A repair and analysis system for smart pipe network topology

Abstract

The present invention discloses a repair and analysis system for a smart pipe network topology structure, which relates to the field of smart pipe network management technology and includes a design and management platform. The design and management platform is communicatively connected to a topology data acquisition module, a topology analysis module, a pipe network fault diagnosis module, a repair solution generation module, and a pipe network operation monitoring module. The present invention integrates multiple sensors and external data sources to construct a pipe network topology structure model, accurately identifies branches, loops, or special-shaped connection structures in complex pipe networks, improves the ability to judge pipe network connection relationships and flow directions, and combines the pipe network topology structure model and flow direction analysis results with a pre-built fault diagnosis model to quickly and accurately diagnose pipe network faults. By integrating multi-source data and extracting fault characteristics, and matching them with preset fault patterns, the present invention can quickly identify the fault type and location, reduce the misdiagnosis rate, and greatly improve the efficiency and accuracy of fault diagnosis.

Description

Technical Field

[0001] The present invention relates to the technical field of smart pipe network management, and in particular to a repair and analysis system for a smart pipe network topology structure. Background Art

[0002] As the global urbanization process continues to advance, the size of cities continues to expand and the population increases dramatically. In order to meet the living needs of urban residents and the needs of industrial production, the scale and complexity of urban pipeline systems are also constantly increasing. Such a large and complex pipeline system, once its topological structure has problems, it will directly affect the normal operation of the city and the quality of life of residents. The pipeline systems in many cities were built early, and after long-term use, problems such as pipeline aging, corrosion, and damage are becoming increasingly serious.

[0003] In the existing technology, it is difficult to accurately identify and analyze the structural characteristics of pipe network topology structures with a large number of branches, loops or special-shaped connections, resulting in unreasonable repair solutions. Therefore, for complex pipe network topology structures, how to use artificial intelligence technology to construct a pipe network topology structure model, accurately judge the connection relationship and flow direction of the pipelines, and improve the ability to identify and analyze complex pipe network topology structures is the problem to be solved by the present invention. To this end, a smart pipe network topology repair and analysis system is proposed. Summary of the Invention

[0004] The purpose of the present invention is to provide a repair and analysis system for a smart pipe network topology structure to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] A smart pipe network topology repair and analysis system includes a design management platform, wherein the design management platform is communicatively connected to a topology data acquisition module, a topology analysis module, a pipe network fault diagnosis module, a repair solution generation module, and a pipe network operation monitoring module;

[0007] The topology data acquisition module is used to integrate multiple sensors and external data sources in the pipe network system to collect and pre-process pipe network operation data and spatial information;

[0008] The topology analysis module is used to analyze the pre-processed pipe network operation data and spatial information using artificial intelligence technology, construct a pipe network topology model, and then analyze the pipeline connection relationship and flow direction;

[0009] The pipeline network fault diagnosis module is used to combine the pipeline network topology model and flow direction analysis results, use the pre-built fault diagnosis model to diagnose pipeline network faults, identify the fault type and fault location, and diagnose pipeline network faults quickly and accurately to reduce the misdiagnosis rate;

[0010] The repair plan generation module is used to generate a repair plan based on the fault diagnosis results, combined with the topological structure and flow direction information of the pipeline network, and simulate the repair process to analyze the effects of different repair plans, optimize the repair strategy, reduce repair costs and construction risks, ensure the feasibility and efficiency of the repair plan, and improve the efficiency and quality of the repair work;

[0011] The pipeline network operation monitoring module is used to continuously monitor the repaired pipeline network, collect pipeline network operation data, adjust the repair plan according to the pipeline network operation status, optimize pipeline network management, and ensure stable operation and efficient management of the system.

[0012] A further improvement of the technical solution of the present invention is that the topology structure data acquisition module specifically includes:

[0013] In the pipe network system, various sensors, including pressure sensors, flow sensors, and liquid level sensors, are deployed based on the pipe network layout and key nodes. The sensor layout is planned based on the pipe network topology and monitoring priorities. Sensors are installed at pipe connection points, branch points, and loop nodes to monitor the pipe network's operating status in real time, obtain pipe network operating data, and ensure full coverage of key nodes and problem-prone areas.

[0014] Integrate external data sources including geographic information systems (GIS) and historical maintenance records to obtain GIS data and organize historical maintenance records of the pipeline network, including maintenance time, fault type and treatment method;

[0015] The collected pipeline network operation data, GIS data and historical maintenance records are integrated to enter the data preprocessing stage. Combined with GIS data and historical maintenance records, the collected data are annotated and supplemented to form complete pipeline network operation data and spatial information, build a complete data set, and then store the preprocessed data in the data warehouse.

[0016] A further improvement of the technical solution of the present invention is that: the topology structure recognition and analysis module includes a topology feature extraction unit, a topology structure modeling unit and a flow direction analysis unit;

[0017] The topological feature extraction unit is used to extract the operation features and spatial features related to the pipe network topology structure from the pre-processed pipe network operation data and spatial information, and integrate them to obtain a pipe network topology structure feature sequence;

[0018] The topology modeling unit is used to analyze the features in the collected pipeline network topology feature sequence using artificial intelligence technology based on graph neural networks, construct a pipeline network topology model, and accurately represent the connection relationship, branch points and loop structure of the pipelines;

[0019] The flow direction analysis unit analyzes the flow direction of the fluid in the pipe network based on the constructed pipe network topology model, and determines the flow direction and flow distribution of the fluid in different pipe sections through computational fluid dynamics simulation.

[0020] A further improvement of the technical solution of the present invention is that the topological feature extraction unit specifically includes:

[0021] From the pre-processed pipeline network operation data, key feature points reflecting the pipeline network topology are screened out, and the locations of the key feature points in the pipeline network spatial layout are determined with reference to spatial information, and a preliminary feature point set is constructed;

[0022] Based on the selected key feature points, the operating characteristics related to the operation status of the pipeline network are extracted, including pressure gradient, flow peak and valley values, and flow change frequency;

[0023] Based on the spatial information of the pipeline network, the spatial features between key feature points are extracted. The actual length of the pipeline is estimated by calculating the Euclidean distance between the feature points, providing geometric parameters for pipeline network modeling. The connection sequence and branching angles of the key feature points are analyzed to determine the direction and branching of the pipeline.

[0024] The extracted operational characteristics and spatial characteristics are integrated, and according to the topological order of the pipeline network, the operational characteristics (pressure gradient, flow peak and valley values, flow change frequency) and spatial characteristics (Euclidean distance, connection order, branch angle) of each key feature point are arranged and combined to form a comprehensive feature sequence, which is the pipeline network topological structure feature sequence.

[0025] A further improvement of the technical solution of the present invention is that the topology structure modeling unit specifically includes:

[0026] Based on the pipeline network topology, the selected key feature points are used as graph nodes. According to the actual connection of the pipeline network, edges are established between the nodes to represent the connection relationship of the pipelines. The graph structure representation of the pipeline network is constructed, and the attributes of each node and each edge are clarified. Among them, each key feature point is regarded as a graph node. The node's feature vector contains its operational characteristics and spatial characteristics. According to the actual connection of the pipeline network, edges are established between the nodes and the edge weights are initialized. The weights are determined based on the physical properties of the pipeline, such as length, material, diameter, and historical operation data such as flow and pressure.

[0027] Utilizing the aggregation mechanism of graph neural networks, the features of each node and its neighboring nodes are aggregated, and the feature information of each node's neighboring nodes is collected. By traversing the graph structure, the directly connected nodes of each node, i.e., neighboring nodes, are determined, and the feature vectors of the neighboring nodes are obtained. The collected neighboring node feature information is then integrated and updated with the node's own features.

[0028] The constructed graph structure is trained to learn the feature representation of nodes and edges. A loss function is defined, and the loss is calculated by comparing the difference between the predicted topology and the actual topology. The gradient descent optimization algorithm is used to update the model parameters. During the training process, the mapping relationship between node features and topology is learned.

[0029] After training is completed, the learned feature representations of nodes and edges are used to generate the final pipeline network topology model. The pipeline topology model represents the connection relationship, branch points and loop structure of the pipelines, and outputs the feature representation of each node, the connection weights between nodes, and the topology diagram of the entire pipeline network.

[0030] A further improvement of the technical solution of the present invention is that the flow direction analysis unit specifically includes:

[0031] Based on the constructed pipe network topology model, the design parameters of the pipe network are initialized, and the initial flow direction and flow distribution assumptions of the fluid in the pipe network are set. At the same time, the type and boundary conditions of each node are clarified, and the inlet pressure and flow rate of the pipe network, as well as the pressure or free outflow condition of the outlet are determined. The types of nodes include water source nodes, user nodes, intersection nodes, etc.

[0032] Select computational fluid dynamics (CFD) software, import the pipe network topology model into it, simulate the fluid flow in the pipe network, initialize the simulation environment in the CFD software, set the simulation time step and number of iterations, apply CFD principles, solve the continuity equation and energy equation, calculate the pressure change and flow distribution of the fluid in different pipe sections, and use the finite difference method to discretize the continuous equation to form a solvable system of equations to predict the flow behavior of the fluid in the pipe network;

[0033] The flow direction results obtained by simulation are compared with the actual pipeline network operation data. The flow rate, flow direction and node pressure data of each pipeline section in the actual pipeline network are collected and compared with the simulation results one by one. The error between the predicted value and the actual value is calculated. The parameters of the model are adjusted using the optimization algorithm of the gradient descent method. Through multiple iterations, the error is continuously reduced so that the flow direction analysis results gradually approach the actual fluid flow situation.

[0034] After multiple iterations of optimization, the flow direction and flow distribution results of the fluid in the pipeline network are output, and a pipeline network flow diagram is generated. The visualization function of the computational fluid dynamics (CFD) software is used to intuitively display the flow direction of the fluid in each pipeline segment, provide a flow distribution table, and list the flow values ​​of each pipeline segment.

[0035] A further improvement of the technical solution of the present invention is that the pipeline network fault diagnosis module specifically includes:

[0036] Integrate the network topology model, flow analysis results, and real-time monitoring data to ensure that all data are aligned in terms of timestamps and spatial locations, remove noise data and outliers, and extract fault characteristics from the integrated data. Fault characteristics include pressure change rate, flow mutation amplitude, node pressure deviation, and flow direction deviation.

[0037] The extracted fault features are input into the pre-built fault diagnosis model, matched with the fault modes preset in the fault diagnosis model, and the similarity between the fault features and each fault mode is calculated to determine whether a fault exists.

[0038] After determining the fault type, the fault location is further located. Combined with the pipeline network topology model, the propagation path and impact range of the fault characteristics in the pipeline network are analyzed, and a fault report is generated. The diagnosed fault type and fault location are output in an intuitive manner. The fault report details the fault type, location, and impact range, and is accompanied by a pipeline network topology diagram with the fault location marked on the diagram.

[0039] A further improvement of the technical solution of the present invention is that the calculation process of the similarity between the fault characteristics and each fault mode is:

[0040] Integrate the network topology model, flow analysis results, and real-time monitoring data to ensure that all data are aligned in terms of timestamp and spatial location. Extract fault characteristics from the integrated data, including pressure change rate, flow mutation amplitude, node pressure deviation, and flow direction deviation.

[0041] For each fault feature, determine its current value and the corresponding characteristic typical value in the fault mode, and then calculate the relative difference between its current value and the corresponding characteristic typical value in the kth fault mode. Sum the squares of the relative differences of all fault features and take the square root to obtain the normalized difference function.

[0042] Calculate the average value of the ratio of the fault feature to the fault mode feature, and substitute the calculated average value into the sine function to obtain the sensitivity adjustment function;

[0043] By combining the standardized difference function and the sensitivity adjustment function, the similarity of the k-th fault mode can be calculated. When the standardized difference function is close to 1, it means that the fault feature and the fault mode have a high degree of match. When the sensitivity adjustment function is close to 1, it means that the overall trend of the fault feature is consistent with the fault mode. That is, the closer the similarity of the k-th fault mode is to 1, the more similar the fault feature is to the k-th fault mode, and the greater the possibility of failure.

[0044] A further improvement of the technical solution of the present invention is that the repair solution generation module specifically includes:

[0045] Based on the fault type, location, and impact range information output by the fault diagnosis module, combined with the pipeline network topology and flow direction information, the key pipeline sections and nodes involved in the fault are analyzed, and a preliminary repair plan suitable for the fault type is screened from the pre-set repair strategy library;

[0046] Initially screened repair options are imported one by one, and the repair process is simulated to assess the impact of the repair on the network operation, including the duration of water supply interruption and short-term fluctuations in water pressure in the surrounding area. The cost and expected repair time of each option are calculated. Based on the simulation results, the initial repair options are optimized, and the cost, construction difficulty, and impact on network operation of different options are compared to select the most optimal repair technology. For complex loop or branch faults, the repair sequence is optimized to avoid secondary faults or construction conflicts, ultimately generating one or more optimized repair options.

[0047] The optimized repair plan is output in an intuitive manner, including detailed construction steps, a list of required materials and equipment, a cost budget, an estimated repair time, and recommendations for temporary measures.

[0048] A further improvement of the technical solution of the present invention is that the pipe network operation monitoring module specifically includes:

[0049] In the repaired pipe network system, integrated pressure sensors, flow sensors, and monitoring equipment collect real-time operating data, including pipeline pressure, flow, and liquid level, to reflect the current operating status of the pipe network. The collected operating data is then imported into the design management platform for comparison and analysis with pre-set normal operating thresholds. The system monitors whether there are abnormal pressure fluctuations and whether the flow meets expectations, thereby determining whether the pipe network is operating normally. If the data exceeds the set normal operating threshold, an alarm is automatically triggered, the abnormality is recorded, and a preliminary analysis report is generated.

[0050] Based on the monitored abnormal conditions, combined with the network topology and flow direction information, the cause of the fault and the scope of impact are analyzed. If new problems are found in the repaired network, the repair plan is adjusted to ensure the stable operation of the network.

[0051] The adjusted repair plan will be re-implemented, and the operation of the pipeline network will be continuously monitored. By comparing the operating data before and after the adjustment, the effectiveness of the optimization plan will be evaluated. At the same time, pipeline network operation reports will be generated regularly to summarize the operating status, problems found and solutions, providing a reference for the long-term management and maintenance of the pipeline network.

[0052] Due to the adoption of the above technical solution, the present invention has the following technical advancements compared to the prior art:

[0053] The present invention provides a repair and analysis system for an intelligent pipe network topology structure. By integrating multiple sensors and external data sources to construct a pipe network topology structure model, the system accurately identifies branches, loops or special-shaped connection structures in complex pipe networks, thereby improving the ability to judge pipe network connection relationships and flow directions. In combination with the pipe network topology structure model and flow direction analysis results, the system utilizes a pre-built fault diagnosis model to quickly and accurately diagnose pipe network faults. Furthermore, by integrating multi-source data and extracting fault features, and matching them with preset fault patterns, the system can quickly identify the fault type and location, reduce the misdiagnosis rate, and greatly improve the efficiency and accuracy of fault diagnosis.

[0054] The present invention provides a repair and analysis system for the topology of an intelligent pipe network. Based on the fault diagnosis results, combined with the topology and flow direction information of the pipe network, a repair plan is generated. The effects of different plans are analyzed by simulating the repair process. The cost, construction difficulty and impact on the operation of the pipe network of different plans can be compared, and the optimal repair strategy can be selected, thereby reducing the repair cost and construction risk. The repaired pipe network is continuously monitored, and the operation data of the pipe network is collected. The repair plan is adjusted according to the operation status. Abnormal conditions in the operation of the pipe network can be discovered in real time, and timely measures can be taken to make adjustments, thereby ensuring the stable operation and efficient management of the pipe network. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0056] Figure 1 It is a schematic diagram of the workflow of the present invention;

[0057] Figure 2 Schematic diagram of the workflow of the pipeline network fault diagnosis module of the present invention. DETAILED DESCRIPTION

[0058] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0059] Example 1, as Figure 1 、 Figure 2 As shown, the present invention provides a smart pipe network topology repair and analysis system, including a design management platform, which is communicatively connected to a topology data acquisition module, a topology analysis module, a pipe network fault diagnosis module, a repair solution generation module, and a pipe network operation monitoring module;

[0060] Topological data acquisition module, used to integrate multiple sensors (pressure sensor, flow sensor, liquid level sensor, etc.) and external data sources (Geographic Information System GIS) in the pipe network system The system collects and preprocesses pipeline network operation data and spatial information, including pipeline data and historical maintenance records. Within the pipeline network system, various sensors, including pressure sensors, flow sensors, and liquid level sensors, are deployed based on the network layout and key nodes. Sensor layout is planned based on the network topology and monitoring priorities. Sensors are installed at pipeline connection points, branch points, and loop nodes to monitor the network's operation status in real time and obtain network operation data, ensuring full coverage of key nodes and problem-prone areas. External data sources, including geographic information systems (GIS) and historical maintenance records, are integrated to obtain GIS data (spatial location information of the pipeline network), including pipeline laying routes and node coordinates. Furthermore, historical maintenance records covering repair times, fault types, and handling methods are collated. The collected pipeline network operation data, GIS data, and historical maintenance records are integrated. The data preprocessing phase involves data cleaning to remove abnormal data caused by sensor failures or environmental interference, formatting the data, and unifying the data format. The collected data is annotated and supplemented, combining GIS data with historical maintenance records, to form complete pipeline network operation data and spatial information, building a complete dataset. The preprocessed data is then stored in a data warehouse.

[0061] The topology analysis module is used to analyze the pre-processed pipe network operation data and spatial information using artificial intelligence technology, construct a pipe network topology model, and then analyze the pipeline connection relationship and flow direction. The topology identification and analysis module includes a topology feature extraction unit, a topology modeling unit, and a flow direction analysis unit;

[0062] Among them, the topological feature extraction unit is used to extract the operation characteristics and spatial characteristics related to the pipeline network topology structure from the preprocessed pipeline network operation data and spatial information, integrate them to obtain the pipeline network topology structure feature sequence, and screen out the key feature points reflecting the pipeline network topology structure from the preprocessed pipeline network operation data, and refer to the spatial information to determine the position of the key feature points in the pipeline network spatial layout, and preliminarily construct a feature point set. Among them, in the pressure data, find the nodes with obvious pressure changes at the pipeline branches or connections as key feature points. In the flow data, identify the points with uneven flow distribution corresponding to the intersection points or loop entrances of the pipeline network as key feature points. For the selected key feature points, extract the operation characteristics related to the pipeline network operation status, including pressure gradient, flow peak and valley values ​​and flow change frequency. Among them, calculate the pressure gradient of each key feature point to reflect the resistance characteristics of the pipeline section, and count the flow peak and valley values ​​and flow change frequency to reflect the water use patterns and pipeline network load fluctuations in different time periods. , capture the dynamic operation characteristics of the pipeline network in the time dimension, extract the spatial characteristics between key feature points based on the spatial information of the pipeline network, among which, by calculating the Euclidean distance between the feature points, the actual length of the pipeline is estimated, and geometric parameters are provided for pipeline network modeling. The connection sequence and branching angle of the key feature points are analyzed to determine the direction and branching of the pipeline, which helps to identify the topological structure of the pipeline network. Combined with geographic coordinate data, the distribution form of the pipeline network in geographic space is further described, and the extracted operation characteristics and spatial characteristics are integrated. According to the topological order of the pipeline network, the operation characteristics (pressure gradient, flow peak and valley value, flow change frequency) and spatial characteristics (Euclidean distance, connection sequence, branch angle) of each key feature point are arranged and combined to form a comprehensive feature sequence, namely the pipeline network topological structure feature sequence. The pipeline network topological structure feature sequence not only includes the dynamic characteristics of the pipeline network in the operating state, but also incorporates the spatial layout information of the pipeline network, and can comprehensively characterize the topological structure of the pipeline network;

[0063] The topology modeling unit is used to analyze the features in the collected pipeline topology feature sequence using artificial intelligence technology based on graph neural networks, construct a pipeline topology model, accurately represent the connection relationship, branch points and loop structure of the pipeline, and use the filtered key feature points as graph nodes according to the pipeline topology. According to the actual connection situation of the pipeline network, edges are established between the nodes to represent the connection relationship of the pipeline, and a graph structure representation of the pipeline network is constructed to clarify the attributes of each node and each edge. Each key feature point is used as a graph node, and the feature vector of the node contains its operating characteristics and spatial characteristics. According to the actual connection situation of the pipeline network, edges are established between the nodes, and the weight of the edge is initialized. The weight is determined according to the physical properties of the pipeline such as length, material, diameter, and historical flow, pressure and other operating data. The aggregation mechanism of the graph neural network is used to aggregate the features of each node and its neighboring nodes, collect the feature information of the neighboring nodes of each node, and determine the directly connected nodes of each node, that is, neighbor nodes, by traversing the graph structure. The feature vector of the node is obtained, and the collected feature information of the neighboring nodes is integrated with the features of the node itself to update the feature vector of each node, so that the feature vector of each node not only contains its own information, but also integrates the features of the surrounding nodes, thereby capturing the local topological structure information and enhancing the node's perception of the surrounding environment. The constructed graph structure is trained to learn the feature representation of nodes and edges, define the loss function, and calculate the loss by comparing the difference between the predicted topology and the actual topology. The parameters of the model are updated using the gradient descent optimization algorithm. During the training process, the mapping relationship between node features and topological structures is learned so that the feature representation of the node can reflect its position and role in the pipeline network topology. Through continuous iteration and optimization, the model gradually masters the characteristic pattern of the pipeline network topology. After the training is completed, the feature representation of the learned nodes and edges is used to generate the final pipeline network topology model. The pipeline topology model represents the connection relationship, branch points and loop structure of the pipeline, and outputs the feature representation of each node, the connection weight between nodes and the topological structure diagram of the entire pipeline network.

[0064] The flow direction analysis unit analyzes the flow direction of the fluid in the pipe network based on the constructed pipe network topology model, and determines the flow direction and flow distribution of the fluid in different pipe sections through computational fluid dynamics simulation. Based on the constructed pipe network topology model, it is initialized according to the design parameters of the pipe network, and the initial flow direction and flow distribution assumptions of the fluid in the pipe network are set. At the same time, the type and boundary conditions of each node are clarified, and the inlet pressure and flow rate of the pipe network, as well as the pressure or free outflow conditions of the outlet are determined. Among them, the types of each node include water source nodes, user nodes, intersection nodes, etc., and computational fluid dynamics (CFD) software is selected to import the pipe network topology model into it to simulate the fluid flow in the pipe network. The simulation environment is initialized in the computational fluid dynamics (CFD) software, and the parameters of the simulation time step and number of iterations are set. By applying the CFD principle, the continuity equation and energy equation are solved. Calculate the pressure changes and flow distribution of the fluid in different pipeline sections, and use the finite difference method to discretize the continuous equations to form a solvable set of equations to predict the flow behavior of the fluid in the pipeline network. Compare the flow direction results obtained by simulation with the actual pipeline network operation data. Collect the flow, flow direction and node pressure data of each pipeline section in the actual pipeline network, compare them one by one with the simulation results, calculate the error between the predicted value and the actual value, and use the gradient descent optimization algorithm to adjust the model parameters. Through multiple iterations, continuously reduce the error so that the flow direction analysis results gradually approach the actual fluid flow situation. After multiple iterative optimizations, output the flow direction and flow distribution results of the fluid in the pipeline network, generate a pipeline network flow diagram, and use the visualization function of the computational fluid dynamics (CFD) software to intuitively display the flow direction of the fluid in each pipeline section, provide a flow distribution table, and list the flow values ​​of each pipeline section.

[0065] The pipeline network fault diagnosis module is used to combine the pipeline network topology model and flow direction analysis results, use the pre-built fault diagnosis model to diagnose pipeline network faults, identify the fault type and fault location, and diagnose pipeline network faults quickly and accurately, reducing the misdiagnosis rate;

[0066] The repair plan generation module is used to generate repair plans based on the fault diagnosis results, combined with the topological structure and flow direction information of the pipeline network. By simulating the repair process, it analyzes the effects of different repair plans, optimizes the repair strategy, reduces repair costs and construction risks, ensures the feasibility and efficiency of the repair plan, and improves the efficiency and quality of the repair work;

[0067] The pipeline network operation monitoring module is used to continuously monitor the repaired pipeline network, collect pipeline network operation data, adjust the repair plan according to the pipeline network operation status, optimize pipeline network management, and ensure stable operation and efficient management of the system.

[0068] Example 2, as Figure 1 、 Figure 2 As shown, based on Example 1, the present invention provides a technical solution: preferably, the pipe network fault diagnosis module specifically includes:

[0069] The network topology model, flow analysis results and real-time monitoring data are integrated to ensure that all data are aligned in timestamp and spatial position, remove noise data and outliers, and extract fault features from the integrated data, where the fault features include pressure change rate, flow mutation amplitude, node pressure deviation and flow deviation. The extracted fault features are input into the pre-built fault diagnosis model, matched with the preset fault mode in the fault diagnosis model, and the similarity between the fault features and each fault mode is calculated to determine whether there is a fault. The connection relationship, length, diameter, material and other information of the pipelines in the pipeline network are collected, the location coordinates, type and other data of the nodes are obtained, the pressure, flow and other historical monitoring data of each node in the pipeline network are collected, and the relevant data when the fault occurs are recorded, including the fault type, fault occurrence time, Fault location, changes in pressure and flow before and after the fault occurs, etc., deploy sensor networks, obtain real-time pressure, flow and other data of each node in the pipeline network, and pre-process the collected data, divide it into training set, validation set and test set. The training set is used to train the model, the validation set is used to adjust the hyperparameters of the model, and the test set is used to evaluate the final performance of the model. The neural network model is trained using the training set, and the extracted fault features are used as model input, and whether there is a fault is used as the model output. By continuously adjusting the parameters of the model, the loss function of the model on the training set is minimized. The validation set is used to evaluate the model during the training process, and the hyperparameters of the model are adjusted according to the evaluation results. The optimal model configuration is selected, and the trained model is evaluated using the test set. The evaluation indicators include accuracy, recall rate, F1 The accuracy rate reflects the proportion of correct classifications by the model, the recall rate reflects the model's ability to identify positive examples (fault samples), and the F1 value is the harmonic mean of the accuracy rate and the recall rate. Taking the performance of both into consideration, the confusion matrix can intuitively show the model's classification of different fault types. Finally, a trained fault diagnosis model is obtained. After determining the fault type, the fault location is further located. Combined with the pipeline network topology model, the propagation path and impact range of the fault characteristics in the pipeline network are analyzed, and a fault report is generated. The diagnosed fault type and fault location are output in an intuitive manner. The fault report details the fault type, location, and impact range, and is accompanied by a pipeline network topology diagram with the fault location marked on the diagram.

[0070] In addition, the calculation process of the similarity between the fault characteristics and each fault mode is as follows:

[0071] The pipeline network topology model, flow analysis results, and real-time monitoring data are integrated to ensure that all data are aligned in terms of timestamp and spatial location. Fault features including pressure change rate, flow mutation amplitude, node pressure deviation, and flow deviation are extracted from the integrated data. For each fault feature, its current value and the corresponding characteristic typical value in the fault mode are determined, and then the relative difference between its current value and the corresponding characteristic typical value in the k-th fault mode is calculated. The relative differences of all fault features are squared and summed, and the square root is taken to obtain a standardized difference function. The average value of the ratio of the fault feature to the fault mode feature is calculated, and the calculated average value is substituted into the sine function to obtain a sensitivity adjustment function. The standardized difference function and the sensitivity adjustment function are combined to calculate the similarity of the k-th fault mode. When the standardized difference function is close to 1, it indicates that the fault feature and the fault mode have a high degree of match. When the sensitivity adjustment function is close to 1, it indicates that the overall trend of the fault feature is consistent with the fault mode. That is, the closer the similarity of the k-th fault mode is to 1, the more similar the fault feature is to the k-th fault mode, and the greater the possibility of failure.

[0072] The calculation expression of the similarity between the fault characteristics and each fault mode is:

[0073] ;

[0074] Where, For the The similarity between the fault mode and the current fault feature, For the extracted Current value of fault characteristics, For the Among the failure modes Typical values ​​of fault characteristics, is the number of fault characteristics, The closer the value is to 1, the more the fault characteristics are consistent with the The more similar the two failure modes are, the more likely it is that the fault represented by the failure mode exists;

[0075] The repair solution generation module specifically includes:

[0076] Based on the fault type, location, and impact range information output by the fault diagnosis module, combined with the network topology and flow direction information, the system analyzes the key pipeline sections and nodes involved in the fault and selects preliminary repair plans suitable for the fault type from a pre-defined repair strategy library. Each of these preliminary repair plans is then imported and the repair process simulated to evaluate the impact of the repair on network operation, including the duration of water supply interruption and short-term fluctuations in water pressure in the surrounding area. The system also calculates the cost and expected repair time for each plan. Based on the simulation results, the preliminary repair plans are optimized, comparing the cost, construction difficulty, and impact on network operation of different plans to select the optimal repair technology. For complex loop or branch faults, the repair sequence is optimized to avoid secondary failures or construction conflicts. Ultimately, one or more optimized repair plans are generated, ensuring an optimal balance between cost, construction risk, and operational impact. The optimized repair plans are then output in an intuitive manner, including detailed construction steps, a list of required materials and equipment, a cost estimate, an expected repair time, and recommended temporary measures. This ensures smooth repair work, improved efficiency and quality, and a rapid and stable return to operation of the network.

[0077] The pipeline network operation monitoring module specifically includes:

[0078] In the repaired pipeline network system, integrated pressure sensors, flow sensors and monitoring equipment are used to collect pipeline network operation data in real time, including pipeline pressure, flow, liquid level, etc., reflecting the current operation status of the pipeline network. The collected operation data is imported into the design management platform and compared with the pre-set normal operation threshold for analysis. The pressure is monitored for abnormal fluctuations and the flow meets expectations, etc., to determine whether the pipeline network is in normal operation. If the data exceeds the set normal operation threshold, an alarm is automatically triggered, the abnormal situation is recorded and a preliminary analysis report is generated. Based on the monitored abnormal situation, combined with the pipeline network topology and flow direction information, the cause of the fault and the scope of impact are analyzed. If new problems are found in the repaired pipeline network, the repair plan is adjusted to ensure the stable operation of the pipeline network. The adjusted repair plan is re-implemented and the operation of the pipeline network is continuously monitored. By comparing the operation data before and after the adjustment, the effect of the optimization plan is evaluated. At the same time, a pipeline network operation report is generated regularly to summarize the operation status, problems found and solutions, providing a reference for the long-term management and maintenance of the pipeline network.

[0079] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A smart pipe network topology repair and analysis system, including a design management platform, characterized by: The design management platform is communicatively connected to a topology data acquisition module, a topology analysis module, a pipe network fault diagnosis module, a repair solution generation module, and a pipe network operation monitoring module; The topology data acquisition module is used to integrate multiple sensors and external data sources in the pipe network system to collect and pre-process pipe network operation data and spatial information; The topology analysis module is used to analyze the pre-processed pipe network operation data and spatial information, build a pipe network topology model, and then analyze the pipeline connection relationship and flow direction; The pipeline network fault diagnosis module is used to diagnose pipeline network faults and identify fault types and locations by combining the pipeline network topology model and flow direction analysis results with a pre-built fault diagnosis model. The pipeline network fault diagnosis module specifically includes: Integrate the pipe network topology model, flow direction analysis results, and real-time monitoring data, and extract fault characteristics from the integrated data. Fault characteristics include pressure change rate, flow mutation amplitude, node pressure deviation, and flow direction deviation. The extracted fault features are input into the pre-built fault diagnosis model, matched with the fault modes preset in the fault diagnosis model, and the similarity between the fault features and each fault mode is calculated to determine whether a fault exists. After determining the fault type, the fault location is further located. Combined with the pipe network topology model, the propagation path and impact range of the fault characteristics in the pipe network are analyzed, and a fault report is generated. The diagnosed fault type and fault location are output in an intuitive manner, accompanied by a pipe network topology diagram with the fault location marked on the diagram; The calculation process of the similarity between the fault characteristics and each fault mode is as follows: Integrate the network topology model, flow analysis results, and real-time monitoring data to ensure that all data are aligned in terms of timestamp and spatial location. Extract fault characteristics from the integrated data, including pressure change rate, flow mutation amplitude, node pressure deviation, and flow direction deviation. For each fault feature, determine its current value and the corresponding characteristic typical value in the fault mode, and then calculate the relative difference between its current value and the corresponding characteristic typical value in the kth fault mode. Sum the squares of the relative differences of all fault features and take the square root to obtain the normalized difference function. Calculate the average value of the ratio of the fault feature to the fault mode feature, and substitute the calculated average value into the sine function to obtain the sensitivity adjustment function; By combining the normalized difference function and the sensitivity adjustment function, the similarity of the k-th fault mode can be calculated; The repair solution generation module is used to generate a repair solution based on the fault diagnosis results, combined with the topological structure and flow direction information of the pipe network, simulate the repair process, analyze the effects of different repair solutions, and optimize the repair strategy; The pipeline network operation monitoring module is used to continuously monitor the repaired pipeline network, collect pipeline network operation data, and adjust the repair plan according to the pipeline network operation status.

2. The intelligent pipe network topology repair and analysis system according to claim 1, characterized in that: The topology structure data acquisition module specifically includes: In the pipe network system, various sensors including pressure sensors, flow sensors, and liquid level sensors are deployed according to the pipe network layout and key nodes. The sensor layout is planned based on the pipe network topology and monitoring priorities. Sensors are installed at pipe connection points, branch points, and loop nodes to monitor the pipe network's operating status in real time and obtain pipe network operation data. Integrate external data sources including geographic information systems and historical maintenance records to obtain GIS data. At the same time, organize historical maintenance records of the pipeline network including maintenance time, fault type and treatment method; The collected pipeline network operation data, GIS data and historical maintenance records are integrated to enter the data preprocessing stage. Combined with GIS data and historical maintenance records, the collected data are annotated and supplemented to form complete pipeline network operation data and spatial information, build a complete data set, and then store the preprocessed data in the data warehouse.

3. The intelligent pipe network topology repair and analysis system according to claim 1, characterized in that: The topology structure recognition and analysis module includes a topology feature extraction unit, a topology structure modeling unit and a flow direction analysis unit; The topological feature extraction unit is used to extract the operation features and spatial features related to the pipe network topology structure from the pre-processed pipe network operation data and spatial information, and integrate them to obtain a pipe network topology structure feature sequence; The topology modeling unit is used to analyze the features in the collected pipeline network topology feature sequence using artificial intelligence technology based on graph neural networks, and construct a pipeline network topology model to represent the connection relationship, branch points and loop structure of the pipelines; The flow direction analysis unit analyzes the flow direction of the fluid in the pipe network based on the constructed pipe network topology model, and determines the flow direction and flow distribution of the fluid in different pipe sections through computational fluid dynamics simulation.

4. The intelligent pipe network topology repair and analysis system according to claim 3, characterized in that: The topological feature extraction unit specifically includes: From the pre-processed pipeline network operation data, key feature points reflecting the pipeline network topology are screened out, and the locations of the key feature points in the pipeline network spatial layout are determined with reference to spatial information, and a preliminary feature point set is constructed; Based on the selected key feature points, the operating characteristics related to the operation status of the pipeline network are extracted, including pressure gradient, flow peak and valley values, and flow change frequency; Based on the spatial information of the pipeline network, the spatial features between key feature points are extracted. The actual length of the pipeline is estimated by calculating the Euclidean distance between the feature points. The connection sequence and branching angles of the key feature points are analyzed to determine the direction and branching of the pipeline. The extracted operation characteristics and spatial characteristics are integrated, and the operation characteristics and spatial characteristics of each key feature point are arranged and combined according to the topological order of the pipeline network to form a comprehensive feature sequence, which is the pipeline network topological structure feature sequence.

5. The intelligent pipe network topology repair and analysis system according to claim 4, characterized in that: The topology structure modeling unit specifically includes: Based on the pipeline network topology, the selected key feature points are used as graph nodes. According to the actual connection of the pipeline network, edges are established between the nodes to represent the connection relationship of the pipelines. The graph structure representation of the pipeline network is constructed, and the attributes of each node and each edge are clarified. According to the actual connection of the pipeline network, edges are established between the nodes and the edge weights are initialized. Utilizing the aggregation mechanism of graph neural networks, the features of each node and its neighboring nodes are aggregated, and the feature information of each node's neighboring nodes is collected. By traversing the graph structure, the directly connected nodes of each node, i.e., neighboring nodes, are determined, and the feature vectors of the neighboring nodes are obtained. The collected neighboring node feature information is then integrated and updated with the node's own features. The constructed graph structure is trained to learn the feature representation of nodes and edges. A loss function is defined, and the loss is calculated by comparing the difference between the predicted topology and the actual topology. The gradient descent optimization algorithm is used to update the model parameters. During the training process, the mapping relationship between node features and topology is learned. After training is completed, the learned feature representations of nodes and edges are used to generate the final pipeline network topology model. The pipeline topology model represents the connection relationship, branch points and loop structure of the pipelines, and outputs the feature representation of each node, the connection weights between nodes, and the topology diagram of the entire pipeline network.

6. The intelligent pipe network topology repair and analysis system according to claim 5, characterized in that: The flow direction analysis unit specifically includes: Based on the constructed pipe network topology model, the design parameters of the pipe network are initialized, the initial flow direction and flow distribution assumptions of the fluid in the pipe network are set, and the type and boundary conditions of each node are clarified, and the inlet pressure and flow rate of the pipe network, as well as the pressure or free outflow condition of the outlet are determined; Select computational fluid dynamics software, import the pipe network topology model into it, simulate the fluid flow in the pipe network, initialize the simulation environment in the computational fluid dynamics software, set the simulation time step and number of iterations, calculate the pressure change and flow distribution of the fluid in different pipe sections by solving the continuity equation and energy equation, and use the finite difference method to discretize the continuous equation to form a solvable system of equations to predict the flow behavior of the fluid in the pipe network; The flow direction results obtained by simulation are compared with the actual pipeline network operation data. The flow rate, flow direction and node pressure data of each pipeline section in the actual pipeline network are collected and compared with the simulation results one by one. The error between the predicted value and the actual value is calculated. The parameters of the model are adjusted using the optimization algorithm of the gradient descent method. Through multiple iterations, the flow direction analysis results are gradually approximated to the actual fluid flow conditions. After multiple iterations of optimization, the flow direction and flow distribution results of the fluid in the pipeline network are output, and a pipeline network flow diagram is generated. The visualization function of the computational fluid dynamics software is used to intuitively display the flow direction of the fluid in each pipeline section, provide a flow distribution table, and list the flow values ​​of each pipeline section.

7. The intelligent pipe network topology repair and analysis system according to claim 1, characterized in that: The repair solution generation module specifically includes: Based on the fault type, location, and impact range information output by the fault diagnosis module, combined with the pipeline network topology and flow direction information, the key pipeline sections and nodes involved in the fault are analyzed, and a preliminary repair plan suitable for the fault type is screened from the pre-set repair strategy library; Import the initially screened repair plans one by one, simulate the repair process, evaluate the impact of the repair on the pipeline network operation, and calculate the cost and expected repair time of each plan. Based on the simulation results, optimize the preliminary repair plans to select the best repair technology, and finally generate one or more optimized repair plans. The optimized repair plan is output in an intuitive manner, including detailed construction steps, a list of required materials and equipment, a cost budget, an estimated repair time, and recommendations for temporary measures.

8. The intelligent pipe network topology repair and analysis system according to claim 7, characterized in that: The pipe network operation monitoring module specifically includes: In the repaired pipe network system, integrated pressure sensors, flow sensors, and monitoring equipment are used to collect real-time operating data. This data is then imported into the design management platform for comparison and analysis with pre-set normal operating thresholds to determine whether the pipe network is operating normally. If data exceeds the set normal operating threshold, an alarm is automatically triggered, the abnormality is recorded, and a preliminary analysis report is generated. Based on the abnormal conditions detected, combined with the network topology and flow information, analyze the cause of the fault and the scope of impact. If new problems are found in the repaired network, adjust the repair plan; The adjusted repair plan will be re-implemented, and the operation of the pipeline network will be continuously monitored. By comparing the operating data before and after the adjustment, the effectiveness of the optimization plan will be evaluated. At the same time, pipeline network operation reports will be generated regularly to summarize the operating status, problems found and solutions.

Images