshold D, get 100 points; for every 10% exceeding the threshold D, deduct 5 points until 0 points;

[0053] NSE score: If ≥ threshold E, get 100 points; for every 10% lower than the threshold E, deduct 5 points until 0 points;

[0054] Error fluctuation score: If the standard deviation of the flow value ≤ threshold F, get 100 points; for every 10% exceeding the threshold F, deduct 2 points until 0 points.

[0055] Comprehensive score of flow values at monitoring points:

[0056] Comprehensive score of flow values at monitoring points = 0.4 × MAE flow value score + 0.4 × NSE flow value score + 0.2 × flow value error fluctuation score

[0057] Overall score of flow values of the hydraulic model system:

[0058]

[0059] Among them is the comprehensive score of the flow value at the k-th monitoring point.

[0060] Preferably, the triggering of the first-level warning includes triggering the first-level warning if the reliability of the monitoring point is low.

[0061] Preferably, if the overall score of the hydraulic model system drops by more than 10% for three consecutive days, a second-level warning is triggered, and it is prompted that the model needs to be calibrated.

[0062] Preferably, it includes a model optimization module. The model optimization module inputs the measured data and the model prediction error into the hydraulic model system, and iteratively optimizes the hydraulic simulation parameters through machine learning algorithms.

[0063] The beneficial effects of this application are as follows: 1. Fault points can be located through reliability assessment, reducing the cost of manual inspection; 2. The daily scoring drives the continuous iteration of the model, improving the accuracy of hydraulic simulation; 3. It provides data support for pipe network pressure regulation and leakage detection, reduces the pipe burst accident rate, and significantly improves the intelligent level of water service operation. Description of the Drawings

[0064] Figure 1 The figure shows a schematic diagram of the system of this application. Detailed Embodiments

[0065] In order to enable those skilled in the art of this technical field to better understand the technical solution of this application, the following further elaborates on the present invention in conjunction with the attached

[0066] drawings and the best embodiments.

[0067] The present invention proposes a digital twin-based water supply pipe network equipment management system, including:

[0068] Data acquisition module: The data acquisition module is used for periodically collecting and preprocessing the pressure and flow data of each monitoring point of the water supply pipe network; the frequency requirements for data acquisition include: only selecting points where both the acquisition frequency and the transmission frequency are ≤ 5 minutes, ensuring that the daily theoretical data volume = 24h × 60min / 5 = 288 pieces, and the hardware equipment includes deploying Internet of Things sensors, including pressure / flow transmitters and sensors. The communication protocols adopted include: transmitting data to the edge computing gateway through NB-IoT, LoRa or 5G to complete data format standardization.

[0069] Reliability assessment module; The reliability assessment module includes calculating the timeliness rate and integrity rate of the pressure and flow data of each monitoring point, and performing reliability classification based on the calculated timeliness rate and integrity rate;

[0070] Data analysis module: The data analysis module comprehensively scores the monitoring points with high reliability, and obtains the comprehensive scores of each monitoring point and the overall score of the hydraulic model system;

[0071] Digital Twin Module: The digital twin module includes a virtual model, a data middle platform, and a visualization module;

[0072] The virtual model includes a water supply network hydraulic model system built based on a GIS map. In the hydraulic model system, each virtual node matches the predicted value of the corresponding monitoring point, and the monitoring data of each monitoring point is synchronized in real time. At the same time, the water supply network hydraulic model system also integrates an equipment ledger, including water supply equipment models, installation times, maintenance records, etc.

[0073] The data middle platform includes a time series database for storing pressure and flow time series data and supporting high-concurrency writing and second-level query, such as using an InfluxDB database; and a metadata module for recording collection frequency, transmission frequency, and equipment type, and for subsequent statistical rule matching.

[0074] The visualization module displays the timeliness rate and integrity rate distribution of each monitoring point in real time, and marks the reliability status with traffic lights, green = high, red = low; at the same time, it displays the curves of the actual measured values and model predicted values of pressure and flow, marks indicators such as MAE and NSE, and displays the calculation errors of pressure and flow.

[0075] The calculation errors of pressure and flow include:

[0076] The error of the pressure value is: simulated value - actual measured value

[0077] The error of the flow value is: (simulated value - actual measured value) / actual measured value

[0078] At the same time, a daily model accuracy report is automatically generated every morning, including the scoring of each point, the overall water supply network hydraulic model system, and early warnings for abnormal points.

[0079] Abnormal Point Early Warning Module: The abnormal point early warning module sets a first-level early warning according to the monitoring points with low reliability, such as sending a text message to notify the operation and maintenance personnel; sets a second-level early warning according to the model score. The trigger of the first-level early warning includes triggering the first-level early warning if the reliability of the monitoring point is low. If the overall score of the water supply network hydraulic model system drops by more than 10% continuously for 3 days, the second-level early warning is triggered, and it is prompted that the model needs to be calibrated. And equipment maintenance work orders are automatically generated, such as sensor failures and transmission link interruptions, associated with the GPS positioning of operation and maintenance personnel, and the processing time limit is recorded.

[0080] The calculation method of the timeliness rate of monitoring data includes:

[0081]

[0082] Among them, R represents the actual amount of data received daily, and T represents the theoretical amount of data, specifically 288 pieces per day;

[0083] The calculation method of the integrity rate of monitoring data includes:

[0084]

[0085] Among them, invalid data are abnormal values such as pressure value ≤ 0 or flow value being negative, etc.

[0086] The reliability grading includes:

[0087] If TR ≥ X and CR ≥ X, then the monitoring point is marked as having high reliability;

[0088] If TR ≤ Y and CR ≤ Y, then the monitoring point is marked as having low reliability;

[0089] Among them, X and Y are set thresholds. The intermediate interval is marked as having medium reliability. At the same time, the threshold X > Y is configured.

[0090] Preferably, the comprehensive score of the pressure value of the monitoring point and the comprehensive score of the flow value of the monitoring point. Among them, the comprehensive score of the pressure value of the monitoring point includes:

[0091] Calculate the mean absolute error of the pressure value of the monitoring points with high reliability:

[0092]

[0093] Among them, represents the predicted value of the i-th sample output by the digital twin module, represents the actual pressure measurement value of the i-th sample, and n represents the number of samples;

[0094] Calculate the mean Nash coefficient of the pressure value of the monitoring points with high reliability:

[0095]

[0096] Among them, represents the average value of all measured pressure data;

[0097] Calculate the error fluctuation of the pressure value of the monitoring points with high reliability:

[0098] , and

[0099] Among them, represents the error between the actual pressure measurement value and the pressure prediction value of the i-th sample, represents the average value of all pressure prediction errors from 1 to n;

[0100] The comprehensive scoring system for the pressure value of the monitoring point:

[0101] Single-index score, with a full score of 100 points:

[0102] MAE score: If ≤ threshold A, get 100 points; for every 0.1 exceeding threshold A, deduct 5 points until 0 points;

[0103] NSE score: If ≥ threshold B, get 100 points; for every 0.1 lower than threshold B, deduct 5 points until 0 points;

[0104] Error fluctuation score: If the standard deviation of the pressure value ≤ threshold C, get 100 points; for every 0.1 exceeding threshold C, deduct 2 points until 0 points; Threshold examples: A = 1, B = 0.7, C = 1, which can be adjusted according to business requirements.

[0105] Comprehensive score of the pressure value at the monitoring point:

[0106] Comprehensive score of the pressure value at the monitoring point = 0.4 × score + 0.4 × score + 0.2 × pressure value error fluctuation score

[0107] Overall score of the pressure value of the hydraulic model system:

[0108]

[0109] Among them, is the comprehensive score of the pressure value at the kth monitoring point, and M is the number of high-reliability points.

[0110] Preferably, the comprehensive score of the flow value at the monitoring point includes:

[0111] Calculate the mean absolute error of the flow value at the monitoring points with high reliability:

[0112]

[0113] Among them, represents the predicted value of the ith sample output by the digital twin module, represents the actual flow measurement value of the ith sample, and n represents the number of samples; Using 's calculation method can eliminate the influence caused by the flow base due to the caliber or scale, make the flow errors at different positions and different scales comparable, and truly reflect the simulation accuracy.

[0114] Calculate the mean Nash coefficient of the flow value at the monitoring points with high reliability:

[0115]

[0116] Among them, represents the average value of all flow measured data from 1 to n;

[0117] Calculate the error fluctuation of the flow value at the monitoring point with high calculation reliability:

[0118] , and

[0119] wherein, represents the error between the actual flow measurement value and the flow prediction value of the i-th sample, represents the average value of all flow prediction errors from 1 to n;

[0120] Comprehensive scoring system for the flow value of the monitoring point:

[0121] Single-index score, with a full score of 100 points:

[0122] MAE score: If ≤ threshold D, get 100 points; for every 10% exceeding the threshold D, deduct 5 points until 0 points;

[0123] NSE score: If ≥ threshold E, get 100 points; for every 10% lower than the threshold E, deduct 5 points until 0 points;

[0124] Error fluctuation score: If the standard deviation of the flow value ≤ threshold F, get 100 points; for every 10% exceeding the threshold F, deduct 2 points until 0 points; Threshold examples: D = 0.2, E = 1, F = 0.1, which can be adjusted according to business requirements.

[0125] Comprehensive score of the flow value of the monitoring point:

[0126] Comprehensive score of the flow value of the monitoring point = 0.4 × MAE flow value score + 0.4 × NSE flow value score + 0.2 × flow value error fluctuation score

[0127] Overall score of the flow value of the hydraulic model system:

[0128]

[0129] where is the comprehensive score of the flow value of the k-th monitoring point.

[0130] At the same time, it also includes a model optimization module, which inputs the measured data and the model prediction error into the digital twin module, and iteratively optimizes the hydraulic simulation parameters through machine learning algorithms.

[0131] Through this system, the full-process digital management of water supply network equipment from data collection - reliability assessment - model analysis - optimization iteration can be realized, significantly improving the intelligent level of water service operation.

[0132] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A digital twin-based water supply network equipment management system, characterized in that, Including: Data acquisition module: The data acquisition module is used to periodically collect and preprocess the pressure and flow data of each monitoring point in the water supply network; Reliability assessment module; The reliability assessment module includes calculating the timeliness rate and completeness rate of the pressure and flow data of each monitoring point, and performing reliability grading based on the calculated timeliness rate and completeness rate; Data analysis module: The data analysis module conducts a comprehensive score for the monitoring points with high reliability, and calculates the comprehensive score of each monitoring point and the overall score of the hydraulic model system; Digital twin module: The digital twin module includes a virtual model, a data middle platform, and a visualization module; The virtual model includes a hydraulic model system of the water supply network built based on a GIS map. In the hydraulic model system, each virtual node matches the predicted value of the corresponding monitoring point, and the monitoring data of each monitoring point is synchronized in real time; The data middle platform includes a time series database for storing pressure and flow time series data and supporting high-concurrency writing and second-level query; and a metadata module for recording collection frequency, transmission frequency, and equipment type, and used for subsequent statistical rule matching; The visualization module displays the distribution of the timeliness rate and completeness rate of each monitoring point in real time, and marks the reliability status with traffic lights; at the same time, it displays the curves of the actual measured values and model predicted values of pressure and flow, and displays the calculation errors of pressure and flow; Abnormal point warning module: The abnormal point warning module sets a first-level warning according to the monitoring points with low reliability; and sets a second-level warning according to the model score.

2. The digital twin-based water supply network equipment management system according to claim 1, wherein The calculation method of the timeliness rate of the monitoring data includes: ； Wherein, R represents the actual daily received data volume, and T represents the theoretical data volume, specifically 288 pieces per day; The calculation method of the completeness rate of the monitoring data includes: 。 3. The digital twin-based water supply network equipment management system according to claim 2, wherein The reliability grading includes: If TR≥X and CR≥X, then this monitoring point is marked as having high reliability; If TR≤Y and CR≤Y, then this monitoring point is marked as having low reliability; Wherein, X and Y are set thresholds.

4. The digital twin-based water supply network equipment management system according to claim 3, wherein The comprehensive score of the monitoring point includes: the comprehensive score of the pressure value of the monitoring point and the comprehensive score of the flow value of the monitoring point. Among them, the comprehensive score of the pressure value of the monitoring point includes: Calculating the mean absolute error of the pressure value of the monitoring points with high reliability: ； Among them, represents the predicted value of the i-th sample output by the digital twin module, represents the actual pressure measurement value of the i-th sample, and n represents the number of samples; Calculating the mean Nash coefficient of the pressure value of the monitoring points with high reliability: ； Among them, is expressed as the average value of all measured pressure data from 1 to n; Calculating the error fluctuation of the pressure value of the monitoring points with high reliability: and ; Among them, represents the error between the actual pressure measurement value and the pressure prediction value of the i-th sample, represents the average value of all pressure prediction errors from 1 to n; Comprehensive score system of the pressure value of the monitoring point: Single-index score, full score 100 points: MAE score: If ≤ threshold A, get 100 points; for every 0.1 exceeding threshold A, deduct 5 points until 0 points; NSE score: If ≥ threshold B, get 100 points; for every 0.1 lower than threshold B, deduct 5 points until 0 points; Error fluctuation score: If the standard deviation of the pressure value ≤ threshold C, get 100 points; for every 0.1 exceeding threshold C, deduct 2 points until 0 points; Comprehensive score of the pressure value of the monitoring point: Comprehensive score of pressure values at monitoring points = 0.4 × Score + 0.4 × Score + 0.2 × Score of pressure value error fluctuation Overall score of the pressure value of the hydraulic model system: ； Among them, is the comprehensive score of the pressure value at the k-th monitoring point, and M is the number of high-reliability points.

5. The digital twin-based water supply network equipment management system according to claim 4, wherein The comprehensive score of the flow value of the monitoring point includes: Calculating the mean absolute error of the flow value of the monitoring points with high reliability: ； Among them, represents the predicted value of the i-th sample output by the digital twin module, represents the actual flow measurement value of the i-th sample, and n represents the number of samples; Calculating the mean Nash coefficient of the flow value of the monitoring points with high reliability: ； Among them, is expressed as the average value of all measured traffic data from 1 to n; Calculating the error fluctuation of the flow value of the monitoring points with high reliability: and ; Among them, represents the error between the actual flow measurement value and the flow prediction value of the i-th sample, represents the average value of all flow prediction errors from 1 to n; Comprehensive score system of the flow value of the monitoring point: Single-index score, full score 100 points: MAE score: If ≤ threshold D, get 100 points; for every 10% exceeding threshold D, deduct 5 points until 0 points; NSE score: If ≥ threshold E, get 100 points; for every 10% lower than threshold E, deduct 5 points until 0 points; Error fluctuation score: If the standard deviation of the flow value ≤ threshold F, get 100 points; for every 10% exceeding threshold F, deduct 2 points until 0 points; Comprehensive score of the flow value of the monitoring point: Comprehensive score of flow values at monitoring points = 0.4 × MAE flow value score + 0.4 × NSE flow value score + 0.2 × flow value error fluctuation score Overall score of flow values in the hydraulic model system: ； Among them is the comprehensive score of the flow value at the k-th monitoring point.

6. The digital twin-based water supply network equipment management system according to claim 4, wherein The triggering of the first-level warning includes triggering the first-level warning if the reliability of the monitoring point is low.

7. The digital twin-based water supply network equipment management system according to claim 4, wherein If the overall score of the hydraulic model system drops by more than 10% for three consecutive days, a second-level warning will be triggered, and it will be prompted that the model needs to be calibrated.

8. The digital twin-based water supply network equipment management system according to claim 1, wherein, It includes a model optimization module. The model optimization module inputs the measured data and the model prediction error into the hydraulic model system, and iteratively optimizes the hydraulic simulation parameters through machine learning algorithms.

Images