Buried pipeline routing recommendation method based on artificial intelligence

Abstract

This invention discloses an artificial intelligence-based method for recommending buried pipeline layout, belonging to the field of intelligent recommendation technology. The method includes: converting the laying environment, construction purpose, and specific parameters of the buried pipeline into input vectors and inputting them into a pre-trained buried pipeline layout recommendation model to obtain an initial recommended layout scheme; mining historical new problems and solutions for each historical pipeline project during the pipeline laying process, forming several problem-strategy pairs; using a hierarchical clustering algorithm to perform cluster analysis on all problem-strategy pairs under all historical pipeline projects; calculating the similarity between the current buried pipeline laying environment and construction purpose and each historical pipeline project based on an autoencoder similarity algorithm to obtain the required retained problem-strategy pairs; and improving the initial recommended layout scheme according to the layout improvement type. The pipeline layout scheme is optimized and improved to generate a layout scheme that meets actual needs while also considering potential risk prevention.

Description

Technical Field

[0001] This invention relates to the field of intelligent recommendation technology, and in particular to an artificial intelligence-based method for recommending the layout of buried pipelines. Background Technology

[0002] As a crucial component of modern urban infrastructure, urban underground integrated buried pipelines serve to centrally lay various municipal pipelines, including power, communication, gas, and water supply and drainage lines. They are of great significance for ensuring urban operational safety and improving urban space utilization. During the construction of buried pipelines, the rationality of the pipeline layout plan directly affects the operation and maintenance efficiency and operational safety of the pipelines.

[0003] Currently, traditional buried pipeline layout designs rely heavily on engineers' experience and manual planning, making it difficult to comprehensively and accurately consider the laying environment, construction objectives, and specific parameters of various pipeline types. Furthermore, the lack of systematic analysis and effective utilization of problems and solutions encountered in past projects during pipeline layout leads to issues such as pipeline space conflicts, high construction difficulty, and inconvenient maintenance during actual construction and operation. Moreover, existing technologies struggle to extract valuable information from historical experience to optimize and improve pipeline layout plans based on current conditions, thus failing to generate layout schemes that meet actual needs while also mitigating potential risks. Summary of the Invention

[0004] To address at least one of the aforementioned technical problems, this invention provides an artificial intelligence-based method for recommending buried pipeline layout.

[0005] In a first aspect, the present invention provides an artificial intelligence-based method for recommending the layout of buried pipelines, the method comprising:

[0006] Collect the current laying environment of the buried pipeline, the purpose of the buried pipeline construction, and obtain the specific parameters of a single type of pipeline applied to the current buried pipeline, and convert them into an input vector;

[0007] The input vector is fed into a pre-trained buried pipeline layout recommendation model to obtain an initial recommended layout scheme;

[0008] The project identifies new historical problems and solutions for each historical pipeline project during the pipeline deployment process, forming several problem-strategy pairs. A hierarchical clustering algorithm is then used to perform cluster analysis on the problem-strategy pairs under all historical pipeline projects.

[0009] The similarity between the current buried pipeline laying environment and the buried pipeline construction purpose and each historical pipeline laying project is calculated based on the autoencoder similarity algorithm.

[0010] Retain the problem-strategy pairs of historical routing projects with a similarity greater than or equal to the preset level;

[0011] Each historical deployment project with a similarity less than a preset degree is determined based on the distribution probability in all cluster analysis results, and the central cluster of the cluster analysis result corresponding to the maximum number of problem-strategy pairs involved in each cluster analysis result of the historical deployment project with the highest probability in the distribution probability is retained;

[0012] Based on all the layout improvement types of the retained results, the initial recommended layout scheme is improved.

[0013] Preferably, the laying environment includes the topography, above-ground attachments, and underground obstacles of each individual pipeline section;

[0014] The specific parameters include: pipe specifications, pipe material, and pipe weight;

[0015] The improved layout scheme includes: the pipe laying sequence under each individual pipe section, the starting pipe laying position, and the pipe laying auxiliary methods for each pipe laying section. The pipe laying auxiliary methods include: buried pipe channel transportation method, hoisting equipment auxiliary method, pulley block method, and / or pipe jacking construction method.

[0016] Preferably, the input vector is converted to include:

[0017] Numerical standardization processing is performed on the numerical data involved in the laying environment, the purpose of buried pipeline construction, and specific parameters.

[0018] Encode non-numerical data;

[0019] The transformation results for the laying environment, the purpose of buried pipeline construction, and the specific parameters are extracted from the numerical standardization and coding results, respectively, and then combined and sorted in order to obtain the input vector.

[0020] Preferably, the similarity between the current buried pipeline's laying environment and construction purpose and each historical pipeline laying project is calculated based on an autoencoder similarity algorithm, including:

[0021] The first and second characteristics of the current buried pipeline laying environment and the buried pipeline construction purpose are obtained respectively. At the same time, the standard combination characteristics of the historical pipeline laying projects and the result combination characteristics after project implementation are obtained.

[0022] Calculate the first Euclidean distance between the first and second representations and the standard combined representation;

[0023] Calculate the second Euclidean distance between the first representation and the second representation and the combined representation of the result;

[0024] Simultaneously, the third Euclidean distance between the standard combination representation and the result combination representation is calculated;

[0025] If the absolute difference between the first Euclidean distance and the second Euclidean distance is less than the third Euclidean distance, determine the average value of the third Euclidean distance and the absolute difference between the distances, and take the result of the ratio of the average value to the maximum preset distance as the similarity.

[0026] Otherwise, the similarity is determined by the ratio of the average distance between the first and second Euclidean distances (1 - the first Euclidean distance and the second Euclidean distance) to the maximum preset distance.

[0027] Preferably, before determining the distribution probability of historical pipeline projects with similarity less than a preset value in all cluster analysis results, the method further includes:

[0028] For each historical deployment project with a similarity less than the preset value, the problem-strategy pair is marked with significance in the cluster analysis results, and the first number in each cluster analysis result is counted;

[0029] Calculate the first ratio between the first quantity and the total number of problem-strategy pairs involved in the corresponding cluster analysis results;

[0030] Calculate the sum of all first ratios, and extract the largest ratio from all first ratios;

[0031] Sort the sums and maximum ratios of all historical pipe laying projects with similarity less than the preset value in descending order, and construct a two-dimensional array, where each value occupies one cell;

[0032] Calculate the first average of the absolute values ​​of the differences between two adjacent sums in the row array related to the sum in the second array, and calculate the second ratio of the first average to the average of all first quantities under the corresponding historical pipeline projects, as the second average.

[0033] Calculate the third average of the absolute values ​​of the differences between two adjacent maximum ratios in the row array related to the maximum ratio in the second array;

[0034] If the second average is greater than or equal to the third average, determine the third ratio between the second average and the third average, and at the same time, determine the fourth ratio between the absolute value of the difference between two adjacent sums and the second average.

[0035] Round down the product of the third and fourth ratios and insert blank cells in the middle of the corresponding two adjacent sums, matching the rounded result.

[0036] Update the index of the rows in the array where blank cells are inserted;

[0037] Based on the total number of blank cells inserted into the row array after the sequence number is updated, starting with one blank cell, the row arrays related to the corresponding maximum ratio are cyclically inserted one blank cell from the middle of adjacent maximum ratios until the number of insertions matches the total number. Then, the insertion stops, and the sequence number is updated.

[0038] Otherwise, leave the two-dimensional array unchanged.

[0039] Preferably, determining the distribution probability of each historical pipeline project with a similarity less than a preset value in all cluster analysis results includes:

[0040] Extract each historical pipe laying project with a similarity less than the preset value from the final array, based on the first index of the first row and the second index of the second row;

[0041] The result of 1 - (first serial number + second serial number) / (2 × total serial number) is taken as the corresponding probability distribution.

[0042] Preferably, based on all retained results, the layout improvement type improves the initial recommended layout scheme, including:

[0043] Map the solution type of each problem-policy pair in all retained results to the structure type of each individual pipeline in the initial recommended layout scheme to obtain the mapping pair for each individual pipeline.

[0044] The mapping pairs are fused with the sub-recommended layout schemes of each individual pipeline based on the rule fusion algorithm to improve the scheme.

[0045] In a second aspect, the present invention also provides an electronic device, comprising: a processor and a memory, the memory being used to store computer program code, the computer program code including computer instructions, wherein when the processor executes the computer instructions, the electronic device executes an artificial intelligence-based buried pipeline layout recommendation method as described in the first aspect above and any possible implementation thereof.

[0046] Thirdly, the present invention also provides a computer-readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor of an electronic device, cause the processor to perform an artificial intelligence-based method for recommending buried pipeline layout as described in the first aspect above and any possible implementation thereof.

[0047] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0048] (1) By collecting the current laying environment, construction purpose and specific parameters of the buried pipeline and converting them into input vectors, and inputting them into the pre-trained buried pipeline layout recommendation model, an initial recommended layout scheme based on the actual situation can be generated. This fully considers the various characteristics of buried pipelines and pipelines, making the layout scheme more in line with actual needs and improving the rationality and scientificity of the scheme.

[0049] (2) Identify problem-strategy pairs from historical pipeline projects and perform cluster analysis. Then, filter relevant information based on the similarity between the current buried pipeline and historical projects. This allows us to learn from past experiences and lessons, avoid repeating similar problems, and provide optimization ideas for the current pipeline layout, thus helping to improve the quality and reliability of the pipeline layout plan.

[0050] (3) Based on the autoencoder similarity algorithm, similarity can be calculated to accurately find historical pipeline projects similar to the current buried pipeline situation and retain relevant problem-strategy pairs. For projects with low similarity, representative cluster analysis center clusters are found by determining the distribution probability, thereby obtaining valuable problem-strategy pairs. This approach can address potential problems that may be encountered in the current pipeline layout in a targeted manner, improving the stability and feasibility of the solution.

[0051] (4) The initial recommended layout scheme is improved by combining all the layout improvement types of the retained results. This realizes the intelligent process of the whole process from data collection and model recommendation to optimization based on historical experience, reduces the influence of manual intervention and subjective factors, improves the efficiency and accuracy of pipeline design, and generates a pipeline scheme that meets actual needs and takes into account potential risk prevention and control.

[0052] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of the present invention or the background art, the accompanying drawings used in the embodiments of the present invention or the background art will be described below.

[0054] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the specification, serve to illustrate the technical solutions of this disclosure.

[0055] Figure 1 A flowchart illustrating an artificial intelligence-based buried pipeline layout recommendation method provided in an embodiment of the present invention;

[0056] Figure 2 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0057] To enable those skilled in the art to better understand the present invention, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0058] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0059] In this document, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" in this document means any combination of at least two of any one or more elements. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0060] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0061] Furthermore, to better illustrate the present invention, numerous specific details are set forth in the following detailed embodiments. Those skilled in the art will understand that the present invention can be practiced without certain specific details. In some instances, methods, means, elements, and circuits well known to those skilled in the art have not been described in detail in order to highlight the spirit of the invention.

[0062] This invention provides an artificial intelligence-based method for recommending buried pipeline layout, such as... Figure 1 As shown, the method includes:

[0063] Step 1: Collect the current laying environment of the buried pipeline, the purpose of the buried pipeline construction, and obtain the specific parameters of the single type of pipeline applied to the current buried pipeline, and convert them into an input vector;

[0064] Currently, buried pipelines refer to buried pipeline projects that are under planning, construction, or evaluation. For example, a newly built underground integrated buried pipeline in a city is used to centrally lay multiple pipelines such as power, communication, and water supply.

[0065] The laying environment refers to the natural and artificial environmental conditions of the area where buried pipelines are laid, including:

[0066] Topography and landforms, such as mountains (slope 20°, elevation difference 100 meters), plains (average elevation 50 meters), and swamps (water content 80%).

[0067] The existing above-ground attachments are above-ground utility tunnels (such as concrete structures, with a foundation depth of 2 meters and a plan dimension of 10m×5m; protection requirements: mechanical excavation is prohibited within 3 meters of the perimeter), buildings, and roads (two-way six lanes, with a traffic flow of 2000 vehicles / hour).

[0068] Information on underground obstacles includes underground pipelines (such as DN300 cast iron water supply pipes, buried at a depth of 1.5 meters), cable trenches (1 meter wide, buried at a depth of 0.8 meters), and underground structures (inspection wells, 2 meters in diameter, buried at a depth of 3 meters).

[0069] The purpose of construction is to meet the core requirements of pipeline laying, such as:

[0070] New municipal sewage pipelines (serving a population of 100,000), expansion of pipelines in old industrial areas (increasing oil transport capacity by 20%), and replacement of emergency rescue pipelines (repairing ruptured gas pipelines).

[0071] Specific parameters for a single type of pipe, such as specifications: DN500 (diameter 500mm), wall thickness 10mm; material: PE (polyethylene), Q235B steel; weight: PE pipe 15kg / meter, steel pipe 40kg / meter; burial depth requirements: sewage pipe burial depth ≥ 3.5 meters (below the antifreeze layer), gas pipe burial depth ≥ 1.2 meters (standard requirements). The input vector converts the above data into a numerical vector recognizable by the model. This is achieved through: numerical standardization using Z-score standardization for terrain elevation (e.g., 50 meters): (50 - average elevation 45) / standard deviation 2.5 = 2; and normalization for burial depth (3.5 meters): 3.5 / maximum burial depth 5 = 0.7.

[0072] The non-numerical coding is [1,0] for the above-ground pipe gallery type "concrete structure" (assuming two types: concrete / steel structure); and [0,1,0] for the crossing method "pipe jacking" (assuming three types: pipe jacking / directional drilling / open excavation).

[0073] Vector combination is concatenated in the order of "laying environment - construction purpose - pipeline parameters", for example: [normalized value of terrain slope, code of above-ground pipe gallery, coordinates of underground obstacles, code of construction purpose, normalized value of pipe diameter, material code, normalized value of burial depth].

[0074] Step 2: Input the input vector into the pre-trained buried pipeline layout recommendation model to obtain the initial recommended layout scheme;

[0075] The pre-trained buried pipeline layout recommendation model is a combination of convolutional neural networks and recurrent neural networks. It utilizes a large amount of historical buried pipeline layout data and adjusts model parameters through backpropagation to optimize model performance, enabling it to accurately output initial plans based on input. The input samples for training this model include: the laying environment of historical buried pipelines, the purpose of buried pipeline construction, and input vectors for specific parameter transformations. The output samples are the pipeline layout plans set by pipeline experts based on the historical buried pipeline laying environment, the purpose of buried pipeline construction, and specific parameter transformations.

[0076] The initial recommended layout scheme includes the pipeline route (e.g., laid 2 meters outside the right side of the road red line), the pipeline laying order (sewage pipe below, water supply pipe above), the initial crossing method (large-scale excavation for road crossing), and the burial depth suggestion (3.8 meters). For example, the initial recommended route for a gas pipeline in an industrial park is to lay along the east side of the main road, and to use directional drilling to cross the intersection, with a burial depth of 1.5 meters.

[0077] Step 3: Discover the historical new problems and historical solutions to the new problems in the deployment process of each historical pipeline project, form several problem-strategy pairs, and use hierarchical clustering algorithm to perform cluster analysis on the problem-strategy pairs under all historical pipeline projects;

[0078] Construction records and issue feedback documents for each historical pipe laying project are extracted from the company's historical project database. Newly added issues and corresponding historical solutions are identified, forming issue-strategy pairs. Hierarchical clustering algorithms are used to calculate the similarity between these pairs based on their textual or numerical features. Similar issue-strategy pairs are grouped together to form clusters. An issue-strategy pair represents a solution to a specific problem encountered in a pipe laying project; the two constitute a combination. For example, "Insufficient pipe space (problem) - Adopt a multi-layer layout (strategy)."

[0079] Natural language processing technology is used to segment, tag, and semantically analyze text information in historical documents to extract key issues and strategy information.

[0080] The algorithm calculates the Euclidean distance between data points based on hierarchical clustering, and merges or splits clusters according to the hierarchy from bottom to top or top to bottom to explore the inherent structure of the data and provide support for subsequent reference.

[0081] Step 4: Calculate the similarity between the current buried pipeline laying environment and the buried pipeline construction purpose and each historical pipeline laying project based on the autoencoder similarity algorithm;

[0082] An autoencoder network is constructed, and data on the current buried pipeline's laying environment and construction purpose, as well as data from each historical pipeline project, are input into the autoencoder to extract low-dimensional feature vectors. The similarity between the feature vectors of the current buried pipeline and those of historical pipeline projects is measured using methods such as Euclidean distance.

[0083] An autoencoder is an unsupervised learning neural network that learns effective feature representations of data through encoding and decoding processes, enabling feature extraction and dimensionality reduction. Euclidean distance reflects the spatial distance between vectors, helping to quantify the similarity between current buried pipelines and historical projects, and providing a basis for selecting valuable historical experience.

[0084] Step 5: Retain the issue-strategy pairs of historical pipe deployment projects with a similarity greater than or equal to the preset level;

[0085] The program uses simple numerical comparison logic to compare similarity with a preset similarity and filter data through conditional statements, with the preset similarity value set to 0.6.

[0086] Step 6: Determine the distribution probability of each historical deployment project with a similarity less than a preset degree based on all cluster analysis results, and retain the central cluster of the cluster analysis result corresponding to the maximum number of problem-strategy pairs involved in each cluster analysis result for the historical deployment project with the highest probability among the distribution probabilities;

[0087] By using probability calculations and statistical analysis, we process "historical pipe laying projects with similarity less than the preset level", and select representative "central clusters of cluster analysis results" based on "distribution probability" to supplement reference information for optimization schemes.

[0088] Step 7: Improve the initial recommended layout based on the layout improvement types of all retained results.

[0089] All retained problem-strategy pairs are analyzed to extract the layout improvement types involved, such as adjusting the arrangement order, changing the starting pipe placement, and selecting appropriate pipe placement assistance methods. Based on these improvement types, the initial recommended layout scheme generated in step 2 is adjusted to form the final improved layout scheme. A rule-based scheme adjustment method is used, through programmed rules, to modify parameters such as the pipe placement order and starting pipe placement position of the initial scheme according to the extracted layout improvement types. "Layout improvement types" are extracted from the retained problem-strategy pairs to optimize the "initial recommended layout scheme," resulting in an improved layout scheme that includes the pipe placement order, starting pipe placement position, and pipe placement assistance methods for each individual pipe segment.

[0090] The layout improvement type is the optimization direction extracted from the retained problem-policy pairs, such as: crossing method adjustment (large-scale excavation → pipe jacking), burial depth optimization (increasing by 0.5 meters to avoid underground cables), route offset (avoiding the foundation of the above-ground utility tunnel). Type mapping associates the solution type of the problem-policy pair with the obstacle type of the current laying environment, such as:

[0091] The obstacle type above-ground utility tunnel foundation is mapped to the solution type trenchless crossing, generating mapping pairs (utility tunnel foundation protection, micro-tunnel excavation).

[0092] The rule fusion algorithm is based on engineering specifications (such as the "Buried Pipeline Construction Specification") to fuse mapping pairs with the initial scheme. The implementation method is as follows: if the initial scheme crosses the road by large-scale excavation, but the mapping pair requires trenchless crossing, then according to the rule in the specification that road traffic flow > 1000 vehicles / hour requires trenchless crossing, it is adjusted to pipe jacking crossing.

[0093] The initial recommended layout scheme was optimized through the complete process of steps 1-7. In the experimental project, 3D laser scanning and data preprocessing technology were used to accurately obtain buried pipeline data to generate input vectors. These vectors were then used to generate an initial scheme via a deep learning model, which was further optimized based on historical experience. Ultimately, spatial conflicts were reduced by 80%, thanks to accurate similarity calculations and reasonable historical experience filtering, which resulted in the selection of appropriate arrangement order and pipe laying assistance methods. Construction efficiency was improved by approximately 30%, with reduced construction obstacles achieved by optimizing the initial pipe laying position and pipe laying assistance methods. The average construction period was shortened by 10%, effectively verifying the effectiveness of this scheme in improving the rationality of pipe laying schemes, reducing potential problems, increasing construction efficiency, and reducing costs.

[0094] Preferably, the laying environment includes the topography, above-ground attachments, and underground obstacles of each individual pipeline section;

[0095] The specific parameters include: pipe specifications, pipe material, and pipe weight;

[0096] The improved layout scheme includes: the pipe laying sequence under each individual pipe section, the starting pipe laying position, and the pipe laying auxiliary methods for each pipe laying section. The pipe laying auxiliary methods include: buried pipe channel transportation method, hoisting equipment auxiliary method, pulley block method, and / or pipe jacking construction method.

[0097] In this embodiment, the transport of buried pipelines mainly relies on railcars, flatbed trucks, and other transport vehicles to move them along tracks or level channels laid inside the buried pipeline. First, a suitable transport vehicle is selected based on the pipeline's size and weight, and the pipeline is secured to the vehicle to ensure stability and prevent swaying during transport. Then, the transport vehicle is moved along a planned route within the buried pipeline channel using manual or electric traction to deliver the pipeline to the designated installation location. During the transport process, dedicated personnel need to monitor the pipeline's condition and transport path in real time to prevent collisions with other facilities inside the buried pipeline.

[0098] Large lifting machinery such as truck cranes and crawler cranes are commonly used for auxiliary lifting operations. Before construction, a site survey is necessary to determine the placement and lifting radius of the lifting equipment, ensuring its safe and stable operation. Appropriate lifting slings and hooks are selected based on the pipeline's weight, size, and installation height, and the pipeline is securely connected to the slings. During lifting, the pipeline is lifted by the boom of the lifting equipment and slowly placed into its designated installation position within the buried pipeline, following the predetermined lifting path and height. Close coordination between professional signalmen and lifting operators is essential during the lifting process to ensure safety and accuracy.

[0099] The pulley system utilizes the labor-saving principle of pulleys to move pipelines within buried underground structures. First, a fixed pulley support frame is installed at a suitable location within the buried pipeline. The pulleys are then mounted on the support frame, with an appropriate number of fixed and movable pulleys configured according to the pipeline's weight and direction of movement. Next, the pipeline is connected to the pulley system via a steel cable. Using manual labor or a small electric winch, the steel cable is pulled, and with the help of the pulley system, the pipeline is slowly moved along a pre-set track or path. During the movement, the tension of the steel cable and the position of the pulley system need to be continuously adjusted to ensure smooth pipeline movement.

[0100] Pipe jacking is a trenchless construction technique. In buried pipeline laying, working shafts and receiving shafts are first set up at the starting and ending points of the buried pipeline, respectively. A pipe jacking machine and jacking equipment are installed in the working shaft. The prefabricated pipe is placed into the working shaft, and the jacking equipment pushes the pipe into the soil along the designed axis. Simultaneously, earthwork excavation and transportation are carried out inside the pipeline, continuously advancing it until it reaches the receiving shaft and installation is completed. During the jacking process, it is necessary to monitor the pipeline's axial deviation and jacking pressure in real time. By adjusting the jacking equipment and excavation methods, the jacking accuracy and construction safety of the pipeline can be ensured.

[0101] The beneficial effects of steps 1 to 7 described above are as follows:

[0102] (1) By collecting the current laying environment, construction purpose and specific parameters of the buried pipeline and converting them into input vectors, and inputting them into the pre-trained buried pipeline layout recommendation model, an initial recommended layout scheme based on the actual situation can be generated. This fully considers the various characteristics of buried pipelines and pipelines, making the layout scheme more in line with actual needs and improving the rationality and scientificity of the scheme.

[0103] (2) Identify problem-strategy pairs from historical pipeline projects and perform cluster analysis. Then, filter relevant information based on the similarity between the current buried pipeline and historical projects. This allows us to learn from past experiences and lessons, avoid repeating similar problems, and provide optimization ideas for the current pipeline layout, thus helping to improve the quality and reliability of the pipeline layout plan.

[0104] (3) Based on the autoencoder similarity algorithm, similarity can be calculated to accurately find historical pipeline projects similar to the current buried pipeline situation and retain relevant problem-strategy pairs. For projects with low similarity, representative cluster analysis center clusters are found by determining the distribution probability, thereby obtaining valuable problem-strategy pairs. This approach can address potential problems that may be encountered in the current pipeline layout in a targeted manner, improving the stability and feasibility of the solution.

[0105] (4) The initial recommended layout scheme is improved by combining all the layout improvement types of the retained results. This realizes the intelligent process of the whole process from data collection and model recommendation to optimization based on historical experience, reduces the influence of manual intervention and subjective factors, improves the efficiency and accuracy of pipeline design, and generates a pipeline scheme that meets actual needs and takes into account potential risk prevention and control.

[0106] This invention provides an artificial intelligence-based method for recommending the layout of buried pipelines, which is converted into an input vector, including:

[0107] Numerical standardization processing is performed on the numerical data involved in the laying environment, the purpose of buried pipeline construction, and specific parameters.

[0108] Encode non-numerical data;

[0109] The transformation results for the laying environment, the purpose of buried pipeline construction, and the specific parameters are extracted from the numerical standardization and coding results, respectively, and then combined and sorted in order to obtain the input vector.

[0110] Numerical data refers to physical quantities or properties that can be directly represented by numbers, including:

[0111] Laying environment related factors: terrain slope (e.g., 20°), altitude (50 meters), burial depth of underground obstacles (1.5 meters), and foundation dimensions of the above-ground utility tunnel (10 meters long and 5 meters wide).

[0112] Pipe parameters: pipe diameter (DN500, i.e., 500mm), wall thickness (10mm), pipe weight (15kg per meter), and design burial depth (3.5 meters).

[0113] Non-numerical data that cannot be directly represented by numbers, including categorical or textual information, such as:

[0114] Laying environment related factors: above-ground pipe gallery type (concrete / steel structure), underground obstacle material (cast iron / PE), terrain type (mountain / plain / swamp);

[0115] Construction purpose related: pipeline use (sewage / gas / water supply), construction type (new construction / expansion / repair);

[0116] Pipeline parameters: material (Q235B steel / PE), crossing method (pipe jacking / directional drilling / open excavation).

[0117] Encoding is the process of converting non-numerical data into numerical vectors that the model can compute. This is achieved through the following methods:

[0118] One-hot encoding is suitable for unordered categorical variables. Each category generates a binary vector. For example, if the type of above-ground pipe gallery is "concrete structure", and the category is [concrete, steel structure], the one-hot encoding is [1,0]; if it is "steel structure", it is [0,1].

[0119] Label encoding is suitable for ordered categorical variables, such as encoding construction priority (high / medium / low) as [2,1,0]. However, it should be noted that the model may misjudge the order relationship.

[0120] Word embedding is suitable for text descriptions (such as "karst landforms") and generates dense vectors through pre-trained models (such as Word2Vec), but the steps focus more on structured classification data.

[0121] The crossing method "pipe jacking" is coded as [0,1,0] (assuming the category is [open excavation, pipe jacking, directional drilling]).

[0122] The pipe material “PE” is coded as [0,1] (assuming the category is [steel, PE]).

[0123] The transformation result extraction involves extracting corresponding features from the standardized and encoded data according to their dimensions, including:

[0124] The laying environment includes standardized values ​​for terrain slope, elevation, above-ground utility tunnel coding vector, and underground obstacle burial depth.

[0125] The purpose of construction is the application coding vector (e.g., sewage pipe coding [1,0,0]) and the construction type coding (new coding 1).

[0126] Pipeline parameters include normalized pipe diameter, material coding vector, normalized weight, and normalized design burial depth.

[0127] Sequential combination sorting involves concatenating vectors in the logical order of "laying environment - construction purpose - pipeline parameters" to form a one-dimensional feature vector for the input model, for example:

[0128] Laying environment section: [Topographic slope Z value (2), elevation Z value (2), above-ground utility tunnel code (1,0), obstacle burial depth normalized value (0.6)];

[0129] Construction Purpose Section: [Purpose Code (1,0,0), Construction Type Label (1)];

[0130] Pipeline parameters: [Normalized pipe diameter (0.5), material code (0,1), weight Z-value (1.2), normalized burial depth (0.4)];

[0131] Final input vector: [2,2,1,0,0.6,1,0,0,1,0.5,0,1,1.2,0.4] (14 dimensions in total).

[0132] The above approach uses Python's pandas library to process data, and the sklearn.preprocessing module performs standardization (StandardScaler) and one-hot encoding (OneHotEncoder). It also uses NumPy arrays to concatenate features of each dimension to ensure that the order is consistent with that during model training.

[0133] After standardization, the values ​​of terrain elevation (50 meters) and pipe diameter (500 mm) are given the same weight, avoiding model bias caused by "large number dominance". For example, without standardization, a pipe diameter of 500 mm (value 500) might mask the influence of an elevation of 50 meters, causing the model to misjudge the impact of terrain on pipe layout. Encoded non-numerical data allows the model to identify differences between categories. For example, the difference in encoded vectors between above-ground pipe gallery types "concrete" and "steel structure" helps the model learn the impact of different structures on pipe layout (e.g., concrete foundations require a larger safety distance). A unified input vector format ensures the stability of model training and inference. For example, the input vector dimension is the same for all projects, avoiding model errors due to inconsistent data formats and improving the accuracy of pipe layout recommendations (e.g., reducing the burial depth recommendation error from ±0.8 meters to ±0.3 meters). Through vector combination, the model can simultaneously process multi-dimensional features such as terrain, obstacles, and pipe parameters to generate more reasonable pipe layout plans. For example, when laying PE pipes in mountainous areas, the model can combine the slope standardization value and material code to recommend a solution of "laying along contour lines + anti-slip supports", which improves construction feasibility by 40% compared to traditional manual design.

[0134] This invention provides an artificial intelligence-based method for recommending buried pipeline layout, which obtains the first and second characteristics of the current buried pipeline laying environment and the buried pipeline construction purpose, and at the same time, obtains the standard combination characteristics of the historical pipeline layout projects and the result combination characteristics after project implementation.

[0135] Calculate the first Euclidean distance between the first and second representations and the standard combined representation;

[0136] Calculate the second Euclidean distance between the first representation and the second representation and the combined representation of the result;

[0137] Simultaneously, the third Euclidean distance between the standard combination representation and the result combination representation is calculated;

[0138] If the absolute difference between the first Euclidean distance and the second Euclidean distance is less than the third Euclidean distance, determine the average value of the third Euclidean distance and the absolute difference between the distances, and take the result of the ratio of the average value to the maximum preset distance as the similarity.

[0139] Otherwise, the similarity is determined by the ratio of the average distance between the first and second Euclidean distances (1 - the first Euclidean distance and the second Euclidean distance) to the maximum preset distance.

[0140] The first representation of the current project (laying environment representation) is a low-dimensional feature vector of the current buried pipeline laying environment extracted by an autoencoder. It includes features such as terrain, above-ground attachments, and underground obstacles. For example, the laying environment of a certain current project is "mountainous (slope 20°, altitude 500 meters), above-ground concrete pipe gallery (foundation depth 2 meters, size 10m×5m), and DN300 water supply pipe at 1.5 meters underground". After being encoded by the autoencoder, the feature vector is [0.8, 0.1, 0.3, 0.2] (assuming 4 dimensions, corresponding to terrain, above-ground pipe gallery, underground obstacles, and soil type, respectively).

[0141] The second representation of the current project (construction purpose representation) is a low-dimensional feature vector of the construction purpose, which includes pipeline use, construction type, etc. For example, the construction purpose is "new municipal sewage pipeline (serving a population of 100,000)", encoded as [0.9,0.1] (assuming 2 dimensions, the first dimension corresponds to "new construction" and the second dimension corresponds to "sewage").

[0142] The standard combination representation of historical projects (pre-construction planning representation) is a feature vector extracted from the pre-construction planning data of historical projects by an autoencoder. It reflects the expected environment and objectives in the design phase. For example, if the pre-construction planning of a certain historical project is "plain terrain, no pipe gallery above ground, laying of DN500 gas pipeline (new construction)", the standard combination representation is [0.2,0,0.8] (3-dimensional, corresponding to terrain, above-ground attachments, and pipeline purpose).

[0143] The combined representation of the results of historical projects (actual representation after construction) is the actual data feature vector after the implementation of historical projects, reflecting the real situation after construction (such as actual route deviation, burial depth adjustment, etc.). For example, the route of the above historical project was adjusted due to underground obstacles after actual construction, and the combined representation of the results is [0.25, 0.05, 0.75] (the terrain features changed slightly and new obstacle features were added).

[0144] The first Euclidean distance (the distance between the current and historical plans) is the spatial distance between the first and second characteristics of the current project and the standard combination characteristics of the historical project. It measures the difference between the current plan and the historical plan. For example: the current laying environment characteristic A=[0.8,0.1,0.3,0.2], the construction purpose characteristic B=[0.9,0.1], and the historical standard combination characteristic C=[0.2,0,0.8,0,0.8] (assuming 5 dimensions, the first 4 dimensions are the environment, and the 5th dimension is the construction purpose).

[0145] The second Euclidean distance (the distance between the current and historical reality) is the distance between the current project representation and the combined representation of historical project results. It measures the difference between the current plan and the historical actual execution results. For example, the historical result combination representation D=[0.25,0.05,0.8,0,0.75], and the current representation splicing vector AB=[0.8,0.1,0.3,0.2,0.9,0.1].

[0146] The third Euclidean distance (the distance between historical planning and actual results) is the distance between the standard combination representation and the result combination representation of a historical project, reflecting the deviations before and after the construction of the historical project (such as design changes and construction errors).

[0147] Example: Historical standard representation C=[0.2,0,0.8,0,0.8], result representation D=[0.25,0.05,0.8,0,0.75].

[0148] It should be noted that the similarity between the current project and historical projects is 0.761, indicating that the experience of historical projects has high reference value for the current solution (if the preset similarity is 0.6, then the problem-strategy pair is retained). Therefore, by calculating different Euclidean distances, the similarity between the current buried pipeline and historical pipeline projects before and after construction, as well as the changes in the historical projects themselves before and after construction, is comprehensively considered, making the similarity calculation more comprehensive and accurate. Based on the difference between the first and second Euclidean distances, different similarity calculation methods are adopted to better adapt to different data characteristics and actual situations, improving the rationality of the similarity calculation. The calculated similarity can provide a reference for the planning, design, and construction of current buried pipelines, helping decision-makers to learn from the experience of historical projects and optimize the construction plan for current buried pipelines.

[0149] This invention provides an artificial intelligence-based method for recommending buried pipeline layouts. Before determining the distribution probability of historical pipeline projects with similarity less than a preset value in all cluster analysis results, the method further includes:

[0150] For each historical deployment project with a similarity less than the preset value, the problem-strategy pair is marked with significance in the cluster analysis results, and the first number in each cluster analysis result is counted;

[0151] Suppose there are 10 problem-strategy pairs under historical pipeline project C1. After analysis using hierarchical clustering algorithm, there are 4 clusters. The first cluster involves 2 problem-strategy pairs from historical pipeline project C1, the second cluster involves 5 problem-strategy pairs from historical pipeline project C1, the third cluster involves 2 problem-strategy pairs from historical pipeline project C1, and the fourth cluster involves 1 problem-strategy pair from historical pipeline project C1. The saliency marker is to mark the corresponding problem-strategy pairs with color for easy viewing, and the first quantities obtained are: 2, 5, 2, 1.

[0152] Calculate the first ratio between the first quantity and the total number of problem-strategy pairs involved in the corresponding cluster analysis results;

[0153] The first ratio = the first number of problem-strategy pairs involved in item C1 under the corresponding cluster / the total number of problem-strategy pairs involved in the corresponding cluster analysis results. For example, if there are 20 problem-strategy pairs under the first cluster, then the first ratio is 2 / 20.

[0154] Calculate the sum of all first ratios, and extract the largest ratio from all first ratios;

[0155] Each historical pipe laying project has a sum and a maximum ratio.

[0156] Sort the sums and maximum ratios of all historical pipe laying projects with similarity less than the preset value in descending order, and construct a two-dimensional array, where each value occupies one cell;

[0157] The cells are set up to hold sums and ratios.

[0158] For example, there are three historical pipeline projects with a similarity score lower than the preset score: project C1, project C6, and project C9. Their corresponding sums are 0.8, 0.6, and 0.9, respectively, and their corresponding maximum ratios are 0.1, 0.2, and 0.6, respectively. In this case, the two-dimensional array would be: In this two-dimensional array, each element occupies one cell, and the element is both the value and the index. The sum is in the first row, and the maximum ratio is in the second row.

[0159] Calculate the first average of the absolute values ​​of the differences between two adjacent sums in the row array related to the sum in the second array, and calculate the second ratio of the first average to the average of all first quantities under the corresponding historical pipeline projects, as the second average.

[0160] The first average of the absolute values ​​of the differences between two adjacent sums: for example: (0.9-0.8+0.8-0.6) / 2.

[0161] The average of all first quantities is the sum of the first quantities of the problem-policy pairs under each cluster and the number of clusters.

[0162] The second average is the average of the first average / the average of all first quantities under the corresponding historical pipe laying projects.

[0163] Calculate the third average of the absolute values ​​of the differences between two adjacent maximum ratios in the row array related to the maximum ratio in the second array;

[0164] Third average: The third average of the absolute values ​​of the difference between two adjacent maximum ratios: for example: (0.6-0.2+0.2-0.1) / 2.

[0165] If the second average is greater than or equal to the third average, determine the third ratio between the second average and the third average, and at the same time, determine the fourth ratio between the absolute value of the difference between two adjacent sums and the second average.

[0166] Round down the product of the third and fourth ratios and insert blank cells in the middle of the corresponding two adjacent sums, matching the rounded result.

[0167] Update the index of the rows in the array where blank cells are inserted;

[0168] Based on the total number of blank cells inserted into the row array after the sequence number is updated, starting with one blank cell, the row arrays related to the corresponding maximum ratio are cyclically inserted one blank cell from the middle of adjacent maximum ratios until the number of insertions matches the total number. Then, the insertion stops, and the sequence number is updated.

[0169] Otherwise, leave the two-dimensional array unchanged.

[0170] ,in, The floor symbol; The number of blank spaces to be inserted.

[0171] for example, and There are a1 blank cells between them. and There are a2 blank cells between them. Therefore, the resulting row array is: .

[0172] Assuming the sum of a1 and a2 is 3, the cyclic insertion process is as follows:

[0173]

[0174] The resulting insertion is as follows: .

[0175] Therefore, by inserting blank cells between adjacent pairs of the sum and the row array and updating the row data with the largest ratio in a loop, the corresponding sequence number update and insertion operations are performed on the two row arrays, and finally the adjusted two-dimensional array is obtained. In this way, the data characteristics can be displayed more intuitively, which helps to analyze the distribution of historical pipeline projects with low similarity in clustering.

[0176] This invention provides an artificial intelligence-based method for recommending buried pipeline layouts, determining the distribution probability of each historical pipeline layout project with a similarity less than a preset value across all clustering analysis results, including:

[0177] Extract each historical pipe laying project with a similarity less than the preset value from the final array, based on the first index of the first row and the second index of the second row;

[0178] The result of 1 - (first serial number + second serial number) / (2 × total serial number) is taken as the corresponding probability distribution.

[0179] Therefore, by calculating the probability distribution, the similarity between historical pipeline projects and current buried pipelines is transformed into a specific numerical value, facilitating quantitative evaluation and comparison of different historical pipeline projects. For example, the closer the probability distribution is to 1, the greater the difference between the historical pipeline project and the current buried pipeline; the closer the probability distribution is to 0, the smaller the difference. This helps to quickly identify historical projects that are similar to or significantly different from current buried pipelines, providing a reference for decision-making. The formula comprehensively considers the first sequence number based on the first row and the second sequence number based on the second row, evaluating historical pipeline projects from multiple dimensions, avoiding the limitations of judging based on only a single factor, and making the evaluation results more comprehensive and accurate. The total sequence number is used as the denominator to normalize the sequence information, ensuring that the probability distribution value ranges between 0 and 1, facilitating understanding and comparison. Integrating sequence information from different dimensions into a unified probability value achieves data standardization and normalization. This allows raw data of different dimensions and ranges to be transformed into comparable probabilistic data, facilitating further data analysis and mining. For example, historical pipeline projects can be sorted and classified based on their distribution probabilities, providing stronger data support for subsequent decision support.

[0180] This invention provides an artificial intelligence-based method for recommending buried pipeline layouts. Based on the layout improvement types of all retained results, the initial recommended layout scheme is improved, including:

[0181] Map the solution type of each problem-policy pair in all retained results to the structure type of each individual pipeline in the initial recommended layout scheme to obtain the mapping pair for each individual pipeline.

[0182] The mapping pairs are fused with the sub-recommended layout schemes of each individual pipeline based on the rule fusion algorithm to improve the scheme.

[0183] The problem-strategy pairs in the retained results are historical problems and corresponding solutions that are retained after similarity filtering, such as ("excessive settlement when crossing railway", "using micro tunnel boring machine"). For example, in a certain historical project, the problem was "excavation within the foundation protection area of ​​the above-ground pipe gallery caused structural cracks", and the strategy was "excavation-free pipe jacking construction + foundation settlement monitoring".

[0184] Solution type is a functional classification of solution strategies, used to abstract the core purpose of the strategy, including:

[0185] Space optimization: Adjusting pipe layout or routing to avoid obstacles (such as "multi-layer pipe layout" or "route offset");

[0186] Trenchless construction: Crossing methods that avoid excavation (such as "pipe jacking" and "directional drilling");

[0187] Material reinforcement: Replace the pipe material or auxiliary materials (such as "high-strength steel pipe" or "anti-corrosion coating").

[0188] The structural types in the initial plan are segment types divided according to the laying environment or pipeline characteristics, reflecting the spatial constraints of pipeline laying, including:

[0189] Obstacle-sensitive areas: areas near the foundations of above-ground utility tunnels and underground cables;

[0190] Complex terrain: steep mountain slopes, swamps;

[0191] Crossing engineering section: Areas that need to cross roads or rivers.

[0192] For example, in a certain initial plan, the pipeline needs to pass through a 3-meter range outside the foundation of the above-ground utility tunnel. This section of the structure is classified as a "sensitive section of the utility tunnel foundation".

[0193] Type mapping and mapping pairs match solution types and structure types according to function, generating (solution type, structure type) tuples as the basis for solution improvement. The association is established through manually defined mapping rules or machine learning models (such as decision trees). For example, the solution type "trenchless construction" is mapped to the structure type "pipe gallery foundation sensitive section", forming a mapping pair ("trenchless construction", "pipe gallery foundation sensitive section").

[0194] The sub-recommended layout scheme is a specific pipe layout suggestion for each structural type in the initial scheme, including:

[0195] Pipeline route coordinates and burial depth;

[0196] Crossing method (open excavation / pipe jacking);

[0197] Pipe fixing method (support type, spacing).

[0198] For example, the sub-scheme for the structural type "sensitive section of pipe gallery foundation" is "the route is laid 5 meters outside the pipe gallery foundation, with a burial depth of 3.5 meters, and a large-scale excavation is used for crossing".

[0199] The rule base is built by storing mapping rules for "structure type + solution type → adjustment method", such as: IF structure type = sensitive section of pipe gallery foundation AND solution type = trenchless construction THEN crossing method = pipe jacking, and matching rules are performed using forward reasoning (from condition to result) or backward reasoning (from goal to condition).

[0200] For example, the rule base contains the rule: "If the structure type is 'sensitive section of pipe gallery foundation' and the solution type is 'trenchless construction,' then the crossing method is changed from open excavation to pipe jacking, and foundation settlement monitoring points are added."

[0201] The specific process of scheme integration is as follows: extract the mapping pairs ("non-excavation construction type", "sensitive section of pipe gallery foundation"), find the matching rules in the rule base, match the corresponding sub-scheme (original large-scale excavation crossing), adjust it to pipe jacking construction according to the rules, add monitoring points, and generate the improved sub-scheme: "pipe jacking crossing, burial depth 4 meters (avoiding the foundation stress zone), and set settlement monitoring points every 10 meters".

[0202] For example: the scenario of improving the laying of sewage pipes under urban roads:

[0203] Installation environment: There is a concrete pipe gallery above ground (foundation depth 2 meters, range 10m×5m), and power cables are located 1.5 meters underground;

[0204] Initial sub-scheme: The section crossing the road will be excavated at a depth of 3 meters, 2 meters away from the foundation of the utility tunnel (within the protection zone).

[0205] Type mapping is a preserved problem-policy pair: ("Cracks caused by excavation within the protection zone of the utility tunnel foundation", "Excavation-free pipe jacking + foundation reinforcement"), solution type: "Excavation-free construction" "Foundation protection", structure type: "Sensitive section of utility tunnel foundation" (within 3 meters of the foundation); mapping pair: ("Excavation-free construction", "Sensitive section of utility tunnel foundation"), ("Foundation protection", "Sensitive section of utility tunnel foundation"). At this point, the rule fusion algorithm is applied: rule base matching:

[0206] Rule 1: Sensitive section of pipe gallery foundation + trenchless construction → crossing method = pipe jacking;

[0207] Rule 2: Sensitive sections of utility tunnel foundations + foundation protection areas → Burial depth ≥ 4 meters (2 meters below the foundation bottom);

[0208] Sub-scheme improvement:

[0209] Crossing method: large-scale excavation → micro-tunneling (pipe jacking);

[0210] Burial depth: 3 meters → 4.2 meters (2 meters below the bottom of the pipe gallery foundation);

[0211] New measure: Foundation grouting reinforcement will be carried out simultaneously during pipe jacking construction.

[0212] Therefore, by using type mapping, the solution logic of historical strategies is applied to the current structure type in a targeted manner. The rule fusion algorithm transforms the scattered normative clauses into executable scheme adjustment rules. By combining historical problem-strategy pairs with the current structure type, risk prevention and control measures are injected in advance. Compared with manual scheme adjustment, this method greatly shortens the improvement time of single pipeline scheme through automated mapping and rule fusion.

[0213] The present invention also provides an electronic device, comprising: a processor, a transmitting device, an input device, an output device, and a memory, wherein the memory is used to store computer program code, the computer program code including computer instructions, and when the processor executes the computer instructions, the electronic device performs a method as described in any of the above possible implementations.

[0214] The present invention also provides a computer-readable storage medium storing a computer program, the computer program including program instructions that, when executed by a processor of an electronic device, cause the processor to perform a method as described in any of the above possible implementations.

[0215] Please see Figure 2 , Figure 2 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present invention.

[0216] The electronic device 2 includes a processor 21, a memory 22, an input device 23, and an output device 24. The processor 21, memory 22, input device 23, and output device 24 are coupled together via connectors, which may include various interfaces, transmission lines, or buses, etc., and are not limited in this embodiment of the invention. It should be understood that in various embodiments of the invention, coupling refers to mutual connection through a specific method, including direct connection or indirect connection through other devices, such as through various interfaces, transmission lines, buses, etc.

[0217] The processor 21 can be one or more graphics processing units (GPUs). If the processor 21 is a GPU, the GPU can be a single-core GPU or a multi-core GPU. Optionally, the processor 21 can be a processor group composed of multiple GPUs, with the multiple processors coupled to each other via one or more buses. Optionally, the processor can also be other types of processors, etc., and this embodiment of the invention is not limited thereto.

[0218] The memory 22 can be used to store computer program instructions, as well as various types of computer program code, including program code for executing the present invention. Optionally, the memory includes, but is not limited to, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or compact disc read-only memory (CD-ROM), which is used for related instructions and data.

[0219] Input device 23 is used to input data and / or signals, and output device 24 is used to output data and / or signals. Input device 23 and output device 24 can be independent devices or an integrated device.

[0220] It is understood that in this embodiment of the invention, the memory 22 can be used not only to store related instructions, but also the specific data stored in the memory is not limited.

[0221] Understandable, Figure 2 This is merely a simplified design of an electronic device. In practical applications, the electronic device may also include other necessary components, including, but not limited to, any number of input / output devices, processors, memories, etc., and all video analysis devices that can implement embodiments of the present invention are within the protection scope of the present invention.

[0222] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0223] Those skilled in the art will readily understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. Those skilled in the art will also readily understand that the various embodiments of the present invention have different focuses, and for the sake of convenience and brevity, the same or similar parts may not be repeated in different embodiments. Therefore, parts not described or not described in detail in one embodiment can be referred to in other embodiments.

[0224] In the embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0225] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0226] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0227] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., digital versatile discs (DVDs)), or semiconductor media (e.g., solid-state disks (SSDs)).

[0228] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as read-only memory (ROM) or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for recommending buried pipeline layout based on artificial intelligence, characterized in that, The method includes: Collect the current laying environment of the buried pipeline, the purpose of the buried pipeline construction, and obtain the specific parameters of a single type of pipeline applied to the current buried pipeline, and convert them into an input vector; The input vector is fed into a pre-trained buried pipeline layout recommendation model to obtain an initial recommended layout scheme; The project identifies new historical problems and solutions for each historical pipeline project during the pipeline deployment process, forming several problem-strategy pairs. A hierarchical clustering algorithm is then used to perform cluster analysis on the problem-strategy pairs under all historical pipeline projects. The similarity between the current buried pipeline laying environment and the buried pipeline construction purpose and each historical pipeline laying project is calculated based on the autoencoder similarity algorithm. Retain the problem-strategy pairs of historical routing projects with a similarity greater than or equal to the preset level; Each historical deployment project with a similarity less than a preset degree is determined based on the distribution probability in all cluster analysis results, and the central cluster of the cluster analysis result corresponding to the maximum number of problem-strategy pairs involved in each cluster analysis result of the historical deployment project with the highest probability in the distribution probability is retained; Based on all the layout improvement types of the retained results, the initial recommended layout scheme is improved.

2. The method for recommending buried pipeline layout based on artificial intelligence according to claim 1, characterized in that, The laying environment includes the topography, above-ground attachments, and underground obstacles of each individual pipeline section; The specific parameters include: pipe specifications, pipe material, and pipe weight; The improved layout scheme includes: the pipe laying sequence of each individual pipe section, the starting pipe laying position, and the pipe laying auxiliary methods for each pipe section. The pipe laying auxiliary methods include: buried pipe channel transportation method, hoisting equipment auxiliary method, pulley block method, and / or pipe jacking construction method.

3. The method for recommending buried pipeline layout based on artificial intelligence according to claim 2, characterized in that, Convert to an input vector, including: Numerical standardization processing is performed on the numerical data involved in the laying environment, the purpose of buried pipeline construction, and specific parameters. Encode non-numerical data; The transformation results for the laying environment, the purpose of buried pipeline construction, and the specific parameters are extracted from the numerical standardization and coding results, respectively, and then combined and sorted in order to obtain the input vector.

4. The method for recommending buried pipeline layout based on artificial intelligence according to claim 1, characterized in that, The similarity between the current buried pipeline's laying environment and construction purpose and each historical pipeline laying project is calculated based on an autoencoder similarity algorithm, including: The first and second characteristics of the current buried pipeline laying environment and the buried pipeline construction purpose are obtained respectively. At the same time, the standard combination characteristics of the historical pipeline laying projects and the result combination characteristics after project implementation are obtained. Calculate the first Euclidean distance between the first and second representations and the standard combined representation; Calculate the second Euclidean distance between the first representation and the second representation and the combined representation of the result; Simultaneously, the third Euclidean distance between the standard combination representation and the result combination representation is calculated; If the absolute difference between the first Euclidean distance and the second Euclidean distance is less than the third Euclidean distance, determine the average value of the third Euclidean distance and the absolute difference between the two distances, and take the result of the ratio of the average value to the maximum preset distance as the similarity. Otherwise, the similarity is determined by the ratio of the average distance between the first and second Euclidean distances (1 - the first Euclidean distance and the second Euclidean distance) to the maximum preset distance.

5. The method for recommending buried pipeline layout based on artificial intelligence according to claim 1, characterized in that, Before determining the distribution probability of historical pipeline projects with similarity less than a preset value in all cluster analysis results, the following steps are also included: For each historical deployment project with a similarity less than the preset value, the problem-strategy pair is marked with significance in the cluster analysis results, and the first number in each cluster analysis result is counted; Calculate the first ratio between the first quantity and the total number of problem-strategy pairs involved in the corresponding cluster analysis results; Calculate the sum of all first ratios, and extract the largest ratio from all first ratios; Sort the sums and maximum ratios of all historical pipe laying projects with similarity less than the preset value in descending order, and construct a two-dimensional array, where each value occupies one cell; Calculate the first average of the absolute values ​​of the differences between two adjacent sums in the row array related to the sum in the two-dimensional array, and calculate the second ratio of the first average to the average of all first quantities under the corresponding historical pipeline projects, as the second average. Calculate the third average of the absolute values ​​of the differences between two adjacent maximum ratios in the row array related to the maximum ratio in the two-dimensional array; If the second average is greater than or equal to the third average, determine the third ratio of the second average to the third average, and at the same time, determine the fourth ratio of the absolute value of the difference between two adjacent sums to the second average. Round down the product of the third and fourth ratios and insert blank cells in the middle of the corresponding two adjacent sums, matching the rounded result. Update the index of the rows in the array where blank cells are inserted; Based on the total number of blank cells inserted into the row array after the sequence number is updated, starting with one blank cell, the row arrays related to the corresponding maximum ratio are cyclically inserted one blank cell from the middle of adjacent maximum ratios until the number of insertions matches the total number. Then, the insertion stops, and the sequence number is updated. Otherwise, leave the two-dimensional array unchanged.

6. The method for recommending buried pipeline layout based on artificial intelligence according to claim 5, characterized in that, Determine the distribution probability of each historical pipe laying project with a similarity less than a preset value in all cluster analysis results, including: Extract each historical pipe laying project with a similarity less than the preset value from the final array, based on the first index of the first row and the second index of the second row; The result of 1 - (first serial number + second serial number) / (2 × total serial number) is taken as the corresponding probability distribution.

7. The method for recommending buried pipeline layout based on artificial intelligence according to claim 1, characterized in that, Based on all retained layout improvement types, the initial recommended layout scheme is improved, including: Map the solution type of each problem-policy pair in all retained results to the structure type of each individual pipeline in the initial recommended layout scheme to obtain the mapping pair for each individual pipeline. The mapping pairs are fused with the sub-recommended layout schemes of each individual pipeline based on the rule fusion algorithm to improve the scheme.

8. An electronic device, characterized in that, include: A processor and a memory, the memory being used to store computer program code, the computer program code including computer instructions, wherein when the processor executes the computer instructions, the electronic device performs the buried pipeline laying recommendation method based on artificial intelligence as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which includes program instructions that, when executed by a processor of an electronic device, cause the processor to perform the artificial intelligence-based buried pipeline layout recommendation method according to any one of claims 1 to 7.

Images