Flow partition intelligent water supply method and system

Abstract

The invention relates to the technical field of intelligent water supply, in particular to a flow partition intelligent water supply method and system. The data sensing layers are installed at the tail ends of the main pipes and the branch pipes of all the flow partitions, partition water supply data are collected in real time, water quality data are subjected to multi-dimensional analysis through a water quality monitoring mechanism, water quality abnormity is accurately recognized, a cleaning instruction is generated, intelligent monitoring and self-cleaning of water quality are achieved, and the water quality monitoring efficiency is improved. The problem of bacterium breeding caused by static water flow is effectively solved, and water safety is guaranteed; through cooperative operation of the global data sensing module, the partition pressure analysis module and the pressure compensation module, real-time pressure, historical fluctuation and external influence data in partition flow data are comprehensively analyzed, a pressure compensation instruction is generated in combination with entropy calculation and weight distribution, pipe network pressure is regulated and controlled in real time, and the pressure compensation efficiency is improved. The problems of water hammer effect and unstable tail end water consumption in peak periods are solved, and the stability of a water supply system is improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent water supply technology, specifically to an intelligent water supply method and system with flow zoned operation. Background Technology

[0002] Urban water supply is a municipal infrastructure system built according to building codes, consisting of three core components: water source intake facilities, water treatment plants, and water transmission and distribution networks. This system adheres to the principle of combining water source development with water conservation, encompassing both public water supply and self-built facility water supply, and implementing a tiered management system. Currently, the main water supply methods are tank-type and tankless water supply. Urban water supply systems rely on water transmission and distribution networks to achieve tiered water supply. Tankless water supply is gaining popularity due to its energy-saving and land-saving advantages. However, during off-peak hours, water remains stagnant in the pipes for extended periods, making it prone to bacterial growth (such as Legionella), especially at the end of risers or in branch pipes, leading to water safety issues. Furthermore, during peak hours, frequent start-stop or delayed speed regulation of distributed pumps causes sudden rises and falls in network pressure, triggering the "water hammer effect" and unstable end-user water supply. Summary of the Invention

[0003] This invention provides a flow-zoned intelligent water supply method and system to solve the above-mentioned technical problems.

[0004] The first aspect of this invention provides a flow-zoned intelligent water supply method and system, comprising the following steps:

[0005] Step 1: Obtain the flow zones corresponding to the water supply area. Based on the preset data acquisition strategy, data is collected through the data sensor layer installed at the end of the main pipe and branch pipe of each flow zone to obtain the zone water supply data. The zone water supply data includes zone water quality data and zone flow data. The zone water quality data includes basic physical parameters, water quality indicators, microbial parameters and pipe attributes.

[0006] Step 2: Obtain the water supply data for each flow zone and identify the zone water supply data to obtain zone water quality data and zone flow data. Monitor the zone water supply data through the data identification unit. When the zone water supply data corresponds to the zone water quality data, proceed to Step 3; when the zone water supply data corresponds to the zone flow data, proceed to Step 4.

[0007] Step 3: Obtain zoned water quality data and transmit the zoned water quality data to the water quality monitoring mechanism. The water quality monitoring mechanism analyzes the zoned water quality data to obtain water quality cleaning instructions. The water quality monitoring mechanism includes water quality monitoring triggering, water quality monitoring analysis, and water quality status analysis.

[0008] As a further improvement of the present invention, the water quality monitoring trigger is specifically as follows: based on the water quality data of the zone, basic physical parameters are obtained; based on the basic physical parameters, the water flow velocity of each main pipe and branch pipe in the flow zone is obtained; the minimum threshold of the water flow velocity corresponding to each main pipe and branch pipe is obtained; when the water flow velocity is lower than the corresponding minimum threshold, the corresponding water flow is marked as a still water flow, and the start information of the timing unit is triggered; the duration of the still water flow in each main pipe and branch pipe is recorded by the timing unit to obtain the still water flow duration; the still water flow duration is multiplied by a pre-set still time risk coefficient to calculate the still water flow risk value; when the still water flow risk value exceeds the pre-set threshold, a water quality monitoring start signal is generated.

[0009] As a further improvement of the present invention, water quality monitoring and analysis specifically involves: when a water quality monitoring start signal is detected in a flow zone, water quality indicators, microbial parameters and pipeline attributes are obtained based on the corresponding zone water quality data.

[0010] Based on water quality indicators, the water quality parameters are turbidity, residual chlorine, and pH value; a continuous Bayesian network model is used to construct definition nodes for each water quality parameter. Where TUR, cl, and pH represent turbidity, residual chlorine, and pH value, respectively, and a Gaussian distribution is used to model the conditional concept of continuous variables. Calculate the joint probability Obtain a pre-designed empirical probability threshold, and compare the joint probability with the empirical probability threshold for analysis. Output indicator outlier S zb Where zb represents the preset anomaly index, which is a constant; θ jy This is expressed as a probability empirical threshold.

[0011] The microbial concentrations of each main pipe and branch pipe were obtained based on microbial parameters. The corresponding time-series data of microbial concentrations were acquired and denoted as WS(t). The microbial concentration time-series data were then discretized to obtain a discrete point set (t). i WS i The Hough transform formula for line detection using the Hough transform model. Where ρ represents the perpendicular distance from the origin to the line, and θ is the angle between the perpendicular line and the x-axis, the data points corresponding to the discrete point set (t) i WS i Substituting into the Hough transform formula above, cumulative voting is performed in the (ρ, θ) parameter space to generate the Hough matrix Hf(ρ, θ); the corresponding peak point (ρ) is obtained by identifying the Hough matrix through the identification unit. peak θ peakThe system obtains a pre-set peak threshold; if the peak value corresponding to the peak point exceeds the peak threshold and the slope of the microbial concentration change at the corresponding data point exceeds the preset normal range, a judgment information of abnormality is generated; and the deviation degree of the peak value corresponding to the judgment information of abnormality is calculated to obtain the microbial abnormal value.

[0012] The attribute information of each main pipe and branch pipe is obtained based on the pipe attributes; the pipe attribute decision table is obtained by constructing a decision table based on the rough set theory model. Where, Uy is the sample set corresponding to the attribute data of different pipelines; As=Cs∪Ds is the attribute set, Cs is the conditional attribute, which corresponds to attribute information, and Ds is the decision attribute, which corresponds to the abnormal information of the pipeline; Vs is the attribute value set, and f represents the information function; through The dependence of the conditional attribute on the decision attribute is calculated. ;in, Let Cs be the positive domain of X; reduce the conditional attributes to obtain the corresponding minimal reduced attribute set Cs-red; generate decision rules based on the reduced attribute set; when the new pipeline attribute data satisfies the anomaly decision rule, determine that the pipeline attribute is anomaly; based on the case that the pipeline attribute is anomaly, calculate the attribute anomaly value according to the rule matching degree.

[0013] The static risk values ​​of water flow, outliers of indicators, outliers of microorganisms, and outliers of attributes are normalized and their values ​​are marked as dimensional outliers, denoted as S. jz S zb S ws S sx By combining the improved DS evidence theory with a simple feedforward neural network, basic probability assignment is performed on the dimensional outliers corresponding to each dimension. Similarly, the basic probability allocations for the remaining dimensions are obtained, and weighted synthesis is performed using the DS evidence synthesis rules. The comprehensive outlier s was calculated. 总 ;in, Represented as the dimensional outliers for each dimension, β i This represents the preset weight factor corresponding to the outlier in the i-th dimension.

[0014] As a further improvement of the present invention, the water quality status analysis specifically involves: obtaining the comprehensive anomaly values ​​corresponding to each main pipe and branch pipe, as well as the corresponding preset anomaly warning thresholds and anomaly allowable values, s 总-min and s 总-max The current comprehensive anomaly values ​​corresponding to each main pipe and branch pipe are matched and analyzed with the anomaly warning threshold and the anomaly allowable value, respectively. When s 总 ≤s 总-minWhen s is reached, the corresponding water quality status is "no abnormalities"; when s is reached... 总-min <s 总 ≤s 总-max When s is present, the corresponding water quality status is generated as water quality abnormality; when s is present... 总 >s 总-max When this occurs, a corresponding water quality status alarm is generated.

[0015] Obtain the water quality status of each main pipe and branch pipe in the flow zone. When the water quality status is normal, there is no water quality cleaning instruction. When the water quality status is abnormal, the water quality cleaning instruction is a water quality cleaning start signal. When the water quality status is an over-limit alarm, the water quality cleaning instruction is a water quality emergency treatment signal.

[0016] Step 4: Obtain the zone flow data corresponding to each flow zone, and input the zone flow data into the water supply monitoring mechanism to obtain the corresponding zone water supply monitoring information. Based on the zone water supply monitoring information of each flow zone, carry out corresponding water supply regulation.

[0017] Step 5: Obtain water quality cleaning instructions for each flow zone and perform self-cleaning execution feedback based on the water quality cleaning instructions.

[0018] As a further improvement of the present invention, self-cleaning execution feedback is performed based on water quality cleaning instructions. Specifically, this involves: monitoring water quality cleaning instructions for each flow zone; when a flow zone has a water quality cleaning instruction, marking the corresponding flow zone as a zone to be cleaned; identifying the water quality cleaning instructions corresponding to each zone to be cleaned; obtaining the main pipe or branch pipe information corresponding to the water quality cleaning instruction; generating water quality self-cleaning execution information when the water quality cleaning instruction corresponds to a water quality abnormality; and generating pipeline isolation execution information when the water quality cleaning instruction corresponds to an over-limit alarm.

[0019] A second aspect of the present invention provides a flow-zoned intelligent water supply system, including a global data sensing module, a zoned pressure analysis module, a pressure compensation module, and a data storage module.

[0020] The global data sensing module is used to collect data from each flow zone of the entire water supply area through a pre-deployed global data sensing layer to obtain the zone flow data, and then send the zone flow data to the data storage module; the zone flow data includes real-time pressure data, historical time series data and external impact data of each flow zone.

[0021] The partition pressure analysis module is used to acquire partition traffic data for each traffic partition, perform partition pressure analysis on the partition traffic data to obtain the partition pressure status, and send the partition pressure status to the pressure compensation module.

[0022] As a further improvement to the present invention, partition pressure analysis is performed on the partition traffic data, and the specific analysis content is as follows:

[0023] Based on the partition traffic data of each traffic partition, real-time pressure data, historical time-series data, and external impact data are obtained;

[0024] Based on real-time pressure data, the real-time pressure values ​​of the main pipe, zone nodes, and terminals of the pipeline in the flow zone are obtained. The real-time pressure value of the main pipe is calculated with the real-time pressure value of the other pipeline sections to obtain two real-time pressure difference values. Similarly, the real-time pressure difference value between the zone node and the terminal is obtained. The average pressure difference value is calculated by combining the three real-time pressure difference values. When the average pressure difference value is greater than the preset threshold, the corresponding average pressure difference value is marked as a pressure anomaly value.

[0025] Based on historical time-series data, a pressure fluctuation curve within a preset historical time period is obtained. Based on the pressure fluctuation curve, a pre-set standard fluctuation range is obtained, and the time axis of the pressure fluctuation curve is divided into multiple time-series intervals according to preset time intervals. The length of the pressure fluctuation curve in each time-series interval that exceeds the standard fluctuation range is obtained and marked as the high fluctuation curve length. The ratio of the high fluctuation curve length to the total length of the pressure fluctuation curve in the time-series interval is calculated to obtain the high fluctuation ratio value. The current time corresponding to the current flow partition is matched with each time-series interval of the corresponding historical pressure fluctuation curve to obtain the corresponding historical time-series interval, and the corresponding high fluctuation ratio value is marked as the historical fluctuation value.

[0026] Based on external influence data, the corresponding external factors are obtained; a decision tree classifier is constructed using a decision tree rule weight model to label each external factor as an input variable and obtain the output variable; for each leaf node rule... Among them, B X and B Y Let be the input and output variables, respectively. Let Ci represent the condition set of the i-th rule. Calculate the rule confidence and support, and multiply the rule confidence and support to obtain the rule weight. If the current external data matches rule Ri, the outlier contribution is the rule weight; otherwise, it is zero. The external outlier value wb is calculated; where α represents the rule weight, and m and M represent the number of matched rules and the total number of rules, respectively.

[0027] Each sub-anomaly value corresponding to the pressure anomaly value, historical fluctuation value, and external anomaly value of the flow zone is normalized and its value is denoted as E. y E s and E wb Substitute each sub-outlier into the entropy calculation formula. The entropy value is calculated to obtain the corresponding entropy value, and then the entropy value corresponding to each sub-outlier is substituted into the weight calculation formula. The corresponding weighting factors are obtained, and the normalized values ​​corresponding to each sub-outlier are weighted using a weighting formula. The total value of partition anomalies is calculated; when the total value of partition anomalies exceeds the preset threshold, the corresponding partition pressure status is generated as an abnormal pressure status.

[0028] The pressure compensation module is used to receive the pressure status of each flow zone and obtain pressure compensation instructions based on the pressure status. Specifically, it monitors the pressure status of each flow zone. When the pressure status is normal, there is no corresponding pressure compensation instruction. When the pressure status is abnormal, a corresponding pressure compensation instruction is generated. The pressure compensation instruction includes pressure compensation through the energy storage compensation module of each zone node and power adjustment compensation through the water pump cluster.

[0029] The data storage module is used to receive and store partition traffic data, as well as partition pressure status and pressure compensation instructions.

[0030] The beneficial effects of the technical solution provided by this invention compared with the prior art are as follows:

[0031] 1. This invention installs data sensing layers at the ends of the main pipes and branch pipes in each flow zone to collect zone water supply data in real time. Through a water quality monitoring mechanism, it performs multi-dimensional analysis of water quality data, accurately identifies water quality anomalies, and generates cleaning instructions. This achieves intelligent monitoring and self-cleaning of water quality, effectively solves the problem of bacterial growth caused by stagnant water flow, and ensures water safety.

[0032] 2. This invention utilizes the coordinated operation of a global data sensing module, a zone pressure analysis module, and a pressure compensation module to comprehensively analyze real-time pressure, historical fluctuations, and external influence data in zone flow data. It combines entropy calculation and weight allocation to generate pressure compensation commands, thereby regulating pipeline pressure in real time, solving the "water hammer effect" during peak hours and the problem of unstable water use at the end of the pipeline, and improving the stability of the water supply system. Attached Figure Description

[0033] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. The following drawings are not deliberately drawn to scale according to the actual size, but are intended to show the main idea of ​​this application.

[0034] Figure 1 This is a flowchart of a flow-zoned intelligent water supply method according to the present invention;

[0035] Figure 2 This is a schematic diagram illustrating the principle of the water quality monitoring mechanism of the present invention.

[0036] Figure 3 This is a schematic diagram of a flow-zoned intelligent water supply system according to the present invention. Detailed Implementation

[0037] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. 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.

[0038] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1-2 In one embodiment of the present invention, a method for intelligent water supply with flow zoning includes:

[0039] Step 1: Flow Zone Data Acquisition: Obtain the flow zone data corresponding to the water supply area. Based on the preset data acquisition strategy, data is collected through the data sensing layer installed at the end of the main pipe and branch pipe of each flow zone to obtain the zone water supply data. The zone water supply data includes zone water quality data and zone flow data. The zone water quality data includes basic physical parameters, water quality indicators, microbial parameters and pipe attributes. The data sensing layer refers to the sensor equipment installed at the end of the main pipe and branch pipe of each flow zone.

[0040] Step 2: Real-time monitoring of water flow status: Acquire water supply data for each flow zone and identify the zone water supply data to obtain zone water quality data and zone flow data. Monitor the zone water supply data through the data identification unit. When the zone water supply data corresponds to zone water quality data, proceed to Step 3; when the zone water supply data corresponds to zone flow data, proceed to Step 4. The data identification unit is a logical module in the system used to distinguish the data types of zone water supply. It identifies the data types in real time and processes them separately, enabling the water quality and flow monitoring processes to operate independently, thereby improving the system's response efficiency and accuracy.

[0041] Step 3: Zoned water quality monitoring: Obtain zoned water quality data and transmit the data to the water quality monitoring mechanism. The water quality monitoring mechanism analyzes the zoned water quality data to obtain water quality cleaning instructions.

[0042] The water quality monitoring mechanism includes water quality monitoring triggering, water quality monitoring analysis, and water quality status analysis.

[0043] The water quality monitoring triggering mechanism is as follows: Based on the zoned water quality data, basic physical parameters are obtained. Based on the basic physical parameters, the water flow velocity of each main pipe and branch pipe in the flow zone is obtained. The minimum threshold of the water flow velocity corresponding to each main pipe and branch pipe is obtained. When the water flow velocity is lower than the corresponding minimum threshold, the corresponding water flow is marked as stagnant water flow, and the start information of the timing unit is triggered. The duration of stagnant water flow in each main pipe and branch pipe is recorded by the timing unit to obtain the water flow stagnant duration. The water flow stagnant duration is multiplied by a pre-set stagnant time risk coefficient to calculate the water flow stagnant risk value. When the water flow stagnant risk value exceeds the pre-set threshold, a water quality monitoring start signal is generated. Through dual judgment of flow velocity and stagnant duration, dead water areas are accurately identified, and the risk of water quality deterioration is prevented in advance.

[0044] Water quality monitoring and analysis specifically involves: when a water quality monitoring activation signal is detected in a flow zone, water quality indicators, microbiological parameters, and pipeline attributes are obtained based on the corresponding zone's water quality data.

[0045] Based on water quality indicators, the water quality parameters are turbidity, residual chlorine, and pH value; a continuous Bayesian network model is used to construct definition nodes for each water quality parameter. Where TUR, cl, and pH represent turbidity, residual chlorine, and pH value, respectively, and a Gaussian distribution is used to model the conditional concept of continuous variables. Calculate the joint probability Obtain a pre-designed empirical probability threshold, and compare the joint probability with the empirical probability threshold for analysis. Output indicator outlier S zb Where zb represents the preset anomaly index, which is a constant; θ jy This is expressed as a probability empirical threshold.

[0046] The microbial concentrations of each main pipe and branch pipe were obtained based on microbial parameters. The corresponding time-series data of microbial concentrations were acquired and denoted as WS(t). The microbial concentration time-series data were then discretized to obtain a discrete point set (t). i WS i The Hough transform formula for line detection using the Hough transform model. Where ρ represents the perpendicular distance from the origin to the line, and θ is the angle between the perpendicular line and the x-axis, the data points corresponding to the discrete point set (t) i WS i Substituting into the Hough transform formula above, cumulative voting is performed in the (ρ, θ) parameter space to generate the Hough matrix Hf(ρ, θ); the corresponding peak point (ρ) is obtained by identifying the Hough matrix through the identification unit. peak θ peakThe system obtains a pre-set peak threshold; if the peak value corresponding to the peak point exceeds the peak threshold and the slope of the microbial concentration change at the corresponding data point exceeds the preset normal range, a judgment information of abnormality is generated; and the deviation degree of the peak value corresponding to the judgment information of abnormality is calculated to obtain the microbial abnormal value.

[0047] Based on pipeline attributes, attribute information for each main and branch pipe is obtained. This attribute information includes, but is not limited to, pipe material, laying period, service life, and pipe wall thickness. A decision table is then constructed from this attribute information using a rough set theory model to obtain the pipeline attribute decision table. Where, Uy is the sample set corresponding to the attribute data of different pipelines; As=Cs∪Ds is the attribute set, Cs is the conditional attribute, which corresponds to attribute information, and Ds is the decision attribute, which corresponds to the abnormal information of the pipeline; Vs is the attribute value set, and f represents the information function; through The dependence of the conditional attribute on the decision attribute is calculated. ;in, Let Cs be the positive domain of X; reduce the conditional attributes to obtain the corresponding minimal reduced attribute set Cs-red; generate decision rules based on the reduced attribute set; when the new pipeline attribute data satisfies the anomaly decision rule, determine that the pipeline attribute is anomaly; based on the case that the pipeline attribute is anomaly, calculate the attribute anomaly value according to the rule matching degree.

[0048] The static risk values ​​of water flow, outliers of indicators, outliers of microorganisms, and outliers of attributes are normalized and their values ​​are marked as dimensional outliers, denoted as S. jz S zb S ws S sx By combining the improved DS evidence theory with a simple feedforward neural network, basic probability assignment is performed on the dimensional outliers corresponding to each dimension. Similarly, the basic probability allocations for the remaining dimensions are obtained, and weighted synthesis is performed using the DS evidence synthesis rules. The comprehensive outlier s was calculated. 总 ;in, Represented as the dimensional outliers for each dimension, β i This represents the preset weight factor corresponding to the outlier of the i-th dimension.

[0049] Water quality status analysis specifically involves: obtaining the comprehensive anomaly values ​​corresponding to each main pipe and branch pipe, as well as the corresponding preset anomaly warning thresholds and anomaly allowable values. 总-min and s 总-max The current comprehensive anomaly values ​​corresponding to each main pipe and branch pipe are matched and analyzed with the anomaly warning threshold and the anomaly allowable value, respectively. When s 总 ≤s总-min At that time, the corresponding water quality status is generated as "no abnormality in water quality";

[0050] When s 总-min <s 总 ≤s 总-max When this occurs, the corresponding water quality status is generated as water quality abnormality;

[0051] When s 总 >s 总-max When this occurs, a corresponding water quality status alarm is generated.

[0052] Obtain the water quality status of each main pipe and branch pipe in the flow zone. When the water quality status is normal, there is no water quality cleaning instruction. When the water quality status is abnormal, the water quality cleaning instruction is a water quality cleaning start signal. When the water quality status is an over-limit alarm, the water quality cleaning instruction is a water quality emergency treatment signal.

[0053] Step 4: Zoned water supply monitoring: Obtain the zoned flow data corresponding to each flow zone, and input the zoned flow data into the water supply monitoring mechanism to obtain the corresponding zoned water supply monitoring information. Based on the zoned water supply monitoring information of each flow zone, perform corresponding water supply regulation.

[0054] Step 5, End-of-line water quality self-cleaning feedback: Obtain water quality cleaning instructions for each flow zone and perform self-cleaning execution feedback based on these instructions. Specifically, monitor the water quality cleaning instructions for each flow zone. When a flow zone has a water quality cleaning instruction, mark the corresponding flow zone as a zone to be cleaned. Identify the water quality cleaning instructions corresponding to each zone to be cleaned and obtain the main or branch pipe information corresponding to the water quality cleaning instruction. When the water quality cleaning instruction corresponds to a water quality abnormality, generate water quality self-cleaning execution information; when the water quality cleaning instruction corresponds to an over-limit alarm, generate pipeline isolation execution information.

[0055] The self-cleaning water quality execution information includes, but is not limited to, waking up the system with low-flow water, draining stagnant water, and local sterilization and disinfection; the pipeline isolation execution information involves isolating the water flow in the corresponding pipeline through an electronic regulating valve and sending a warning message to maintenance personnel. The electronic regulating valve is used to perform pipeline isolation operations, and the maintenance warning message is used to notify personnel to handle water quality issues exceeding limits.

[0056] Please see Figure 3 The present invention also provides a flow-zoned intelligent water supply system, comprising:

[0057] The global data sensing module collects data from each flow zone of the entire water supply area through a pre-deployed global data sensing layer to obtain zone flow data, and sends the zone flow data to the data storage module. The zone flow data includes real-time pressure data, historical time series data and external impact data of each flow zone. The global data sensing layer is a sensor network covering the water supply area, used to collect real-time pressure, historical time series and external impact data.

[0058] The partition pressure analysis module acquires the partition traffic data of each traffic partition, performs partition pressure analysis on the partition traffic data to obtain the partition pressure status, and sends the partition pressure status to the pressure compensation module.

[0059] Partition pressure analysis is performed on partitioned traffic data, and the specific analysis content is as follows:

[0060] Based on the partition traffic data of each traffic partition, real-time pressure data, historical time-series data, and external impact data are obtained;

[0061] Based on real-time pressure data, the real-time pressure values ​​of the main pipe, zone nodes, and terminals of the pipeline in the flow zone are obtained. The real-time pressure value of the main pipe is calculated with the real-time pressure value of the other pipeline sections to obtain two real-time pressure difference values. Similarly, the real-time pressure difference value between the zone node and the terminal is obtained. The average pressure difference value is calculated by combining the three real-time pressure difference values. When the average pressure difference value is greater than the preset threshold, the corresponding average pressure difference value is marked as a pressure anomaly value.

[0062] Based on historical time-series data, a pressure fluctuation curve within a preset historical time period is obtained. Based on the pressure fluctuation curve, a pre-set standard fluctuation range is obtained, and the time axis of the pressure fluctuation curve is divided into multiple time-series intervals according to preset time intervals. The length of the pressure fluctuation curve in each time-series interval that exceeds the standard fluctuation range is obtained and marked as the high fluctuation curve length. The ratio of the high fluctuation curve length to the total length of the pressure fluctuation curve in the time-series interval is calculated to obtain the high fluctuation ratio value. The current time corresponding to the current flow partition is matched with each time-series interval of the corresponding historical pressure fluctuation curve to obtain the corresponding historical time-series interval, and the corresponding high fluctuation ratio value is marked as the historical fluctuation value.

[0063] Based on external impact data, the current corresponding external factors are identified. These external factors include, but are not limited to, temperature, holidays, and emergencies. A decision tree classifier is constructed using a decision tree rule weight model to label each external factor as an input variable and obtain the output variable. For example, input variables {temperature, holidays, emergencies}, output variables {abnormal pressure, normal}. For each leaf node rule... Among them, B X and B YLet be the input and output variables, respectively. Let Ci represent the condition set of the i-th rule. Calculate the rule confidence and support, and multiply the rule confidence and support to obtain the rule weight. If the current external data matches rule Ri, the outlier contribution is the rule weight; otherwise, it is zero. The external outlier wb is calculated; where α represents the rule weight, m ​​and M represent the number of matched rules and the total number of rules, respectively. The decision tree rule weight model quantifies the impact of external factors on pipeline pressure by constructing a decision tree classifier.

[0064] Each sub-anomaly value corresponding to the pressure anomaly value, historical fluctuation value, and external anomaly value of the flow zone is normalized and its value is denoted as E. y E s and E wb Substitute each sub-outlier into the entropy calculation formula. The entropy value is calculated to obtain the corresponding entropy value, and then the entropy value corresponding to each sub-outlier is substituted into the weight calculation formula. The corresponding weighting factors are obtained, and the normalized values ​​corresponding to each sub-outlier are weighted using a weighting formula. The total value of partition anomalies is calculated; when the total value of partition anomalies exceeds the preset threshold, the corresponding partition pressure status is generated as an abnormal pressure status.

[0065] The pressure compensation module receives the pressure status of each flow zone and obtains pressure compensation instructions based on the pressure status. Specifically, it monitors the pressure status of each flow zone. When the pressure status is normal, there is no corresponding pressure compensation instruction. When the pressure status is abnormal, a corresponding pressure compensation instruction is generated. The pressure compensation instruction includes pressure compensation through the energy storage compensation module of each zone node and power adjustment compensation through the water pump cluster.

[0066] The data storage module receives and stores partition traffic data, as well as partition pressure status and pressure compensation instructions. Persistent data storage provides historical data for system operation status analysis and strategy optimization, thereby improving the system's iteration capabilities.

[0067] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A flow-zoned intelligent water supply method, characterized in that, The steps include the following: Step 1: Obtain the flow zones corresponding to the water supply area. Based on the preset data acquisition strategy, obtain the zoned water supply data through the pre-deployed data sensing layer. Step 2: Obtain the water supply data for each flow zone and identify the zone water supply data to obtain zone water quality data and zone flow data. Monitor the zone water supply data through the data identification unit. When the zone water supply data corresponds to the zone water quality data, proceed to Step 3; when the zone water supply data corresponds to the zone flow data, proceed to Step 4. Step 3: Obtain zoned water quality data and transmit the zoned water quality data to the water quality monitoring mechanism. The water quality monitoring mechanism analyzes the zoned water quality data to obtain water quality cleaning instructions. Step 4: Obtain the zone flow data corresponding to each flow zone, and input the zone flow data into the water supply monitoring mechanism to obtain the corresponding zone water supply monitoring information. Based on the zone water supply monitoring information of each flow zone, carry out corresponding water supply regulation. Step 5: Obtain water quality cleaning instructions for each flow zone and perform self-cleaning execution feedback based on the water quality cleaning instructions.

2. The intelligent water supply method with flow zoned distribution according to claim 1, characterized in that, The water quality monitoring mechanism includes water quality monitoring triggering, water quality monitoring analysis, and water quality status analysis; the water quality monitoring triggering is used to obtain a water quality monitoring start signal by analyzing the water quality data of each zone; the water quality monitoring analysis is used to obtain a comprehensive anomaly value by monitoring and analyzing the water quality data of each zone; and the water quality status analysis is used to obtain a water quality cleaning instruction by analyzing the comprehensive anomaly value.

3. The intelligent water supply method with flow zoned distribution according to claim 2, characterized in that, The specific analysis process for triggering water quality monitoring is as follows: Based on the water quality data of the zone, basic physical parameters are obtained. Based on the basic physical parameters, the water flow velocity of each main pipe and branch pipe in the flow zone is obtained. The minimum threshold of the water flow velocity corresponding to each main pipe and branch pipe is obtained. When the water flow velocity is lower than the corresponding minimum threshold, the corresponding water flow is marked as a static water flow, and the start information of the timing unit is triggered. The duration of the static water flow in each main pipe and branch pipe is recorded by the timing unit to obtain the static water flow duration. The static water flow duration is multiplied by the preset static time risk coefficient to calculate the static water flow risk value. When the static water flow risk value exceeds the preset threshold, a water quality monitoring start signal is generated.

4. The intelligent water supply method with flow zoning according to claim 3, characterized in that, The water quality monitoring and analysis specifically involves: when a water quality monitoring start signal is detected in a flow zone, water quality indicators, microbial parameters, and pipeline attributes are obtained based on the corresponding zone water quality data. Based on water quality indicators, the water quality parameters are turbidity, residual chlorine and pH value. The water quality parameters are defined by constructing definition nodes through a continuous Bayesian network model. The joint probability is calculated for the definition nodes by using the conditional concept of Gaussian distribution to model continuous variables. The pre-designed probability empirical threshold is obtained. The joint probability is compared with the probability empirical threshold to output the index outliers. Microbial concentrations in each main and branch pipe are obtained based on microbial parameters. Time-series data of microbial concentrations are acquired and discretized to obtain a discrete point set. The Hough transform model is used to obtain the Hough transform formula for line detection. Data points corresponding to the discrete point set are substituted into the Hough transform formula, and accumulated voting is performed in the (ρ, θ) parameter space to generate the Hough matrix Hf(ρ, θ). The Hough matrix is ​​then identified by a recognition unit to obtain the corresponding peak points, and a pre-set peak threshold is obtained. If the peak value of a peak point exceeds the peak threshold, and the slope of the microbial concentration change at the corresponding data point exceeds the preset normal range, an abnormal judgment is generated. The deviation of the peak value corresponding to the abnormal judgment information is calculated to obtain the microbial abnormal value; The attribute information of each main pipe and branch pipe is obtained based on the pipe attributes; Based on the rough set theory model, a decision table is constructed from attribute information to obtain the pipeline attribute decision table; The dependence of conditional attributes on decision attributes is calculated using the formula of the rough set theory model; decision rules are generated based on the reduced attribute set; when new pipeline attribute data satisfies the anomaly decision rules, it is determined that the pipeline attribute is anomaly; based on the case of anomaly in the pipeline attribute, the attribute anomaly value is calculated according to the degree of rule matching. The static risk values ​​of water flow, index anomalies, microbial anomalies, and attribute anomalies are normalized and their values ​​are marked as dimensional anomalies. The improved DS evidence theory is combined with a simple feedforward neural network to perform basic probability allocation on the dimensional anomalies corresponding to each dimension. The comprehensive anomaly value is obtained by weighted synthesis calculation using the DS evidence synthesis rules.

5. The intelligent water supply method with flow zoned distribution according to claim 4, characterized in that, The water quality status analysis specifically involves: obtaining the comprehensive anomaly values ​​corresponding to each main pipe and branch pipe, as well as the corresponding preset anomaly warning thresholds and anomaly allowable values, and recording them as s. 总-min and s 总-max The current comprehensive anomaly values ​​corresponding to each main pipe and branch pipe are matched and analyzed with the anomaly warning threshold and the anomaly allowable value, respectively. When s 总 ≤s 总-min At that time, the corresponding water quality status is generated as "no abnormality in water quality"; When s 总-min <s 总 ≤s 总-max When s is present, the corresponding water quality status is generated as water quality abnormality; when s is present... 总 >s 总-max When this occurs, a corresponding water quality status alarm is generated. Obtain the water quality status of each main pipe and branch pipe in the flow zone. When the water quality status is normal, there is no water quality cleaning instruction. When the water quality status is abnormal, the water quality cleaning instruction is a water quality cleaning start signal. When the water quality status is an over-limit alarm, the water quality cleaning instruction is a water quality emergency treatment signal.

6. The intelligent water supply method with flow zoning according to claim 1, characterized in that, The self-cleaning execution feedback based on water quality cleaning instructions specifically involves: monitoring water quality cleaning instructions for each flow zone; marking the corresponding flow zone as a zone to be cleaned when a water quality cleaning instruction is present; identifying the water quality cleaning instructions corresponding to each zone to be cleaned; obtaining the main or branch pipe information corresponding to the water quality cleaning instruction; generating water quality self-cleaning execution information when the water quality cleaning instruction corresponds to a water quality abnormality; and generating pipeline isolation execution information when the water quality cleaning instruction corresponds to an over-limit alarm.

7. A flow-zoned intelligent water supply system, characterized in that, It includes a global data sensing module, a zone pressure analysis module, a pressure compensation module, and a data storage module, so that the flow zone intelligent water supply system can perform the flow zone intelligent water supply method as described in any one of claims 1-6.

8. A flow-zoned intelligent water supply system according to claim 7, characterized in that, The global data sensing module is used to collect data from each flow zone of the entire water supply area through a pre-deployed global data sensing layer to obtain zone flow data, and then send the zone flow data to the data storage module. The partition pressure analysis module is used to obtain the partition traffic data of each traffic partition, perform partition pressure analysis on the partition traffic data to obtain the partition pressure status, and send the partition pressure status to the pressure compensation module. The pressure compensation module is used to receive the pressure status of each flow zone and obtain pressure compensation instructions based on the pressure status of the zone. The data storage module is used to receive and store partition traffic data, as well as partition pressure status and pressure compensation instructions.

9. A flow-zoned intelligent water supply system according to claim 8, characterized in that, The partition traffic data is subjected to partition pressure analysis, and the specific analysis content is as follows: Based on the partition traffic data of each traffic partition, real-time pressure data, historical time-series data, and external impact data are obtained; Based on real-time pressure data, the real-time pressure values ​​of the main pipe, zone nodes, and terminals of the pipeline section in the flow zone are obtained. The real-time pressure value of the main pipe is calculated with the real-time pressure value of the other pipeline sections to obtain two real-time pressure difference values. Similarly, the real-time pressure difference value between the zone node and the terminal is obtained. The average pressure difference value is calculated by the three real-time pressure difference values. When the average pressure difference value is greater than the preset threshold, the corresponding average pressure difference value is marked as a pressure anomaly value. Based on historical time-series data, a pressure fluctuation curve within a preset historical time period is obtained. Based on the pressure fluctuation curve, a preset standard fluctuation range is obtained, and the time axis of the pressure fluctuation curve is divided into multiple time-series intervals according to preset time intervals. The length of the pressure fluctuation curve in the time-series interval that exceeds the standard fluctuation range is obtained and marked as the high fluctuation curve length. The ratio of the high fluctuation curve length to the total length of the pressure fluctuation curve in the time-series interval is calculated to obtain the high fluctuation ratio value. The current time corresponding to the current flow partition is matched with each time-series interval of the corresponding historical pressure fluctuation curve to obtain the corresponding historical time-series interval, and the corresponding high fluctuation ratio value is marked as the historical fluctuation value. Based on the external influence data, the corresponding external factors are obtained. A decision tree classifier is constructed using a decision tree rule weight model to label each external factor as an input variable and obtain the output variable. For each leaf node rule... Among them, B X and B Y Let be the input variables and the output variables, respectively. Let Ci be the condition set of the i-th rule. Calculate the rule confidence and support, and multiply the rule confidence and support to obtain the rule weight. If the current external data matches rule Ri, the outlier contribution is the rule weight; otherwise, it is zero. Calculate the external outliers. The sub-anomalies of the pressure anomaly, historical fluctuation value, and external anomaly corresponding to the traffic partition are normalized and their values ​​are taken. Each sub-anomaly is substituted into the entropy value calculation formula to calculate the corresponding entropy value. The entropy value of each sub-anomaly is then substituted into the weight calculation formula to obtain the corresponding weight factor. The normalized values ​​of each sub-anomaly are then used to calculate the total anomaly value of the partition using a weighted formula. When the total anomaly value of the partition is greater than a preset threshold, the corresponding partition pressure status is generated as an anomaly.

10. A flow-zoned intelligent water supply system according to claim 8, characterized in that, The specific analysis of obtaining pressure compensation instructions based on the pressure status of each zone is as follows: the pressure status of each flow zone is monitored. When the pressure status of the zone is normal, there is no corresponding pressure compensation instruction. When the pressure status of the zone is abnormal, a corresponding pressure compensation instruction is generated. The pressure compensation instruction includes pressure compensation through the energy storage compensation module of each zone node and power adjustment compensation of the water pump cluster.

Images