Underground pipe network intelligent management method and system integrating surveying and mapping geographic information and remote sensing

Abstract

The invention discloses an underground pipe network intelligent management method and system fusing surveying and mapping geographic information and remote sensing, and relates to the technical field of underground pipe network intelligent management. The method comprises the following steps that the leakage type of the underground pipe network is obtained, and a pipeline aging leakage characteristic factor and a pipeline sudden leakage characteristic factor are determined; surveying and mapping geographic information and remote sensing data are obtained through surveying and mapping and remote sensing respectively, an underground pipe network topological structure is constructed, and pipeline aging leakage characteristic factor data and pipeline sudden leakage characteristic factor data are obtained; respectively constructing a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model by combining Transform and an LSTM network, so as to obtain a pipeline aging leakage prediction result and a pipeline sudden leakage prediction result; and an ant colony algorithm is adopted to generate an inspection path containing a repair strategy so as to perform intelligent management on the underground pipe network. According to the invention, accurate analysis of the leakage mechanism is realized, the prediction result of leakage is improved, and the real-time performance of response is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent management of underground pipe networks, and specifically relates to an intelligent management method and system for underground pipe networks that integrates surveying and mapping geographic information and remote sensing. Background Art

[0002] With the acceleration of the urbanization process, as an important part of the urban lifeline, the safe operation and maintenance of underground pipe networks face increasingly severe challenges. In recent years, the introduction of Internet of Things, remote sensing, and artificial intelligence technologies has promoted the evolution of pipe network management towards intelligence, but there are still significant technical bottlenecks in core links such as leakage risk prediction and operation and maintenance decision-making, as follows:

[0003] (1) Existing methods do not distinguish between pipeline aging leakage and sudden leakage, and use a single model for unified processing. However, in the actual process, the physical mechanisms of these two types of leakage are significantly different: aging leakage is strongly correlated with soil humidity and the corrosion rate of pipe materials, while sudden leakage is mostly caused by pressure mutation and external construction disturbance. Therefore, using a hybrid model for the above two situations in existing methods will cause the model to be unable to accurately capture the differential features, the prediction results deviate from the actual situation, resulting in a lag in response and an increase in high operation and maintenance costs.

[0004] (2) Existing methods rely on isolated data sources, either only using GIS static topology data or only analyzing sensor time-series signals, lacking the deep integration of surveying and mapping geographic information and remote sensing dynamic data. Data fragmentation leads to the lack of key environmental parameters and limited input dimensions of the model, thus increasing the prediction error.

[0005] (3) Existing methods use a single network model to handle leakage prediction and are unable to effectively model the actual spatio-temporal state of underground pipe networks. The limitations of the model structure directly result in low spatio-temporal resolution of the prediction results, which cannot support refined operation and maintenance. Summary of the Invention

[0006] In view of the above deficiencies in the prior art, the present invention provides an intelligent management method and system for underground pipe networks that integrates surveying and mapping geographic information and remote sensing.

[0007] To achieve the above invention purpose, the technical solution adopted by the present invention is:

[0008] An intelligent management method for underground pipe networks that integrates surveying and mapping geographic information and remote sensing, comprising the following steps:

[0009] Collect historical leakage data of the underground pipe network, obtain the leakage types of the underground pipe network according to the historical leakage data of the underground pipe network. The leakage types of the underground pipe network include pipeline aging leakage and pipeline sudden leakage, and determine the pipeline aging leakage characteristic factors and pipeline sudden leakage characteristic factors respectively according to the pipeline aging leakage and the pipeline sudden leakage;

[0010] Obtain surveying and mapping geographic information and remote sensing data respectively using surveying and remote sensing. Taking inspection wells as nodes and pipeline connection relationships as edges, construct the topological structure of the underground pipeline network. According to the topological structure of the underground pipeline network, fuse the surveying and mapping geographic information and remote sensing data to obtain the multi-source fusion data of the underground pipeline network. According to the multi-source fusion data of the underground pipeline network, obtain the data of pipeline aging leakage characteristic factors and the data of pipeline sudden leakage characteristic factors;

[0011] Combine the Transformer and LSTM networks to construct a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model respectively, and train the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model according to the historical leakage data of the underground pipeline network. Use the trained pipeline aging leakage prediction model and the pipeline aging leakage characteristic factor data to obtain the pipeline aging leakage prediction result, and use the trained pipeline sudden leakage prediction model and the pipeline sudden leakage characteristic factor data to obtain the pipeline sudden leakage prediction result;

[0012] According to the topological structure of the underground pipeline network, the pipeline aging leakage prediction result and the pipeline sudden leakage prediction result, use the ant colony algorithm to generate an inspection path including repair strategies to intelligently manage the underground pipeline network.

[0013] Furthermore, the pipeline aging leakage characteristic factors include the pipeline service time, the pipeline corrosion rate, the average soil humidity around the pipeline, and the number of historical pipeline repairs; the pipeline sudden leakage characteristic factors include the surface deformation amount around the pipeline and the surface temperature change value around the pipeline.

[0014] Furthermore, obtain surveying and mapping geographic information and remote sensing data respectively using surveying and remote sensing. Taking inspection wells as nodes and pipeline connection relationships as edges, construct the topological structure of the underground pipeline network. According to the topological structure of the underground pipeline network, fuse the surveying and mapping geographic information and remote sensing data to obtain the multi-source fusion data of the underground pipeline network, including the following steps:

[0015] A1. Use surveying to obtain the location coordinates of inspection wells, pipeline connection relationships, and pipeline attributes. Taking inspection wells as nodes and pipeline connection relationships as edges, construct the topological structure of the underground pipeline network. The pipeline attributes include pipeline material, pipeline service time, and the number of historical pipeline repairs;

[0016] A2. Use remote sensing to obtain the time-series deformation data of the ground surface around the pipeline, the thermal infrared remote sensing data, and the microwave remote sensing data of the soil around the pipeline. Match the time-series deformation data of the ground surface around the pipeline, the thermal infrared remote sensing data, and the microwave remote sensing data of the soil around the pipeline to the spatial accuracy of the underground pipe network topology through Kriging spatial interpolation method, and unify the timestamps of the surveyed geographic information and remote sensing data for spatio-temporal alignment. Based on the time-series deformation data and the thermal infrared remote sensing data of the ground surface around the pipeline after spatio-temporal alignment, obtain the ground surface deformation amount around the pipeline and the ground surface temperature change value around the pipeline, and invert the microwave remote sensing data of the soil around the pipeline after spatio-temporal alignment to obtain the average soil moisture around the pipeline;

[0017] A3. Construct an indirect solution model for the corrosion rate of the underground pipe network, and calculate the pipeline corrosion rate based on the pipeline material, the average soil moisture around the pipeline, and the indirect solution model for the corrosion rate of the underground pipe network;

[0018] A4. According to the underground pipe network topology, fuse the pipeline service time, the pipeline corrosion rate, the average soil moisture around the pipeline, the pipeline historical maintenance times, the ground surface deformation amount around the pipeline, and the ground surface temperature change value around the pipeline to obtain the multi-source fusion data of the underground pipe network.

[0019] Further, in step A3, the indirect solution model for the corrosion rate is expressed as:

[0020]

[0021] Where: is the corrosion rate, is the pre-exponential factor of the corrosion rate, which is related to the pipeline material, is the average soil moisture around the pipeline, is the non-linear sensitivity coefficient of humidity to the corrosion rate, is the activation energy, which is related to the pipeline material, is the gas constant, is the absolute temperature of the ground surface around the pipeline.

[0022] Further, combine the Transformer and LSTM networks to construct a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model respectively. The specific process is as follows: Combine the Transformer and LSTM networks, and use the sequentially connected input layer, position encoding layer, Transformer encoding layer, attention weight optimization layer, LSTM decoding layer, and output layer to construct a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model respectively;

[0023] The input layer is used to input the leakage feature factor data and transmit it to the position encoding layer;

[0024] The position encoding layer is used to receive the leakage feature factor data, perform position encoding on the leakage feature factor data, and transmit the leakage feature factor data with position encoding to the Transformer encoding layer;

[0025] The Transformer encoding layer is used to perform dot product operations on the leakage feature factor data with position encoding to obtain enhanced leakage feature factor data, and transmit it to the attention weight optimization layer;

[0026] The attention weight optimization layer is used to store the memory items of the enhanced leakage feature factor data, obtain the optimized attention weight coefficients according to the memory items of the enhanced leakage feature factor data and the enhanced leakage feature factor data, and transmit the enhanced leakage feature factor data and the optimized attention weight coefficients to the LSTM decoding layer;

[0027] The LSTM decoder includes a first linear layer, a long short-term memory network layer, a second linear layer, and a fully connected layer connected in sequence; the LSTM decoding layer is used to serially process the enhanced leakage feature factor data according to the optimized attention weight coefficients to obtain the leakage prediction result, and transmit it to the output layer;

[0028] The output layer is used to output the leakage prediction result.

[0029] Furthermore, the data processing process of the attention weight optimization layer includes the following steps:

[0030] B1. According to the memory items of the enhanced leakage feature factor data and the enhanced leakage feature factor data, construct the initial attention weight coefficients, expressed as:

[0031]

[0032] Where: is the initial attention weight coefficient at the th time step, is the exponential function, is the cosine similarity between the enhanced leakage feature factor output by the Transformer encoding layer at the th time step and the memory item of the enhanced leakage feature factor data at the th time step, is the cosine similarity between the enhanced leakage feature factor output by the Transformer encoding layer at the th time step and the memory item of the enhanced leakage feature factor data at the th time step, is the number of time steps of the enhanced leakage feature factor data;

[0033] B2. Optimize the initial attention weight coefficient to obtain the optimized attention weight coefficient, expressed as:

[0034]

[0035] Where: is the optimized attention weight coefficient at the -th time step, is the RELU activation function, is the attention weight constraint threshold, which is set to , is a constant greater than 0.

[0036] Furthermore, the long short-term memory network layer includes an input gate sub-structure, a forget gate sub-structure, an output gate sub-structure, and a cell state gate sub-structure;

[0037] The data processing process of the input gate sub-structure is expressed as:

[0038] ,

[0039]

[0040] Where: is the output of the input gate sub-structure at the -th time step, is the sigmoid activation function, is the trainable weight of the optimized input gate sub-structure at the -th time step, is the concatenation symbol, is the hidden state at the -th time step, is the enhanced leakage feature factor output by the Transformer encoding layer at the -th time step, is the trainable bias of the input gate sub-structure, is the candidate value of the cell state at the -th time step, is the hyperbolic tangent function, is the trainable weight of the optimized candidate value of the cell state at the -th time step, is the trainable bias of the candidate value of the cell state;

[0041] The data processing process of the forget gate sub-structure is expressed as:

[0042]

[0043] Where: is the output of the forget gate sub-structure at the The output at a time step is the trainable weight of the optimized forget gate sub-structure at the th time step, is the concatenation symbol, is the trainable bias of the forget gate sub-structure;

[0044] The data processing process of the output gate sub-structure is expressed as:

[0045]

[0046] where: is the output of the output gate sub-structure at the th time step, is the trainable weight of the optimized output gate sub-structure at the th time step, is the trainable bias of the output gate sub-structure;

[0047] The data processing process of the cell state gate sub-structure is expressed as:

[0048] ,

[0049]

[0050] where: is the long-term memory output of the cell state gate sub-structure at the th time step, is the output of the cell state gate sub-structure at the th time step, is the short-term memory output of the cell state gate sub-structure at the th time step.

[0051] Furthermore, training the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model respectively according to the historical leakage data of the underground pipeline network includes the following steps:

[0052] C1. Determine the training set of pipeline aging leakage characteristic factor data and the training set of pipeline sudden leakage characteristic factor data respectively according to the historical leakage data of the underground pipeline network;

[0053] C2. Construct the loss function models of the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model, expressed as:

[0054]

[0055] where: is the total loss of the leakage prediction model, is the number of time steps of the training set, is the The actual leakage result of one time step, is the predicted leakage result of the th time step, is the L2 norm, is the hyperparameter of the loss function, is at the th time step, the trainable weight numbered after optimization, = 1, 2, 3, 4, respectively representing the forget gate substructure, input gate substructure, candidate value of cell state, and output gate substructure, is the logarithmic function;

[0056] C3. According to the training set of pipeline aging leakage characteristic factor data, the training set of pipeline sudden leakage characteristic factor data, and the loss function models of the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model, the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model are trained respectively.

[0057] Furthermore, according to the underground pipe network topology, the pipeline aging leakage prediction result, and the pipeline sudden leakage prediction result, the ant colony algorithm is used to generate an inspection path including repair strategies to intelligently manage the underground pipe network. The specific process is as follows: The pipeline aging leakage prediction result and the pipeline sudden leakage prediction result are weighted and fused to obtain the comprehensive pipeline leakage prediction result, and repair strategies are formulated based on the comprehensive pipeline leakage prediction result. Based on the comprehensive pipeline leakage prediction result and the underground pipe network topology, inspection path data modeling is carried out to generate a graph structure including the location coordinates of inspection wells, pipeline distances, and the comprehensive pipeline leakage prediction result. The ant colony algorithm is initialized, the pheromone and heuristic factor are set, the pheromone is associated with the comprehensive leakage prediction result, the heuristic factor is associated with the pipeline distance, and based on the repair strategy, an inspection path including the repair strategy is generated.

[0058] An underground pipe network intelligent management system integrating mapping geographic information and remote sensing applied to the above method, including a leakage analysis module, a data acquisition module, a leakage prediction module, and an intelligent management module connected in sequence;

[0059] The leakage analysis module is used to collect the historical leakage data of the underground pipe network, obtain the leakage types of the underground pipe network according to the historical leakage data of the underground pipe network. The leakage types of the underground pipe network include pipeline aging leakage and pipeline sudden leakage, and determine the pipeline aging leakage characteristic factors and pipeline sudden leakage characteristic factors according to the pipeline aging leakage and the pipeline sudden leakage respectively;

[0060] The data acquisition module is used to obtain surveying and mapping geographic information and remote sensing data respectively by means of surveying and mapping and remote sensing. Taking inspection wells as nodes and pipeline connection relationships as edges, it constructs the topological structure of the underground pipe network, fuses the surveying and mapping geographic information and remote sensing data according to the topological structure of the underground pipe network to obtain the multi-source fusion data of the underground pipe network, and obtains the data of pipeline aging leakage characteristic factors and pipeline sudden leakage characteristic factors according to the multi-source fusion data of the underground pipe network;

[0061] The leakage prediction module is used to construct a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model respectively by combining the Transformer and LSTM networks, and train the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model respectively according to the historical leakage data of the underground pipe network. It uses the trained pipeline aging leakage prediction model and the pipeline aging leakage characteristic factor data to obtain the pipeline aging leakage prediction result, and uses the trained pipeline sudden leakage prediction model and the pipeline sudden leakage characteristic factor data to obtain the pipeline sudden leakage prediction result;

[0062] The intelligent management module is used to generate an inspection path including repair strategies by using the ant colony algorithm according to the topological structure of the underground pipe network, the pipeline aging leakage prediction result and the pipeline sudden leakage prediction result, so as to intelligently manage the underground pipe network.

[0063] The beneficial effects of the present invention are as follows:

[0064] (1) By determining that the leakage types of the underground pipe network include pipeline aging leakage and pipeline sudden leakage, and respectively determining the pipeline aging leakage characteristic factors and pipeline sudden leakage characteristic factors according to the pipeline aging leakage and pipeline sudden leakage, then using surveying and mapping and remote sensing to obtain surveying and mapping geographic information and remote sensing data respectively, fusing the surveying and mapping geographic information and remote sensing data to obtain the multi-source fusion data of the underground pipe network, and combining the Transformer and LSTM networks to construct a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model respectively to obtain the pipeline aging leakage prediction result and the pipeline sudden leakage prediction result respectively, this process realizes the accurate analysis of the leakage mechanism, improves the leakage prediction result, further improves the real-time response, and reduces the operation and maintenance cost;

[0065] (2) By combining the surveying and mapping geographic information and remote sensing data, constructing the topological structure of the underground pipe network, and fusing the pipeline service time, pipeline corrosion rate, average soil humidity around the pipeline, historical pipeline repair times, ground deformation around the pipeline and ground temperature change value around the pipeline to obtain the multi-source fusion data of the underground pipe network, it realizes the deep fusion of spatial, attribute and environmental data, and further provides a data basis for the prediction of the two leakage types of the underground pipe network;

[0066] (3) The present invention constructs a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model by combining the Transformer and LSTM networks respectively. The pipeline aging leakage prediction result is obtained by using the trained pipeline aging leakage prediction model and the pipeline aging leakage characteristic factor data, and the pipeline sudden leakage prediction result is obtained by using the trained pipeline sudden leakage prediction model and the pipeline sudden leakage characteristic factor data, so as to improve the ability to capture time series features.

[0067] (4) The present invention constructs an indirect solution model for the corrosion rate, which can update the corrosion rate prediction in real time, break through the spatio-temporal limitations of traditional manual detection, and realize dynamic monitoring of the corrosion risk of the entire network in large-scale underground pipe networks, especially in hidden areas that are difficult to directly detect. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 Schematic diagram of the process of the intelligent management method for underground pipe networks integrating surveying and mapping geographic information and remote sensing;

[0069] Figure 2 Schematic diagram of the structure of the intelligent management system for underground pipe networks integrating surveying and mapping geographic information and remote sensing. DETAILED DESCRIPTION OF THE INVENTION

[0070] The following describes the specific embodiments of the present invention to facilitate those skilled in the art of the present technology to understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those of ordinary skill in the art of the present technology, as long as various changes are within the spirit and scope of the present invention defined and determined by the appended claims, these changes are obvious, and all inventions and creations using the concept of the present invention are within the scope of protection.

[0071] As Figure 1 shown, the intelligent management method for underground pipe networks integrating surveying and mapping geographic information and remote sensing includes steps S1-S4, which are specifically as follows:

[0072] S1. Collect the historical leakage data of the underground pipe network, obtain the leakage types of the underground pipe network according to the historical leakage data of the underground pipe network. The leakage types of the underground pipe network include pipeline aging leakage and pipeline sudden leakage. Determine the pipeline aging leakage characteristic factors and the pipeline sudden leakage characteristic factors respectively according to the pipeline aging leakage and the pipeline sudden leakage.

[0073] In an alternative embodiment of the present invention, the historical leakage data of the underground pipe network includes leakage type, pipeline service time, pipeline corrosion rate, average soil humidity around the pipeline, historical maintenance times of the pipeline, ground deformation around the pipeline, and ground temperature change value around the pipeline.

[0074] Specifically, the pipeline aging leakage characteristic factors determined by the present invention include the pipeline service time, the pipeline corrosion rate, the average soil humidity around the pipeline, and the number of historical pipeline repairs; the pipeline sudden leakage characteristic factors determined by the present invention include the surface deformation amount around the pipeline and the surface temperature change value around the pipeline.

[0075] S2. Use surveying and remote sensing to obtain surveying and mapping geographical information and remote sensing data respectively. Taking inspection wells as nodes and pipeline connection relationships as edges, construct an underground pipe network topology. According to the underground pipe network topology, fuse the surveying and mapping geographical information and remote sensing data to obtain multi-source fusion data of the underground pipe network. According to the multi-source fusion data of the underground pipe network, obtain pipeline aging leakage characteristic factor data and pipeline sudden leakage characteristic factor data.

[0076] In an alternative embodiment of the present invention, the present invention uses surveying and remote sensing to obtain surveying and mapping geographical information and remote sensing data respectively. Taking inspection wells as nodes and pipeline connection relationships as edges, construct an underground pipe network topology. According to the underground pipe network topology, fuse the surveying and mapping geographical information and remote sensing data to obtain multi-source fusion data of the underground pipe network, including the following steps:

[0077] A1. Use surveying to obtain the position coordinates of inspection wells, pipeline connection relationships, and pipeline attributes. Taking inspection wells as nodes and pipeline connection relationships as edges, construct an underground pipe network topology. Pipeline attributes include pipeline material, pipeline service time, and the number of historical pipeline repairs.

[0078] A2. Use remote sensing to obtain the deformation time series data of the surface around the pipeline, thermal infrared remote sensing data, and microwave remote sensing data of the soil around the pipeline. Match the deformation time series data of the surface around the pipeline, thermal infrared remote sensing data, and microwave remote sensing data of the soil around the pipeline to the spatial accuracy of the underground pipe network topology through Kriging spatial interpolation method, and unify the timestamps of the surveying and mapping geographical information and remote sensing data for spatio-temporal alignment. Based on the deformation time series data and thermal infrared remote sensing data of the surface around the pipeline after spatio-temporal alignment, obtain the surface deformation amount around the pipeline and the surface temperature change value around the pipeline, and invert the microwave remote sensing data of the soil around the pipeline after spatio-temporal alignment to obtain the average soil humidity around the pipeline.

[0079] A3. Construct an indirect solution model for the corrosion rate of the underground pipe network, and calculate the pipeline corrosion rate based on the pipeline material, the average soil humidity around the pipeline, and the indirect solution model for the corrosion rate of the underground pipe network.

[0080] The indirect solution model for the corrosion rate is expressed as:

[0081]

[0082] Where: is the corrosion rate, is the pre-exponential factor of the corrosion rate, which is related to the pipeline material. When the pipeline material is cast iron, the value is 0.15 mm / year. When the pipeline material is low-carbon steel the value range is 0.3 mm / year. When the pipeline material is PE pipe the value range is 0.01 mm / year, is the average soil humidity around the pipeline, is the non-linear sensitivity coefficient of humidity to the corrosion rate, which is obtained by fitting corrosion experiments under different humidity conditions, is the activation energy, which is related to the pipeline material, is the gas constant, and the value in the present invention is 8.314, is the absolute temperature of the ground surface around the pipeline.

[0083] When the present invention determines the area around the pipeline, taking the pipeline as the center, a surrounding area unit with a length of 10 m and a circumferential thickness of 1 m around the pipe wall is selected.

[0084] By constructing an indirect solution model for the corrosion rate, the present invention can update the corrosion rate prediction in real time, break through the spatio-temporal limitations of traditional manual detection, and realize dynamic monitoring of the corrosion risk of the entire network in large-scale underground pipe networks, especially in hidden areas that are difficult to directly detect.

[0085] A4. According to the topological structure of the underground pipe network, fuse the pipeline service time, pipeline corrosion rate, average soil humidity around the pipeline, historical pipeline repair times, ground surface deformation around the pipeline, and ground surface temperature change value around the pipeline to obtain multi-source fusion data of the underground pipe network.

[0086] S3. Combine the Transformer and LSTM networks to construct a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model respectively, and train the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model according to the historical leakage data of the underground pipe network. Use the trained pipeline aging leakage prediction model and pipeline aging leakage characteristic factor data to obtain the pipeline aging leakage prediction result, and use the trained pipeline sudden leakage prediction model and pipeline sudden leakage characteristic factor data to obtain the pipeline sudden leakage prediction result.

[0087] In an alternative embodiment of the present invention, the present invention combines the Transformer and LSTM networks to construct a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model respectively. The specific process is as follows: combining the Transformer and LSTM networks, using an input layer, a position encoding layer, a Transformer encoding layer, an attention weight optimization layer, an LSTM decoding layer, and an output layer connected in sequence to construct a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model respectively. Both the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model include an input layer, a position encoding layer, a Transformer encoding layer, an attention weight optimization layer, an LSTM decoding layer, and an output layer connected in sequence.

[0088] The input layer is used to input the leakage feature factor data and transmit it to the position encoding layer. Since the structures of the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model are the same, the present invention simplifies the description and generalizes the pipeline aging leakage feature factor data and the pipeline sudden leakage feature factor data as leakage feature factor data. The present invention inputs the pipeline aging leakage feature factor data and the pipeline sudden leakage feature factor data into the input layer of the pipeline aging leakage prediction model and the input layer of the pipeline sudden leakage prediction model respectively. The pipeline aging leakage feature factor data includes data corresponding to the pipeline service time, the pipeline corrosion rate, the average soil humidity around the pipeline, and the number of historical repairs of the pipeline. The pipeline sudden leakage feature factor data includes data corresponding to the ground deformation amount around the pipeline and the ground temperature change value around the pipeline.

[0089] The position encoding layer is used to receive the leakage feature factor data, perform position encoding on the leakage feature factor data, and transmit the leakage feature factor data with position encoding to the Transformer encoding layer. The specific processing process of the position encoding layer is as follows: calculating the odd-position encoding data of the leakage feature factor data; calculating the even-position encoding data of the leakage feature factor data; obtaining the leakage feature factor data with position encoding according to the leakage feature factor data, the odd-position encoding data of the leakage feature factor data, and the even-position encoding data of the leakage feature factor data, and transmitting it to the Transformer encoding layer.

[0090] The Transformer encoding layer is used to perform dot product operations on the leakage feature factor data with position encoding to obtain enhanced leakage feature factor data and transmit it to the attention weight optimization layer. The Transformer encoder includes a multi-head attention mechanism, a first normalization layer, a first linear addition layer, a feed-forward layer, a second normalization layer, and a second linear addition layer connected in sequence; the first linear addition layer is connected to the input of the Transformer encoder; the feed-forward layer is connected to the second linear addition layer. In the multi-head attention mechanism of the present invention, dot product operations are performed on the leakage feature factor data with position encoding. The specific process is as follows: Calculate the keyword vector, value vector, and query vector of the leakage feature factor data with position encoding; perform dot product operations on the keyword vector, value vector, and query vector of the leakage feature factor data with position encoding to obtain the enhanced leakage feature factor data of a single attention head; perform parallel calculations on the enhanced leakage feature factor data of a single attention head to obtain the initial enhanced leakage feature factor data output by the multi-head attention mechanism. Then, the initial enhanced leakage feature factor data output by the multi-head attention mechanism is processed in sequence by the first normalization layer, the first linear addition layer, the feed-forward layer, the second normalization layer, and the second linear addition layer to obtain the enhanced leakage feature factor data. The feed-forward layer includes two linear transformation layers connected by ReLU activation functions.

[0091] The attention weight optimization layer is used to store the memory items of the enhanced leakage feature factor data, obtain the optimized attention weight coefficients according to the memory items of the enhanced leakage feature factor data and the enhanced leakage feature factor data, and transmit the enhanced leakage feature factor data and the optimized attention weight coefficients to the LSTM decoding layer.

[0092] The data processing process of the attention weight optimization layer includes the following steps:

[0093] B1. According to the memory items of the enhanced leakage feature factor data and the enhanced leakage feature factor data, construct the initial attention weight coefficients, expressed as:

[0094]

[0095] Where: is the initial attention weight coefficient at the th time step, is the exponential function, is the cosine similarity between the enhanced leakage feature factor output by the Transformer encoding layer at the th time step and the memory item of the enhanced leakage feature factor data at the th time step, is the Transformer encoding layer at the The enhanced leakage feature factor output at a time step and the memory term of the enhanced leakage feature factor data at the time step cosine similarity, where is the number of time steps of the enhanced leakage feature factor data.

[0096] B2. Optimize the initial attention weight coefficient to obtain the optimized attention weight coefficient, expressed as:

[0097]

[0098] where: is the optimized attention weight coefficient at the time step, is the RELU activation function, is the attention weight constraint threshold, which is set to , is a constant greater than 0.

[0099] The LSTM decoder includes a first linear layer, a long short-term memory network layer, a second linear layer, and a fully connected layer connected in sequence; the LSTM decoding layer is used to serially process the enhanced leakage feature factor data according to the optimized attention weight coefficient to obtain the leakage prediction result and transmit it to the output layer.

[0100] The long short-term memory network layer includes an input gate substructure, a forget gate substructure, an output gate substructure, and a cell state gate substructure.

[0101] The data processing process of the input gate substructure, expressed as:

[0102] ,

[0103]

[0104] where: is the output of the input gate substructure at the time step, is the sigmoid activation function, is the trainable weight of the optimized input gate substructure at the time step, is the concatenation symbol, is the hidden state at the time step, is the enhanced leakage feature factor output by the Transformer encoding layer at the time step, is the trainable bias of the input gate substructure, is at the The candidate value of the cell state at a time step, is the hyperbolic tangent function, is the trainable weight of the optimized candidate value of the cell state at the th time step, is the trainable bias of the candidate value of the cell state.

[0105] The data processing process of the forget gate substructure is expressed as:

[0106]

[0107] where: is the output of the forget gate substructure at the th time step, is the trainable weight of the optimized forget gate substructure at the th time step, is the concatenation symbol, is the trainable bias of the forget gate substructure.

[0108] The data processing process of the output gate mechanism is expressed as:

[0109]

[0110] where: is the output of the output gate substructure at the th time step, is the trainable weight of the optimized output gate mechanism at the th time step, is the trainable bias of the output gate mechanism.

[0111] The data processing process of the cell state gate substructure is expressed as:

[0112] ,

[0113]

[0114] where: is the long-term memory output of the cell state gate substructure at the th time step, is the output of the cell state gate substructure at the th time step, is the short-term memory output of the cell state gate substructure at the th time step.

[0115] Before the output result of the reconstructed LSTM decoding layer, the present invention introduces a fully connected layer in the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model to improve the robustness of the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model and the reliability of the prediction data result.

[0116] The output layer is used to output the leakage prediction result.

[0117] The present invention trains the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model respectively according to the historical leakage data of the underground pipe network, including the following steps:

[0118] C1. Determine the training set of pipeline aging leakage characteristic factor data and the training set of pipeline sudden leakage characteristic factor data respectively according to the historical leakage data of the underground pipe network.

[0119] C2. Construct the loss function models of the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model, expressed as:

[0120]

[0121] Where: is the total loss of the leakage prediction model, is the number of time steps of the training set, is the actual leakage result at the th time step, is the leakage prediction result at the th time step, is the L2 norm, is the hyperparameter of the loss function, is the optimized number at the th time step numbered trainable weight, = 1, 2, 3, 4, respectively representing the forget gate substructure, the input gate substructure, the candidate value of the cell state, and the output gate substructure, is the logarithmic function.

[0122] C3. Train the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model respectively according to the training set of pipeline aging leakage characteristic factor data, the training set of pipeline sudden leakage characteristic factor data, and the loss function models of the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model.

[0123] S4. According to the underground pipe network topology, the pipeline aging leakage prediction result and the pipeline sudden leakage prediction result, use the ant colony algorithm to generate an inspection path including repair strategies to intelligently manage the underground pipe network.

[0124] In an alternative embodiment of the present invention, the present invention generates an inspection path including repair strategies by using the ant colony algorithm according to the underground pipe network topology, the predicted results of pipeline aging leakage, and the predicted results of pipeline sudden leakage, so as to perform intelligent management on the underground pipe network. The specific process is as follows: The predicted results of pipeline aging leakage and the predicted results of pipeline sudden leakage are weighted and fused to obtain the comprehensive pipeline leakage prediction results, and repair strategies are formulated based on the comprehensive pipeline leakage prediction results. Based on the comprehensive pipeline leakage prediction results and the underground pipe network topology, inspection path data modeling is performed to generate a graph structure including the location coordinates of manholes, the pipeline distance, and the comprehensive pipeline leakage prediction results. The ant colony algorithm is initialized, the pheromone and the heuristic factor are set, the pheromone is associated with the comprehensive leakage prediction results, the heuristic factor is associated with the pipeline distance, and an inspection path including repair strategies is generated based on the repair strategies.

[0125] The present invention initializes the ant colony algorithm, sets the pheromone and the heuristic factor, associates the pheromone with the comprehensive leakage prediction results, associates the heuristic factor with the pipeline distance, and generates an inspection path including repair strategies based on the repair strategies. The specific process is as follows: The ant colony algorithm is initialized, the pheromone and the heuristic factor are set. If the risk in the comprehensive leakage prediction results is high, the pheromone is strong, and thus the pheromone is associated with the comprehensive leakage prediction results. If the pipeline distance is short, the heuristic value is large, and thus the heuristic factor is associated with the pipeline distance. An association function of the node selection probability with the pipeline distance and the comprehensive leakage prediction results is established to obtain the node selection probability. Then, the ants in the ant colony algorithm select the next manhole according to the node selection probability to generate an inspection path, and an inspection path including repair strategies is generated based on the repair strategies.

[0126] As Figure 2 shown, the intelligent management system for underground pipe networks integrating surveying and mapping geographic information and remote sensing applied to the above method includes a leakage analysis module, a data acquisition module, a leakage prediction module, and an intelligent management module connected in sequence.

[0127] In an alternative embodiment of the present invention, the leakage analysis module is used to collect the historical leakage data of the underground pipe network, obtain the leakage types of the underground pipe network according to the historical leakage data of the underground pipe network. The leakage types of the underground pipe network include pipeline aging leakage and pipeline sudden leakage. The pipeline aging leakage characteristic factors and the pipeline sudden leakage characteristic factors are determined respectively according to the pipeline aging leakage and the pipeline sudden leakage.

[0128] In an alternative embodiment of the present invention, the data acquisition module is used to respectively obtain surveying and mapping geographical information and remote sensing data by means of surveying and mapping and remote sensing, construct an underground pipe network topological structure with inspection wells as nodes and pipe connection relationships as edges, fuse the surveying and mapping geographical information and remote sensing data according to the underground pipe network topological structure to obtain multi-source fusion data of the underground pipe network, and obtain pipeline aging leakage characteristic factor data and pipeline sudden leakage characteristic factor data according to the multi-source fusion data of the underground pipe network.

[0129] In an alternative embodiment of the present invention, the leakage prediction module is used to respectively construct a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model by combining the Transformer and LSTM networks, train the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model respectively according to the historical leakage data of the underground pipe network, obtain the pipeline aging leakage prediction result by using the trained pipeline aging leakage prediction model and the pipeline aging leakage characteristic factor data, and obtain the pipeline sudden leakage prediction result by using the trained pipeline sudden leakage prediction model and the pipeline sudden leakage characteristic factor data.

[0130] In an alternative embodiment of the present invention, the intelligent management module is used to generate an inspection path including repair strategies by using the ant colony algorithm according to the underground pipe network topological structure, the pipeline aging leakage prediction result and the pipeline sudden leakage prediction result, so as to perform intelligent management on the underground pipe network.

[0131] Those of ordinary skill in the art will realize that the embodiments described herein are for helping the reader understand the principles of the present invention, and it should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific deformations and combinations that do not depart from the essence of the present invention based on these technical revelations disclosed in the present invention, and these deformations and combinations are still within the protection scope of the present invention.

Claims

1. An intelligent management method for underground pipe networks integrating surveying and mapping geographic information and remote sensing, characterized in that, It includes the following steps: Collect historical leakage data of the underground pipe network, obtain the leakage types of the underground pipe network according to the historical leakage data of the underground pipe network. The leakage types of the underground pipe network include pipeline aging leakage and pipeline sudden leakage. Determine the pipeline aging leakage characteristic factors and pipeline sudden leakage characteristic factors respectively according to the pipeline aging leakage and pipeline sudden leakage; Use surveying and remote sensing to obtain surveying geographical information and remote sensing data respectively. Take the inspection wells as nodes and the pipeline connection relationships as edges to construct the topological structure of the underground pipe network. Integrate the surveying geographical information and remote sensing data according to the topological structure of the underground pipe network to obtain the multi-source fusion data of the underground pipe network. Obtain the pipeline aging leakage characteristic factor data and pipeline sudden leakage characteristic factor data according to the multi-source fusion data of the underground pipe network; Construct a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model by combining the Transformer and LSTM networks respectively, and train the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model respectively according to the historical leakage data of the underground pipe network. Use the trained pipeline aging leakage prediction model and the pipeline aging leakage characteristic factor data to obtain the pipeline aging leakage prediction result, and use the trained pipeline sudden leakage prediction model and the pipeline sudden leakage characteristic factor data to obtain the pipeline sudden leakage prediction result; According to the topological structure of the underground pipe network, the pipeline aging leakage prediction result and the pipeline sudden leakage prediction result, use the ant colony algorithm to generate an inspection path including repair strategies to perform intelligent management on the underground pipe network.

2. The intelligent management method for underground pipe networks integrating surveying and mapping geographic information and remote sensing according to claim 1, characterized in that, The pipeline aging leakage characteristic factors include the pipeline service time, the pipeline corrosion rate, the average soil humidity around the pipeline, and the number of historical pipeline repairs; the pipeline sudden leakage characteristic factors include the surface deformation amount around the pipeline and the surface temperature change value around the pipeline.

3. The intelligent management method for underground pipe networks integrating surveying and mapping geographic information and remote sensing according to claim 1, characterized in that, Use surveying and remote sensing to obtain surveying geographical information and remote sensing data respectively. Take the inspection wells as nodes and the pipeline connection relationships as edges to construct the topological structure of the underground pipe network. Integrate the surveying geographical information and remote sensing data according to the topological structure of the underground pipe network to obtain the multi-source fusion data of the underground pipe network, including the following steps: A1. Use surveying to obtain the location coordinates of the inspection wells, the pipeline connection relationships and the pipeline attributes. Take the inspection wells as nodes and the pipeline connection relationships as edges to construct the topological structure of the underground pipe network. The pipeline attributes include the pipeline material, the pipeline service time and the number of historical pipeline repairs; A2. Use remote sensing to obtain the deformation time series data of the surface around the pipeline, the thermal infrared remote sensing data and the microwave remote sensing data of the soil around the pipeline. Match the deformation time series data of the surface around the pipeline, the thermal infrared remote sensing data and the microwave remote sensing data of the soil around the pipeline to the spatial accuracy of the topological structure of the underground pipe network through the Kriging spatial interpolation method, and unify the timestamps of the surveying geographical information and the remote sensing data for spatio-temporal alignment. Obtain the surface deformation amount around the pipeline and the surface temperature change value around the pipeline based on the deformation time series data of the surface around the pipeline and the thermal infrared remote sensing data after spatio-temporal alignment, and perform inversion on the microwave remote sensing data of the soil around the pipeline after spatio-temporal alignment to obtain the average soil humidity around the pipeline; A3. Construct an indirect solution model for the corrosion rate of the underground pipe network, and calculate the pipe corrosion rate based on the pipe material, the average soil moisture around the pipe, and the indirect solution model for the corrosion rate of the underground pipe network; A4. According to the topological structure of the underground pipe network, fuse the pipe service time, the pipe corrosion rate, the average soil moisture around the pipe, the historical maintenance times of the pipe, the ground deformation amount around the pipe, and the ground temperature change value around the pipe to obtain the multi-source fusion data of the underground pipe network.

4. The intelligent management method for underground pipe networks integrating surveying and mapping geographic information and remote sensing according to claim 3, characterized in that, In step A3, the indirect solution model for the corrosion rate is expressed as: ; Wherein: is the corrosion rate, is the pre-exponential factor of the corrosion rate, which is related to the pipeline material, is the average soil humidity around the pipeline, is the non-linear sensitivity coefficient of humidity to the corrosion rate, is the activation energy, which is related to the pipeline material, is the gas constant, is the absolute temperature of the ground surface around the pipeline.

5. The intelligent management method for underground pipe networks integrating surveying and mapping geographical information and remote sensing according to claim 1, characterized in that, Combine the Transformer and LSTM networks to construct a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model respectively. The specific process is as follows: Combine the Transformer and LSTM networks, and use the sequentially connected input layer, position encoding layer, Transformer encoding layer, attention weight optimization layer, LSTM decoding layer, and output layer to construct a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model respectively; The input layer is used to input the leakage feature factor data and transmit it to the position encoding layer; The position encoding layer is used to receive the leakage feature factor data, perform position encoding on the leakage feature factor data, and transmit the leakage feature factor data with position encoding to the Transformer encoding layer; The Transformer encoding layer is used to perform dot product operations on the leakage feature factor data with position encoding to obtain enhanced leakage feature factor data and transmit it to the attention weight optimization layer; The attention weight optimization layer is used to store the memory items of the enhanced leakage feature factor data, obtain the optimized attention weight coefficient according to the memory items of the enhanced leakage feature factor data and the enhanced leakage feature factor data, and transmit the enhanced leakage feature factor data and the optimized attention weight coefficient to the LSTM decoding layer; The LSTM decoder includes a first linear layer, a long short-term memory network layer, a second linear layer, and a fully connected layer connected in sequence; the LSTM decoding layer is used to serially process the enhanced leakage feature factor data according to the optimized attention weight coefficient to obtain the leakage prediction result and transmit it to the output layer; The output layer is used to output the leakage prediction result.

6. The intelligent management method of underground pipe network integrating surveying and mapping geographic information and remote sensing according to claim 5, wherein, The data processing process of the attention weight optimization layer includes the following steps: B1. According to the memory items of the enhanced leakage feature factor data and the enhanced leakage feature factor data, construct the initial attention weight coefficient, which is expressed as: Wherein: is the initial attention weight coefficient at the th time step, is the exponential function, is the cosine similarity between the enhanced leakage feature factor output by the Transformer encoding layer at the th time step and the memory item of the enhanced leakage feature factor data at the th time step, is the cosine similarity between the enhanced leakage feature factor output by the Transformer encoding layer at the th time step and the memory item of the enhanced leakage feature factor data at the th time step, is the number of time steps of the enhanced leakage feature factor data; B2. Optimize the initial attention weight coefficient to obtain the optimized attention weight coefficient, which is expressed as: Wherein: is the optimized attention weight coefficient at the th time step, is the RELU activation function, is the attention weight constraint threshold, which is set to , is a constant greater than 0.

7. The intelligent management method for underground pipe networks integrating surveying and mapping geographic information and remote sensing according to claim 6, wherein The long short-term memory network layer includes an input gate sub-structure, a forget gate sub-structure, an output gate sub-structure, and a cell state gate sub-structure; The data processing process of the input gate sub-structure is expressed as: ， Wherein: is the output of the input gate structure at the th time step, is the sigmoid activation function, is the trainable weight of the optimized input gate structure at the th time step, is the concatenation symbol, is the th hidden state at the time step, is the enhanced leakage feature factor output by the Transformer encoding layer at the th time step, is the trainable bias of the input gate structure, is at the th candidate value of the cell state at the time step, is the hyperbolic tangent function, is the trainable weight of the optimized candidate value of the cell state at the th time step, is the trainable bias of the candidate value of the cell state; The data processing process of the forget gate sub-structure is expressed as: Wherein: is the output of the forget gate substructure at the -th time step, is the trainable weight of the optimized forget gate substructure at the -th time step, is the concatenation symbol, is the trainable bias of the forget gate substructure; The data processing process of the output gate mechanism is expressed as: Wherein: is the output of the output gate sub-structure at the th time step, is the trainable weight of the optimized output gate mechanism at the th time step, is the trainable bias of the output gate mechanism; The data processing process of the cell state gate sub-structure is expressed as: ， Wherein: is the long-term memory output of the unit state gate sub-structure at the th time step, is the output of the unit state gate sub-structure at the th time step, is the short-term memory output of the unit state gate sub-structure at the th time step.

8. The intelligent management method for underground pipe networks integrating surveying and mapping geographical information and remote sensing according to claim 1, wherein Train the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model respectively according to the historical leakage data of the underground pipe network, including the following steps: C1. Determine the training set of pipeline aging leakage characteristic factor data and the training set of pipeline sudden leakage characteristic factor data respectively according to the historical leakage data of the underground pipe network; C2. Construct the loss function models of the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model, expressed as: Wherein: is the total loss of the leakage prediction model; is the number of time steps of the training set; is the actual leakage result at the -th time step; is the predicted leakage result at the -th time step; is the L2 norm; is the hyperparameter of the loss function; is the -th optimized trainable weight numbered ; = 1, 2, 3, 4, respectively representing the forget gate sub-structure, input gate sub-structure, candidate value of the cell state, and output gate sub-structure; is the logarithmic function; C3. Train the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model respectively according to the training set of pipeline aging leakage characteristic factor data, the training set of pipeline sudden leakage characteristic factor data, and the loss function models of the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model.

9. The intelligent management method for underground pipe networks integrating surveying and mapping geographic information and remote sensing according to claim 1, characterized in that According to the topological structure of the underground pipe network, the pipeline aging leakage prediction result, and the pipeline sudden leakage prediction result, use the ant colony algorithm to generate an inspection path including repair strategies to perform intelligent management on the underground pipe network. The specific process is as follows: perform weighted fusion on the pipeline aging leakage prediction result and the pipeline sudden leakage prediction result to obtain the comprehensive pipeline leakage prediction result, formulate repair strategies based on the comprehensive pipeline leakage prediction result, perform inspection path data modeling based on the comprehensive pipeline leakage prediction result and the topological structure of the underground pipe network, and generate a graph structure including the location coordinates of inspection wells, the pipeline distance, and the comprehensive pipeline leakage prediction result; initialize the ant colony algorithm, set the pheromone and heuristic factor, associate the pheromone with the comprehensive leakage prediction result, associate the heuristic factor with the pipeline distance, and generate an inspection path including repair strategies based on the repair strategies.

10. An intelligent management system for underground pipe networks integrating surveying and mapping geographic information and remote sensing, which is applied to the method according to any one of claims 1-9, characterized in that It includes a leakage analysis module, a data acquisition module, a leakage prediction module, and an intelligent management module connected in sequence; The leakage analysis module is used to collect the historical leakage data of the underground pipe network, obtain the leakage types of the underground pipe network according to the historical leakage data of the underground pipe network. The leakage types of the underground pipe network include pipeline aging leakage and pipeline sudden leakage, and determine the pipeline aging leakage characteristic factors and the pipeline sudden leakage characteristic factors respectively according to the pipeline aging leakage and the pipeline sudden leakage; The data acquisition module is used to obtain surveying and mapping geographic information and remote sensing data respectively by using surveying and mapping and remote sensing, construct the topological structure of the underground pipe network with inspection wells as nodes and pipeline connection relationships as edges, fuse the surveying and mapping geographic information and the remote sensing data according to the topological structure of the underground pipe network to obtain the multi-source fusion data of the underground pipe network, and obtain the pipeline aging leakage characteristic factor data and the pipeline sudden leakage characteristic factor data according to the multi-source fusion data of the underground pipe network; The leakage prediction module is used to construct a pipeline aging leakage prediction model and a pipeline sudden leakage prediction model respectively by combining the Transformer and LSTM networks, train the pipeline aging leakage prediction model and the pipeline sudden leakage prediction model respectively according to the historical leakage data of the underground pipe network, obtain the pipeline aging leakage prediction result by using the trained pipeline aging leakage prediction model and the pipeline aging leakage characteristic factor data, and obtain the pipeline sudden leakage prediction result by using the trained pipeline sudden leakage prediction model and the pipeline sudden leakage characteristic factor data; The intelligent management module is used to generate an inspection path including repair strategies by using the ant colony algorithm according to the underground pipe network topology, the prediction results of pipeline aging and leakage, and the prediction results of pipeline sudden leakage, so as to conduct intelligent management of the underground pipe network.

Images