Intelligent analysis processing and decision support method for mass data of intelligent water affair integrated platform based on GIS and Internet of Things

Abstract

The invention relates to the technical field of water affair pipe network operation and maintenance, in particular to an intelligent analysis processing and decision support method for mass data of an intelligent water affair integrated platform based on a GIS (Geographic Information System) and the Internet of Things. According to the method, GIS topology and data of the internet of things are fused, the problems that the estimation deviation of the influence range of traditional pipe explosion is large, and the valve closing instruction lacks accuracy are solved, and the valve closing effectiveness is verified. The pipe explosion emergency response efficiency is improved, the water resource waste is reduced, the water supply stability is guaranteed, and the method is suitable for urban water refinement management.

Description

Technical Field

[0001] The present invention relates to the field of water pipe network operation and maintenance technology, and specifically to a method for intelligent analysis and processing of massive data and decision support based on a smart water integration platform based on GIS and the Internet of Things. Background Art

[0002] Water supply and pipeline network operation and maintenance is a critical technology. In urban water management, the stable operation of water supply networks is directly related to residents' lives and social production. Constructing the spatial topology of the pipeline network through GIS and combining it with the Internet of Things to collect real-time pressure and other data are the core technical means to achieve leakage control, pipe burst emergency response, and resource optimization.

[0003] This technology is crucial for improving water management efficiency, reducing operating costs, and ensuring water supply security. Traditional management methods that rely on manual inspections or static models are no longer able to cope with the dynamic operational demands of complex pipe networks. However, traditional smart water management platforms suffer from a core issue: insufficient emergency response accuracy in pipe network troubleshooting. Existing solutions only use a single sensor's pressure anomaly value to roughly define the impact area when a pipe burst occurs. These solutions fail to integrate GIS with the network topology and dynamic analysis of pressure wave propagation. Once a pipe burst occurs, the pressure anomaly propagates back along the pipe network. Traditional methods fail to track the temporal sequence of pressure changes at upstream nodes and fail to consider the impact of pipe material and diameter on pressure attenuation, resulting in significant errors in the impact area estimation. This also results in a lack of precise boundaries and priority for valve closing instructions. This can lead to valves not closing, causing increased leakage, or excessive valve closing affecting normal water supply. This not only wastes water resources and increases maintenance costs, but also leads to user complaints due to uncontrolled water outages, making it difficult to meet the requirements of refined urban water management. To address this technical issue, we provide a massive data intelligent analysis and processing and decision support method for an integrated smart water platform based on GIS and the Internet of Things. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for intelligent analysis and processing of massive data and decision support for a smart water management integration platform based on GIS and the Internet of Things, so as to solve the problems raised in the above background technology.

[0005] 1. Traditional solutions rely on a single sensor to define the impact range of a pipe burst, without incorporating the pipe network topology and pressure wave patterns, resulting in large deviations in range estimation. Therefore, this case reversely traces upstream nodes, constructs a pressure anomaly propagation map, and dynamically calculates a pressure wave velocity correction model to accurately determine the impact range and reduce leakage.

[0006] 2. Because traditional valve closing instructions lack precise boundaries and priorities, they are prone to missed or excessive valve closing. Therefore, this case calculates the pressure decay gradient to generate valve closing boundaries, and generates graded valve closing instructions based on connectivity weight and distance, which can optimize valve closing operations and ensure stable water supply. To achieve the above objectives, one of the objectives of the present invention is to provide a method for intelligent analysis and processing of massive data and decision support for a smart water management integration platform based on GIS and the Internet of Things, comprising the following steps: S1. Real-time monitoring of pressure sensor data. When the pressure value of the target sensor exceeds the threshold and continues to decrease within a preset period, the corresponding pipe network node is marked as a pipe burst monitoring point. Based on the coordinates and pipe diameter data of this point in the GIS topology network, an initial impact domain is generated. S2: Starting from the burst monitoring point, trace upstream nodes along the pipe network water flow in reverse within the initial impact domain, extract time series data from associated sensors within the domain, and dynamically calculate the pressure wave velocity and modify the hydraulic model by comparing the time difference between the burst monitoring point and the upstream sensor pressure anomaly, combined with the spatial length of the pipe section GIS. This constructs a pressure anomaly propagation map in the spatiotemporal dimension. S3. Based on the pressure anomaly propagation map, the first sensor that detects the pressure anomaly is identified as the attenuation calculation origin. The pressure attenuation gradient value is calculated layer by layer along the pipeline network topology path. When the pressure attenuation gradient value decays to a preset critical threshold, a valve closing boundary range is automatically generated at the corresponding pipeline network topology location. Within the valve closing boundary range, a hierarchical valve closing instruction set is generated based on the valve's connectivity weight in the GIS topology network and its spatial geometric distance from the attenuation calculation origin. S4. After executing the graded valve closing instruction set, monitor the change in the recovery rate of the residual pressure value of the target sensor in the initial impact domain in real time, and verify the validity of the valve closing boundary range by analyzing the degree of deviation between the rate and the hydraulic model prediction reference curve.

[0007] Compared with the prior art, the present invention has the following beneficial effects: The present invention takes the burst pipe monitoring point as the starting point to reversely trace the upstream node, combines the pressure anomaly time difference and the spatial length of the pipe section to dynamically calculate the pressure wave velocity, and modifies the hydraulic model to generate a time-space propagation map, which can clearly present the diffusion path of the pressure anomaly along the pipe network. At the same time, the segmented compensation adjustment of the diffusion radius based on the pipe diameter and material solves the range estimation deviation caused by the traditional single sensor data, provides an accurate spatial boundary reference for the subsequent valve closing operation, reduces the risk of leakage expansion or excessive valve closing, and determines the valve closing boundary range through the pressure attenuation gradient value. Combined with the valve connection weight and the spatial distance coefficient, the operation priority is calculated to divide the valve into groups. It is divided into three operating levels to ensure that key valves are closed first to avoid aggravated leakage caused by leaky valves, while reducing the impact of unnecessary valve closures on normal water supply, balancing fault control and people's water needs. After closing the valve, the pressure recovery rate is monitored and the degree of deviation from the predicted curve is monitored to verify the effectiveness of the valve closing boundary. If valve closing is invalid, the critical threshold is automatically lowered to recalculate the boundary and generate supplementary instructions. At the same time, the historical parameter library is updated to realize model self-learning. This closed-loop optimization process enhances the system's adaptability to complex pipeline network conditions, ensures the continued effectiveness of decision-making, reduces water resource waste and maintenance costs, and improves the level of refined management of urban water affairs. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 It is the overall workflow diagram of the present invention. DETAILED DESCRIPTION

[0009] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0010] See also Figure 1 As shown, this embodiment provides a method for intelligent analysis and processing of massive data and decision support for a smart water integration platform based on GIS and the Internet of Things, including the following steps: S1. Real-time monitoring of pressure sensor data. When the pressure value of the target sensor exceeds the threshold and continues to decrease within a preset period, the corresponding pipe network node is marked as a pipe burst monitoring point. Based on the coordinates and pipe diameter data of this point in the GIS topology network, an initial impact domain is generated. Generating the initial impact domain includes: After marking the burst monitoring point, an initial impact domain must be generated to accurately define the immediate impact range of the initial burst. This process combines GIS pipe network topology and pipeline attributes to achieve refined segmentation. Based on the pipe diameter data of the burst monitoring point, a preset pipe diameter-diffusion radius mapping table is queried. This table, constructed based on historical pipe burst data and hydraulic simulation results, determines the basic diffusion radius and defines a preliminary range for the impact domain to ensure that the diffusion range matches the pipeline's transportation capacity. All hydraulically connected pipe sections centered on the monitoring point are extracted from the GIS pipe network topology. When extending along the pipeline to a tee or cross-way branch node, the boundary must be automatically segmented to form polygon vertices. The automatic segmentation uses the "branch priority segmentation algorithm": When extending to a branch node, first identify the diameter of each branch pipe at the node. The main pipe diameter is larger than the branch pipe. It is preferentially extended along the main pipe direction by 1.5 times the basic diffusion radius. Then, the extension length is allocated to each branch pipe according to the pipe diameter ratio. For example, the main pipe is extended by 5 meters, and the branch pipe is extended by 4 meters according to the pipe diameter ratio of 0.8. The vertex is marked at each extension end and branch node, and the node type is recorded at the same time. The tee is recorded as a T-type vertex and the cross is recorded as a cross-type vertex. By dynamically allocating the extension distance according to the importance of the pipeline, the boundary segmentation is more in line with the actual hydraulic impact path of the pipeline network, avoiding the range distortion caused by simple equal segmentation. Afterwards, the basic diffusion radius is segmented and compensated according to the permeability coefficient corresponding to the pipe segment material type. The specific process is as follows: The connecting pipe section is divided into cast iron pipe, steel pipe, PE pipe and other sections according to the material, and the permeability coefficient of each section is retrieved from the material-permeability coefficient comparison table. The larger the permeability coefficient, the worse the pipeline sealing is, and the easier it is to spread the impact of the pipe burst. The compensation coefficient is calculated = (permeability coefficient of the section / permeability coefficient of the cast iron pipe) × 0.3 + 1 (based on the cast iron pipe, the compensation coefficient range is 1.0-1.5). The basic diffusion radius is multiplied by the compensation coefficient to obtain the actual diffusion radius of the section. For example, the permeability coefficient of the PE pipe is 1.2 times that of the cast iron pipe, the compensation coefficient is 1.06, and the actual diffusion radius = basic diffusion radius × 1.06. Through segmented adjustment, the diffusion range can reflect the differences in leakage characteristics of pipelines of different materials, thereby improving the impact. To improve the accuracy of the domain, finally, all vertices are connected to form a closed polygon. With the burst monitoring point as the center, the boundary distance of each pipe section is determined according to the adjusted actual diffusion radius. The vertices at the branch nodes and the extension end points are connected according to the pipe network topology order. For non-straight pipe sections, Bezier curves are used to smoothly connect adjacent vertices to form an irregular polygon covering the direct impact area of ​​the burst pipe. This is marked as the initial impact domain in the GIS system. This polygon includes the direct leakage area around the burst point and reflects the constraints of the pipe network structure on the impact range through branch segmentation and material compensation. It provides an accurate spatial benchmark for subsequent pressure anomaly propagation analysis and effectively avoids the range estimation deviation caused by ignoring pipe network details in traditional methods.

[0011] S2: Starting from the burst monitoring point, trace upstream nodes along the pipe network water flow in reverse within the initial impact domain, extract time series data from associated sensors within the domain, and dynamically calculate the pressure wave velocity and modify the hydraulic model by comparing the time difference between the burst monitoring point and the upstream sensor pressure anomaly, combined with the spatial length of the pipe section GIS. This constructs a pressure anomaly propagation map in the spatiotemporal dimension. Reverse tracing of upstream nodes specifically includes: After generating the initial impact domain, in order to track the propagation path of pressure anomalies in the pipe network, it is necessary to reversely trace the upstream nodes from the burst monitoring point. This process achieves precise positioning by combining GIS topology and sensor data, and loads the preset water flow direction arrow data in the GIS pipe network topology network. These arrows are based on the water supply direction preset during pipe network design, and clearly mark the path of water flow from the water plant to the user. Starting from the burst monitoring point, follow the arrow in the opposite direction to identify the directly connected upstream nodes, such as the previous valve or three-way node connected to the monitoring point is the first hop upstream node, and the node connected to the node before it is the second hop, and so on. In this way, the upstream area that may be affected can be traced back in the opposite direction of the water flow, which provides clear information for subsequent pressure anomaly analysis. Confirm the scope and verify the hydraulic connectivity between nodes within the boundary of the initial impact domain. Hydraulic connectivity means that there is a continuous pipeline path that can transmit water flow between two nodes, and the pipeline is in a normal water flow state and is not completely closed or broken. During verification, first retrieve the pipeline connection relationship between nodes from the GIS topology database to confirm the existence of physical connection, and then query the real-time switch status of the valve on the connecting pipe section through the Internet of Things platform. Fully open or half-open is considered connected, and fully closed is considered blocked. At the same time, combined with historical hydraulic model data, determine whether the pipe section has lost its flow capacity due to long-term siltation. Only when the three conditions of physical connection, valve not fully closed, and normal flow capacity are met at the same time, can it be determined that there is hydraulic connectivity between nodes. This refinement Verification can avoid including blocked nodes in the tracking range, ensuring that the nodes for subsequent analysis are all valid nodes that actually participate in pressure propagation. After determining the valid upstream nodes, all pressure sensor devices installed on these nodes are screened out. These sensors collect pressure data in the pipe section in real time and are key devices for capturing pressure anomalies. Through the IoT data bus, the data transmission channel used to connect each sensor to the platform supports real-time data retrieval, and the original sampling data sequence of the selected sensor within the pressure anomaly time window is retrieved in real time. The time window is set from 10 minutes before the first pressure anomaly at the burst monitoring point to the current time, ensuring coverage of the complete process before and after the anomaly occurs. The first recorded pressure value of each sensor falling below The timestamp of the safety threshold is set according to the normal operating pressure range of the pipe section, and it is used as the key time series mark point. These mark points can accurately reflect the time when the pressure anomaly reaches each upstream node, providing a time series basis for the subsequent calculation of the pressure wave velocity. By loading the water flow direction data to determine the tracking path, refining and verifying the hydraulic connectivity to ensure the validity of the node, retrieving the sensor data and extracting the time series mark points, the reverse tracking process not only relies on the GIS topology to clarify the spatial association, but also uses the sensor time series data to capture the pressure changes in the time dimension, laying a dual time and space foundation for constructing the pressure anomaly propagation map, making the subsequent pressure wave velocity calculation and model correction more accurate, and effectively making up for the limitation of the traditional method that relies only on a single node data.

[0012] Dynamic calculation of pressure wave velocity includes: After completing the reverse tracing of the upstream nodes and extracting the key timing mark points, in order to accurately grasp the propagation speed of the pressure anomaly in the pipe network, it is necessary to dynamically calculate the pressure wave velocity. This process realizes quantitative analysis by combining the time difference and spatial distance, and calculates the difference between the pressure anomaly timestamps of the burst monitoring point and each upstream sensor as the pressure wave transmission delay time. Specifically, the timestamp of the first pressure anomaly at the burst monitoring point (denoted as T0) and the timestamp of the first pressure anomaly at each upstream sensor (denoted as Ti, i represents different upstream sensors) are extracted from the key timing mark points. Since the pressure wave propagates from the burst point to the upstream, Ti must be later than T0, so This conduction delay is the difference between Ti and T0. For example, if T0 is 10:00:00 and the Ti of an upstream sensor is 10:00:05, the delay is 5 seconds. This calculation can intuitively reflect the time required for the pressure wave to propagate from the pipe burst point to each upstream node, providing a quantitative basis for the time dimension for subsequent wave velocity calculations. The three-dimensional broken line path length of the pipe section between the pipe burst monitoring point and the corresponding upstream node is retrieved from the GIS spatial database as the spatial reference distance. This length is the sum of the three-dimensional paths calculated based on the actual direction of the pipeline network, including horizontal turns and vertical elevation changes, rather than a straight-line distance, which can more truly reflect the actual propagation path length of the pressure wave. Dividing the spatial reference distance by the propagation delay yields the measured pressure velocity for that pipe segment. For example, if the path length is 500 meters and the delay is 5 seconds, the measured velocity is 100 meters per second. The measured pressure velocity is then compared with the velocity previously preset in the hydraulic model, and the deviation rate is calculated (deviation rate = |measured velocity - preset velocity| ÷ preset velocity × 100%). When the deviation rate exceeds the set tolerance threshold, the preset velocity no longer accurately reflects the current pipe network conditions. In this case, the hydraulic model's velocity parameters are updated using a time-weighted moving average algorithm. This algorithm assigns a time-weighted weight to the five most recent measured velocity values, with more recent values ​​receiving a higher weight, such as the most recent value receiving a weight of 0. 3, the last time was 0.25, and so on. The weighted average value is calculated as the new wave velocity parameter, and the correction time, wave velocity values ​​before and after correction, deviation rate and other information are synchronously recorded in the correction log to trace the model adjustment process. The delay duration is obtained by accurately calculating the time difference, the measured wave velocity is obtained by combining the three-dimensional path length, and the model parameters are dynamically corrected according to the deviation rate. This process not only quantifies the propagation characteristics of the pressure wave, but also ensures the consistency between the hydraulic model and the actual operating status of the pipeline network through real-time correction. It provides reliable wave velocity data support for the subsequent construction of the pressure anomaly propagation map, effectively improves the accuracy of the pressure anomaly propagation analysis, and avoids the prediction deviation caused by the solidification of model parameters.

[0013] The construction of pressure anomaly propagation map includes: After dynamically calculating the pressure wave velocity and correcting the hydraulic model, in order to intuitively present the propagation law of pressure anomalies in the pipe network, it is necessary to construct a pressure anomaly propagation map. This map can integrate spatial topology and time series characteristics to provide a visualization basis for subsequent analysis. The GIS pipe network topology structure is used as the spatial coordinate system base, that is, the existing pipe network spatial coordinate system in the GIS system is used as the spatial reference frame of the map to ensure that the spatial position of the map is completely consistent with the actual pipe network. This will enable the map to accurately reflect the spatial path of pressure anomaly propagation and provide a reliable benchmark for spatial topology analysis. The spatial geographic coordinates of each pressure sensor (obtained from the sensor's metadata, including specific latitude and longitude information) are used as the spatial coordinate system of the pressure sensor. ) is directly mapped to the node position in the atlas. Each node represents the pipe network node where a pressure sensor is located. The size of the node can be displayed differently according to the monitoring accuracy level of the sensor. This operation can clearly mark the key points involved in pressure anomaly monitoring, which is convenient for quickly locating the position of each sensor in the atlas. The direction of pressure anomaly conduction is represented by the time-series connection line with an arrow. Since the pressure anomaly propagates upstream along the pipe network, the direction of the arrow uniformly points to the upstream sensor node. At the same time, the thickness of the connection line is dynamically adjusted according to the size of the pressure wave velocity, so as to intuitively reflect the difference in propagation speed. Such a design can clearly show the conduction path and direction of pressure anomaly in space. , the thickness of the combined line can quickly determine the characteristics of the propagation speed, and embed dynamic parameters such as the percentage of pressure drop, the conduction delay time in seconds, and the measured wave velocity value of the pipe section in the node attribute table. The percentage of pressure drop is (initial pressure - minimum pressure) ÷ initial pressure × 100%, the conduction delay time is the difference between Ti and T0 calculated in the previous article, and the measured wave velocity value of the pipe section is the ratio of the spatial reference distance to the delay time. The embedding of these parameters can make each node not only a spatial marker, but also carry rich timing and quantitative information, providing data support for timing backtracking analysis. At the same time, the pipe section material code and pipe diameter size value are marked in the connection line attributes. This information is directly related The physical properties of pipe sections can help analyze the impact of material and pipe diameter on pressure anomaly propagation and enhance the analytical depth of the map. Through the above steps, the final pressure anomaly propagation map is a multidimensional dynamic network model that supports spatial topological analysis and time series backtracking. It not only retains the spatial topological relationship of the GIS pipe network, but also incorporates the time series parameters and pipe section properties of pressure anomaly propagation. Detailed parameters can be viewed by clicking on nodes or connecting lines, and the pressure anomaly propagation status at different times can be traced back by sliding the time axis. This provides an intuitive and comprehensive visualization tool for subsequent determination of the attenuation calculation origin and analysis of pressure attenuation laws, effectively improving the efficiency and accuracy of pressure anomaly propagation analysis.

[0014] S3. Based on the pressure anomaly propagation map, the first sensor that detects the pressure anomaly is identified as the attenuation calculation origin. The pressure attenuation gradient value is calculated layer by layer along the pipeline network topology path. When the pressure attenuation gradient value decays to a preset critical threshold, a valve closing boundary range is automatically generated at the corresponding pipeline network topology location. Within the valve closing boundary range, a hierarchical valve closing instruction set is generated based on the valve's connectivity weight in the GIS topology network and its spatial geometric distance from the attenuation calculation origin. Calculating the pressure decay gradient value includes: After constructing the pressure anomaly propagation map, the pressure attenuation gradient value needs to be calculated to accurately determine the attenuation boundary of the pressure anomaly. This process is achieved by combining the pipe network topology and pipe segment properties to achieve quantitative analysis. The attenuation calculation origin is identified from the pressure anomaly propagation map. This origin is the first sensor node in the map that detects the pressure anomaly, that is, the starting point of the pressure anomaly propagation. Starting from this origin, the adjacent nodes are expanded layer by layer along the pipe network topology path. Specifically, the "breadth-first hierarchical traversal method" is used: The origin of the attenuation calculation is set as the 0th layer, and all the adjacent nodes directly connected to it are traversed and marked as the 1st layer. Then, starting from the 1st layer node, the unmarked adjacent nodes of each node are traversed and marked as the 2nd layer, and so on, until the initial influence domain boundary is expanded. The nodes of each layer are sorted in ascending order according to the topological distance from the origin. This hierarchical expansion method can clearly reflect the hierarchical relationship of the pressure anomaly propagating from the origin to the outside, and provide an ordered node sequence for subsequent gradient calculation. When expanding to a new node, the material type of the pipe section to which the node belongs is first queried, and the preset attenuation coefficient benchmark table is called. The table presets the basic attenuation coefficient based on the pressure conduction characteristics of different materials, and then combines the pipe diameter size compensation coefficient and the service life reduction factor to calculate the absolute value of the attenuation of the current layer. The pipe diameter size compensation coefficient is calculated as follows: Taking the standard pipe diameter (200mm) as the benchmark, take the square root of the ratio of the actual pipe diameter to the standard pipe diameter (the compensation coefficient for a pipe diameter of 300mm is √(300 / 200)≈1.22). The larger the pipe diameter, the slower the pressure decay and the larger the compensation coefficient. The service life reduction factor is calculated as: 1-(service life / design service life)×0.3 (for example, for a pipeline with a design service life of 30 years, the reduction factor after 10 years of use is 1-(10 / 30)×0.3=0.9). The longer the service time, the faster the pressure decay due to pipeline aging, and the smaller the reduction factor. The influence of pipe diameter and service life is quantified as a dynamic coefficient rather than a fixed value, so that the attenuation calculation is more in line with the actual state of the pipeline. The final absolute value of attenuation = basic attenuation coefficient × pipe diameter size compensation coefficient × service life reduction factor × pipe section length, where the pipe section is the length of the pipeline between the current node and the previous level node. For example, a pipe section is a cast iron pipe (basic coefficient 0.05), a pipe diameter of 300mm (compensation coefficient 1.22), a service life of 10 years (reduction factor 0.9), and a length of 50 meters. The absolute value of attenuation = 0.05×1.22×0.9×50≈2.75. After calculating the attenuation of the current level, the gradient value of the previous node is subtracted from the attenuation as the gradient value of the current node. Initially, the gradient value of the origin of the attenuation calculation is set to the initial amplitude of the pressure anomaly, representing the maximum amplitude of the pressure drop. The gradient value of the first-layer node = the gradient value of the origin - the attenuation of the first layer , the gradient value of the second-layer node = the gradient value of the first-layer node - the attenuation of the second layer, and so on. The gradient value of each node is calculated in sequence according to the level, and then these gradient values ​​are connected in the order of the pipeline network topology path to form a gradient value curve continuously distributed along the pipeline network path. The horizontal axis is the topological distance from the origin, and the vertical axis is the gradient value. Finally, the gradient values ​​of all nodes are stored in the attribute set of the GIS topology node to facilitate subsequent query and analysis. By traversing the nodes according to the level to clarify the calculation order, combining the dynamic coefficient to accurately calculate the attenuation, and deducing the gradient curve layer by layer, the calculation process of the pressure attenuation gradient value not only reflects the attenuation law of the pressure anomaly with the propagation distance, but also incorporates the influence of actual factors such as pipeline material, pipe diameter, and service life, so that the gradient value can truly reflect the propagation attenuation characteristics of the pressure anomaly, providing a quantitative basis for the subsequent generation of the valve closing boundary, and effectively improving the accuracy of the boundary division.

[0015] Generate valve closing boundary range including: After calculating the pressure decay gradient value and forming a continuously distributed gradient curve, the valve closing boundary range needs to be generated to determine the effective valve closing operation range. This process achieves precise division by tracking the gradient value changes and combining the pipeline network topology characteristics. First, the pipeline network node where the pressure decay gradient value first falls below the preset critical threshold is located from the gradient curve. The critical threshold is set based on the pipeline network's safe operation pressure gradient. A value below this value indicates that the impact of the pressure anomaly has been reduced to a controllable range. The boundary starting anchor point is set at this node location. The specific setting process is as follows: The spatial coordinates of the node are marked through the GIS system, and the material, diameter and other properties of the pipe section to which it belongs are recorded. A "starting anchor point" label is added to the node in the topological network. The logic of the setting is that the node where the gradient value is lower than the critical threshold for the first time is the turning point where the pressure anomaly impact changes from significant to weak. Using this as the boundary starting point can ensure that the valve closing range covers the affected area while avoiding excessive expansion. The search continues from the starting anchor point along the topological path toward the upstream direction of the water flow. Because the pressure anomaly propagates upstream, there may still be affected areas upstream. The gradient value changes of the nodes along the way are monitored in real time. When the gradient value of a node is detected to rise above the critical threshold, the boundary termination anchor point is set at the node. The setting method is the same as the starting anchor point, and the coordinates are marked in the same way. Attributes are added and the "end anchor point" tag is added. All pipe network paths between the start anchor point and the end anchor point are connected. Specifically, all connected pipe segments between the two points are extracted through GIS topology analysis functions. The endpoints of these pipe segments are connected in sequence to form a closed polygon, that is, a hydraulic isolation closed loop. The function of this closed loop is to form a hydraulic isolation area by closing the valves within the loop, preventing the pressure anomaly from spreading further. Its formation logic utilizes the connectivity of the pipe network topology to ensure that the closed loop can completely surround the upstream area affected by the pressure anomaly. When the closed loop passes through the pipe network ring structure, such as the circulation path in the ring pipe network, in order to reduce the impact of valve closure on normal water supply, the optimal segmentation algorithm is automatically activated to select the segmentation path with the least interruption to the main pipe. This algorithm introduces the "main pipe priority index": First, identify the main pipe levels of all pipe sections in the closed loop (determined by pipe diameter, water transport volume, etc., for example, pipes with a diameter ≥ 500mm and that are responsible for the main water supply of the area are first-level main pipes, 300-500mm are second-level, and below are branch pipes). Assign a priority index of 3 to the first-level main pipe, 2 to the second-level, and 1 to the branch pipe. Traverse all possible segmentation paths of the closed loop through the ring structure and calculate the sum of the priority indexes of the main pipes interrupted on each path. A path interrupts one first-level main pipe and one second-level main pipe, and the sum is 3+2=5. Select the path with the smallest sum as the optimal segmentation path and integrate the optimal segmentation path with the original closed loop. The final valve closing boundary range is generated, that is, the boundary line of the closed loop is updated in the GIS system to extend along the optimal segmentation path, ensuring that all valves that need to be closed are included in the boundary range, while minimizing the impact on the main pipeline. By setting the start and end anchor points to determine the boundary range, connecting the paths to form a closed loop for isolation, and using the optimal segmentation algorithm to optimize the boundary path in the ring pipe network, the generated valve closing boundary range can not only accurately surround the area affected by the pressure anomaly, but also reduce the interference with the normal water supply, providing a scientific spatial basis for the subsequent generation of the graded valve closing instruction set, and effectively improving the accuracy and feasibility of the valve closing operation.

[0016] Connectivity weight calculation includes: After generating the valve closing boundary range, in order to clarify the priority of the valve closing operation, it is necessary to calculate the valve connectivity weight. This weight comprehensively reflects the importance of the valve in the pipeline network and provides a quantitative basis for graded valve closing. In the GIS topology network, the number of pipelines directly connected to each valve device is counted as the basic connectivity value. The basic connectivity value reflects the control range of the valve on the pipeline network branch. The more pipelines connected, the more significant the valve's hub role in the pipeline network topology. The specific process is as follows: Through the topological analysis function of the GIS system, the position of each valve in the pipe network is located, and the directly connected pipes are traversed, excluding indirectly connected pipes. The number of pipes is counted. For example, if a valve is directly connected to three pipes, the basic connectivity value is 3. This value intuitively reflects the physical connection scale of the valve and lays the foundation for subsequent weight calculation. The hierarchical depth of the pipe network partition where the valve is located in the global topology is identified and an exponential weighting coefficient is assigned. The hierarchical depth reflects the importance of the valve's position in the water supply path. The closer to the water plant, the deeper the level, and the greater the impact on the overall water supply. The specific process is as follows: Taking the water supply starting point of the water plant as the root node, the network is divided into levels according to the topological structure. For example, the root node is the first level, and the partition to which the pipe directly connected to the root node belongs is the second level. And so on. The path analysis function of the GIS system is used to determine the level depth of the partition where each valve is located (denoted as n). Then, the coefficient is calculated according to the formula "exponential weighting coefficient = 2^(n-1)" (for example, the weighting coefficient of the third level is 2^(3-1)=4). The deeper the level, the faster the coefficient grows, thereby highlighting the importance of deep-level valves. This exponential growth design can more accurately reflect the different impacts of different levels on the global water supply and avoid the insufficient distinction of importance caused by linear weighting. Then, the number of water meters affected downstream after the valve is closed is calculated as the impact factor. The impact factor reflects the scope of the impact of the valve closing on the user's water use. The larger the number, the more cautious the valve should be closed. The specific process is as follows: Through GIS topology, all connected pipe sections downstream of the valve are traced, the water meters connected to these sections are located, and the total number of water meters is counted (for example, if there are 500 water meters downstream of a valve, the impact factor is 500). Water meters are weighted according to water use type (for example, industrial water meters have a weight of 1.5, and residential water meters have a weight of 1). The weighted total number of water meters is calculated as the final impact factor (for example, 300 residential water meters + 200 industrial water meters, the impact factor = 300 × 1 + 00 × 1.5 = 600). This makes the impact factor more consistent with the actual water use impact. The basic connectivity value, the hierarchical weighting coefficient, and the impact factor are integrated through a linear normalization algorithm to generate a standardized comprehensive connectivity weight value. The specific process is as follows: The three indicators are normalized separately (i.e. (indicator value - minimum value) ÷ (maximum value - minimum value), so that the numerical range is unified to 0-1), and then the comprehensive value is calculated according to the weight ratio of "basic connectivity value × 0.3 + hierarchical weighting coefficient × 0.4 + impact factor × 0.3". The weight ratio is set according to the importance of the indicator to the valve closing decision. The higher the obtained comprehensive connectivity weight value, the higher the priority of the valve in the valve closing operation. The physical connection scale is reflected by the statistical basic connectivity value, the global impact is highlighted by assigning an exponential coefficient according to the hierarchical depth, and the user impact is reflected by combining the weighted number of water meters. The comprehensive weight is then obtained through normalization and fusion. This process comprehensively considers the topological position and actual impact of the valve, makes the division of valve closing priority more scientific and reasonable, provides a reliable basis for the subsequent generation of accurate hierarchical valve closing instruction sets, and effectively balances the needs of fault control and user water supply guarantee.

[0017] Generating a staged valve closing instruction set includes: After calculating the comprehensive connectivity weight value of each valve, in order to ensure that the valve closing operation can quickly control the impact of pipe bursts and minimize interference with normal water supply, it is necessary to generate a hierarchical valve closing instruction set within the valve closing boundary range, traverse all valve devices within the valve closing boundary range, and give priority to selecting the valve with the highest comprehensive connectivity weight value as the first-level valve closing target. This is because the valve with a higher comprehensive connectivity weight value plays a more significant hub role in the pipe network topology, controls more pipeline branches, and has a greater impact on the global water supply. Prioritizing the closure of such valves can quickly cut off the key path for the propagation of pressure anomalies, curb the expansion of leakage from the source, and buy time for subsequent valve closing operations. Taking the spatial coordinates of the attenuation calculation origin as the benchmark, the Euclidean distance of the latitude and longitude coordinates of each valve within the boundary is calculated and normalized into a distance coefficient. The specific process is as follows: The latitude and longitude coordinates of the attenuation calculation origin (denoted as (x0, y0)) and the latitude and longitude coordinates of each valve (denoted as (xi, yi)) are extracted from the GIS system. The straight-line distance between each valve and the origin is calculated using the Euclidean distance formula. The maximum distance (denoted as Dmax) and minimum distance (denoted as Dmin) from the origin are then found among all valves. Normalization is performed using the "distance coefficient = (distance - Dmin) / (Dmax - Dmin)" formula, so that the distance coefficient ranges from 0 to 1. The closer the distance to the origin, the closer the coefficient is to 0; the farther the distance, the closer it is to 1. This coefficient reflects the spatial association between the valve and the source of the pressure anomaly, providing a spatial dimension reference for the priority score. The valve operation priority score is calculated based on the decision model of "comprehensive connectivity weight × priority factor A + distance coefficient × priority factor B", where priority factors A and B are set according to the valve closing strategy. The specific operation process is as follows: Multiply each valve's comprehensive connectivity weight (normalized, range 0-1) by factor A, multiply the distance coefficient by factor B, and then add the two products to obtain a priority score (for example, if a valve's comprehensive connectivity weight is 0.9 and its distance coefficient is 0.2, then the score = 0.9 × 0.7 + 0.2 × 0.3 = 0.63 + 0.06 = 0.69). This score combines the valve's topological importance and spatial distance, making the priority ranking more consistent with actual valve closing requirements. Valves are divided into three operating levels in descending order of score. The specific process is as follows: All valves are ranked by priority from high to low, the total score is calculated, and the grading thresholds are determined at a ratio of 30%, 50%, and 20% (the top 30% of valves with the total score are level one, the middle 50% are level two, and the bottom 20% are level three). The level one valve closing target is the valve with the highest score, and the closing operation must be performed first to quickly control the core impact area. The level two valve has the second highest score and is closed in sequence after the level one valve is closed to further narrow the scope of impact. The level three valve has the lowest score and can be closed in the final stage or flexibly adjusted according to actual conditions. This grading method can ensure that key valves are operated first, while avoiding inefficiency caused by chaotic valve closing sequence, making valve closing instructions more organized and executable, effectively improving the accuracy and timeliness of pipe burst emergency response, and further balancing the needs of fault control and water supply stability.

[0018] S4. After executing the graded valve closing instruction set, monitor the changes in the recovery rate of the residual pressure value of the target sensor in the initial impact domain in real time. By analyzing the degree of deviation between this rate and the hydraulic model prediction benchmark curve, the effectiveness of the valve closing boundary range is verified.

[0019] In the integrated smart water management platform based on GIS and the Internet of Things, each step, from marking pipe burst monitoring points to generate the initial impact domain, to reversely tracing upstream nodes, constructing a pressure anomaly propagation map, and finally calculating the pressure decay gradient and generating the valve shutoff boundary range, is closely linked, forming a logically coherent pipe burst emergency response process. The generation of the initial impact domain defines the base range for subsequent analysis, while reverse tracing upstream nodes within this range traces the propagation path of the pressure anomaly in the opposite direction of the water flow. By verifying the hydraulic connectivity between nodes, it ensures that all tracked nodes are valid nodes that actually participate in pressure propagation. This lays the foundation for accurately capturing the temporal characteristics of pressure anomalies. Dynamically calculating pressure wave velocity and correcting the hydraulic model combines pressure anomaly propagation data in the temporal dimension with pipe length data in the spatial dimension, making the model more accurate to the actual operating status of the pipeline network and providing reliable wave velocity parameters for constructing the pressure anomaly propagation map. The pressure anomaly propagation map integrates the spatial information of the GIS topology with the temporal data of pressure anomalies, intuitively demonstrating the propagation patterns of pressure anomalies in the pipeline network and providing a clear visualization basis for determining the origin of pressure decay calculations. On this basis, the pressure attenuation gradient value is calculated. By traversing the nodes layer by layer and quantifying the attenuation in combination with the pipe section attributes, the resulting gradient curve can accurately reflect the attenuation trend of the pressure anomaly, which naturally leads to the generation of the valve closing boundary range. Anchor points are set based on the critical change of the gradient value, and the connection path forms a hydraulic isolation closed loop to ensure that the valve closing range can cover the affected area while reducing interference with normal water supply. After the valve closing boundary range is determined, calculating the valve connectivity weight becomes the key. By comprehensively considering the physical connection scale of the valve, the hierarchical depth in the pipeline network, and the impact on users, a quantitative standard is provided for the division of valve closing priorities. Finally, a hierarchical valve closing instruction set is generated based on the connectivity weight and distance coefficient, so that the valve closing operation can be executed in order according to priority, from quickly controlling the core affected area to gradually narrowing the scope, step by step, effectively improving the emergency response efficiency and processing accuracy of the smart water platform to pipe burst incidents.

[0020] The present invention marks pipe burst monitoring points and generates an initial impact domain, reversely traces upstream nodes, constructs a pressure anomaly propagation map and modifies the hydraulic model, calculates the pressure decay gradient to generate the valve closing boundary, generates a graded valve closing instruction set, verifies the valve closing effectiveness and dynamically optimizes it. This method integrates GIS topology and Internet of Things data to solve the problems of large deviations in traditional pipe burst impact range estimation and lack of accuracy in valve closing instructions, improves the efficiency of pipe burst emergency response, reduces water resource waste, ensures water supply stability, and is suitable for refined management of urban water services.

[0021] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely preferred examples of the present invention and are not intended to limit the present invention. Various changes and improvements may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and improvements fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent analysis and processing of massive data and decision support based on a GIS and IoT-based smart water management integrated platform, characterized by: The following steps are involved: S1. Real-time monitoring of pressure sensor data. When the pressure value of the target sensor exceeds the threshold and continues to decrease within a preset period, the corresponding pipe network node is marked as a pipe burst monitoring point. Based on the coordinates and pipe diameter data of this point in the GIS topology network, an initial impact domain is generated. S2: Starting from the burst monitoring point, trace upstream nodes along the pipe network water flow in reverse within the initial impact domain, extract time series data from associated sensors within the domain, and dynamically calculate the pressure wave velocity and modify the hydraulic model by comparing the time difference between the burst monitoring point and the upstream sensor pressure anomaly, combined with the spatial length of the pipe section GIS. This constructs a pressure anomaly propagation map in the spatiotemporal dimension. S3. Based on the pressure anomaly propagation map, the first sensor that detects the pressure anomaly is identified as the attenuation calculation origin. The pressure attenuation gradient value is calculated layer by layer along the pipeline network topology path. When the pressure attenuation gradient value decays to a preset critical threshold, a valve closing boundary range is automatically generated at the corresponding pipeline network topology location. Within the valve closing boundary range, a hierarchical valve closing instruction set is generated based on the valve's connectivity weight in the GIS topology network and its spatial geometric distance from the attenuation calculation origin. S4. After executing the graded valve closing instruction set, monitor the change in the recovery rate of the residual pressure value of the target sensor in the initial impact domain in real time, and verify the validity of the valve closing boundary range by analyzing the degree of deviation between the rate and the hydraulic model prediction reference curve.

2. The method for intelligent analysis and processing of massive data and decision support for the smart water integration platform based on GIS and the Internet of Things according to claim 1 is characterized in that: Generating the initial impact domain includes: Based on the pipe diameter data of the pipe burst monitoring point, the preset pipe diameter-diffusion radius mapping table is queried to determine the basic diffusion radius. All hydraulically connected pipe sections centered at the monitoring point are extracted from the GIS pipe network topology. When extending along the pipeline to the three-way or four-way branch node, the boundary is automatically segmented to form polygon vertices. The basic diffusion radius is compensated and adjusted in sections based on the permeability coefficient corresponding to the pipe segment material type. Finally, an irregular polygonal influence domain covering the direct impact area of ​​the pipe burst is generated and marked as the initial influence domain.

3. The method for intelligent analysis and processing of massive data and decision support for the smart water integration platform based on GIS and the Internet of Things according to claim 1 is characterized in that: The reverse tracing of the upstream node specifically includes: Load the preset water flow direction arrow data in the GIS pipe network topology network, identify the directly connected upstream nodes hop by hop starting from the pipe burst monitoring point, verify the hydraulic connectivity between nodes within the initial impact domain boundary, screen out all pressure sensor devices with continuous water flow paths to the monitoring point, and retrieve the original sampling data sequence of the selected sensor within the pressure anomaly time window in real time through the Internet of Things data bus. Extract the timestamp of the first time the pressure value recorded by each sensor falls below the set safety threshold as the key timing mark point.

4. The method for intelligent analysis and processing of massive data and decision support for the smart water integration platform based on GIS and the Internet of Things according to claim 1 is characterized in that: The dynamic calculation of pressure wave velocity includes: The difference between the pressure anomaly timestamps of the pipe burst monitoring point and each upstream sensor is calculated as the pressure wave transmission delay time. The three-dimensional polyline path length of the corresponding pipe section is retrieved from the GIS spatial database as the spatial reference distance. The spatial reference distance is divided by the transmission delay time to obtain the measured pressure wave velocity value of the pipe section. The deviation rate of the measured pressure wave velocity is compared with the preset wave velocity of the hydraulic model. When the deviation rate exceeds the tolerance threshold, the time-weighted sliding average algorithm is used to update the wave velocity parameters of the hydraulic model and record the correction log.

5. The method for intelligent analysis and processing of massive data and decision support for the smart water integration platform based on GIS and the Internet of Things according to claim 1 is characterized in that: The pressure anomaly propagation map construction includes: Taking the GIS pipe network topology structure as the spatial coordinate system, the spatial geographic coordinates of each pressure sensor are mapped to the node position of the map. The direction of pressure anomaly transmission is represented by a time-series connecting line with an arrow, and the arrow points to the upstream sensor node. The dynamic parameters of the pressure drop percentage, transmission delay time in seconds, and measured wave velocity value of the pipe section are embedded in the node attribute table. The pipe section material code and pipe diameter size value are annotated in the connecting line attributes, forming a multidimensional dynamic network model that supports spatial topology analysis and time-series backtracking as a pressure anomaly propagation map.

6. The method for intelligent analysis and processing of massive data and decision support for the smart water integration platform based on GIS and the Internet of Things according to claim 1 is characterized in that: Calculating the pressure decay gradient value includes: Starting from the attenuation calculation origin identified by the pressure anomaly propagation map, the system expands outwards along the pipeline network topology path layer by layer to traverse adjacent nodes. Every time a new node is expanded, the material type of the pipe section to which the node belongs is queried and the preset attenuation coefficient benchmark table is called. The absolute value of the attenuation at the current level is calculated by combining the pipe diameter compensation coefficient and the service life reduction factor. The gradient value of the previous node is subtracted from the attenuation of the current level as the gradient value of this node, forming a gradient value curve continuously distributed along the pipeline network path and stored in the topological node attribute set.

7. The method for intelligent analysis and processing of massive data and decision support for the smart water integration platform based on GIS and the Internet of Things according to claim 1 is characterized in that: The generated valve closing boundary range includes: A boundary starting anchor point is set at the location of the pipeline node where the pressure decay gradient value first falls below the preset critical threshold, and the search continues to expand along the topological path toward the upstream direction of the water flow. When a node is detected where the gradient value rises above the critical threshold, a boundary ending anchor point is set, and all pipeline paths between the starting anchor point and the ending anchor point are connected to form a hydraulic isolation closed loop. When the closed loop passes through the pipeline ring structure, the optimal segmentation algorithm is automatically started to select the segmentation path with the least interruption to the main pipe and generate the final valve closing boundary range.

8. The method for intelligent analysis and processing of massive data and decision support for the smart water integration platform based on GIS and the Internet of Things according to claim 1 is characterized in that: The connectivity weight calculation includes: In the GIS topology network, the number of pipes directly connected to each valve device is counted as the basic connectivity value. The hierarchical depth of the pipe network partition where the valve is located in the global topology is identified and an exponential weighting coefficient is assigned. The number of water meters affected downstream after the valve is closed is calculated as the influencing factor. The basic connectivity value, hierarchical weighting coefficient and influencing factor are fused through a linear normalization algorithm to generate a standardized comprehensive connectivity weight value.

9. The method for intelligent analysis and processing of massive data and decision support for a smart water services integration platform based on GIS and the Internet of Things according to claim 1 is characterized in that: The generating of the staged valve closing instruction set comprises: All valve devices within the valve closing boundary are traversed, and the valve with the highest comprehensive connectivity weight value is preferentially selected as the first-level valve closing target. The Euclidean distance of the latitude and longitude coordinates of each valve within the boundary is calculated based on the spatial coordinates of the attenuation calculation origin and normalized into a distance coefficient. The valve operation priority score is calculated according to the decision model of "comprehensive connectivity weight × priority factor A + distance coefficient × priority factor B", and the valves are divided into three operation levels in descending order of score.

10. The method for intelligent analysis and processing of massive data and decision support for the smart water integration platform based on GIS and the Internet of Things according to claim 9 is characterized in that: When verifying the validity of the valve closing boundary range, if it is found that the valve closing is ineffective, the optimization mechanism will be automatically triggered, the pressure decay critical threshold will be gradually lowered and the gradient value calculation will be re-executed. Based on the new threshold, the expanded valve closing boundary range is generated, and the connectivity weight calculation model and distance coefficient analysis process are reused within the expanded boundary. The supplementary valve closing instruction set for the newly added valve is dynamically generated, and the optimized threshold parameters and execution effects are fed back to the historical parameter library of the gradient calculation model to realize self-learning evolution.

Images