Drainage pipe network sensor optimal arrangement method facing whole system operation state perception

Abstract

The application discloses a kind of drainage pipe network sensor optimization arrangement methods for whole system operation state perception, comprising: according to drainage pipe network topological structure graph model is constructed;Based on graph model, the k-order neighbor node set of each node is determined, and the depth perception set of all nodes is obtained;The depth perception set is de-redundant, and the candidate node set is obtained;The candidate node set is executed greedy algorithm, and the sensor optimization arrangement node set is obtained;Based on the sensor optimization arrangement node set, Edge-STGCN model is constructed, and the training method is used to train Edge-STGCN model;The whole system operation state is estimated using trained Edge-STGCN model, and the optimization arrangement scheme of drainage pipe network sensor is determined.The application provides a more systematic and operable drainage pipe network sensor optimization arrangement framework, which helps to improve the monitoring performance and robustness of the designed sensor network in practical engineering applications.

Description

Technical Field

[0001] This application relates to the field of sensor optimization layout for drainage pipe networks, and more specifically, to a method for optimizing the arrangement of sensors in drainage pipe networks for sensing the overall system operation status. Background Technology

[0002] Drainage networks are an integral part of urban infrastructure, responsible for collecting and transporting wastewater and rainwater, and are crucial for ensuring urban safety and function. Regular monitoring of the drainage network's operation helps prevent urban flooding, control water pollution, and facilitate timely maintenance of key areas, thereby improving the overall system's operational efficiency and safety. However, due to limitations such as installation costs, site conditions, and maintenance complexity, the actual number of sensors available for installation is limited. Therefore, optimizing the deployment of sensors in the drainage network is essential to effectively reflect the overall system status and reduce monitoring costs.

[0003] Existing methods for optimizing drainage pipe network layouts largely rely on hydraulic model simulations to generate complete hydraulic state data, including genetic algorithms and clustering algorithms. However, such methods have the following drawbacks: (1) The construction process of high-precision hydraulic models is complex. Not only is the modeling process cumbersome, but the calibration is also difficult and affected by a variety of uncertain factors.

[0004] (2) Existing sensor network layouts are mostly limited to range perception for specific tasks and lack the ability to accurately locate anomalies. Taking infiltration or siltation events as an example, although their occurrence can be identified, it is difficult to further determine the specific location of the affected pipeline.

[0005] In view of this, the present invention provides a sensor optimization layout method for sensing the overall system operation status. Based on topology analysis, an optimized sensor layout is proposed, and a sensing model for monitoring the overall system operation status is built. This enables dynamic sensing of the hydraulic status of all nodes in the drainage network based on the layout results, with broad performance assurance and applicability to multiple services and functions for different objectives. Summary of the Invention

[0006] To address the aforementioned issues, this application provides a method for optimizing the layout of sensors in a drainage network for sensing the operational status of the entire system. This method aims to resolve the technical bottlenecks caused by relying on hydraulic model simulation data for sensor optimization in existing technologies, as well as the resource waste caused by the single task of the sensor network.

[0007] The first aspect of this invention provides a method for optimizing the layout of sensors in a drainage pipe network for sensing the overall system operation status, comprising: Construct a graphical model based on the topology of the drainage pipe network; Based on the graph model, the set of k-order neighbor nodes of each node is determined, and the depth perception set of all nodes is obtained. The depth-sensing set is subjected to redundancy removal processing to obtain a candidate node set; A greedy algorithm is executed on the candidate node set to obtain the optimal sensor node set; The Edge-STGCN model is constructed based on the optimized node set of the aforementioned sensors, and the Edge-STGCN model is trained using a masking training method. The trained Edge-STGCN model is used to estimate the overall system operating status and determine the optimal layout scheme of the drainage network sensors.

[0008] In one alternative implementation, the specific steps for determining the set of k-order neighbor nodes for each node based on the graph model are as follows: Initialize the set of 0th-order neighbors of node i:

[0009] in, Let be the set of neighbor nodes of node i when the search depth is 0; based on Recursive computation node The set of k-order neighbor nodes:

[0010] in, Let be the directly connected upstream and downstream nodes of node u. For nodes At a search depth of The set of neighboring nodes at that time.

[0011] In one optional implementation, the redundancy removal preprocessing of the depth-sensing set includes the following steps: In a deep perception set, if nodes neighbor node set The set of all neighboring nodes of node j If it contains, then the node and its neighbor node set Remove from the depth-sensing set; The candidate node set obtained after redundancy processing consists of all neighbor node sets that are not completely contained by any other neighbor node set.

[0012] In one optional implementation, the greedy algorithm is performed on the candidate node set to obtain the optimal sensor node set, and the specific steps are as follows: Initialize the selected sensor set and the set of covered nodes It is an empty set; Calculate the size of the neighbor set of each candidate node in the candidate node set. ; Select the candidate node with the largest neighbor node set. As a sensor node, add the sensor node to the selected sensor set. Set the neighboring nodes of this node The node is added to the set of covered nodes. ; Remove the sensor node and its neighbor node set from the candidate node set. The nodes in the list are selected, and the remaining candidate node set is updated, with any nodes that have been covered removed:

[0013] in, For a set of nodes, ; Repeat the selection and update steps until the candidate node set is empty, then output the selected sensor set. As a set of nodes for optimized sensor placement.

[0014] In one alternative implementation, when selecting a sensor node, if multiple candidate nodes have the same size set of neighbor nodes, then the following further steps are performed: Calculate the shortest path length between these candidate nodes and nodes in the selected sensor set S; The candidate node with the largest shortest path length to the selected sensor set S is selected as the sensor node.

[0015] In one alternative implementation, prior to constructing the Edge-STGCN model, the following is also included: Acquire node monitoring data and pipeline edge attribute data of the sensor optimized layout node set; the pipeline edge attribute data includes pipe diameter, pipe length, upstream bottom elevation of the pipeline, downstream bottom elevation of the pipeline, and pipeline drainage capacity; The node monitoring data and edge attribute data are normalized separately. Based on the normalized edge attribute data, an edge feature vector is constructed to characterize the pipeline properties.

[0016] In one optional implementation, the Edge-STGCN model includes a node feature encoding layer, an edge feature encoding layer, a spatial information processing layer, a temporal information processing layer, and an information decoding layer. The spatial information processing layer adopts the GINEConv layer, which is configured to combine normalized node monitoring data and edge feature vectors to aggregate node information.

[0017] In one alternative implementation, the loss function used when training the Edge-STGCN model includes mean squared error loss and physical constraint loss, as shown in the following formula:

[0018]

[0019]

[0020] in, This refers to the loss function used during model training. Let the mean squared error loss function be . Let be the physical constraint loss function, and λ be an adjustable hyperparameter that balances data fidelity and physical consistency. For the set of monitoring nodes, The number of monitoring nodes selected. These are the actual monitoring values. Output values ​​for the model. The bottom elevation of the inspection well. This refers to the ground elevation of the inspection well.

[0021] A second aspect of the present invention provides an electronic device, characterized in that it includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements a method for optimizing the arrangement of drainage network sensors for sensing the overall system operation status.

[0022] A third aspect of the present invention provides a computer-readable storage medium, characterized in that a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, a method for optimizing the arrangement of drainage pipe network sensors for sensing the overall system operation status.

[0023] This application has at least the following advantages or beneficial effects: The method provided by this invention utilizes only the topology of the drainage pipe network for optimized sensor layout, avoiding reliance on hydraulic model simulation data. Furthermore, it establishes a sensing model for monitoring the overall system operation status using this optimized sensor layout, enabling dynamic sensing of the hydraulic status of all nodes in the drainage pipe network based on the layout results. By accurately locating the position of a specific emergency based on the hydraulic status of all nodes across the entire system, rather than being limited to estimating the range of a single incident, this invention bridges the "layout-sensing-application" chain, providing a more systematic and operable framework for optimizing the layout of drainage pipe network sensors. This helps improve the monitoring performance and robustness of the designed sensor network in practical engineering applications. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a flowchart of a method for optimizing the layout of sensors in a drainage network for sensing the overall system operation status, as proposed in an embodiment of this application. Figure 2 This is a schematic diagram of a sensor sensing set proposed in an embodiment of this application; Figure 3 This is a schematic diagram of an electronic device according to this application. Detailed Implementation

[0026] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0027] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a method for optimizing the arrangement of sensors in a drainage pipe network based on the overall system operation status perception, as proposed in an embodiment of this application. Figure 1 As shown, a method for optimizing the layout of sensors in a drainage pipe network for sensing the overall system operation status includes: S100: Construct a graphical model based on the drainage network topology; In this embodiment, based on the drainage network topology diagram file (such as a CAD file or GIS file), inspection wells, forebays, and drainage outlets are abstracted as nodes, while pipes and pumps are abstracted as edges. Node set and edge set Composition diagram .

[0028] S200: Based on the graph model, determine the set of k-order neighbor nodes for each node to obtain the depth perception set of all nodes; In this embodiment, based on the obtained diagram Find the connection relationships between nodes in the graph. Specified search depth k-order range neighbor node set This forms a deep perception set for all nodes. .

[0029] Specifically, in one embodiment of this application, the steps for determining the set of k-order neighbor nodes for each node based on the graph model are as follows: S210: Initialize the set of 0-order neighbor nodes of node i:

[0030] in, Let be the set of neighbor nodes of node i when the search depth is 0; S220: Based on Recursive computation node The set of k-order neighbor nodes:

[0031] in, Let be the directly connected upstream and downstream nodes of node u. For nodes At a search depth of The set of neighboring nodes at that time.

[0032] In this embodiment, , representing the upstream and downstream nodes directly connected to node u. Represents a node The set of neighbor nodes at a search depth of 0 contains only nodes. itself;

[0033] S300: Perform redundancy removal processing on the depth sensing set to obtain a candidate node set; In this embodiment, the depth sensing set obtained in step 2... Perform redundancy removal preprocessing to obtain the redundancy-removed depth sensing set. This refers to the candidate node perception set, which narrows the search space and speeds up the selection process of the subsequent greedy algorithm.

[0034] Specifically, in one embodiment of this application, the redundancy removal preprocessing of the depth-sensing set includes the following steps: S310: In the depth-sensing set, if node neighbor node set The set of all neighboring nodes of node j If it contains, then the node and its neighbor node set Remove from the depth-sensing set; The candidate node set obtained after redundancy processing consists of all neighbor node sets that are not completely contained by any other neighbor node set.

[0035] In this embodiment, the depth-sensing set Perform redundancy removal preprocessing, i.e., in the deep sensing set, such as nodes neighbor node set The set of all neighboring nodes of node j If it contains, then the node and its neighbor node set From depth perception set Delete, and only retain the "representative" set that is not completely contained by any other set of neighboring nodes. That is, the set of candidate nodes.

[0036] S400: Execute a greedy algorithm on the candidate node set to obtain the optimal sensor node set; In this embodiment, based on the obtained candidate node set The algorithm employs a greedy iterative selection process, using the maximum value of the neighbor node set as the selection criterion to obtain the optimal sensor placement node set. .

[0037] Specifically, in one embodiment of this application, the steps of performing a greedy algorithm on the candidate node set to obtain the optimal sensor node set are as follows: S410: Initialize the selected sensor set and the set of covered nodes It is an empty set; S420: Calculate the size of the neighbor set of each candidate node in the candidate node set. ; S430: Select the candidate node with the largest set of neighboring nodes. As a sensor node, add the sensor node to the selected sensor set. Set the neighboring nodes of this node The node is added to the set of covered nodes. ; S431: When selecting sensor nodes, if multiple candidate nodes have the same size of their neighbor node sets, then further execution is performed: In this embodiment, when there are multiple candidate nodes with the same neighbor node set size, the set of each identical candidate node and the set of already selected sensor nodes are further calculated. The shortest path length, defined as: the path length of the graph... Viewed as an undirected graph, from candidate nodes To sensor node The number of nodes in the path with the fewest nodes among all possible paths is denoted as . ,in Indicates from arrive The set of all paths Representing a path The number of nodes included, and the shortest path length from the candidate node to the selected sensor set are: Prioritize those that meet the requirements. Candidate nodes are used as sensors.

[0038] Calculate the shortest path length between these candidate nodes and nodes in the selected sensor set S; The candidate node with the largest shortest path length to the selected sensor set S is selected as the sensor node.

[0039] S440: Remove the sensor node and its neighbor node set from the candidate node set. The nodes in the list are selected, and the remaining candidate node set is updated, with any nodes that have been covered removed:

[0040] in, For a set of nodes, ; S450: Repeat the selection and update steps until the candidate node set is empty, then output the selected sensor set. As a set of nodes for optimized sensor placement.

[0041] S500: Based on the optimized arrangement of the sensor node set, construct the Edge-STGCN model, and train the Edge-STGCN model using the masking training method; In this embodiment, a masked training method is employed. Half of the node monitoring data from the selected monitoring node set in step 4 is used as input, and the liquid levels of all nodes in the drainage network are used as output. An Edge-STGCN model is constructed based on the topology of the drainage network. Through model training, the operating status of all nodes in the drainage network can be estimated based on the sensor network. During the loss calculation during training, only the node output values ​​from the monitoring node set are calculated.

[0042] S600: The trained Edge-STGCN model is used to estimate the overall system operating status and determine the optimal layout scheme of the drainage network sensors.

[0043] Furthermore, before constructing the Edge-STGCN model, the following steps are also included: Acquire node monitoring data and pipeline edge attribute data of the sensor optimized layout node set; the pipeline edge attribute data includes pipe diameter, pipe length, upstream bottom elevation of the pipeline, downstream bottom elevation of the pipeline, and pipeline drainage capacity; In this embodiment, node monitoring data from the monitoring node set is collected, and pipe edge attribute data, including pipe diameter, is obtained based on the topology of the drainage pipe network. Manager Pipe bottom elevation and pipeline drainage capacity The drainage capacity of the pipeline is calculated using the Manning formula, as follows:

[0044] in, The flow rate of the pipeline. The roughness coefficient is Manning's coefficient. The cross-sectional area of ​​the pipe. For hydraulic radius, This is the hydraulic gradient (generally taken as the pipe gradient).

[0045] Representing nodes in vector form The recorded monitoring data is in the following format:

[0046] in, For monitoring nodes Arranged in chronological order One monitoring value.

[0047] The monitoring data of all nodes in the drainage pipe network form a time series matrix. The monitoring data of unmonitored nodes is replaced with 0.

[0048]

[0049] in, This indicates the number of nodes contained in the drainage network. This indicates the matrix transpose.

[0050] The node monitoring data and edge attribute data are normalized separately. In this embodiment, the time series matrix Normalization is performed, mapping it to the [0,1] interval to eliminate dimensional differences, specifically as follows:

[0051] in, This represents the original node monitoring value. and This indicates the minimum and maximum values ​​of the attribute. This represents the normalization result.

[0052]

[0053] According to the diagram Construct an adjacency matrix to determine the connection relationships between nodes in the drainage pipe network. ,specific:

[0054] in , When node and nodes When there are connecting pipes otherwise .

[0055] Opposite side attributes, including pipe diameter Manager Pipe bottom elevation and pipeline drainage capacity Normalization is performed, mapping it to the [0,1] interval to eliminate dimensional differences, specifically as follows:

[0056] in, Represents the original edge feature attribute value. and This indicates the minimum and maximum values ​​of the attribute. This represents the normalization result.

[0057] Based on the normalized edge attribute data, an edge feature vector is constructed to characterize the pipeline properties.

[0058] In this embodiment, the normalized edge feature attribute values ​​are concatenated to form the feature vector of each edge, in the following form:

[0059] in Represents a node and nodes The edge feature vectors between them These are the normalized values ​​for pipe diameter, pipe length, pipe inlet elevation, and pipe volume, respectively. Include and The two values ​​represent the upstream and downstream bottom elevations of the pipeline, respectively. Water flows from... Flow direction Then with node The bottom elevation of the connected pipes is called the upstream, and it is related to the node. The bottom elevation of connected pipes is called downstream, and vice versa.

[0060] Furthermore, the Edge-STGCN model includes a node feature encoding layer, an edge feature encoding layer, a spatial information processing layer, a temporal information processing layer, and an information decoding layer; The spatial information processing layer adopts the GINEConv layer, which is configured to combine normalized node monitoring data and edge feature vectors to aggregate node information.

[0061] In this embodiment, based on the specified search depth , build The spatial information processing layer (GINEConvlayer) combines the input features of nodes with edge features and processes them according to the connection relationships between nodes. The information is aggregated and transmitted at each layer to transmit the monitoring information of the monitoring nodes to the unmonitored nodes, and to perform hydraulic state estimation by combining edge feature attributes.

[0062] The masking training method operates as follows: In each round of training, half of the monitoring nodes are randomly selected from the monitoring node set for masking, i.e., the monitoring value of the corresponding node in the time matrix is ​​replaced with a zero value, and then used as the input of the model. The model output is the operating status value of all nodes in the drainage network. When calculating the loss during training, only the output value of the nodes in the monitoring node set is calculated.

[0063] Furthermore, when training the Edge-STGCN model, the loss functions used include mean squared error loss and physical constraint loss, as shown in the following formulas:

[0064]

[0065]

[0066] in, This refers to the loss function used during model training. Let the mean squared error loss function be . Let be the physical constraint loss function, and λ be an adjustable hyperparameter that balances data fidelity and physical consistency. For the set of monitoring nodes, The number of monitoring nodes selected. These are the actual monitoring values. Output values ​​for the model. The bottom elevation of the inspection well. This refers to the ground elevation of the inspection well.

[0067] In this embodiment, MSE is the mean squared deviation between the true and estimated values ​​of a specified node, providing a direct assessment of the reconstruction accuracy in the data space. Besides data fidelity, it is also essential to ensure that the estimated hydraulic state values ​​output by the model conform to the physical constraints of the drainage network system. Specifically, the estimated hydraulic state values ​​of the nodes must be maintained at the bottom elevation of the manhole and... Ground elevation Within the physically feasible range.

[0068] This disclosure also provides an electronic device, please refer to... Figure 3 , Figure 3 This is a schematic diagram of an electronic device illustrated in an embodiment of this disclosure. For example... Figure 3 As shown, the electronic device 100 includes a memory 110 and a processor 120. The memory 110 and the processor 120 are connected via a bus for communication. The memory 110 stores a computer program that can run on the processor 120 to implement the steps in the drainage network sensor optimization layout method for sensing the overall system operation status disclosed in this embodiment.

[0069] The disclosed embodiments also provide a computer-readable storage medium, wherein when the instructions in the computer-readable storage medium are executed by a processor of a computer device, the computer device is able to perform the steps in the drainage network sensor optimization layout method for overall system operation status perception as described in the embodiments of this disclosure.

[0070] For example, taking a rainwater pipe network of a drainage system in Shanghai as an example, this paper illustrates the process of optimizing the layout and setting of sensors using the topological relationship of the drainage pipe network.

[0071] The drainage network system comprises 406 nodes (including 404 inspection wells, 1 pump station forebay, and 1 river outlet), 412 pipe sections, and 3 pumps. The drainage capacity of the terminal pump station is 6.9 m³ / s. The system covers a total area of ​​124 hectares, with a pipe length of 10,880 m and pipe diameters ranging from DN300 to DN2400.

[0072] like Figure 2 As shown, obtain the set of neighbor nodes for all candidate nodes. In the embodiment of the node Specified search depth The set of neighboring nodes (see) Figure 2 The set of neighbor nodes of all candidate nodes. Composition of depth perception set .

[0073] For depth sensing set Perform redundancy removal preprocessing, i.e., in the deep sensing set, such as nodes neighbor node set The set of all neighboring nodes of node j Includes (e.g.) Figure 2 As shown, the neighbor set of nodes 97 and 96 is contained within the neighbor set of node 98, so the neighbor set of nodes 97 and 96 will be removed from the depth-sensing set. If deleted, the node will be... and its neighbor node set From depth perception set Delete, and only retain the "representative" set that is not completely contained by any other set of neighboring nodes. That is, the set of candidate nodes.

[0074] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, apparatuses, electronic devices, and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0075] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0076] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0077] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0078] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0079] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0080] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0081] The above provides a detailed description of a method for optimizing the arrangement of sensors in a drainage network for sensing the overall system operation status, as provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and its core ideas. At the same time, those skilled in the art will recognize that there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for optimizing the arrangement of sensors in a drainage pipe network for sensing the overall system operation status, characterized in that, include: Construct a graphical model based on the drainage pipe network topology; Based on the graph model, the set of k-order neighbor nodes of each node is determined, and the depth perception set of all nodes is obtained. The depth-sensing set is subjected to redundancy removal processing to obtain a candidate node set; A greedy algorithm is executed on the candidate node set to obtain the optimal sensor node set; The Edge-STGCN model is constructed based on the optimized node set of the aforementioned sensors, and the Edge-STGCN model is trained using a masking training method. The trained Edge-STGCN model is used to estimate the overall system operating status and determine the optimal layout scheme of the drainage network sensors.

2. The method for optimizing the layout of drainage pipe network sensors for sensing the overall system operation status as described in claim 1, characterized in that, The specific steps for determining the set of k-order neighbor nodes for each node based on the graph model are as follows: Initialize the set of 0th-order neighbors of node i: in, Let be the set of neighbor nodes of node i when the search depth is 0; based on Recursive computation node The set of k-order neighbor nodes: in, Let be the directly connected upstream and downstream nodes of node u. For nodes At a search depth of The set of neighboring nodes at that time.

3. The method for optimizing the layout of drainage pipe network sensors for sensing the overall system operation status as described in claim 2, characterized in that, The specific steps for performing redundancy removal preprocessing on the depth-sensing set are as follows: In a deep perception set, if nodes neighbor node set The set of all neighboring nodes of node j If it contains, then the node and its neighbor node set Remove from the depth-sensing set; The candidate node set obtained after redundancy processing consists of all neighbor node sets that are not completely contained by any other neighbor node set.

4. The method for optimizing the layout of drainage pipe network sensors for sensing the overall system operation status as described in claim 1, characterized in that, The greedy algorithm is applied to the candidate node set to obtain the optimal sensor node set. The specific steps are as follows: Initialize the selected sensor set and the set of covered nodes It is an empty set; Calculate the size of the neighbor set of each candidate node in the candidate node set. ; Select the candidate node with the largest neighbor node set. As a sensor node, add the sensor node to the selected sensor set. Set the neighboring nodes of this node The node is added to the set of covered nodes. ; Remove the sensor node and its neighbor node set from the candidate node set. The nodes in the list are selected, and the remaining candidate node set is updated, with any nodes that have been covered removed: in, For a set of nodes, ; Repeat the selection and update steps until the candidate node set is empty, then output the selected sensor set. As a set of nodes for optimized sensor placement.

5. The method for optimizing the layout of drainage pipe network sensors for sensing the overall system operation status as described in claim 4, characterized in that, When selecting sensor nodes, if multiple candidate nodes have the same size set of neighbor nodes, then further steps are performed: Calculate the shortest path length between these candidate nodes and nodes in the selected sensor set S; The candidate node with the largest shortest path length to the selected sensor set S is selected as the sensor node.

6. The method for optimizing the layout of drainage pipe network sensors for sensing the overall system operation status as described in claim 1, characterized in that, Before building the Edge-STGCN model, the following is also included: Acquire node monitoring data and pipeline edge attribute data of the sensor optimized layout node set; the pipeline edge attribute data includes pipe diameter, pipe length, upstream bottom elevation of the pipeline, downstream bottom elevation of the pipeline, and pipeline drainage capacity; The node monitoring data and edge attribute data are normalized separately. Based on the normalized edge attribute data, an edge feature vector is constructed to characterize the pipeline properties.

7. The method for optimizing the layout of drainage pipe network sensors for sensing the overall system operation status as described in claim 6, characterized in that, The Edge-STGCN model includes a node feature encoding layer, an edge feature encoding layer, a spatial information processing layer, a temporal information processing layer, and an information decoding layer. The spatial information processing layer adopts the GINEConv layer, which is configured to combine normalized node monitoring data and edge feature vectors to aggregate node information.

8. The method for optimizing the layout of drainage pipe network sensors for sensing the overall system operation status as described in claim 1, characterized in that, When training the Edge-STGCN model, the loss functions used include mean squared error loss and physical constraint loss, as shown in the following formulas: in, This refers to the loss function used during model training. Let the mean squared error loss function be . Let be the physical constraint loss function, and λ be an adjustable hyperparameter that balances data fidelity and physical consistency. For the set of monitoring nodes, The number of monitoring nodes selected. These are the actual monitoring values. Output values ​​for the model. The bottom elevation of the inspection well. This refers to the ground elevation of the inspection well.

9. An electronic device, characterized in that, include: A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the drainage network sensor optimization layout method for sensing the overall system operation status as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the drainage network sensor optimization layout method for sensing the overall system operation status as described in any one of claims 1 to 8.

Images