Multi-level water supply and drainage balance monitoring method and system for water affair industry

Abstract

The invention discloses a multi-level water supply and drainage balance monitoring method and system for the water affair industry, and relates to the technical field of water affair monitoring. The method comprises the following steps: deploying a multi-level monitoring network; the method comprises the following steps: collecting real-time data of a multi-level monitoring network, executing space-time alignment operation, and preprocessing the data; constructing a quarterly three-dimensional water balance equation, and dynamically estimating through Kalman filtering; dBSCAN clustering is implemented based on pipe network topology, and leakage positioning is carried out to generate a leakage probability thermodynamic diagram; and establishing a hierarchical response mechanism to carry out multi-level early warning linkage, and rendering a multi-layer perspective view in a GIS (Geographic Information System) engine. According to the method, a quarterly three-dimensional water balance equation is constructed through a city-region-enterprise three-level dynamic weight model and a hierarchical penetration algorithm, and dynamic estimation is performed through Kalman filtering; dBSCAN clustering is implemented based on the pipe network topology, leakage positioning is performed, a leakage probability thermodynamic diagram is generated, the water affair monitoring efficiency is improved, and the manual troubleshooting workload is reduced.

Description

Technical Field

[0001] The present invention belongs to the field of water monitoring technology, and in particular relates to a multi-level water supply and drainage balance monitoring method and system for the water industry. By constructing a spatiotemporal coupled multi-level monitoring system, the invention innovatively integrates intelligent algorithms with water expertise to form a smart water solution with autonomous diagnostic capabilities. Background Art

[0002] With the acceleration of my country's urbanization process, the supply and drainage balance management of water systems faces severe challenges. Traditional water monitoring technology has the following prominent problems:

[0003] (1) Fragmentation of monitoring levels: Existing systems mostly operate independently at a single level (such as a municipal platform or an enterprise DCS system), lacking a dynamic linkage mechanism for data at the municipal, district, and enterprise levels. For example, when a deviation occurs between municipal water supply and discharge, manual checks of district-level pipe networks and enterprise water meter data are required, resulting in poor response timeliness (usually taking 3-5 days).

[0004] (2) Extensive temporal dimension: Conventional statistics are based on an annual cycle, which makes it difficult to capture the characteristics of seasonal water volume fluctuations.

[0005] (3) Lack of indicator system: The existing method only calculates the water supply and sewage volume, ignoring the dynamic impact of external water volume (such as groundwater infiltration and inter-regional water transfer), resulting in deviations in the calculation of supply and discharge balance.

[0006] (4) Insufficient algorithm intelligence: Water supply leakage detection mostly relies on manual inspections and zone metering, which cannot identify continuous leakage below 5%;

[0007] (4.1) Drainage performance evaluation focuses only on treatment volume, ignoring the coupling relationship between energy consumption and water quality compliance, making it difficult to truly reflect the operating efficiency of the treatment plant;

[0008] (4.2) The balance analysis uses a static threshold method, which cannot adapt to dynamic factors such as pipeline network aging and population mobility.

[0009] There is an urgent need for a multi-level water supply and drainage balance monitoring method and system in the water industry. This patent enhances the refined management and control capabilities of water systems by constructing a spatiotemporally coupled quarterly balance model and a hierarchical penetration mechanism, combined with an improved machine learning algorithm. Summary of the Invention

[0010] The purpose of the present invention is to provide a multi-level water supply and drainage balance monitoring method and system for the water industry. Through the three-level dynamic weight model of "city-district-enterprise" and the hierarchical penetration algorithm, a quarterly three-dimensional water balance equation is constructed, and dynamic estimation is performed through Kalman filtering; DBSCAN clustering is implemented based on the pipeline network topology to locate leakage and generate a leakage probability heat map, which solves the problems of insufficient refined control of existing water systems, lack of linkage in the monitoring layer, and insufficient intelligence.

[0011] To solve the above technical problems, the present invention is achieved through the following technical solutions:

[0012] The present invention is a multi-level water supply and drainage balance monitoring method for the water industry, comprising the following steps:

[0013] Step S1: deploy a multi-level monitoring network;

[0014] Step S2: Collect real-time data from the multi-level monitoring network, perform spatiotemporal alignment operations, and pre-process the data;

[0015] Step S3: Calculate the information entropy value according to the hierarchy and generate an adaptive weight matrix for dynamic weight allocation;

[0016] Step S4: constructing a quarterly three-dimensional water balance equation and dynamically estimating it through Kalman filtering;

[0017] Step S5: Implement DBSCAN clustering based on the pipe network topology, superimpose the nighttime minimum flow analysis, perform leakage location, and generate a leakage probability heat map;

[0018] Step S6: Establish a hierarchical response mechanism for multi-level early warning linkage and send abnormal information to the mobile inspection terminal;

[0019] Step S7: Rendering the multi-layer perspective view in the GIS engine.

[0020] As a preferred technical solution, in step S1, the multi-level monitoring network includes: deploying a cloud platform server cluster at the municipal water dispatching center, installing pressure transmitters and electromagnetic flowmeters at key nodes of the district and county pipeline networks, configuring ultrasonic water meters with RTU modules at the water intakes of key enterprises, and deploying water quality spectrometers at the inlets and outlets of sewage treatment plants.

[0021] As a preferred technical solution, in step S2, the specific process of performing the spatiotemporal alignment operation is as follows:

[0022] Step S21, dynamic time series adjustment: define a time series cost function, synchronize batch data through a sliding window, output an alignment path, and generate a unified time series matrix;

[0023] Step S22, spatial topology reconstruction: constructing a graph structure of the pipe network GIS topology and performing spatial interpolation;

[0024] Step S23, abnormal data collaborative filtering: establish a hybrid detection model and generate a three-dimensional abnormality confidence cloud map;

[0025] Step S24, spatiotemporal benchmark verification: perform a cross-validation mechanism for the time dimension and the space dimension, and output an alignment quality report; the alignment quality report includes: a maximum time deviation distribution histogram, a spatial reconstruction error heat map, and an abnormal data repair log.

[0026] As a preferred technical solution, in step S21, the specific process of dynamic timing adjustment is as follows:

[0027] Step S211: constructing a cumulative distance matrix to allow the time series of different monitoring devices to be aligned under nonlinear paths, solving the problems of clock drift and communication delay between devices;

[0028] Step S212: Introduce a Gaussian weighted time deviation penalty term into the distance calculation:

[0029]

[0030] Where γ is the penalty intensity coefficient, σ is the control time tolerance, and D(i,j) is the minimum cumulative alignment cost between the first i points of sequence x and the first j points of sequence y;

[0031] Step S213: Set the main window according to the quarterly data span and superimpose a floating sub-window of ±2 hours to adapt to the device clock synchronization error, limit the fluctuation of the regularized path within the sub-window, and avoid cross-quarter data misalignment; perform composite feature calculations, integrating multi-dimensional parameters such as pressure and flow in addition to water volume data;

[0032] Step S214: reversely search for the minimum cost path from the lower right corner of the accumulation matrix to generate an aligned unified timestamp sequence;

[0033] Step S215: using cubic spline interpolation to repair data faults at continuous jump points in the path;

[0034] Step S216: Calculate the root mean square error of the aligned sequence, which is required to be ≤3%. The processing time for calculating single-quarter data is ≤15 minutes, meeting the real-time requirement.

[0035] As a preferred technical solution, in step S22, the spatial topology reconstruction includes spatial alignment and temporal alignment. The spatial alignment is performed by constructing a water affairs digital twin map in the EPSG coordinate system, loading the pipe network topology structure vector layer, and using the Delaunay triangulation algorithm to perform spatial interpolation on discrete monitoring points. The specific formula is as follows:

[0036]

[0037] Where w i Dynamic adjustment of the spatial density of detection points;

[0038] Time alignment is achieved by creating a quarterly timeline and using a sliding window mechanism to handle device clock deviations. The specific formula is as follows:

[0039]

[0040] Where, σ is 1.5 hours;

[0041] The NTP time synchronization service is activated to perform secondary timestamp re-marking on delayed data packets.

[0042] In step S23, a hybrid detection model is established to generate a three-dimensional anomaly confidence cloud map, and an XGBoost classifier is trained to identify anomaly patterns:

[0043] Input features: pressure fluctuation variance, flow rate change rate, water quality parameter gradient;

[0044] Output: abnormal probability value P∈[0,1];

[0045] Sliding window validation (window length = 24h, step length = 1h) was used to ensure model stability.

[0046] Construct a three-dimensional anomaly confidence field: C(x, y, t) = w1δ + w2P; where w1 and w2 are the statistical model weight coefficient and the machine learning weight coefficient, respectively; δ is the quarterly water balance deviation, and P represents the anomaly probability value; the generation of the three-dimensional anomaly confidence cloud map uses WebGL technology to achieve browser-side three-dimensional rendering.

[0047] As a preferred technical solution, in step S3, information entropy values ​​are calculated according to the hierarchy to generate an adaptive weight matrix for dynamic weight allocation;

[0048] The specific formula for calculating information entropy value according to the hierarchy is as follows:

[0049]

[0050] In the formula, m=3 corresponds to the city, district and enterprise levels respectively, p ij is the standardized indicator value, that is, the standardized value of the i-th indicator at the j-th level; when the entropy value E1 of the city is greater than 0.8, it is determined that the chaos of the city-level data is high, and its weight ratio is automatically reduced.

[0051] As a preferred technical solution, in step S4, the specific steps of constructing the quarterly three-dimensional water balance are as follows:

[0052] Step S41: construct a multi-source data cube, align the multi-source data, perform spatial interpolation on the meteorological satellite data, and match the coordinates of the pipe network nodes;

[0053] The data dimensions include time axis, space axis and element axis; the time axis is quarterly slices (Q1-Q4), with a time resolution of ±2 hours; the space axis is the three-level topology structure of city-district-enterprise (based on GIS pipe network node coding); the element axis is water supply, sewage volume, and external water volume;

[0054] Step S42: Perform dynamic evaluation of external water volume; the specific formula is as follows:

[0055]

[0056] Where, is the total water supply of the city, Q ext is the amount of external water, α is the sewage treatment rate correction coefficient, is the sewage volume in area d, β is the seasonal discharge factor, and ΔS is the change in the capacity of the discharge facility;

[0057] The specific formula for dynamic estimation through Kalman filtering is as follows:

[0058] X t =AX t-1 +B(u t )+w t ;

[0059] Where, X t is the estimated value of external water at time t, B(u t ) is the real-time rainfall infiltration, w t is the process noise, which is dynamically adjusted with the groundwater monitoring data, and A and B are weight coefficients;

[0060] Step S43: Dynamic calibration of equilibrium equation parameters;

[0061] Sewage treatment rate correction coefficient α: Where C out is the real-time concentration of effluent COD, C std is the emission standard value;

[0062] Step S44: three-dimensional balance deviation analysis;

[0063] Time dimension: Calculate the difference between the balance value of this quarter and the same period last year:

[0064]

[0065] Spatial dimension: Analyzing regional imbalance aggregation through the Moran index:

[0066]

[0067] Where x i is the balance deviation value of the i-th district-level node, w ij is the spatial weight matrix based on the hydraulic connectivity of the pipe network;

[0068] Factor dimension: The formula for calculating the contribution of each factor is as follows:

[0069]

[0070] Where δ is the total deviation;

[0071] Step S45: Dynamic feedback optimization, establishing a quarterly rolling optimization window, executing at the end of each quarter, and using the quantum particle swarm optimization algorithm (QPSO) for rapid optimization.

[0072] As a preferred technical solution, in step S5, the specific implementation process is as follows:

[0073] Step S51: Construct a pipe network hydraulic model based on the GIS topology, import basic pipe network parameters such as pipe diameter, material, burial depth, and other attribute parameters, and establish a pressure fluctuation transmission equation. The specific formula is:

[0074]

[0075] Where ΔP is the total pressure drop between two points in the pipeline, ρ is the fluid density, v1 and v2 are the average flow velocities at the front and end of the pipeline respectively, Δh is the elevation difference between the two ends of the pipeline, f is the Darcy-Weisbach friction coefficient, L is the actual length of the pipeline, and D is the nominal diameter of the pipeline.

[0076] Step S52: Adjust the DBSCAN search radius according to the real-time pipe diameter, construct a spatiotemporal constraint matrix, use 15-minute time slices in the time dimension, and establish accessibility constraints based on pipeline connectivity in the spatial dimension;

[0077] Step S53: Establish a dynamic baseline model and calculate the traffic baseline using the exponentially weighted moving average method. The specific formula is as follows:

[0078] Q t =λQ t-1 +(1-λ)Q obs ;

[0079] Where λ = 0.85 is the seasonal adjustment factor;

[0080] Step S54: Construct a leakage probability model, generate a heat map, and render the probability distribution on the official website GIS map; the specific formula of the leakage probability model is:

[0081] P leak =0.6PDBSCAN +0.3P flow +0.1P vib ;

[0082] Where, P DBSCAN is the confidence of clustering results, P flow is the flow deviation mapping value, P vib It is the acoustic vibration characteristic of fusion.

[0083] As a preferred technical solution, in step S6, the hierarchical response mechanism includes red warning, orange warning and yellow warning; when the three-level indicators are abnormal and the joint probability is greater than 75%, the red warning is triggered and the emergency plan generation is started; when the second-level indicators are abnormal and the balance deviation is greater than 8%, the orange warning is triggered and the pipeline pressure adjustment is initiated; when a single-level indicator is abnormal, the yellow indicator warning is triggered and an equipment maintenance work order is sent to the mobile inspection terminal.

[0084] The present invention is a multi-level water supply and drainage balance monitoring system for the water industry, comprising a data acquisition layer, a data processing layer, an intelligent analysis layer, and an application interaction layer;

[0085] The data acquisition layer includes multi-source sensing terminals and heterogeneous data gateways; the multi-source sensing terminals include smart water meters, pipe network pressure transmitters, and water quality analyzers; the smart water meters are deployed on water supply pipeline nodes to transmit water consumption data via the NB-IoT network; the pipe network pressure transmitters are deployed at the sudden changes in the water supply pipeline, the end of the main pipeline, valves, and pipeline outlets to collect pipeline pressure fluctuation data and dynamically generate pressure field thermal maps; the water quality analyzers are deployed at the water inlet and outlet to monitor water quality indicators through video spectral analysis; the heterogeneous data gateway is used to access multi-source sensing terminal devices through the Modbus / OPCUA protocol to build a unified data channel; the heterogeneous data gateway can also complete data cleaning, filtering, and compression locally to reduce the amount of data uploaded to the cloud (such as filtering low-value data and reducing bandwidth consumption); perform real-time analysis (such as equipment status monitoring and energy efficiency calculation), generate early warning signals or control instructions, and shorten response delays to milliseconds;

[0086] The data processing layer includes a spatiotemporal database, a data cleaning engine, and a water balance calculator; the spatiotemporal database is used to store quarterly dimension data using a time series database and establish a three-dimensional GIS spatial index; the data cleaning engine is used to automatically identify abnormal monitoring values ​​and perform data cleaning; the water balance calculator is used to execute the three-dimensional balance equation in real time and generate a dynamic balance thermodynamic map;

[0087] The intelligent analysis layer includes a multi-level cascade diagnosis center, an early warning center, and a decision optimization platform; the multi-level cascade diagnosis center includes a hierarchical penetration analysis module, a leakage location module, and an efficiency evaluation submodule; the early warning center integrates the Prophet time series model and the XGBoost classifier; the decision optimization platform is used to provide a multi-objective scheduling solution;

[0088] The application interaction layer includes a three-dimensional visualization terminal and a mobile inspection terminal; the three-dimensional visualization terminal is used to display a quarterly balance index radar chart; the mobile inspection terminal is used to overlay the health status of the pipeline network in real time.

[0089] The hierarchical penetration analysis module is used to establish a vertical penetration mechanism for monitoring data at the city, district, and enterprise levels. When city-level indicators are abnormal, the weight model is used to automatically trigger lower-level data tracing. The leakage location module is used to integrate GIS topological data with real-time sensor information to accurately identify pipeline leakage points. The efficiency evaluation submodule is used to calculate the treatment energy efficiency ratio (treatment volume × compliance rate / design volume × energy consumption), optimize the operation mode of the sewage treatment plant, construct a quarterly balance index radar chart, compare the LSTM prediction curve with the measured value, and identify seasonal production capacity deviations. The Copula coupling model is used to analyze the matching degree between water supply reliability and drainage carrying capacity, and warn of system imbalance risks.

[0090] As an optimal technical solution, the hierarchical penetration analysis module is used to establish a vertical penetration mechanism for monitoring data at the city, district, and enterprise levels. When city-level indicators are abnormal, the weight model is used to automatically trigger lower-level data tracing; the leakage positioning module is used to integrate GIS topology data and real-time sensor information to accurately identify pipeline leakage points; the efficiency evaluation submodule is used to calculate the treatment energy efficiency ratio (treatment volume × compliance rate / design volume × energy consumption), optimize the operation mode of the sewage treatment plant, construct a quarterly balance index radar chart, compare the LSTM prediction curve with the measured value, and identify seasonal production capacity deviations; and analyze the matching degree of water supply reliability and drainage carrying capacity through the Copula coupling model.

[0091] As a preferred technical solution, the above.

[0092] The present invention has the following beneficial effects:

[0093] (1) The present invention constructs a quarterly three-dimensional water balance equation through a three-level dynamic weight model of "city-district-enterprise" and a hierarchical penetration algorithm, and dynamically estimates it through Kalman filtering; implements DBSCAN clustering based on the pipe network topology, locates leakage and generates a leakage probability heat map, thereby improving water monitoring efficiency and reducing the workload of manual investigation.

[0094] (2) The present invention realizes nonlinear time regularization through dynamic windows, adapts to device clock drift and communication delay, introduces pipeline flow direction parameters in IDW interpolation, improves the calculation accuracy of spatial correlation of pressure data, integrates isolated forest rapid screening with spatiotemporal correlation Z-score fine screening, actively shields testing and periodic verification, establishes quantifiable alignment quality evaluation indicators, and improves data consistency.

[0095] (3) The present invention applies hydraulic transient pressure fluctuation comparison technology and combines it with an improved DBSCAN algorithm to perform cluster radius adaptive pipe diameter adjustment, which can detect tiny leaks of more than 0.5% of the pipe diameter, improve the leakage location accuracy, and reduce water resource waste.

[0096] (4) The present invention uses the supply and drainage balance map and AR inspection system integrated into the GIS engine to improve managers' perception of abnormal quarterly indicators and shorten the time for auxiliary decision-making. In particular, when responding to extreme weather such as heavy rain, the timeliness of emergency dispatch plan generation is improved.

[0097] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0098] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0099] Figure 1 This is a flow chart of a multi-level water supply and drainage balance monitoring method for the water industry of the present invention;

[0100] Figure 2 This is a structural schematic diagram of a multi-level water supply and drainage balance monitoring system for the water industry according to the present invention. DETAILED DESCRIPTION

[0101] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0102] In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0103] In order to make the purpose, technical solutions and advantages of this application more clear, the following Figure 1-2 It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application.

[0104] Example 1

[0105] See also Figure 1 As shown, the present invention is a multi-level water supply and drainage balance monitoring method for the water industry, comprising the following steps:

[0106] Step S1: deploy a multi-level monitoring network;

[0107] Step S2: Collect real-time data from the multi-level monitoring network, perform spatiotemporal alignment operations, and pre-process the data;

[0108] Step S3: Calculate the information entropy value according to the hierarchy and generate an adaptive weight matrix for dynamic weight allocation;

[0109] Step S4: constructing a quarterly three-dimensional water balance equation and dynamically estimating it through Kalman filtering;

[0110] Step S5: Implement DBSCAN clustering based on the pipe network topology, superimpose the nighttime minimum flow analysis, perform leakage location, and generate a leakage probability heat map;

[0111] Step S6: Establish a hierarchical response mechanism for multi-level early warning linkage and send abnormal information to the mobile inspection terminal;

[0112] Step S7: Rendering the multi-layer perspective view in the GIS engine.

[0113] In step S1, the multi-level monitoring network includes: deploying a cloud platform server cluster at the municipal water dispatching center, installing pressure transmitters and electromagnetic flowmeters at key nodes of the district and county pipe networks, configuring ultrasonic water meters with RTU modules at the water intakes of key enterprises, and deploying water quality spectrometers at the inlets and outlets of sewage treatment plants.

[0114] In step S2, the specific process of performing the spatiotemporal alignment operation is as follows:

[0115] Step S21, dynamic time series adjustment: define a time series cost function, synchronize batch data through a sliding window, output an alignment path, and generate a unified time series matrix;

[0116] Step S22, spatial topology reconstruction: constructing a graph structure of the pipe network GIS topology and performing spatial interpolation;

[0117] Step S23, abnormal data collaborative filtering: establish a hybrid detection model and generate a three-dimensional abnormality confidence cloud map;

[0118] Step S24, spatiotemporal benchmark verification: perform a cross-validation mechanism for the time dimension and the space dimension, and output an alignment quality report; the alignment quality report includes: a maximum time deviation distribution histogram, a spatial reconstruction error heat map, and an abnormal data repair log.

[0119] In step S21, the specific process of dynamic timing adjustment is as follows:

[0120] Step S211: constructing a cumulative distance matrix to allow the time series of different monitoring devices to be aligned under nonlinear paths, solving the problems of clock drift and communication delay between devices;

[0121] Step S212: Introduce a Gaussian weighted time deviation penalty term into the distance calculation:

[0122]

[0123] Where γ is the penalty intensity coefficient, σ is the control time tolerance, and D(i,j) is the minimum cumulative alignment cost between the first i points of sequence x and the first j points of sequence y;

[0124] Step S213: Set the main window according to the quarterly data span and superimpose a floating sub-window of ±2 hours to adapt to the device clock synchronization error, limit the fluctuation of the regularized path within the sub-window, and avoid cross-quarter data misalignment; perform composite feature calculations, integrating multi-dimensional parameters such as pressure and flow in addition to water volume data;

[0125] Step S214: reversely search for the minimum cost path from the lower right corner of the accumulation matrix to generate an aligned unified timestamp sequence;

[0126] Step S215: using cubic spline interpolation to repair data faults at continuous jump points in the path;

[0127] Step S216: Calculate the root mean square error of the aligned sequence, which is required to be ≤3%. The processing time for calculating single-quarter data is ≤15 minutes, meeting the real-time requirement.

[0128] In step S22, spatial topology reconstruction includes spatial alignment and temporal alignment. Spatial alignment is achieved by constructing a water affairs digital twin map in the EPSG coordinate system, loading the pipeline network topology structure vector layer, and using the Delaunay triangulation algorithm to perform spatial interpolation on discrete monitoring points. The specific formula is as follows:

[0129]

[0130] Where w i Dynamic adjustment of the spatial density of detection points;

[0131] Time alignment creates a quarterly timeline and uses a sliding window mechanism to handle device clock deviations. The specific formula is as follows:

[0132]

[0133] Where, σ is 1.5 hours;

[0134] The NTP time synchronization service is activated to perform secondary timestamp re-marking on delayed data packets.

[0135] In step S23, a hybrid detection model is established to generate a three-dimensional anomaly confidence cloud map, and an XGBoost classifier is trained to identify anomaly patterns:

[0136] Input features: pressure fluctuation variance, flow rate change rate, water quality parameter gradient;

[0137] Output: abnormal probability value P∈[0,1];

[0138] Sliding window validation (window length = 24h, step length = 1h) was used to ensure model stability.

[0139] Construct a three-dimensional anomaly confidence field: C(x, y, t) = w1δ + w2P; where w1 and w2 are the statistical model weight coefficient and the machine learning weight coefficient, respectively; δ is the quarterly water balance deviation, and P represents the anomaly probability value; the generation of the three-dimensional anomaly confidence cloud map uses WebGL technology to achieve browser-side three-dimensional rendering.

[0140] In step S3, the information entropy value is calculated according to the hierarchy, and an adaptive weight matrix is ​​generated for dynamic weight allocation;

[0141] The specific formula for calculating information entropy value according to the hierarchy is as follows:

[0142]

[0143] In the formula, m=3 corresponds to the city, district and enterprise levels respectively, p ij is the standardized indicator value, that is, the standardized value of the i-th indicator at the j-th level; when the entropy value E1 of the city is greater than 0.8, it is determined that the chaos of the city-level data is high, and its weight ratio is automatically reduced.

[0144] In step S4, the specific steps for constructing the quarterly three-dimensional water balance are as follows:

[0145] Step S41: construct a multi-source data cube, align the multi-source data, perform spatial interpolation on the meteorological satellite data, and match the coordinates of the pipe network nodes;

[0146] The data dimensions include time axis, space axis and element axis; the time axis is quarterly slices (Q1-Q4), with a time resolution of ±2 hours; the space axis is the three-level topology structure of city-district-enterprise (based on GIS pipe network node coding); the element axis is water supply, sewage volume, and external water volume;

[0147] Step S42: Perform dynamic evaluation of external water volume; the specific formula is as follows:

[0148]

[0149] Where, is the total water supply of the city, Q ext is the amount of external water, α is the sewage treatment rate correction coefficient, is the sewage volume in area d, β is the seasonal discharge factor, and ΔS is the change in the capacity of the discharge facility;

[0150] The specific formula for dynamic estimation through Kalman filtering is as follows:

[0151] X t =AX t-1 +B(u t )+w t ;

[0152] Where, X t is the estimated value of external water at time t, B(u t ) is the real-time rainfall infiltration, w t is the process noise, which is dynamically adjusted with the groundwater monitoring data, and A and B are weight coefficients;

[0153] Step S43: Dynamic calibration of equilibrium equation parameters;

[0154] Sewage treatment rate correction coefficient α: Where C out is the real-time concentration of effluent COD, C std is the emission standard value;

[0155] Step S44: three-dimensional balance deviation analysis;

[0156] Time dimension: Calculate the difference between the balance value of this quarter and the same period last year:

[0157]

[0158] Spatial dimension: Analyzing regional imbalance aggregation through the Moran index:

[0159]

[0160] Where x i is the balance deviation value of the i-th district-level node, w ijis the spatial weight matrix based on the hydraulic connectivity of the pipe network;

[0161] Factor dimension: The formula for calculating the contribution of each factor is as follows:

[0162]

[0163] Where δ is the total deviation;

[0164] Step S45: Dynamic feedback optimization, establishing a quarterly rolling optimization window, executing at the end of each quarter, and using the quantum particle swarm optimization algorithm (QPSO) for rapid optimization.

[0165] In step S5, the specific implementation process is as follows:

[0166] Step S51: Construct a pipe network hydraulic model based on the GIS topology, import basic pipe network parameters such as pipe diameter, material, burial depth, and other attribute parameters, and establish a pressure fluctuation transmission equation. The specific formula is:

[0167]

[0168] Where ΔP is the total pressure drop between two points in the pipeline, ρ is the fluid density, v1 and v2 are the average flow velocities at the front and end of the pipeline respectively, Δh is the elevation difference between the two ends of the pipeline, f is the Darcy-Weisbach friction coefficient, L is the actual length of the pipeline, and D is the nominal diameter of the pipeline.

[0169] Step S52: Adjust the DBSCAN search radius according to the real-time pipe diameter, construct a spatiotemporal constraint matrix, use 15-minute time slices in the time dimension, and establish accessibility constraints based on pipeline connectivity in the spatial dimension;

[0170] Step S53: Establish a dynamic baseline model and calculate the traffic baseline using the exponentially weighted moving average method. The specific formula is as follows:

[0171] Q t =λQ t-1 +(1-λ)Q obs ;

[0172] Where λ = 0.85 is the seasonal adjustment factor;

[0173] Step S54: Construct a leakage probability model, generate a heat map, and render the probability distribution on the official website GIS map; the specific formula of the leakage probability model is:

[0174] P leak =0.6P DBSCAN +0.3P flow +0.1P vib ;

[0175] Where, P DBSCANis the confidence of clustering results, P flow is the flow deviation mapping value, P vib It is the acoustic vibration characteristic of fusion.

[0176] In step S6, the hierarchical response mechanism includes red warning, orange warning and yellow warning; when the three-level indicators are abnormal and the joint probability is greater than 75%, the red warning is triggered and the emergency plan generation is started; when the second-level indicators are abnormal and the balance deviation is greater than 8%, the orange warning is triggered and the pipeline pressure adjustment is initiated; when a single-level indicator is abnormal, the yellow indicator warning is triggered and an equipment maintenance work order is sent to the mobile inspection terminal.

[0177] Example 2

[0178] See Figure 2 As shown, the present invention is a multi-level water supply and drainage balance monitoring system for the water industry, which can be used to implement the method content of Example 1 of the present invention, including: a data acquisition layer, a data processing layer, an intelligent analysis layer and an application interaction layer;

[0179] The data collection layer includes multi-source sensing terminals and heterogeneous data gateways. Multi-source sensing terminals include smart water meters, pipe network pressure transmitters, and water quality analyzers. Smart water meters are deployed at water supply pipeline nodes to transmit water consumption data via the NB-IoT network. Pipe network pressure transmitters are deployed at sudden changes in water supply pipelines, at the end of main pipelines, at valves, and at pipeline outlets to collect pipeline pressure fluctuation data and dynamically generate pressure field thermal maps. Water quality analyzers are deployed at water inlets and outlets to monitor water quality indicators through video spectral analysis. The heterogeneous data gateway is used to access multi-source sensing terminal devices via the Modbus / OPCUA protocol to build a unified data channel. The heterogeneous data gateway can also perform data cleaning, filtering, and compression locally to reduce the amount of data uploaded to the cloud (such as filtering low-value data and reducing bandwidth consumption). It performs real-time analysis (such as equipment status monitoring and energy efficiency calculation), generates warning signals or control instructions, and shortens response delays to milliseconds.

[0180] The data processing layer includes a spatiotemporal database, a data cleaning engine, and a water balance calculator. The spatiotemporal database uses a time series database to store quarterly dimension data and establish a three-dimensional GIS spatial index. The data cleaning engine automatically identifies abnormal monitoring values ​​and performs data cleaning. The water balance calculator executes the three-dimensional balance equation in real time and generates a dynamic balance heat map.

[0181] The intelligent analysis layer includes a multi-level diagnostic center, an early warning center, and a decision optimization platform. The multi-level diagnostic center includes a hierarchical penetration analysis module, a leakage location module, and an efficiency evaluation submodule. The early warning center integrates the Prophet time series model and the XGBoost classifier. The decision optimization platform is used to provide multi-objective scheduling solutions.

[0182] The application interaction layer includes a three-dimensional visualization terminal and a mobile inspection terminal; the three-dimensional visualization terminal is used to display the quarterly balance index radar chart; the mobile inspection terminal is used to overlay the health status of the pipeline network in real time.

[0183] The hierarchical penetration analysis module is used to establish a vertical penetration mechanism for monitoring data at the city, district, and enterprise levels. When city-level indicators are abnormal, the weight model automatically triggers the tracing of lower-level data. The leakage location module is used to integrate GIS topological data with real-time sensor information to accurately identify leakage points in the pipeline network. The efficiency evaluation submodule is used to calculate the treatment energy efficiency ratio (treatment volume × compliance rate / design volume × energy consumption), optimize the operation mode of the sewage treatment plant, construct a quarterly balance index radar chart, compare the LSTM prediction curve with the measured value, and identify seasonal production capacity deviations. The Copula coupling model is used to analyze the matching degree between water supply reliability and drainage carrying capacity, and warn of the risk of system imbalance.

[0184] In this embodiment,

[0185] (1) The locations where smart water meters are deployed include:

[0186] Water meter well in residential areas: at the entrance of the water supply pipeline in residential buildings, covering water consumption monitoring of household users;

[0187] Commercial building pipeline room: water supply main pipeline node of public buildings such as office buildings and shopping malls;

[0188] Industrial park water inlet: the access point of the factory production water main pipeline;

[0189] Farmland irrigation hub: the main water pipeline or branch valve in the agricultural irrigation area;

[0190] Specific effects: Minute-by-minute water usage data can be transmitted back through the NB-IoT network, replacing manual meter reading and enabling remote meter reading and data management. AI algorithms are used to analyze water consumption curves, identify abnormal minimum flow at night, and locate hidden leaks (with an accuracy of >99%) for leakage monitoring and early warning. Remote valve control and fee prepayment are supported, and tiered water pricing rules are automatically implemented, enabling prepaid and tiered water pricing. User water usage profiles are generated to provide data support for water-saving strategy formulation and enable water behavior analysis.

[0191] (2) The locations where the pipe network pressure transmitters are deployed include:

[0192] Pipeline network pressure-sensitive areas: places where the water flow direction of the water supply main pipe changes suddenly (such as elbows and tees);

[0193] Water-intensive areas: the ends of main pipes in areas with concentrated water loads, such as residential and commercial areas;

[0194] Pipeline network intersection node: key valve or pump station outlet for multi-channel water flow distribution;

[0195] Low-pressure / high-pressure risk areas: areas with significant elevation differences or long-distance water pipeline sections;

[0196] Specific effects:

[0197] Collect pipeline network pressure fluctuation data, dynamically generate pressure field thermal maps, and realize real-time pressure monitoring; identify pipe burst accidents through pressure drop characteristics, and combine GIS to locate abnormal points to realize pipe burst warning and location; coordinate water supply pump stations to adjust output pressure, reducing energy consumption by 10%-20%; provide measured pressure data for pipeline network simulation, and improve model prediction accuracy.

[0198] (3) The locations where water quality analyzers are deployed include:

[0199] Key nodes of water plants: raw water inlet, sedimentation tank outlet, clear water tank and other process sections;

[0200] Secondary water supply facilities: community pressure pump room, rooftop water tank outlet;

[0201] Rivers / sewage outlets: key river sections and sewage treatment plant outlets;

[0202] Emergency monitoring points: sites of sudden pollution incidents or water source protection areas;

[0203] Specific effects:

[0204] Through video spectral analysis, turbidity, color, suspended solids and other indicators are detected to achieve real-time monitoring of water quality parameters; abnormal events such as oil pollution, algae proliferation, floating garbage, etc. are automatically identified; the dosing system is linked to dynamically adjust the coagulant dosage, improving treatment efficiency by 20%; combined with water flow trajectory analysis, the pollution source is quickly located to achieve pollution tracing and early warning.

[0205] It is worth noting that in the above system embodiment, the various units included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0206] In addition, those skilled in the art will appreciate that all or part of the steps in the above-mentioned embodiments can be accomplished by instructing related hardware through a program, and the corresponding program can be stored in a computer-readable storage medium.

[0207] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to the specific embodiments described. Obviously, many modifications and variations are possible based on the content of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A multi-level water supply and drainage balance monitoring method for the water industry, characterized in that: The steps include: Step S1: deploy a multi-level monitoring network; Step S2: Collect real-time data from the multi-level monitoring network, perform spatiotemporal alignment operations, and pre-process the data; Step S3: Calculate the information entropy value according to the hierarchy and generate an adaptive weight matrix for dynamic weight allocation; Step S4: constructing a quarterly three-dimensional water balance equation and dynamically estimating it through Kalman filtering; Step S5: Implement DBSCAN clustering based on the pipe network topology, superimpose the nighttime minimum flow analysis, perform leakage location, and generate a leakage probability heat map; Step S6: Establish a hierarchical response mechanism for multi-level early warning linkage and send abnormal information to the mobile inspection terminal; Step S7: Rendering the multi-layer perspective view in the GIS engine.

2. A multi-level water supply and drainage balance monitoring method for the water industry according to claim 1, characterized in that: In step S1, the multi-level monitoring network includes: deploying a cloud platform server cluster at the municipal water dispatching center, installing pressure transmitters and electromagnetic flowmeters at key nodes of the district and county pipe networks, configuring ultrasonic water meters with RTU modules at the water intakes of key enterprises, and deploying water quality spectrometers at the inlets and outlets of sewage treatment plants.

3. A multi-level water supply and drainage balance monitoring method for the water industry according to claim 1, characterized in that: In step S2, the specific process of performing the spatiotemporal alignment operation is as follows: Step S21, dynamic time series adjustment: define a time series cost function, synchronize batch data through a sliding window, output an alignment path, and generate a unified time series matrix; Step S22, spatial topology reconstruction: constructing a graph structure of the pipe network GIS topology and performing spatial interpolation; Step S23, abnormal data collaborative filtering: establish a hybrid detection model and generate a three-dimensional abnormality confidence cloud map; Step S24, spatiotemporal benchmark verification: perform a cross-validation mechanism for the temporal and spatial dimensions and output an alignment quality report; The alignment quality report includes: maximum temporal deviation distribution histogram, spatial reconstruction error heatmap, and abnormal data repair log.

4. A multi-level water supply and drainage balance monitoring method for the water industry according to claim 3, characterized in that: In step S21, the specific process of dynamic timing adjustment is as follows: Step S211: constructing a cumulative distance matrix to allow the time series of different monitoring devices to be aligned under nonlinear paths; Step S212: introducing a Gaussian weighted time deviation penalty term into the distance calculation; Step S213: Setting the main window according to the quarterly data span and superimposing a floating sub-window of ±2 hours to limit the regularized path to fluctuate within the sub-window; Step S214: reversely search for the minimum cost path from the lower right corner of the accumulation matrix to generate an aligned unified timestamp sequence; Step S215: using cubic spline interpolation to repair data faults at continuous jump points in the path; Step S216: Calculate the root mean square error of the aligned sequences and the data processing time for a single quarter.

5. A multi-level water supply and drainage balance monitoring method for the water industry according to claim 3, characterized in that: In step S22, the spatial topology reconstruction includes spatial alignment and temporal alignment; the spatial alignment is performed by constructing a water affairs digital twin map in the EPSG coordinate system, loading the pipeline network topology structure vector layer, and using the Delaunay triangulation algorithm to perform spatial interpolation on discrete monitoring points; the temporal alignment is performed by creating a quarterly timeline, using a sliding window mechanism to process device clock deviations, activating the NTP time synchronization service, and performing secondary timestamp re-marking on delayed data packets.

6. A multi-level water supply and drainage balance monitoring method for the water industry according to claim 1, characterized in that: In step S4, the specific steps of constructing the quarterly three-dimensional water balance are as follows: Step S41: construct a multi-source data cube, align the multi-source data, perform spatial interpolation on the meteorological satellite data, and match the coordinates of the pipe network nodes; Step S42: Perform dynamic evaluation of external water volume; Step S43: Dynamic calibration of equilibrium equation parameters; Step S44: three-dimensional balance deviation analysis; Step S45: Dynamic feedback optimization.

7. A multi-level water supply and drainage balance monitoring method for the water industry according to claim 1, characterized in that: In step S5, the specific implementation process is as follows: Step S51: constructing a pipe network hydraulic model based on GIS topology, importing basic pipe network parameters, and establishing a pressure fluctuation transmission equation; Step S52: adjusting the DBSCAN search radius according to the real-time pipe diameter and constructing a spatiotemporal constraint matrix; Step S53: Establish a dynamic baseline model and calculate the flow baseline using the exponentially weighted moving average method; Step S54: Build a leakage probability model, generate a heat map and render the probability distribution on the official website GIS map.

8. A multi-level water supply and drainage balance monitoring method for the water industry according to claim 1, characterized in that: In step S6, the hierarchical response mechanism includes red warning, orange warning and yellow warning; when the three-level indicators are abnormal and the joint probability is greater than 75%, a red warning is triggered and the generation of an emergency plan is started; When the secondary indicators are abnormal and the balance deviation is greater than 8%, an orange warning is triggered and pipeline pressure adjustment is initiated; when a single-level indicator is abnormal, a yellow indicator warning is triggered and an equipment maintenance work order is sent to the mobile inspection terminal.

9. A multi-level water supply and drainage balance monitoring system for the water industry, comprising a data acquisition layer, a data processing layer, an intelligent analysis layer, and an application interaction layer, characterized by: The data acquisition layer includes multi-source perception terminals and heterogeneous data gateways; the multi-source perception terminals include smart water meters, pipe network pressure transmitters, and water quality analyzers; the smart water meters are deployed on water supply pipeline nodes and are used to transmit water consumption back through the NB-IoT network; the pipe network pressure transmitters are deployed at the sudden changes in the water supply pipelines, the ends of the main pipelines, valves, and pipeline outlets to collect pipeline pressure fluctuation data and dynamically generate pressure field thermal maps; the water quality analyzers are deployed at the water inlet and outlet to monitor water quality indicators through video spectral analysis; the heterogeneous data gateway is used to access multi-source perception terminal devices through the Modbus / OPCUA protocol to build a unified data channel; The data processing layer includes a spatiotemporal database, a data cleaning engine, and a water balance calculator; the spatiotemporal database is used to store quarterly dimension data using a time series database and establish a three-dimensional GIS spatial index; the data cleaning engine is used to automatically identify abnormal monitoring values ​​and perform data cleaning; the water balance calculator is used to execute the three-dimensional balance equation in real time and generate a dynamic balance thermodynamic map; The intelligent analysis layer includes a multi-level cascade diagnosis center, an early warning center, and a decision optimization platform; the multi-level cascade diagnosis center includes a hierarchical penetration analysis module, a leakage location module, and an efficiency evaluation submodule; the early warning center integrates the Prophet time series model and the XGBoost classifier; the decision optimization platform is used to provide a multi-objective scheduling solution; The application interaction layer includes a three-dimensional visualization terminal and a mobile inspection terminal; The three-dimensional visualization terminal is used to display the quarterly balance index radar chart; the mobile inspection terminal is used to overlay the health status of the pipeline network in real time.

10. The multi-level water supply and drainage balance monitoring system for the water industry according to claim 1, characterized in that: The hierarchical penetration analysis module is used to establish a vertical penetration mechanism for monitoring data at the city, district, and enterprise levels. When city-level indicators are abnormal, the weight model is used to automatically trigger lower-level data tracing. The leakage location module is used to integrate GIS topology data with real-time sensor information to identify pipeline leakage points. The efficiency evaluation submodule is used to calculate the treatment energy efficiency ratio, optimize the operation mode of the sewage treatment plant, construct a quarterly balance index radar chart, compare the LSTM prediction curve with the measured value, and identify seasonal production capacity deviations. The Copula coupling model is used to analyze the matching degree between water supply reliability and drainage carrying capacity.

Images