Water supply energy-saving safe scheduling and purifying intelligent control system

Abstract

The invention discloses a water supply energy-saving safe scheduling and purification intelligent control system, and relates to the field of water supply energy-saving purification, and the system comprises a sensing module, a calculation module, an acquisition terminal, a scheduling module and a purification module.A final water demand prediction value is obtained by constructing a time-sequence-dependent water use feature set and performing prediction and error correction on an LSTM model; performing water supply energy-saving safety scheduling analysis, generating a water supply pipeline equipment control parameter combination, performing pipeline state abnormity and pipeline fluid pressure analysis, calculating a pipeline state abnormity value, performing pressure loss state change analysis when the pipeline fluid flows to obtain a pipeline fluid pressure loss value, and performing image construction to obtain a water pipeline pressure difference value; the electronic valve works according to the adjusted valve opening value through the scheduling module, and the calculation module receives the water quality data obtained by the sensing module and performs corresponding valve control and purification treatment, so that responsibility complaints of users are greatly reduced, and resource waste is reduced.

Description

Technical Field

[0001] The present invention relates to the field of water supply energy-saving purification, and in particular to an intelligent control system for water supply energy-saving safety scheduling and purification. Background Art

[0002] With the accelerating pace of urbanization, traditional water supply systems face challenges such as high energy consumption, low scheduling efficiency, and water quality safety. Existing technologies rely on empirical scheduling and manual inspections, making it difficult to respond to pipeline aging and sudden pollution incidents in real time. Furthermore, they lack precise dynamic energy-saving control and water usage prediction. Consequently, intelligent control systems for water supply energy-saving, safety scheduling, and purification have emerged.

[0003] When the existing water supply energy-saving safety scheduling and purification intelligent control system is in operation, it is unable to predict water demand from multiple dimensions and angles, nor can it judge the water supply pipeline equipment, let alone predict the water pollution path to implement timely interception and discharge, and cannot perform purification control in time. As a result, it often receives complaints from users, coupled with ineffective purification energy consumption, resulting in serious waste of resources.

[0004] In order to solve the above-mentioned defects, a technical solution is now provided. Summary of the Invention

[0005] In order to solve the technical problems raised by the above background technology, the present invention is proposed. The embodiments of the present invention provide an intelligent control system for water supply energy saving, safety scheduling and purification.

[0006] The objectives of the present invention can be achieved through the following technical solutions: an intelligent control system for energy-saving, safe scheduling and purification of water supply, comprising a sensing module, a computing module, a collection terminal, a scheduling module, and a purification module. The scheduling module includes an electronic valve, which divides the entire water supply area into several independent sub-areas based on the city's industrial water use area, residential distribution density, and functional zoning factors. An electronic valve is set at the boundary of each sub-area to independently control the water supply of each sub-area. The electronic valve receives the calculated value from the computing module and operates according to the adjusted valve opening value. The purification module includes a water distribution pipeline and purification facilities, receives the signal from the computing module, and performs corresponding purification operations.

[0007] The collection terminal collects user water use related information and water supply pipeline equipment control information and transmits it to the calculation module;

[0008] The sensing module includes water quality monitoring sensors, which are arranged at key nodes of the pipeline network, such as branch pipe intersections and user ends, and transmit the obtained water quality data to the calculation module; the calculation module is used to construct a time-dependent water use feature set, analyze water demand forecast values, and construct a combination of control parameters for water supply pipeline equipment. Based on the combination of control parameters for water supply pipeline equipment, a fractional-order degraded diffusion process control model for water supply energy-saving scheduling is created, and pressure loss state changes during pipeline fluid flow are analyzed. Water quality anomalies are analyzed and interception and purification signals are generated to deduce pollution paths.

[0009] The following steps are also included:

[0010] Step S01: Solve the final water demand forecast value. The calculation module obtains the water use-related information of the end users in the sub-area, constructs a time-dependent water use feature set, predicts and corrects the error of the long short-term memory network (LSTM) model, and finally adjusts the water use ladder to obtain the final water demand forecast value.

[0011] Step S02: Water supply pipeline equipment control analysis: The calculation module obtains the pipeline physical parameter set, equipment degradation parameter set, and environmental impact parameter set of the water supply pipeline equipment, and performs water supply energy-saving and safety scheduling analysis on the pipeline physical parameter set, equipment degradation parameter set, and environmental impact parameter set to generate a water supply pipeline equipment control parameter combination;

[0012] Step S03: Analyze pipeline state anomalies and pipeline fluid pressure. The calculation module creates a fractional-order degenerate diffusion process control model for water supply energy-saving scheduling based on the control parameter combination of the water supply pipeline equipment. The model analyzes and calculates pipeline state anomalies and analyzes the pressure loss state changes during pipeline fluid flow to obtain the pipeline fluid pressure loss value.

[0013] Step S04: Valve opening scheduling: the calculation module obtains the water pipe pressure difference value of each area at each time based on the image construction, and uses the scheduling module to adjust the electronic valve to work according to the valve opening value;

[0014] Step S05: shut-off valve control and purification processing: the calculation module receives the water quality data obtained by the sensor module, analyzes various states and sends corresponding signals, and performs corresponding valve control and purification processing.

[0015] Furthermore, the steps for analyzing the final water demand forecast value are as follows:

[0016] Step S103: The calculation module inputs the time-dependent water use feature set into the long short-term memory network LSTM model. The LSTM model architecture is that the input layer receives the time-dependent water use feature set, the hidden layer uses the cell state as a long-term memory carrier, and uses the forget gate to selectively retain key information such as seasonal water use patterns. The input gate integrates current temperature, scene water use and other features to update the memory, and the output gate extracts the patterns related to the future time period, and finally generates the initial value of water demand prediction. The output layer outputs the initial value of water demand prediction for the future time period. The initial value of water demand prediction output by the LSTM model is optimized, and the error correction function expression is E(t)=λ×(Y actu (t)-Y pre (t))+α×E(t-1),Y actu (t)-Y pre (t) is the deviation between the actual water consumption at time t and the model prediction value, which is used to reflect the instantaneous error of the current prediction. λ is the weight coefficient of the current error. E(t) and E(t-1) are the correction errors at time t and t-1 respectively. ɑ is the weight coefficient of the historical error. The optimal values ​​of the weight coefficients λ and ɑ are substituted into the error correction function to make the prediction value closer to the actual water consumption. The iterative calculation is initialized with E(0) = 0. Starting from t = 1, the error correction function is substituted into each moment in turn to calculate the current correction error E(t). The predicted value is corrected based on E(t) to obtain Y corre (t) = Y pre (t)+E(t), the iteration continues until the correction error E(t) is less than the minimum threshold ∈, then the termination condition is output and the corrected water demand forecast value Y is corre (t);

[0017] Furthermore, the optimal value solution steps for weight coefficients λ and ɑ are as follows:

[0018] The calculation module obtains Y of the historical period actu (t), Y pre (t) data, objective function Take partial derivatives of λ and ɑ and set the derivatives to zero, solve the equations to obtain the optimal parameters of weight coefficients λ and ɑ;

[0019] Step S104: Adjust the water demand forecast value, obtain the user water demand level at each moment, the level is the first level, the second level, the third level, and the corresponding adjustment coefficient β, and adjust the water demand forecast value Y after correction. corre (t) is multiplied by the adjustment coefficient β to obtain the final water demand forecast value Y final (t).

[0020] Furthermore, the steps for analyzing the time-dependent water usage feature set are as follows:

[0021] Step S101: Establishing a water use-related parameter set. The calculation module obtains a user water use-related parameter set from the collection terminal in the sub-region. The parameter set includes: time-series water use archives, household configuration parameters, environmental field variables, and water intelligent equipment data. The time-series water use archives include time period water consumption, monthly total water consumption, and multi-dimensional scenario water use maps. The household configuration parameters include population size, age structure, water consumption for pet breeding, and water consumption for planting green plants. The environmental field variables include regional temperature, precipitation probability, sunshine duration, and air humidity. The water intelligent equipment data includes data from intelligent water devices.

[0022] Step S102: The calculation module constructs a time-dependent water use feature set. Based on the time-series water use archive, the time period water use and monthly total are decomposed into timestamp features, including hour code, work and rest period type, and monthly seasonal basic time dimensions. The average water use over the past 24 hours and the extreme water use over the past week are calculated using a sliding window to capture intraday fluctuations and cyclical patterns. The water use time distribution, water use proportion, and water use correlation characteristics between scenarios are extracted for each scenario in the multidimensional scenario water use map. The scenario features are integrated into the basic time series. The changes in population size in household configuration parameters and the seasonal changes in water use for pet breeding and green plant cultivation are time-series processed and embedded into the time series framework of the integrated scenario. The regional temperature, precipitation probability, sunshine duration, and air humidity in the environmental field variables are aligned with the corresponding water use data timestamps. The change features at different timestamps are extracted and superimposed on the time series feature set. Abnormal water use event markers are extracted from the water intelligent equipment data. The abnormal features are marked at the time series nodes corresponding to the scenario and superimposed on the time series feature set to form a complete feature set, and finally a time-series-dependent water use feature set is constructed.

[0023] Furthermore, the steps for analyzing the control parameter combination of the water supply pipeline equipment are as follows:

[0024] (1) The calculation module obtains the set of pipeline physical parameters of the collection terminal in each sub-area. The set of pipeline physical parameters includes: geometric characteristic parameters, material characteristic parameters and fluid migration parameters. The geometric characteristic parameters include equivalent diameter, flow channel complexity factor and pipeline length. The material characteristic parameters include interface energy attenuation coefficient and roughness coefficient. The fluid migration parameters include water hardness and flow velocity.

[0025] (2) Monitor the operation of the water pump flow channel to obtain a set of equipment degradation parameters, which include: water pump flow channel deformation coefficient and seal viscoelastic loss factor;

[0026] (3) Conducting pipeline equipment environmental monitoring on the water pump flow channel to obtain a set of environmental impact parameters, which include: corrosive medium activity and soil pH;

[0027] (4) Conduct water supply energy-saving and safety scheduling analysis based on the pipeline physical parameter set, equipment degradation parameter set, and environmental impact parameter set to generate a combination of water supply pipeline equipment control parameters.

[0028] Furthermore, the pipeline status abnormal value analysis steps are as follows:

[0029] The calculation module creates a fractional-order degenerate diffusion process control model for water supply energy-saving scheduling based on the control parameter combination of water supply pipeline equipment. The fractional-order degenerate diffusion process control model includes: Among them D α is the Caputo fractional derivative, S(t) is the abnormal value of pipeline status at time t, X i (t) is the current actual value of the influencing factor in the water supply pipeline equipment control parameter combination at time t, i is the serial number of the influencing factor, which is a positive integer with a maximum value of 11. is the diffusion coefficient, χ i is the weight of influencing factor i, is the memory kernel integral term, K×(t-γ) is the decay memory kernel, It means that the Laplace operation is performed on the pipeline state abnormal value S(t) to obtain the pipeline state abnormal value at time t;

[0030] The weight χ of the influencing factor i i The steps to solve are as follows:

[0031] To X i (t) The data is standardized to eliminate the dimension effect, and the information entropy E of each factor i is calculated by the calculation formula i , the weight χ is calculated according to the formula i .

[0032] Furthermore, the pipeline fluid pressure loss value analysis steps are as follows:

[0033] The Darcy-Weisbach equation is used to control the pressure loss state of the fractional-order degenerate diffusion process model during pipeline fluid flow, and the pipeline fluid pressure loss value ΔP(t) is obtained. The Darcy-Weisbach equation includes: , where f is the pipeline roughness, LG is the pipeline length, the physical length of the pipeline through which the fluid flows, DG is the pipeline diameter, the flow cross-sectional characteristics of the calculated fluid flow, ρ is the fluid density, v is the fluid flow velocity, and k S is the outlier amplification factor.

[0034] Furthermore, the steps of adjusting the valve opening value for work analysis are as follows:

[0035] The calculation module processes the final water demand forecast value, pipeline status abnormal value, and pipeline fluid pressure loss value of each sub-area through graphical construction to obtain the water pipeline pressure abnormal value of each area at each time;

[0036] Set the water pipe pressure difference value, each water pipe pressure difference value corresponds to a value range, match the water pipe pressure difference value corresponding to each sub-area at each time point with the value range, if the set water pipe pressure difference value falls within the value range, then the electronic valve operating opening value corresponding to the value range is marked as the adjusted valve opening value; the scheduling module receives the calculated value of the calculation module and operates the electronic valve according to the adjusted valve opening value.

[0037] Furthermore, the purification analysis steps are as follows:

[0038] Step S501: Water quality analysis and purification processing. The calculation module receives the water quality data obtained by the sensor module and sends signal one, signal two, and signal three accordingly. If signal one is included in the sent signal, a shut-off valve control signal is sent. If two or three signal ones are included in the sent signal, a water quality abnormality and serious pollution signal is sent and sent to the purification module. The purification module performs purification measure one. If one signal one or one signal two is included in the sent signal, a water quality abnormality and slight pollution signal is sent and sent to the purification module. The purification module performs purification measure two.

[0039] Furthermore, the valve control analysis steps are as follows:

[0040] Step S502: Dynamic hydrodynamic modeling: The calculation module performs valve control analysis after receiving the interception and discharge signal. The dynamic hydrodynamic modeling is based on the pipeline network drawings and 3D scanning data to establish a digital twin model of the pipeline network. Parameters such as the length, diameter, and roughness of each pipe segment are annotated. Combined with real-time flow data, the hydraulic state of the pipeline network is calculated using a pressure-driven model. The water flow rate is calculated using the aforementioned Darcy-Weisbach equation and Manning's formula.

[0041] Step S503: Pollutant transport prediction, establishing a one-dimensional convection-diffusion equation model;

[0042] Step S504: Pollution path deduction and valve control, based on the graph theory algorithm, integrating hydrodynamic modeling and pollutant transport prediction results, identifying the priority paths of pollution propagation, marking sensitive nodes that need to be protected, and remotely closing the electric valves on the pollution path through the system to cut off the spread of pollution. First, close the valve closest to the pollution source, and then cut off the pollution upstream step by step.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. The present invention obtains water consumption information collected from terminal users in a sub-area through a calculation module, constructs a time-dependent water consumption feature set, predicts and corrects the error of the long short-term memory network LSTM model, and finally obtains the final water demand forecast value based on the water use ladder adjustment. The calculation module obtains the pipeline physical parameter set, equipment degradation parameter set and environmental impact parameter set of the water supply pipeline equipment, and performs water supply energy-saving and safety scheduling analysis on the pipeline physical parameter set, equipment degradation parameter set and environmental impact parameter set, generates a water supply pipeline equipment control parameter combination, pipeline state abnormality and pipeline fluid pressure analysis, and the calculation module creates a fractional-order degradation diffusion process control model for water supply energy-saving scheduling based on the water supply pipeline equipment control parameter combination, analyzes and calculates pipeline state abnormality values, and performs pressure loss state change analysis when the pipeline fluid flows to obtain the pipeline fluid pressure loss value. It can predict water demand from multiple dimensions and angles, and can also make judgments on water supply pipeline equipment.

[0045] 2. The present invention obtains the water pipe pressure difference values ​​in each area at each time based on image construction through the calculation module, and adjusts the valve opening value through the scheduling module. The calculation module receives the water quality data obtained by the sensor module, analyzes various states and sends corresponding signals, and performs corresponding valve control and purification processing. It can predict the water pollution path and implement interception and discharge in time, and can perform purification control in time, greatly reducing users' responsible complaints, reducing purification energy consumption, and reducing resource waste. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] 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. The following drawings are not intentionally scaled to the actual size, and the focus is on illustrating the main purpose of the present invention.

[0047] Figure 1 is a system block diagram of the present invention;

[0048] Figure 2 is a flow chart of the method of the present invention;

[0049] Figure 3 This is a flow chart of the shut-off valve control and purification process of the present invention. DETAILED DESCRIPTION

[0050] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts also fall within the scope of protection of the present invention.

[0051] like Figure 1As shown, a water supply energy-saving, safe scheduling and purification intelligent control system includes a sensing module, a computing module, a collection terminal, a scheduling module, and a purification module.

[0052] The dispatching module includes electronic valves. Based on the city's industrial water use areas, residential density, and functional zoning factors, the entire water supply area is divided into several independent sub-areas. Electronic valves are set at the boundaries of each sub-area to independently control the water supply of each sub-area. The electronic valves receive the calculated values ​​from the calculation module and operate according to the valve opening values. The purification module includes water distribution pipelines and purification facilities. It receives signals from the calculation module and performs corresponding purification operations.

[0053] The collection terminal collects user water use related information and water supply pipeline equipment control information and transmits it to the calculation module;

[0054] The sensing module includes water quality monitoring sensors, which are deployed at key nodes in the pipeline network, such as branch pipe intersections and user terminals, and transmit the acquired water quality data to the computing module. The computing module is used to construct a time-dependent water use feature set, analyze water demand forecasts, and build a combination of control parameters for water supply pipeline equipment. Based on this combination of control parameters, a fractional-order degenerate diffusion process control model for water supply energy-saving scheduling is created. The module also analyzes changes in pressure loss during pipeline fluid flow, analyzes water quality anomalies, generates interception and purification signals, and deduces pollution paths.

[0055] like Figure 2 As shown, the following steps are also included:

[0056] Step S01: Solve the final water demand forecast value. The calculation module obtains the water use-related information of the end users in the sub-area, constructs a time-dependent water use feature set, predicts and corrects the error of the long short-term memory network (LSTM) model, and finally adjusts the water use ladder to obtain the final water demand forecast value.

[0057] Step S02: Water supply pipeline equipment control analysis: The calculation module obtains the pipeline physical parameter set, equipment degradation parameter set, and environmental impact parameter set of the water supply pipeline equipment, and performs water supply energy-saving and safety scheduling analysis on the pipeline physical parameter set, equipment degradation parameter set, and environmental impact parameter set to generate a water supply pipeline equipment control parameter combination;

[0058] Step S03: Analyze pipeline state anomalies and pipeline fluid pressure. The calculation module creates a fractional-order degenerate diffusion process control model for water supply energy-saving scheduling based on the control parameter combination of the water supply pipeline equipment. The model analyzes and calculates pipeline state anomalies and analyzes the pressure loss state changes during pipeline fluid flow to obtain the pipeline fluid pressure loss value.

[0059] Step S04: Valve opening scheduling: the calculation module obtains the water pipe pressure difference value of each area at each time based on the image construction, and uses the scheduling module to adjust the electronic valve to work according to the valve opening value;

[0060] Step S05: shut-off valve control and purification processing: the calculation module receives the water quality data obtained by the sensor module, analyzes various states and sends corresponding signals, and performs corresponding valve control and purification processing.

[0061] In a specific embodiment, the process of executing step S01 may specifically include the following steps:

[0062] Step S101: Establishing a water use-related parameter set. The calculation module obtains a user water use-related parameter set from the collection terminal in the sub-region. The parameter set includes: time-series water use archives, household configuration parameters, environmental field variables, and water intelligent equipment data. The time-series water use archives include time period water consumption, monthly total water consumption, and multi-dimensional scenario water use maps. The household configuration parameters include population size, age structure, water consumption for pet breeding, and water consumption for planting green plants. The environmental field variables include regional temperature, precipitation probability, sunshine duration, and air humidity. The water intelligent equipment data includes data from intelligent water devices.

[0063] Specifically, the multi-dimensional water usage map shows the proportion and time distribution of water usage in bathroom flushing, household cleaning, kitchen cooking, and laundry, accurately characterizing user water usage behavior. Smart water device data includes the frequency of use of smart toilets' water-saving mode and the automated water usage records of smart irrigation systems.

[0064] Step S102: The calculation module constructs a time-dependent water use feature set. Based on the time-series water use archive, the time period water use and the monthly total are decomposed into timestamp features, including hour code, work and rest period type, and monthly seasonal basic time dimension. The average water use in the past 24 hours and the extreme water use in the past week are calculated through a sliding window to capture intraday fluctuations and periodic patterns. The water use time distribution, usage proportion, and water use correlation characteristics of each scene in the multi-dimensional scene water use map are extracted. The scene features are integrated into the basic time series, and the changes in population size in household configuration parameters, and the seasonal changes in water use for pet breeding and green plant planting are integrated into the time series. Perform time serialization processing and embed the time series framework of the fused scene. Align the regional temperature, precipitation probability, sunshine duration, and air humidity in the environmental field variables with the corresponding water use data timestamps. Extract the change features at different timestamps and add them to the time series feature set. Extract abnormal water use event tags from the water intelligent equipment data. The abnormal water use event is when the scene water consumption exceeds the historical average by 3 times. In this case, the abnormal feature is marked at the time series node corresponding to the scene and added to the time series feature set. The "normal scene time series law + abnormal event" is supplemented to form a complete feature set, and finally a time series dependent water use feature set is constructed.

[0065] Specifically, the types of work and rest periods are divided into weekdays, weekends, and holidays;

[0066] Step S103: The calculation module inputs the time-dependent water use feature set into the long short-term memory network LSTM model to perform water demand trend learning. The LSTM model architecture is that the input layer receives the time-dependent water use feature set, the hidden layer uses the cell state as a long-term memory carrier, and uses the forget gate to selectively retain key information such as seasonal water use patterns. The input gate integrates current temperature, scene water use and other features to update the memory, and the output gate extracts patterns related to future time periods, and finally generates the initial value of water demand prediction. The output layer outputs the initial value of water demand prediction for the future time period. The initial value of water demand prediction output by the LSTM model is optimized, and the error correction function expression is E(t)=λ×(Y actu (t)-Y pre (t))+α×E(t-1),Y actu (t)-Y pre (t) is the deviation between the actual water consumption at time t and the model prediction value, which is used to reflect the instantaneous error of the current prediction. λ is the weight coefficient of the current error. E(t) and E(t-1) are the correction errors at time t and t-1 respectively. ɑ is the weight coefficient of the historical error. The optimal values ​​of the weight coefficients λ and ɑ are substituted into the error correction function to make the prediction value closer to the actual water consumption. The iterative calculation is initialized with E(0) = 0, and starting from t = 1, the error correction function is substituted into each moment in turn to calculate the current correction error E(t), and the prediction value is corrected based on E(t) to obtain Y corre(t) = Y pre (t)+E(t), the iteration continues until the correction error E(t) is less than the minimum threshold ∈, then the termination condition is output and the corrected water demand forecast value Y is corre (t);

[0067] Furthermore, the optimal value solution steps for weight coefficients λ and ɑ are as follows:

[0068] The calculation module obtains Y of the historical period actu (t), Y pre (t) data, objective function Take partial derivatives of λ and ɑ and set the derivatives to zero, solve the equations to obtain the optimal parameters of weight coefficients λ and ɑ;

[0069] Step S104: Adjust the water demand forecast value, obtain the user water use level at each moment, the level is the first level, the second level, the third level, and the corresponding adjustment coefficients are 1, 0.98 and 0.95, respectively, marked as adjustment coefficient β, and the corrected water demand forecast value Y is obtained. corre (t) is multiplied by the adjustment coefficient β to obtain the final water demand forecast value Y final (t).

[0070] In a specific embodiment, the process of executing step S02 may specifically include the following steps:

[0071] (1) The calculation module obtains the set of pipeline physical parameters of the collection terminal in each sub-area. The set of pipeline physical parameters includes: geometric characteristic parameters, material characteristic parameters and fluid migration parameters. The geometric characteristic parameters include equivalent diameter, flow channel complexity factor and pipeline length. The material characteristic parameters include interface energy attenuation coefficient and roughness coefficient. The fluid migration parameters include water hardness and flow velocity.

[0072] (2) Monitor the operation of the water pump flow channel to obtain a set of equipment degradation parameters, which include: water pump flow channel deformation coefficient and seal viscoelastic loss factor;

[0073] (3) Conducting pipeline equipment environmental monitoring on the water pump flow path to obtain a set of environmental impact parameters, including: corrosive medium activity and soil pH;

[0074] (4) Conduct water supply energy-saving and safety scheduling analysis based on the pipeline physical parameter set, equipment degradation parameter set, and environmental impact parameter set to generate a combination of water supply pipeline equipment control parameters;

[0075] Specifically, the equivalent diameter is used to represent the diameter of a circular pipe, as changes in the inner diameter caused by scaling and corrosion are equivalent to these changes. These changes are dynamically updated to accurately reflect the actual flow capacity within the pipeline. The flow complexity factor is a dimensionless index calculated from geometric parameters such as elbow angle, number, valve type, and layout. It is used to quantify the tortuosity of the pipeline system. Specifically, the corresponding values ​​for 45-degree elbows, 90-degree elbows, gate valves, and ball valves are 0.5, 1, 0.3, and 0.6, respectively. The values ​​for elbows of the same angle and different valve types in the subregion are multiplied and summed. The interfacial energy attenuation coefficient is calculated by measuring the change in the contact angle of the liquid on the material surface and using Young's equation to calculate the solid-liquid interfacial energy. Its rate of change with time is the interfacial energy attenuation rate. The pump flow channel deformation coefficient is obtained by fitting the pump body vibration signal collected by a vibration sensor, combining it with flow sensor data, and analyzing the relationship between vibration frequency and flow attenuation using Fourier transform. The seal viscoelastic loss factor is calculated by analyzing the seal vibration response through ultrasonic spectroscopy, extracting the attenuation coefficient and phase difference, and calculating the viscoelastic loss factor. The activity of the corrosive medium is determined by electrochemical measurement.

[0076] In a specific embodiment, the process of executing step S03 may specifically include the following steps:

[0077] (1) The calculation module creates a fractional-order degenerate diffusion process control model for water supply energy-saving scheduling based on the control parameter combination of water supply pipeline equipment. The fractional-order degenerate diffusion process control model includes: Among them D α is the Caputo fractional derivative, ɑ takes the value of (0, 1), which describes the historical dependence of degradation, S(t) is the abnormal value of the pipeline state at time t, X i (t) is the current actual value of the influencing factor in the water supply pipeline equipment control parameter combination at time t, i is the serial number of the influencing factor, which is a positive integer with a maximum value of 11. is the diffusion coefficient, χ i is the weight of influencing factor i, is the memory kernel integral term, K×(t-γ) is the decay memory kernel, Indicates that the Laplace operation is performed on the pipeline state abnormal value S(t), which is used to characterize the change trend of the pipeline state abnormality in the spatial dimension and obtain the pipeline state abnormal value at time t;

[0078] It should be noted that the weight of influencing factor i is i The steps to solve are as follows:

[0079] To X i (t) The data is standardized to eliminate the dimension effect and calculate the information entropy E of each factor i i , The probability of the i-th factor at each moment is X it To obtain the i-th factor at different times t1, t2, ..., t m The monitoring value is calculated to get the weight χ i ,

[0080] (2) The Darcy-Weisbach equation is used to control the fractional-order degenerate diffusion process model to calculate the pressure loss state change during pipeline fluid flow, and the pipeline fluid pressure loss value ΔP(t) is obtained. The Darcy-Weisbach equation includes: Where f is the pipe roughness, LG is the pipe length, the physical length of the pipe through which the fluid flows, DG is the pipe diameter, and the flow cross-sectional characteristics of the fluid flow are calculated. ρ is the fluid density, v is the fluid velocity, and k S is the outlier amplification factor.

[0081] In a specific embodiment, the process of executing step S04 may specifically include the following steps:

[0082] The calculation module normalizes the final water demand forecast value, pipeline status anomaly value, and pipeline fluid pressure loss value of each sub-region using a standardized formula to eliminate dimensional differences. The module uses a regular triangular pyramid as the basic spatial model, defines the base as an equilateral triangle, and the height is perpendicular to the base. The normalized water demand forecast value is mapped to the side length of the base equilateral triangle, and the pipeline status anomaly value is mapped to the height of the triangular pyramid. A sphere is constructed on the basic spatial model, where the center of the sphere coincides with the upper vertex of the regular triangular pyramid. The pipeline fluid pressure loss value is used as the radius of the sphere. The volume formed by the sphere and the regular triangular pyramid is identified and marked as the water pipeline pressure anomaly value at each region and time.

[0083] Set r water pipe pressure difference values, each of which corresponds to a value range, G1(0, g1), G2(g1, g2), G3(g2, g3), ..., G r (g r-1 , g r ), where G1, G2, G3, ..., G r Respectively represent the segment interval numbers of the water pipe pressure difference values, g1, g2, g3, ..., g r They respectively represent the corresponding electronic valve operating opening values, and match the water pipe pressure difference values ​​corresponding to each sub-area at each time point with the value range. If the set water pipe pressure difference value falls within the value range, the electronic valve operating opening value corresponding to the value range is marked as the adjusted valve opening value; the scheduling module receives the calculated value of the calculation module and operates the electronic valve according to the adjusted valve opening value.

[0084] It should be noted that the greater the pressure difference in the water pipes in each area, the larger the opening value of the electronic valve, which can realize intelligent analysis, measurement and operation control, and operate with minimum energy consumption.

[0085] In a specific embodiment, the process of executing step S05 may specifically include the following steps: Figure 3 As shown:

[0086] Step S501: Water quality analysis and purification processing. The calculation module receives the water quality data obtained by the sensor module. If the total value of the total bacteria, total coliform bacteria, and total heat-resistant coliform bacteria content in the water quality data exceeds the set threshold τ01, signal one is issued, otherwise signal three is issued. If the pH, heavy metal content, and organic matter content in the water quality data all exceed the corresponding set thresholds, signal one is issued. If one or two of them exceed the corresponding set thresholds, signal two is issued. If none of them exceed the corresponding set thresholds, signal three is issued. If the residual chlorine in the water quality data exceeds the set threshold τ02, signal one is issued, otherwise signal three is issued.

[0087] If there is signal one in the sent signal, a shut-off valve control signal is issued; if there are two or three signal ones in the sent signal, a water quality abnormality and severe pollution signal is issued and sent to the purification module, and the purification module automatically switches the pipeline in the water quality pollution area to the water distribution pipeline of the clean water source; if there is one signal one or one signal two in the sent signal, a water quality abnormality and slight pollution signal is issued and sent to the purification module, and the purification module activates the distributed water quality purification facilities in the water quality pollution sub-area for real-time treatment and disposal. No corresponding operation is performed in other cases;

[0088] Step S502: Dynamic hydrodynamic modeling. The calculation module performs valve control analysis after receiving the interception and discharge signal. The dynamic hydrodynamic modeling is based on the pipeline network drawings and 3D scanning data to establish a digital twin model of the pipeline network. The length, diameter, roughness and other parameters of each pipe section are marked. Combined with real-time flow data, the hydraulic state of the pipeline network is calculated through the pressure-driven model, and the above-mentioned Darcy-Weisbach equation is applied. And Manning formula: v = 1 / n × R 2 / 3 ×S 1 / 2 Calculate water flow velocity, where n is the Manning roughness coefficient, R is the hydraulic radius, and S represents the hydraulic slope. Pollution path analysis provides flow velocity data support;

[0089] Step S503: Pollutant transport prediction, establishing a one-dimensional convection-diffusion equation model: C represents the pollutant concentration, x spatial coordinate, a one-dimensional spatial variable along the direction of water flow, D is the molecular diffusion coefficient, k is the pollutant attenuation coefficient, is the convection term, describing the migration of pollutants with water flow, is the diffusion term, reflecting the diffusion driven by the concentration gradient, and -k×C is the attenuation term, simulating the natural attenuation process of pollutants. It should be noted that different diffusion parameters are set for yellow water and odorous water, which are specifically obtained from the database;

[0090] Step S504: Pollution path deduction and valve control. Based on a graph theory algorithm, the hydrodynamic modeling and pollutant transport prediction results are integrated to identify the priority pollution propagation paths, mark sensitive nodes that need protection, and remotely close the electric valves on the pollution path, such as butterfly valves and gate valves, through the system to cut off the spread of pollution. The valve closest to the pollution source is closed first, and then the pollution is cut off step by step upstream.

[0091] The above is an illustration of the present invention and should not be considered as limiting thereof. Although several exemplary embodiments of the present invention have been described, it will be readily understood by those skilled in the art that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present invention. Therefore, all such modifications are intended to be included within the scope of the present invention as defined by the claims. It should be understood that the above is an illustration of the present invention and should not be considered as being limited to the specific embodiments disclosed, and modifications to the disclosed embodiments and other embodiments are intended to be included within the scope of the appended claims. The present invention is defined by the claims and their equivalents.

Claims

1. A water supply energy-saving, safe scheduling and purification intelligent control system, including a sensor module, a calculation module, a collection terminal, a scheduling module and a purification module, characterized in that: The dispatching module includes electronic valves. Based on the city's industrial water use areas, residential density, and functional zoning factors, the entire water supply area is divided into several independent sub-areas. Electronic valves are set at the boundaries of each sub-area to independently control the water supply of each sub-area. The electronic valves receive the calculated values ​​from the calculation module and operate according to the adjusted valve opening values. The purification module includes water distribution pipelines and purification facilities. It receives signals from the calculation module and performs corresponding purification operations. The collection terminal collects user water use related information and water supply pipeline equipment control information and transmits it to the calculation module; The sensing module includes water quality monitoring sensors, which are arranged at key nodes of the pipeline network and transmit the obtained water quality data to the computing module; the computing module is used to construct a time-dependent water use feature set, analyze water demand forecast values, construct a combination of water supply pipeline equipment control parameters, and create a fractional-order degraded diffusion process control model for water supply energy-saving scheduling based on the combination of water supply pipeline equipment control parameters. It also analyzes the changes in pressure loss state during pipeline fluid flow, analyzes water quality anomalies and generates interception and purification signals, and deduces pollution paths.

2. The water supply energy-saving, safe scheduling and purification intelligent control system according to claim 1 is characterized in that: The following steps are also included: Step S01: Calculating the final water demand forecast: The calculation module obtains the water use-related information of the end users in the sub-region, constructs a time-dependent water use feature set, predicts and corrects the error using the long short-term memory network (LSTM) model, and finally adjusts the water use ladder to obtain the final water demand forecast; Step S02: Water supply pipeline equipment control analysis: The calculation module obtains a set of pipeline physical parameters, a set of equipment degradation parameters, and an environmental impact parameter set of the water supply pipeline equipment, and performs a water supply energy-saving and safety scheduling analysis on the set of pipeline physical parameters, the set of equipment degradation parameters, and the set of environmental impact parameters to generate a combination of water supply pipeline equipment control parameters; Step S03: Analysis of pipeline state abnormalities and pipeline fluid pressure: The calculation module creates a fractional-order degenerate diffusion process control model for water supply energy-saving scheduling based on the control parameter combination of the water supply pipeline equipment, analyzes and calculates pipeline state abnormalities, and analyzes the pressure loss state changes during pipeline fluid flow to obtain the pipeline fluid pressure loss value; Step S04: Valve opening scheduling: The calculation module obtains the water pipe pressure difference value of each area at each time based on the image construction, and uses the scheduling module to adjust the electronic valve to work according to the valve opening value; Step S05: shut-off valve control and purification process: the calculation module receives the water quality data obtained by the sensor module, analyzes various states and sends corresponding signals, and performs corresponding valve control and purification process.

3. The water supply energy-saving, safe scheduling and purification intelligent control system according to claim 2 is characterized in that: The steps for analyzing the final water demand forecast value are as follows: Step S103: The calculation module inputs the time-dependent water use feature set into the long short-term memory network LSTM model. The LSTM model architecture is that the input layer receives the time-dependent water use feature set, the hidden layer uses the cell state as a long-term memory carrier, and uses the forget gate to selectively retain the key information of the seasonal water use law. The input gate integrates the current temperature and scene water use features to update the memory, and the output gate extracts the law related to the future time period, and finally generates the initial value of water demand prediction. The output layer outputs the initial value of water demand prediction for the future time period. The initial value of water demand prediction output by the LSTM model is optimized, and the error correction function expression is E(t)=λ×(Y actu (t)-Y pre (t))+α×E(t-1),Y actu (t)-Y pre (t) is the deviation between the actual water consumption at time t and the model prediction value, which is used to reflect the instantaneous error of the current prediction. λ is the weight coefficient of the current error. E(t) and E(t-1) are the correction errors at time t and t-1 respectively. ɑ is the weight coefficient of the historical error. The optimal values ​​of the weight coefficients λ and ɑ are substituted into the error correction function to make the prediction value closer to the actual water consumption. The iterative calculation is initialized with E(0) = 0. Starting from t = 1, the error correction function is substituted into each moment in turn to calculate the current correction error E(t). The predicted value is corrected based on E(t) to obtain Y corre (t) = Y pre (t)+E(t), the iteration continues until the correction error E(t) is less than the minimum threshold ∈, then the termination condition is output and the corrected water demand forecast value Y is corre (t); The steps for solving the optimal values ​​of the weight coefficients λ and ɑ are as follows: The calculation module obtains Y of the historical period actu (t), Y pre (t) data, objective function Take partial derivatives of λ and ɑ and set the derivatives to zero, solve the equations to obtain the optimal parameters of weight coefficients λ and ɑ; Step S104: Adjust the water demand forecast value, obtain the user water demand level at each moment, the level is the first level, the second level, the third level, and the corresponding adjustment coefficient β, and adjust the water demand forecast value Y after correction. corre (t) is multiplied by the adjustment coefficient β to obtain the final water demand forecast value Y final (t).

4. The water supply energy-saving, safe scheduling and purification intelligent control system according to claim 3 is characterized in that: The steps for analyzing the time-dependent water use feature set are as follows: Step S101: Establishing a water use-related parameter set. The calculation module obtains a user water use-related parameter set from the collection terminal in the sub-region. The parameter set includes: time-series water use archives, household configuration parameters, environmental field variables, and water-connected intelligent equipment data. The time-series water use archives include time period water consumption, monthly total water consumption, and multi-dimensional scenario water use maps. The household configuration parameters include population size, age structure, water consumption for pet breeding, and water consumption for planting plants. The environmental field variables include regional temperature, precipitation probability, sunshine duration, and air humidity. The water-connected intelligent equipment data includes data from intelligent water devices. Step S102: The calculation module constructs a time-dependent water use feature set. Based on the time-series water use archive, the time period water use and monthly total are decomposed into timestamp features, including hour code, work and rest period type, and monthly seasonal basic time dimensions. The average water use over the past 24 hours and the extreme water use over the past week are calculated using a sliding window to capture intraday fluctuations and cyclical patterns. The water use time distribution, water use proportion, and water use correlation characteristics between scenarios are extracted for each scenario in the multidimensional scenario water use map. The scenario features are integrated into the basic time series. The changes in population size in household configuration parameters and the seasonal changes in water use for pet breeding and green plant cultivation are time-series processed and embedded into the time series framework of the integrated scenario. The regional temperature, precipitation probability, sunshine duration, and air humidity in the environmental field variables are aligned with the corresponding water use data timestamps. The change features at different timestamps are extracted and superimposed on the time series feature set. Abnormal water use event markers are extracted from the water intelligent equipment data. The abnormal features are marked at the time series nodes corresponding to the scenario and superimposed on the time series feature set to form a complete feature set, and finally a time-series-dependent water use feature set is constructed.

5. The water supply energy-saving, safe scheduling and purification intelligent control system according to claim 2 is characterized in that: The steps for analyzing the control parameter combination of the water supply pipeline equipment are as follows: (1) The calculation module obtains the set of pipeline physical parameters of the collection terminal in each sub-area. The set of pipeline physical parameters includes: geometric characteristic parameters, material characteristic parameters and fluid migration parameters. The geometric characteristic parameters include equivalent diameter, flow channel complexity factor and pipeline length. The material characteristic parameters include interface energy attenuation coefficient and roughness coefficient. The fluid migration parameters include water hardness and flow velocity. (2) Monitor the operation of the water pump flow channel to obtain a set of equipment degradation parameters, which include: water pump flow channel deformation coefficient and seal viscoelastic loss factor; (3) Conducting pipeline equipment environmental monitoring on the water pump flow channel to obtain a set of environmental impact parameters, which include: corrosive medium activity and soil pH; (4) Conduct water supply energy-saving and safety scheduling analysis based on the pipeline physical parameter set, equipment degradation parameter set, and environmental impact parameter set to generate a combination of water supply pipeline equipment control parameters.

6. The water supply energy-saving, safe scheduling and purification intelligent control system according to claim 2 is characterized in that: The pipeline status abnormal value analysis steps are as follows: The calculation module creates a fractional-order degenerate diffusion process control model for water supply energy-saving scheduling based on the control parameter combination of water supply pipeline equipment. The fractional-order degenerate diffusion process control model includes: Among them D α is the Caputo fractional derivative, S(t) is the abnormal value of pipeline status at time t, X i (t) is the current actual value of the influencing factor in the water supply pipeline equipment control parameter combination at time t, i is the serial number of the influencing factor, which is a positive integer with a maximum value of 11. is the diffusion coefficient, χ i is the weight of influencing factor i, is the memory kernel integral term, K×(t-γ) is the decay memory kernel, It means that the pipeline state abnormal value S(t) is subjected to Laplace operation to obtain the pipeline state abnormal value at time t; The weight χ of the influencing factor i i The steps to solve are as follows: To X i (t) The data is standardized to eliminate the dimension effect, and the information entropy E of each factor i is calculated by the calculation formula i , the weight χ is calculated according to the formula i .

7. The water supply energy-saving, safe scheduling and purification intelligent control system according to claim 2 is characterized in that: The pipeline fluid pressure loss value analysis steps are as follows: The Darcy-Weisbach equation is used to calculate the pressure loss state change of the pipeline fluid during the flow of the fractional-order degenerate diffusion process control model, and the pipeline fluid pressure loss value ΔP(t) is obtained. The Darcy-Weisbach equation includes: Where f is the pipe roughness, LG is the pipe length, the physical length of the pipe through which the fluid flows, DG is the pipe diameter, and the flow cross-sectional characteristics of the fluid flow are calculated. ρ is the fluid density, v is the fluid velocity, and k S is the outlier amplification factor.

8. The water supply energy-saving, safe scheduling and purification intelligent control system according to claim 2 is characterized in that: The steps for adjusting the valve opening value to perform work analysis are as follows: The calculation module processes the final water demand forecast value, pipeline status abnormal value, and pipeline fluid pressure loss value of each sub-area through graphical construction to obtain the water pipeline pressure abnormal value of each area at each time; Set the water pipe pressure deviation value. Each water pipe pressure deviation value corresponds to a value range. Match the water pipe pressure deviation value corresponding to each sub-area at each time point with the value range. If the set water pipe pressure deviation value falls within the value range, the electronic valve operation opening value corresponding to the value range is marked as the adjusted valve opening value. The scheduling module receives the calculated value from the calculation module and operates the electronic valve according to the adjusted valve opening value.

9. The water supply energy-saving, safe scheduling and purification intelligent control system according to claim 2 is characterized in that: The purification process analysis steps are as follows: Step S501: Water quality analysis and purification processing. The calculation module receives the water quality data obtained by the sensor module and sends signal one, signal two, and signal three accordingly. If signal one is included in the sent signal, a shut-off valve control signal is sent. If two or three signal ones are included in the sent signal, a water quality abnormality and serious pollution signal is sent and sent to the purification module. The purification module performs purification measure one. If one signal one or one signal two is included in the sent signal, a water quality abnormality and slight pollution signal is sent and sent to the purification module. The purification module performs purification measure two.

10. The water supply energy-saving, safe scheduling and purification intelligent control system according to claim 2 is characterized in that: The valve control analysis steps are as follows: Step S502: Dynamic hydrodynamic modeling: The calculation module performs valve control analysis after receiving the interception and discharge signal. The dynamic hydrodynamic modeling is based on the pipeline network drawings and 3D scanning data to establish a digital twin model of the pipeline network. Parameters such as the length, diameter, and roughness of each pipe segment are annotated. Combined with real-time flow data, the hydraulic state of the pipeline network is calculated using a pressure-driven model. The water flow rate is calculated using the aforementioned Darcy-Weisbach equation and Manning formula. Step S503: Pollutant transport prediction, establishing a one-dimensional convection-diffusion equation model; Step S504: Pollution path deduction and valve control, based on the graph theory algorithm, integrating hydrodynamic modeling and pollutant transport prediction results, identifying the priority paths of pollution propagation, marking sensitive nodes that need to be protected, and remotely closing the electric valves on the pollution path through the system to cut off the spread of pollution. First, close the valve closest to the pollution source, and then cut off the pollution upstream step by step.

Images