Digital asset life cycle management and intelligent water consumption energy-saving system for water supply enterprise
By introducing heterogeneous data acquisition, edge fusion, leakage location, and scheduling optimization modules into the water supply system, and combining sensor arrays and LSTM neural networks, the problems of leakage control and energy-saving scheduling in the water supply system were solved, achieving accurate location of leakage points and energy-saving water consumption.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING HUAIXIN IOT TECH CO LTD
- Filing Date
- 2026-01-14
- Publication Date
- 2026-04-24
AI Technical Summary
The existing water supply system is inadequate in terms of leakage control and energy-saving scheduling. It cannot quickly locate leakage points and its energy consumption remains high, resulting in water waste and water supply interruptions.
By employing a heterogeneous data acquisition module, an edge fusion module, a leakage location module, and a scheduling optimization module, combined with a sensor array and an LSTM neural network, the system enables real-time monitoring and optimized management of the water supply system, quickly locates leakage points, optimizes water load prediction, and reduces energy consumption.
It enables precise location of leakage points, reduces water waste, improves water use efficiency, lowers energy consumption, and ensures the stable operation and sustainable development of the water supply system.
Smart Images
Figure CN121920773A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of water supply data management technology, specifically to a digital asset lifecycle management and smart water-saving system for water supply enterprises. Background Technology
[0002] With the continuous advancement of urbanization, the urban population is increasing rapidly, and urban construction and economic activities are becoming increasingly frequent. This has led to a surge in the demand for water resources. To ensure the normal operation of residents' lives, industrial production, and various public services, water supply systems need large-scale expansion and upgrades. Therefore, water supply companies need to introduce digital asset lifecycle management technology and intelligent water-saving systems, efficiently develop new water sources, and add water intake facilities to meet the ever-increasing water demand, thereby ensuring the stable operation and sustainable development of the water supply system.
[0003] However, most current digital asset lifecycle management and smart water and energy-saving systems rely heavily on manual leak detection or regional pressure monitoring for leakage control, which cannot quickly locate the leak point, resulting in water waste and water supply interruption. At the same time, in terms of energy-saving scheduling, they are often based on fixed rules or manual experience, which cannot adapt to dynamically changing water demand, resulting in high energy consumption. Summary of the Invention
[0004] To address these issues, this invention provides a digital asset lifecycle management and intelligent water-saving system for water supply enterprises.
[0005] To achieve the above objectives, the present invention provides the following technical solution:
[0006] The digital asset lifecycle management and smart water use and energy saving system for water supply enterprises includes a heterogeneous data acquisition module, an edge fusion module, a leakage location module, a scheduling optimization module, and a control module.
[0007] The heterogeneous data acquisition module is able to collect raw water supply data based on the sensor array deployed on the asset nodes;
[0008] The edge fusion module can sequentially perform data cleaning, feature extraction and compression processing on the raw water supply data, and extract features from the cleaned sound wave, vibration, pressure and flow data to obtain a set of asset operation parameters.
[0009] The leakage location module continuously compares the asset operation parameter set with preset thresholds to filter the list of leaking pipe sections in the pipeline network area where the abnormal node is located, and performs data acquisition and calculation based on high-frequency acoustic sensors at both ends of the leaking pipe section to determine the location of the leak. ;
[0010] The scheduling optimization module can receive the asset operation parameter set in real time, integrate external data sources as input to the water load prediction model, output the water load prediction curve, and generate the corresponding water-saving solution by combining the objective function.
[0011] The control module can establish digital periodic archives based on water-saving and energy-saving schemes, water load prediction curves, leak locations, and users' historical water usage data.
[0012] Furthermore, the raw water supply data includes acoustic signals, vibration signals, pressure data, and flow rate data.
[0013] Furthermore, the specific content of the edge blending module is as follows:
[0014] 1) Input raw water supply data through the Internet of Things (IoT) communication protocol and perform data cleaning on the raw water supply data;
[0015] 2) The cleaned data is subjected to feature extraction and compression according to type to obtain the asset operation parameter set.
[0016] Furthermore, the feature extraction includes calculating the energy percentage P of a specific leakage characteristic frequency band and the frequency band energy of each sub-band of the acoustic signal. This yields the first eigenvector set of the energy distribution of the signal in the time-frequency domain;
[0017] For vibration signals, spectral analysis is also performed. After obtaining the spectrum of the vibration signal through spectral analysis, the frequency point corresponding to the characteristic frequency of the rotating part of the equipment is found in the spectrum diagram. The amplitude and phase information at the frequency point are read to form a second feature vector set characterizing the mechanical health status of the equipment.
[0018] For stress data, record the time interval. Instantaneous internal pressure value and calculation of pressure fluctuation rate and pressure descent gradient This forms the third feature vector set;
[0019] For the flow data, instantaneous flow, cumulative flow and flow change rate features are extracted, and the start and end times of the water use pattern are identified to obtain the fourth feature vector set.
[0020] Furthermore, the pressure volatility The calculation formula is as follows:
[0021]
[0022] in, Time interval Internally recorded instantaneous pressure values, This represents the maximum value among the instantaneous pressure values in this set. To represent the minimum value among these instantaneous pressure values, The number of instantaneous pressure values recorded. For the first Each pressure value.
[0023] Furthermore, the leakage location module includes an anomaly monitoring submodule, a screening submodule, and a location submodule;
[0024] The anomaly monitoring submodule can detect whether each feature vector parameter in the asset operation parameter set continuously exceeds a preset threshold. If a feature vector parameter exceeds the preset threshold, the screening submodule is activated and marked as an abnormal node.
[0025] The screening submodule is built based on a simulation calculation model and can convert the abnormal node information marked by the abnormal monitoring submodule into a specific list of suspected leaking pipe sections.
[0026] The location submodule can activate the high-frequency acoustic sensors at both ends of each suspected pipe segment in the list of suspected leak areas to enter a high sampling rate working mode, synchronously collect pipe vibration signals, and calculate the leak location based on the one-dimensional coordinates projected from the two high-frequency acoustic sensors. .
[0027] Furthermore, the working principle of the screening submodule is as follows:
[0028] 1) Retrieve the complete pipeline topology of the area where the abnormal node is located from the asset database, as well as the physical parameters and asset information of the relevant pipe sections, and build a simulation calculation model;
[0029] 2) Simulated pressure values based on leakage calculation nodes ;
[0030] 3) Simulate pressure values Measured pressure values at abnormal nodes Matching degree calculate;
[0031] 4) Output matching degree The first few hypothetical leak locations with the smallest values are used to identify the corresponding pipe sections as a preliminary list of suspected leak areas.
[0032] Furthermore, the degree of matching The calculation formula is as follows:
[0033]
[0034] in, For the number of nodes, and All are preset weighting coefficients. For nodes The standard deviation of the pressure measurement values.
[0035] Furthermore, the specific content of the positioning submodule is as follows:
[0036] 1) Based on the list of suspicious pipe sections, the high-frequency acoustic sensors deployed at both ends of the suspicious pipe sections are instructed to enter a high sampling rate synchronous acquisition mode to capture the vibration signal of the suspicious pipes;
[0037] 2) Analyze the time-domain signals acquired by the two sensors and calculate the time shift corresponding to the maximum cross-correlation coefficient. and time shift The maximum value is determined to be the time difference between the arrival of the leakage signal at the two high-frequency acoustic sensors. Time shift The calculation formula is as follows:
[0038]
[0039] in, This is the time-domain signal of the vibration signal acquired by the first sensor. This is the time-domain signal of the vibration signal acquired by the second sensor. For signal Relative to signal Time delay;
[0040] 3) Calculate the sound wave propagation velocity within the suspected pipe section. ;
[0041] 4) Project the coordinates of the two high-frequency acoustic sensors onto the pipe axis, convert them into one-dimensional linear coordinates along the pipe direction, and then calculate the location of the leak. Leakage location The calculation formula is as follows:
[0042]
[0043] in, and The coordinates are the one-dimensional coordinates of the two high-frequency acoustic sensors projected onto the pipe axis.
[0044] This invention has the following advantages: It improves data quality and transmission efficiency by processing and extracting features from raw data such as sound waves and pressure using an edge fusion module; the leak location module quickly identifies suspicious pipe sections using a simulation calculation model, and then calculates the leak location using the position coordinates of two high-frequency sound wave sensors, thus facilitating accurate leak location. This shortens the time for leak detection and location, reduces water loss, and effectively prevents water supply interruptions caused by leak expansion.
[0045] Meanwhile, the prediction of water load is realized based on LSTM neural network, which makes it convenient to sum up and minimize the power consumption of all pump groups, so as to achieve the goal of minimizing the total energy consumption, thereby achieving the purpose of water conservation and improving the overall water use efficiency. It also realizes the unified management of asset health, water supply security and energy conservation.
[0046] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. Attached Figure Description
[0047] To more intuitively illustrate the prior art and this application, exemplary drawings are provided below. It should be understood that the specific shapes and structures shown in the drawings should not generally be regarded as limiting conditions for implementing this application; for example, based on the technical concept disclosed in this application and the exemplary drawings, those skilled in the art are able to easily make conventional adjustments or further optimizations to the addition / reduction / classification, specific shapes, positional relationships, connection methods, size ratios, etc. of certain units (components).
[0048] Figure 1 This is a block diagram of the digital asset lifecycle management and smart water-saving system for water supply enterprises according to the present invention. Detailed Implementation
[0049] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. It should be understood that these embodiments are merely for further explanation of the present invention and should not be construed as limiting the scope of protection of the present invention. Technical engineers in the field can make some non-essential improvements and adjustments to the present invention based on the above-described content. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0050] Please see Figure 1 The digital asset lifecycle management and smart water-saving system for water supply enterprises includes a heterogeneous data acquisition module, an edge fusion module, a leakage location module, a scheduling optimization module, and a control module.
[0051] The heterogeneous data acquisition module can collect raw water supply data based on sensor arrays deployed on asset nodes. This raw data includes acoustic signals, vibration signals, pressure data, and flow data, facilitating the provision of raw data for subsequent leakage identification. Asset nodes include water sources, water plants, pipeline networks, and end users.
[0052] The sensor array includes a high-frequency acoustic sensor, a triaxial vibration sensor, a pressure transmitter, and an electromagnetic flow meter;
[0053] High-frequency acoustic sensors are deployed at key nodes in water supply and distribution networks to actively collect wide-band (typically >40kHz) acoustic signals generated by pipeline leaks, friction, or cavitation, providing a core data source for subsequent acoustic leakage identification.
[0054] Triaxial vibration sensors are installed on rotating machinery (such as water pumps, motor bearing housings, and machine casings) and large valve bodies to collect vibration acceleration signals of the equipment in order to monitor the mechanical balance of the equipment, component wear, and the health of the installation foundation.
[0055] Pressure transmitters are deployed at water source outlets, water plant inlet and outlet mains, branch network boundaries, network ends, and secondary water supply pump stations to continuously monitor static and dynamic water pressure data.
[0056] Electromagnetic flow meter: Installed in the water plant's outlet pipeline, regional water inlet, and end-user water inlet, it is used to measure instantaneous flow and cumulative flow, and provide flow data.
[0057] The edge fusion module can sequentially perform data cleaning, feature extraction, and compression on the raw water supply data. It can also extract features from the cleaned sound wave, vibration, pressure, and flow data to obtain a set of asset operation parameters, providing a unified data foundation for subsequent leakage location, scheduling optimization analysis, and other related tasks.
[0058] The specific contents of the edge blending module are as follows:
[0059] 1) Input raw water supply data through the Internet of Things (IoT) communication protocol and perform data cleaning on the raw water supply data;
[0060] By combining a pre-defined rule base built with expert knowledge and an adaptive statistical algorithm (such as the 3σ criterion) with a sliding time window, noise points and outliers caused by environmental electromagnetic interference, sensor momentary failure, communication packet loss, or abnormal operating conditions are dynamically identified and eliminated.
[0061] The aforementioned adaptive statistical algorithm refers to an intelligent algorithm that can dynamically adjust the identification threshold based on the real-time data distribution characteristics. Its purpose is to identify and filter noise and outliers to cope with different sensor data types and changing patterns, and to ensure that the cleaning rules can be adaptively optimized according to the system's operating status.
[0062] 2) After cleaning, the data is categorized by type for feature extraction and compression to obtain a set of asset operation parameters: For acoustic signals, Fast Fourier Transform (FFT) is applied to transform them from the time domain to the frequency domain, and the energy proportion P of a specific leakage characteristic frequency band (e.g., 200Hz-800Hz) is calculated; simultaneously, wavelet packet decomposition is used to extract the energy distribution characteristics of the signal in the time-frequency domain to distinguish between leakage sound and environmental noise. The formula for calculating the energy proportion P is as follows:
[0063] in, This refers to the frequency index corresponding to the starting frequency of a specific frequency band. This is the frequency index corresponding to the end frequency of a specific frequency band. N is the frequency domain sequence obtained after the signal undergoes a Fast Fourier Transform (FFT), where N is the number of sampling points.
[0064] Wavelet packet decomposition can simultaneously preserve the local features of a signal in both time and frequency. For the cleaned acoustic signal, wavelet packet decomposition is used to decompose it into multiple sub-bands, each containing energy information of the signal within a specific time and frequency range. The frequency band energy of each sub-band is then calculated. This yields the first feature vector set of the signal's energy distribution in the time-frequency domain. Since the energy distribution patterns of leak sound and environmental noise in the time-frequency domain are typically different, analyzing these energy distribution characteristics can further distinguish leak sound from environmental noise, improving the accuracy of leak detection. (Frequency band energy) The calculation formula is as follows:
[0065]
[0066] in, Here, L represents the wavelet packet coefficients, and L is the length of the wavelet packet coefficients in this sub-band. For the index of the frequency domain sequence, =0,1, , L-1.
[0067] For example, the sound of water leakage may have concentrated energy in certain specific sub-bands and time periods, while the energy distribution of environmental noise is relatively dispersed.
[0068] For vibration signals, spectral analysis is also performed. After obtaining the spectrum of the vibration signal through spectral analysis, the frequency point corresponding to the characteristic frequency (rotation frequency, bearing failure frequency, etc.) of the rotating parts of the equipment is found in the spectrum diagram. The amplitude and phase information at the frequency point are read to form a second feature vector set that characterizes the mechanical health status of the equipment.
[0069] For stress data, record the time interval. Instantaneous internal pressure value and calculation of pressure fluctuation rate and pressure descent gradient This facilitates subsequent location of leakage, forming a third feature vector set. Pressure volatility The calculation formula is as follows:
[0070]
[0071] in, Time interval Internally recorded instantaneous pressure values, This represents the maximum value among the instantaneous pressure values in this set. This represents the minimum value among these instantaneous pressure values. The number of instantaneous pressure values recorded. For the first Each pressure value.
[0072] Pressure descent gradient The calculation formula is as follows:
[0073]
[0074] in, The pressure coefficient, This is the initial pressure value.
[0075] For flow data, instantaneous flow, cumulative flow, and flow change rate features are extracted, and the start and end times of water usage patterns are identified to obtain the fourth feature vector set.
[0076] The first feature vector set, the second feature vector set, the third feature vector set, and the fourth feature vector set are combined to form the asset operation parameter set.
[0077] The leak location module continuously compares the asset operation parameter set with preset thresholds to filter the list of leaking pipe sections in the pipeline network area where the abnormal node is located, and uses high-frequency acoustic sensors at both ends of the leaking pipe section to collect and calculate data to determine the location of the leak. This facilitates shorter response time for leak location and improves location accuracy.
[0078] The leakage location module includes an anomaly monitoring submodule, a screening submodule, and a location submodule. The anomaly monitoring submodule can detect whether each feature vector parameter in the asset operation parameter set continuously exceeds a preset threshold. If a feature vector parameter exceeds the preset threshold, the screening submodule is activated and the point is marked as an anomaly node.
[0079] The screening submodule is built on a simulation model and can convert the abnormal node information marked by the anomaly monitoring submodule into a specific list of suspected leaking pipe sections. Its working principle is as follows:
[0080] 1) Retrieve the complete pipeline topology of the area where the abnormal node is located from the asset database, along with the physical parameters and asset information of the relevant pipe segments, to construct a simulation model of the area. In the model, each node and pipe segment is laid out according to its actual connection relationship, and the simulation program and corresponding technical parameters are preset to prepare for subsequent calculations. The simulation program can sequentially assume the existence of a virtual leak point for each pipeline node in the model. An assumed leakage rate is assigned to each virtual leak point. The value of the leakage rate can be set based on actual conditions and experience.
[0081] The asset database includes the pipeline topology and technical parameters, procurement standards, and supplier information for assets such as pumping stations, water meters, and valves.
[0082] 2) Simulated pressure values based on leakage calculation nodes ;
[0083]
[0084] in, The density of water, It is the acceleration due to gravity. This represents the leakage volume at the virtual leak point. This refers to the head loss of the pipe section at the network node.
[0085]
[0086] in, For the length of the pipe section, For the flow rate of the pipe section, For pipeline coefficients, The diameter is the pipe diameter.
[0087] 3) Simulate pressure values Measured pressure values at abnormal nodes Matching degree The calculations facilitate finding the most accurate assumptions about the location and volume of leaks in actual pipeline operations. The smaller the value, the higher the degree of matching between the simulation results and the measured data, and the more reasonable the corresponding hypothetical scenario. Degree of Matching The calculation formula is as follows:
[0088]
[0089] in, For the number of nodes, and All are preset weighting coefficients. For nodes The standard deviation of the pressure measurement values.
[0090] 4) Output matching degree The first few hypothetical leak locations with the smallest values are used to identify the corresponding pipe segments as a preliminary list of suspected leak areas. This list allows for narrowing down the potential leak area from the entire monitoring area to a limited number of specific pipe segments, providing a clear and highly suspicious target range for subsequent location analysis.
[0091] The location submodule can activate the high-frequency acoustic sensors at both ends of each suspected pipe segment in the list of suspected leak areas to enter high sampling rate mode, synchronously collect pipe vibration signals, and calculate the leak location based on the one-dimensional coordinates projected from the two high-frequency acoustic sensors. The specific details are as follows:
[0092] 1) Based on the list of suspicious pipe sections, the high-frequency acoustic sensors deployed at both ends of the suspicious pipe sections are instructed to enter a high sampling rate synchronous acquisition mode to capture the vibration signal of the suspicious pipes;
[0093] 2) Analyze the time-domain signals acquired by the two sensors and calculate the time shift corresponding to the maximum cross-correlation coefficient. The maximum value was determined as the time difference between the arrival of the leakage signal at the two high-frequency acoustic sensors. Time shift The calculation formula is as follows:
[0094]
[0095] in, This is the time-domain signal of the vibration signal acquired by the first sensor. This is the time-domain signal of the vibration signal acquired by the second sensor. For signal Relative to signal The time delay.
[0096] 3) Calculate the sound wave propagation velocity within the suspected pipe section. Speed of sound propagation The calculation formula is as follows:
[0097]
[0098] in, The speed of sound wave propagation in free water. Let be the bulk modulus of water. For pipe diameter, The elastic modulus of the pipe. This refers to the pipe wall thickness.
[0099] 4) Project the coordinates of the two high-frequency acoustic sensors onto the pipe axis, converting them into one-dimensional linear coordinates along the pipe direction. Let the pipe's starting point be the origin, and the pipe's extension direction be... If the axis is positive, then the one-dimensional coordinates of the two sensors after projection are respectively... and Then calculate the location of the leak. Leakage location The calculation formula is as follows:
[0100]
[0101] in, and The coordinates are the one-dimensional coordinates of the two high-frequency acoustic sensors projected onto the pipe axis.
[0102] Specifically, by combining the start and end coordinates of the pipe segment in the asset database, a spatial geometric projection algorithm is used to map the three-dimensional coordinates of the sensors onto a straight axis defined by the start and end points of the pipe. A one-dimensional linear coordinate system is established with the pipe start point as the origin and the pipe extension direction as the positive direction. The position of each sensor on this axis is obtained and converted into scalar coordinate values along the pipe direction. and This is for subsequent time difference-based and the speed of sound wave propagation This provides a data foundation for leak detection.
[0103] The scheduling optimization module can receive the set of asset operation parameters processed by the edge fusion module in real time, and integrate external data sources as input to the water load prediction model, and output water load prediction curves. The external data sources include weather forecasts and calendar information (weekdays, holidays, and large events).
[0104] The water load prediction model is built on an LSTM neural network. The LSTM neural network can process historical sequence data cyclically through its internal gating mechanism (input gate, forget gate, output gate) to learn the complex nonlinear variation patterns of water load in daily, weekly, and externally influenced periods. Historical sequence data includes historical water usage sequences, i.e., the fourth feature vector set corresponding to the time in the past few weeks; real-time feature sequences, i.e., the latest feature vectors output by the edge fusion module; and external factor sequences, i.e., the numerical external data sources corresponding to the time in the past few weeks.
[0105] The hidden state of the last time step of the water load forecasting model is fed into the fully connected layer and decoded into water consumption forecasts for each zone at one-hour intervals for the next 24 hours. The output is a 24×N matrix, where N is the number of zones, thus generating hourly forecast curves for each zone.
[0106] Based on the prediction results, the scheduling optimization module can sum up and minimize the power consumption of all pump groups to achieve the goal of minimizing total energy consumption, which facilitates dynamic optimization of water plant capacity strategy.
[0107] First, a multi-objective optimization problem is constructed, with the constraint of ensuring the minimum service pressure of the pipeline network and the optimization objective of minimizing total operating energy consumption. In each scheduling cycle (e.g., 15 minutes), the objective function is solved, and the corresponding water-saving scheme is given. The water-saving scheme includes the recommended water output of each water plant in the future, the start / stop status and recommended speed of each pump in each pumping station, and the opening setting value of key pressure regulating valves.
[0108] objective function Specifically as follows:
[0109]
[0110] in, For the first Power consumption of each pump unit. The calculation formula is as follows:
[0111]
[0112] in, The density of water, It is the acceleration due to gravity. For the first The flow rate of each pump unit For the first The head of each pump unit For the first The efficiency of each pump unit.
[0113] For example: during periods of low water usage, reduce the setpoint of the pipeline pressure to reduce the reactive power loss of the pumping station; before peak water usage, schedule and reserve water in advance, and optimize the start-up and shutdown combinations of pump sets to ensure water supply stability while reducing overall power consumption; combine real-time pressure and flow data to adjust pump frequency and valve opening to achieve hydraulic balance and energy-saving operation.
[0114] The management and control module can establish a digital lifecycle archive based on water-saving and energy-saving solutions, water load prediction, abnormal monitoring information reported by the leakage location module, and historical water usage data of users. It records in detail the asset's entire lifecycle information from the initial deployment stage to various status information during use, and finally to its scrapping, thereby gaining a comprehensive understanding of the asset's status.
[0115] In addition, the control module will promptly address any leaks detected based on the leak location results to prevent water waste. Simultaneously, it will monitor the entire water supply system in real time according to water conservation plans.
[0116] This invention improves data quality and transmission efficiency by processing and extracting features from raw data such as sound waves and pressure using an edge fusion module. The leak location module quickly identifies suspicious pipe sections using a simulation model and then calculates the leak location using the position coordinates of two high-frequency acoustic sensors, facilitating precise leak detection. This shortens the time required for leak detection and location, reduces water loss, and effectively prevents water supply interruptions caused by leak expansion.
[0117] Meanwhile, the prediction of water load is realized based on LSTM neural network, which makes it convenient to sum up and minimize the power consumption of all pump groups, so as to achieve the goal of minimizing the total energy consumption, thereby achieving the purpose of water conservation and improving the overall water use efficiency. It also realizes the unified management of asset health, water supply security and energy conservation.
[0118] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A digital asset lifecycle management and intelligent water-saving system for water supply enterprises, characterized in that: It includes a heterogeneous data acquisition module, an edge fusion module, a leakage location module, a scheduling optimization module, and a management and control module; The heterogeneous data acquisition module is able to collect raw water supply data based on the sensor array deployed on the asset nodes; The edge fusion module can sequentially perform data cleaning, feature extraction and compression processing on the raw water supply data, and extract features from the cleaned sound wave, vibration, pressure and flow data to obtain a set of asset operation parameters. The leakage location module continuously compares the asset operation parameter set with preset thresholds to filter the list of leaking pipe sections in the pipeline network area where the abnormal node is located, and performs data acquisition and calculation based on high-frequency acoustic sensors at both ends of the leaking pipe section to determine the location of the leak. ; The scheduling optimization module can receive the asset operation parameter set in real time, integrate external data sources as input to the water load prediction model, output the water load prediction curve, and generate the corresponding water-saving solution by combining the objective function. The control module can establish digital periodic archives based on water-saving and energy-saving schemes, water load prediction curves, leak locations, and users' historical water usage data.
2. The digital asset lifecycle management and intelligent water-saving system for water supply enterprises according to claim 1, characterized in that, The raw water supply data includes acoustic signals, vibration signals, pressure data, and flow rate data.
3. The digital asset lifecycle management and intelligent water-saving system for water supply enterprises according to claim 1, characterized in that, The specific details of the edge blending module are as follows: 1) Input raw water supply data through the Internet of Things (IoT) communication protocol and perform data cleaning on the raw water supply data; 2) The cleaned data is subjected to feature extraction and compression according to type to obtain the asset operation parameter set.
4. The digital asset lifecycle management and intelligent water-saving system for water supply enterprises according to claim 3, characterized in that, The feature extraction includes calculating the energy percentage P of a specific leakage characteristic frequency band and the frequency band energy of each sub-band of the acoustic signal. This yields the first eigenvector set of the energy distribution of the signal in the time-frequency domain; For vibration signals, spectral analysis is also performed. After obtaining the spectrum of the vibration signal through spectral analysis, the frequency point corresponding to the characteristic frequency of the rotating part of the equipment is found in the spectrum diagram. The amplitude and phase information at the frequency point are read to form a second feature vector set characterizing the mechanical health status of the equipment. For stress data, record the time interval. Instantaneous internal pressure value and calculation of pressure fluctuation rate and pressure descent gradient This forms the third feature vector set; For the flow data, instantaneous flow, cumulative flow and flow change rate features are extracted, and the start and end times of the water use pattern are identified to obtain the fourth feature vector set.
5. The digital asset lifecycle management and intelligent water-saving system for water supply enterprises according to claim 4, characterized in that, The pressure fluctuation rate The calculation formula is as follows: ； in, Time interval Internally recorded instantaneous pressure values, This represents the maximum value among the instantaneous pressure values in this set. To represent the minimum value among these instantaneous pressure values, The number of instantaneous pressure values recorded. For the first Each pressure value.
6. The digital asset lifecycle management and intelligent water-saving system for water supply enterprises according to claim 1, characterized in that, The leakage location module includes an anomaly monitoring submodule, a screening submodule, and a location submodule; The anomaly monitoring submodule can detect whether each feature vector parameter in the asset operation parameter set continuously exceeds a preset threshold. If a feature vector parameter exceeds the preset threshold, the screening submodule is activated and marked as an abnormal node. The screening submodule is built based on a simulation calculation model and can convert the abnormal node information marked by the abnormal monitoring submodule into a specific list of suspected leaking pipe sections. The location submodule can activate the high-frequency acoustic sensors at both ends of each suspected pipe segment in the list of suspected leak areas to enter a high sampling rate working mode, synchronously collect pipe vibration signals, and calculate the leak location based on the one-dimensional coordinates projected from the two high-frequency acoustic sensors. .
7. The digital asset lifecycle management and intelligent water-saving system for water supply enterprises according to claim 6, characterized in that, The screening submodule works as follows: 1) Retrieve the complete pipeline topology of the area where the abnormal node is located from the asset database, as well as the physical parameters and asset information of the relevant pipe sections, and build a simulation calculation model; 2) Simulated pressure values based on leakage calculation nodes ; 3) Simulate pressure values Measured pressure values at abnormal nodes Matching degree calculate; 4) Output matching degree The first few hypothetical leak locations with the smallest values are used to identify the corresponding pipe sections as a preliminary list of suspected leak areas.
8. The digital asset lifecycle management and intelligent water-saving system for water supply enterprises according to claim 7, characterized in that, The degree of matching The calculation formula is as follows: ； in, For the number of nodes, and All are preset weighting coefficients. For nodes The standard deviation of the pressure measurement values.
9. The digital asset lifecycle management and intelligent water-saving system for water supply enterprises according to claim 7, characterized in that, The specific contents of the positioning submodule are as follows: 1) Based on the list of suspicious pipe sections, the high-frequency acoustic sensors deployed at both ends of the suspicious pipe sections are instructed to enter a high sampling rate synchronous acquisition mode to capture the vibration signal of the suspicious pipes; 2) Analyze the time-domain signals acquired by the two sensors and calculate the time shift corresponding to the maximum cross-correlation coefficient. and time shift The maximum value is determined to be the time difference between the arrival of the leakage signal at the two high-frequency acoustic sensors. Time shift The calculation formula is as follows: ； in, This is the time-domain signal of the vibration signal acquired by the first sensor. This is the time-domain signal of the vibration signal acquired by the second sensor. For signal Relative to signal Time delay; 3) Calculate the sound wave propagation velocity within the suspected pipe section. ; 4) Project the coordinates of the two high-frequency acoustic sensors onto the pipe axis, convert them into one-dimensional linear coordinates along the pipe direction, and then calculate the location of the leak. Leakage location The calculation formula is as follows: ； in, and The coordinates are the one-dimensional coordinates of the two high-frequency acoustic sensors projected onto the pipe axis.