Method and system for comprehensive detection, management and repair of inspection well, computer equipment and storage medium

Abstract

The invention relates to a comprehensive inspection, management and repair method and system for an inspection well, computer equipment and a storage medium. The method comprises the following steps: acquiring a target position and a fault type of a to-be-repaired target inspection well; obtaining a starting point position of the repairing personnel, and based on the starting point position and the target position, performing multi-target inspection well path planning by adopting a genetic algorithm to obtain a target path; the target path is used for indicating the repairing personnel to pass through each target inspection well according to the optimal path; when the position information of the repairing personnel is matched with the target position, a downhole operation instruction is generated according to the fault type; the well descending operation instruction is used for indicating the repairing personnel to repair the target inspection well according to the fault type. The multi-target inspection well path planning is carried out by comprehensively considering factors such as road conditions and traffic rules through a genetic algorithm, so that the purposes of accurately positioning target inspection wells and faults, planning an optimal repair path to improve the working efficiency, ensuring that the inspection wells can be recovered to be normal in time and ensuring the stable operation of related infrastructure systems are achieved.

Description

Technical Field

[0001] The invention belongs to the technical field of petrochemical facility management, and in particular relates to a comprehensive detection, management and repair method for an inspection well. Background Art

[0002] In the petrochemical industry, manholes are numerous and widely distributed, and are a key part of the infrastructure of petrochemical enterprises. However, there are many problems with the current management of manholes.

[0003] On the one hand, manhole status diagnosis relies on traditional sensors collecting data and then tediously analyzing it with professionals, lacking intelligent diagnosis. In the event of a minor harmful gas leak or pipeline leakage, the lack of intelligent diagnosis makes it difficult to accurately locate the fault in a timely manner, and small problems can easily worsen.

[0004] Furthermore, the planning of inspection routes for repair personnel often relies on experience or rough zoning, without considering factors such as the spacing between manholes and traffic conditions. For example, in busy areas, these inappropriate routes often leave workers stuck in traffic, significantly reducing inspection efficiency.

[0005] Given the shortcomings of existing technologies in status diagnosis and path planning, an innovative and comprehensive inspection, management and repair method for manholes is needed to improve the intelligence, efficiency and reliability of manhole management in petrochemical enterprises and ensure the stable operation of petrochemical enterprise facilities. Summary of the Invention

[0006] Based on this, it is necessary to provide a method, system, computer equipment and storage medium for comprehensive inspection, management and repair of inspection wells in order to improve the intelligence, efficiency and reliability of inspection well management in the petrochemical industry in response to the above technical problems.

[0007] In a first aspect, the present application provides a comprehensive inspection, management and repair method for inspection wells, comprising:

[0008] Obtain the target location and fault type of the target inspection well to be repaired;

[0009] The starting point of the repair personnel is obtained, and based on the starting point and target positions, a genetic algorithm is used to perform multi-target inspection well path planning to obtain the target path; the target path is used to instruct the repair personnel to pass through each target inspection well along the optimal route;

[0010] When the location information of the repair personnel matches the target location, a downhole operation instruction is generated according to the fault type; the downhole operation instruction is used to instruct the repair personnel to repair the target inspection well according to the fault type.

[0011] In a second aspect, the present application also provides a comprehensive inspection, management and repair system for inspection wells, including:

[0012] A data acquisition module is used to obtain the target location and fault type of the target inspection well to be repaired;

[0013] The path planning module is used to obtain the starting point location of the repair personnel and, based on the starting point location and target location, use a genetic algorithm to perform multi-target inspection well path planning to obtain the target path; the target path is used to instruct the repair personnel to pass through each target inspection well along the optimal route;

[0014] The position matching module is used to generate a downhole operation instruction according to the fault type when the position information of the repair personnel matches the position information of the target inspection well; the downhole operation instruction is used to instruct the repair personnel to repair the target inspection well according to the fault type.

[0015] In a third aspect, the present application further provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, it implements a comprehensive inspection, management and repair method for inspection wells as in the first aspect.

[0016] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a comprehensive inspection, management and repair method for inspection wells as in the first aspect.

[0017] The above-mentioned comprehensive inspection, management and repair method, system, computer equipment and storage medium for inspection wells improve the management efficiency of inspection wells by intelligently planning inspection routes. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 A schematic flow chart of a comprehensive inspection, management and repair method for inspection wells provided by the present invention;

[0020] Figure 2 This is a structural schematic diagram of a comprehensive inspection, management and repair system for inspection wells provided by the present invention. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0022] refer to Figure 1 , which presents a flow chart of a comprehensive inspection, management and repair method for inspection wells provided by this application, the method comprising the following steps:

[0023] S101. Obtain the target location and fault type of the target inspection well to be repaired.

[0024] Specifically, a complete inspection well information database can be established or the location information of the target inspection well can be determined through special detection equipment and systems. The location information can be accurately obtained through geographic information system (GIS) coordinates, map annotations, or numbering and positioning rules based on specific areas to ensure that the location of the target inspection well to be repaired in the actual geographic space can be accurately located. To determine the type of fault, a combination of multiple detection methods can be used. For example, visual detection can be performed using downhole cameras to check whether the structure in the well is damaged, whether there is leakage at the pipe connection, whether there is foreign object blockage, etc.; sensors can also be used to detect abnormal changes in the water level in the well, abnormal gas composition and concentration, etc. Through comprehensive analysis and judgment of these detection data, the type of fault in the target inspection well can be accurately determined, such as structural damage, functional failure (poor drainage, water leakage, etc.) or other specific types of fault conditions.

[0025] S102. Obtain the starting point location of the repair personnel, and based on the starting point location and the target location, use a genetic algorithm to perform multi-target inspection well path planning to obtain a target path; the target path is used to instruct the repair personnel to pass through each target inspection well along the optimal route.

[0026] Specifically, the location of the repair personnel's current starting point is obtained by exchanging positioning information with their mobile devices (such as mobile phones, professional operation terminals, etc.). A genetic algorithm is an optimization algorithm that simulates the natural evolutionary process. In this step, it is used to solve the multi-objective manhole path planning problem. Its basic principle is to first generate a set of initial path plan populations, each of which represents a possible manhole visit sequence and route. These plans are then evaluated based on the set optimization objectives (such as shortest total path, minimum time consumption, priority of passing through specific areas, etc.), and the fitness value of each plan is calculated. The fitness value reflects the degree to which the plan meets the multi-objective requirements. Next, through genetic operations such as selection, crossover, and mutation, excellent plans are selected from the current population for combination and mutation, generating a new generation of populations. This process is repeated. After multiple generations of evolution and screening, a path plan with the optimal fitness value is ultimately obtained, namely the target path. In this process, various practical factors can be considered, such as road conditions (whether it is congested, whether there are traffic restrictions, etc.), traffic rules, changes in road conditions in different time periods, etc., and these factors can be incorporated into the calculation of the fitness function to ensure that the planned target path has high feasibility and efficiency in the actual implementation process, so that repair personnel can pass through each target inspection well in turn according to the optimal route, reducing unnecessary travel and time waste, and improving the overall efficiency of the repair operation.

[0027] S103. When the location information of the repair personnel matches the target location, a downhole operation instruction is generated according to the fault type; the downhole operation instruction is used to instruct the repair personnel to repair the target inspection well according to the fault type.

[0028] Specifically, high-precision positioning technology (such as GPS combined with base station positioning or indoor positioning technology, if the manhole is located in a special area such as a building) can be used to monitor the location of repair personnel in real time. When the repair personnel approach the target manhole and their location information is within an acceptable error range (for example, within a certain radius from the target manhole), the system automatically generates downhole operation instructions based on the previously determined fault type. Downhole operation instructions should include detailed repair steps and requirements. For example, if the fault is structural, the instructions will clearly specify the required repair materials (such as specific types of cement, bricks, etc.), the specific repair location and process (such as the method of repairing the damaged manhole wall and reinforcement measures). If the fault is a drainage pipe blockage, the instructions will inform the repair personnel of the unclogging tools (such as high-pressure water guns, unclogging machines, etc.) and the unclogging process and precautions (such as controlling water pressure and unclogging direction). The instructions may also include safety precautions, reminding the repair personnel to wear necessary safety equipment (such as hard hats, safety belts, and gas masks).

[0029] The following uses a construction environment as an example to illustrate the repair process of the inspection well:

[0030] Construction environment:

[0031] During a sewage inspection well renovation project at a petrochemical enterprise, the groundwater level in the construction area was high, the soil was relatively soft, and contained a large amount of fine sand. This resulted in leaks at pipe joints, and there were no misalignments or disconnections between the pipe joints. Grouting was used to plug the leaking joints, but it was discovered that due to the pressure of the groundwater level and the influence of the fine sand, conventional grouting materials were easily diluted and washed away, making it difficult to form an effective seal. Furthermore, the strength and durability of the joints after grouting were insufficient, making them prone to recurrence of leakage.

[0032] Repair process:

[0033] S1. Pipe section sealing: First, pre-treat an annular area inside the inspection well to be inspected and repaired, and ensure that the height of the annular area is larger than the diameter of the annular partition airbag. Apply glue to the pre-treated annular area (the surface of the annular area is required to be smooth, clean and free of soft layers), and apply glue to the outside of the annular partition airbag of the sealing cover. Install the sealing cover in the inspection well, and inflate the annular partition airbag to make it close to the inner wall of the inspection well. Then connect the water inlet pipe and the drainage pipe on both sides of the annular partition airbag, and adjust the water level in the inspection well through the water inlet pipe and the drainage pipe until the water level in the inspection well drops to the preset distance, and then maintain the water level in the inspection well by controlling the air pressure and air volume.

[0034] S2. The waterproof lining and HDPE film are pre-laid, and the dimensional data of the inner surface of the inspection well is first obtained by a laser ranging sensor, and then the waterproof lining and HDPE film are cut to a shape that matches the inner surface of the inspection well. The cut HDPE film and waterproof lining are then fitted and laid on the inner surface of the inspection well. The waterproof lining and HDPE film are then hot-melt welded, and at the same time, the HDPE film and waterproof lining are welded to the inner side of the annular partition airbag. A metal ring is then used to tighten the inner side of the connection of the annular partition airbag, thereby sealing the connection between the drain pipe and the water inlet pipe and the inner surface of the annular partition airbag.

[0035] S3. Leakage detection in pipes: Use the detection camera lens to enter the detection pipeline and inspection well for detection. The movement trajectory of the camera lens is located on the axis of the pipeline to accurately determine the leakage location. Without blocking and drainage, the inner wall of the inspection well can also be inspected through a periscope.

[0036] S4. Cleaning of pipe joint gaps: Clean the surface of pipe joint gaps, remove impurities and loose particles, and use a high-pressure air spray gun to blow dry to ensure that the gap surface is dry and clean.

[0037] S5. Pipeline leakage repair:

[0038] Ⅰ) When it is determined that the leakage is at the joint between the sewage well and the pipeline, the sewage well cleaning and leak plugging repair method is used to repair the leakage. The sewage well cleaning and leak plugging repair method specifically includes:

[0039] A) Use a high-pressure water gun to flush the well wall to ensure that the base layer is clean and solid. Check the roughness of the well wall. If the well wall is smooth, it needs to be roughened.

[0040] B) Use an electric pick to chisel open the joint between the sewage well and the pipeline;

[0041] C) Then use electric hammer drilling to embed water stop needles, and use high-pressure grouting machine to grout the gap after excavation;

[0042] D) After the final inspection to ensure there is no leakage, remove the water-stop needle and use the plugging material to repair the hole. Grouting reinforcement of the soil and well wall reinforcement may be carried out.

[0043] II) When the leakage is judged to be from a collapsed part of the pipeline, the ground solidification pretreatment is first carried out to consolidate the soft soil layer and quicksand layer around the deformed and collapsed part. Then the pipe segments at the deformed and collapsed part are cut to remove impurities at the collapsed part and restore the space inside the pipeline. The inner lining steel pipe segments are used to support and maintain the pipe position space. The pipeline is then cleaned. Finally, ultraviolet light curing repair technology is used to pull the glass fiber reinforced hose into the pipe section to be repaired, inflate it until it fits the inner wall of the pipe, and cure it under the action of ultraviolet light to complete the pipeline repair. After the repair is completed, check for leakage.

[0044] III) When it is determined that the leakage is caused by the pipe joint and the large-diameter pipe is misaligned, the soil around the leaking gap of the pipe joint is pre-reinforced first, and cement-water glass double liquid slurry is used for low-pressure and slow grouting to form a relatively stable consolidation layer in the soil within a certain range around the pipe joint, reducing the impact of groundwater pressure and fine sand flow on subsequent grouting; then the double expansion ring treatment method is used to clean the pipe and interface gaps, perform water-stopping treatment, fill the gaps until they are flush with the pipe wall, use rubber sealing tape to seal the pipe connection interface gaps, and use a special hydraulic expander to apply pressure to the stainless steel retaining belt to form a fixed ring;

[0045] For leakage in inspection wells, spraying equipment can be used for spray repair construction. At the same time, the sealed space formed by the circular partition airbag and the sealing cover plate can be used to lower the water level in the inspection well through an inflation device to facilitate detection and repair operations.

[0046] The cement-water glass dual-liquid slurry in step S5 (III) is a special anti-seepage and anti-scour grouting material. Its components include high-strength cement, polymer emulsion, waterproofing agent, and an appropriate amount of fine aggregate, which gives the grouting material excellent adhesion, anti-seepage, and stability under groundwater pressure. Low-pressure, slow grouting is a segmented and layered grouting process. A small amount of grouting material is first injected at the bottom of the pipe joint gap. After initial solidification, grouting is gradually carried out in layers upward. The thickness of each grouting is controlled to be 5-10 cm to ensure that the grouting material evenly fills the gap and is free of bubbles and voids. After grouting is completed, a layer of polymer waterproof mortar with a thickness of approximately 2-3 cm is applied to the surface of the pipe joint gap to further enhance the waterproof performance and durability of the gap and protect the grouting material from external erosion. The repaired pipe joint is cured by covering it with a moist geotextile and regularly sprinkling water to keep it moist. The curing time is no less than 7 days to ensure that the grouting material and waterproof mortar are fully cured and reach the designed strength.

[0047] In the leakage repair step, for the repair of the collapsed part of the pipeline, the land solidification pretreatment also includes grouting to reinforce the soft soil layer and quicksand layer around the collapsed part to improve the repair effect; for leaks at the pipe joints but there is no misalignment or detachment, grouting is used to fill the leaking gaps in the pipe joints for repair; but for the spray repair of the inspection well, quick-drying glue is used to ensure the sealing effect between the sealing cover and the inner wall of the inspection well to improve the repair efficiency.

[0048] The above-mentioned comprehensive inspection, management and repair method for inspection wells can obtain the inspection well location and multiple inspection equipment to determine the fault type through the geographic information system (GIS), and use the positioning interaction with the repair personnel's mobile device to obtain the starting point location, and use the genetic algorithm to comprehensively consider road conditions, traffic rules and other factors to perform multi-target inspection well path planning, thereby achieving the effect of accurately locating the target inspection well and fault, and planning the optimal repair path to improve work efficiency; at the same time, relying on high-precision positioning technology to monitor the position of the repair personnel in real time, and generate detailed and targeted downhole operation instructions according to the fault type, effectively ensuring the standardization, safety and effectiveness of the repair operation, thereby comprehensively improving the quality and efficiency of the comprehensive inspection, management and repair work of the inspection well, reducing manpower and time costs, ensuring that the inspection well can resume normal operation functions in time, and ensuring the stable operation of related infrastructure systems.

[0049] In an optional embodiment, obtaining the target location and fault type of the target inspection well to be repaired includes the following steps:

[0050] Get the status data of all inspection wells;

[0051] Extract features from the state data to obtain a state feature vector;

[0052] Input each state feature vector into the trained intelligent state diagnosis model to obtain the state diagnosis result of each inspection well; the state diagnosis result is normal state or fault type;

[0053] The inspection well corresponding to the state characteristic vector whose state diagnosis result is a fault type is taken as the target inspection well, and the target position of the target inspection well is obtained.

[0054] Specifically, in order to comprehensively and accurately obtain the status data of all inspection wells, multiple data sources can be integrated. On the one hand, various sensors installed inside or near the inspection wells can be used to collect data. For example, water level sensors are used to monitor the changes in the water level in the well in real time. By continuously collecting water level values ​​and their fluctuations over time, they can reflect relevant information such as whether the inspection well is draining smoothly; pressure sensors can detect the pressure of water flow in the pipeline. If the pressure fluctuates abnormally, it may indicate problems such as blockage or leakage in the pipeline. On the other hand, regular or irregular manual inspection records can also be used. Inspectors can record the appearance of the inspection well (such as whether there are cracks in the well wall, whether the well cover is intact, etc.), odor conditions (whether there is any odor, which may indicate sewage leakage or other pollution problems) and other information on site through observation and simple measurements, and these manually recorded data can be digitized.

[0055] Reasonable collection frequencies should be set for different sensors and data types. For data with relatively fast dynamic changes, such as water level and pressure, a higher collection frequency can be set, such as collecting data every few minutes or even tens of seconds, to capture possible instantaneous abnormalities. For some relatively stable characteristic data, such as the basic dimensions and construction materials of inspection wells, the collection frequency can be relatively low, and may only need to be collected when first entering the system or regularly updated. By setting the collection frequency in this way, we can ensure that sufficiently detailed and timely status data is obtained, and avoid data redundancy and waste of storage resources caused by excessive collection.

[0056] When extracting features from the rich and diverse status data obtained, we can determine which data features are most critical based on the operating principles of the inspection well, common failure modes, and the needs of subsequent diagnostic models. For example, to determine whether an inspection well has a leakage problem, features such as the rate of change of the water level in the well, humidity data, and changes in the humidity of the surrounding soil are very important; while to determine the stability of the pipeline connection, features such as the sudden change of pressure data at the connection and the vibration frequency near the connection require special attention. By screening and extracting these key data closely related to different aspects of the inspection well status, the actual status of the inspection well can be more accurately portrayed.

[0057] After determining the key features, they need to be quantified and encoded to form state feature vectors that can be accepted by the intelligent state diagnosis model. For example, the water level is represented by a specific numerical value, and dimensional information such as the length and width of the well wall cracks is also converted into numerical values. The numerical values ​​of different features are then encoded and combined according to specific rules. In this way, the originally complex and diverse state data is converted into state feature vectors with a unified format and containing key information. Each vector represents a specific combination of state features of a manhole at a specific moment.

[0058] Trained intelligent state diagnosis models are typically built using machine learning or deep learning algorithms. For example, a neural network model with multiple hidden layers can be used. Trained with extensive historical data, the model learns the mapping between different state feature vectors and their corresponding true states (normal or various fault types). When a new state feature vector is input, the model determines the manhole state represented by the vector based on the learned mapping.

[0059] After each state feature vector is input into the model, the model will output the corresponding state diagnosis result. This result will either indicate that the inspection well is in a normal state, meaning that all detection indicators are within a reasonable range and the inspection well is operating normally; or it will clearly indicate the specific fault type, such as pipeline blockage, well wall damage, leakage, etc. The output fault type can be further refined. For example, pipeline blockage can be divided more finely based on the degree of blockage (mild, moderate, severe) and the location of the blockage (near the wellhead, in the middle of the pipeline, near the bottom of the well, etc.), so as to provide more targeted guidance for subsequent repair work.

[0060] Based on the status diagnosis results output by the intelligent status diagnosis model, those manholes identified as faulty are screened. These manholes are designated as targets for repair. This model-based diagnosis approach can quickly and accurately identify problematic manholes among numerous manholes, avoiding the inefficiencies and misjudgments that can result from manual, one-by-one inspections.

[0061] For the selected target inspection wells, their target locations can be clarified again, which can be achieved by using the location information recorded in the previous stage of obtaining the status data of all inspection wells.

[0062] In an optional embodiment, the training set of the intelligent state diagnosis model is obtained by the following method:

[0063] Obtain the historical data of the inspection well, which includes: historical status data set A and historical fault records, where A = (a1, a2, ..., a N ), a1, a2, …, aN is the historical status data of different samples, N is the number of samples;

[0064] Extract the features of the historical state data and obtain the historical state feature vector set B=(b1,b2,…,b N );

[0065] According to the historical fault records, the normal state or fault type corresponding to each historical state feature vector is determined, the normal state is marked, and different fault types are marked differently, and the state type set Z = (z1, z2, ..., z N ); where z1, z2, …, z N For b1, b2, …, b N The corresponding status type label;

[0066] Based on the historical state feature vector set B and the state type set P, the training set E={(b1,z1),(b2,z2),…,(b N ,z N )}.

[0067] Specifically, we collect a wide range of historical status data sets covering various information such as water level changes and pipeline flow rates, as well as historical fault records that record past failure events (such as manhole covers being washed away, pipeline ruptures, etc.) in detail, to ensure the completeness and accuracy of the data and to identify and pre-process abnormal data.

[0068] Mathematical transformations (such as Fourier transform) and statistical methods are used to extract key features, such as the slope of water level change and the frequency of flow velocity fluctuation, and combine them to form a feature vector set that can represent the historical status of the inspection well.

[0069] Accurately mark the status based on historical fault records, and classify and mark the normal status or specific fault type (such as pipeline blockage to varying degrees, well wall damage, etc.) to ensure accurate and consistent marking and establish an audit mechanism.

[0070] Pair feature vectors with state labels to construct a training set, provide model learning samples, ensure that the training set is large enough and representative, and covers various states. Stratified sampling and other methods can be used for optimization.

[0071] In an optional embodiment, based on the starting point location information of the repair personnel and the location information of each target inspection well, a genetic algorithm is used to perform multi-target inspection well path planning to obtain the target path, including the following steps:

[0072] S1: Record the starting point of the repairer as coordinate S(x s ,y s ); all target inspection wells are recorded as set Q = {q1,q2,…,q n}, determine the target inspection well q according to the location information of the target inspection well i The coordinates of (x i ,y i ), i = 1, 2, ..., n, n is the number of target inspection wells.

[0073] S2: Calculate the coordinates S to each target inspection well q i distance Calculate the distance between each target inspection well Where j = 1, 2, ..., n, d i,j Indicates the target inspection well q i Go to the target inspection well q j The distance between them.

[0074] S3: Determine the distance from coordinate S to each target inspection well q i Traffic condition weight on the road w s,i ; According to the target inspection well q i Go to the target inspection well q j Real-time traffic congestion data on the road C i,j and historical average traffic data H i,j Determine the target inspection well q i Go to the target inspection well q j Traffic condition weight on the road w i,j , where w i,j =α×C i,j +(1-α)×H i,j , 0<α<1.

[0075] S4: Calculate the distance from coordinate S to each target inspection well q i The comprehensive cost m s,i =w s,i ×d s,i ; Calculate the target inspection well q i Go to the target inspection well q j The comprehensive cost m i,j =d i,j ×w i,j .

[0076] S5: Randomly generate P individuals as a population, each individual k represents a target manhole arrival sequence with coordinate S as the starting point, where

[0077] S6: Calculate the total path cost L for each individual k k and congestion time penalty T k ;in, Indicates the distance from coordinate S to the target inspection well The comprehensive cost, Indicates the inspection well from the target Go to the target inspection well The time penalty coefficient, Indicates the coordinate from coordinate S to the target inspection well The time penalty coefficient.

[0078] S7: According to the total path cost L k and congestion time penalty T k Constructing the fitness function F k , where F k =γ×L k +δ×T k , γ is the total path cost L k The weight coefficient, δ is the congestion time penalty T k The weight coefficient is , and γ+δ=1.

[0079] S8: According to the fitness function F k The selection probability of each individual is calculated by And based on the selection probability p of each individual k , the roulette wheel selection strategy is used to select the parent individuals from the individuals of the population; where F l is the fitness value of the lth individual in the population.

[0080] S9: Perform a crossover operation on the parent individuals to generate offspring individuals, and perform a mutation operation on the offspring individuals with a preset probability; the offspring individual with the highest fitness value is taken as the optimal individual of this generation.

[0081] S10: Repeat S8 to S9 until the preset number of iterations is reached; by comparing the fitness values ​​of the best individuals of previous generations, the offspring individual with the highest fitness value is taken as the global optimal individual, and the arrival order of the target inspection well represented by the global optimal individual is the target path.

[0082] Specifically, the starting point position information of the repairer is converted into the coordinate form S(x s ,y s This coordinate can be accurately obtained by the positioning device (such as GPS device) carried by the restoration personnel, which clearly defines the starting point of the restoration work.

[0083] Consider all the target inspection wells that need to be repaired as a set Q = {q1,q2,…,q n For each target inspection well q i , according to its location information, its coordinates are determined as (x i ,y i ), where i = 1, 2, ..., n, where n is the number of target inspection wells. These coordinates can also be obtained through precise measurement or positioning systems.

[0084] Use the distance formula between two points to calculate the coordinates S of the repair personnel's starting point to each target inspection well q i The distance d s,i The formula is This calculation can intuitively reflect the straight-line distance that the repair personnel need to reach each target inspection well. The longer the distance, the higher the time and cost required to reach the well. The distance d between each target inspection well is calculated using the distance formula between two points. i,j , the formula is where j = 1, 2, ..., n, d i,j Indicates the distance between target inspection wells.

[0085] Determine the coordinates from the starting point S of the repair personnel to each target inspection well q i Traffic condition weight on the road w s,i The determination of this weight can take into account a variety of factors, including road type (such as main roads, secondary roads, and side roads), historical traffic flow data, and current traffic control conditions. For example, a main road with heavy traffic but relatively stable speeds might have a moderate weight; a side road with light traffic but prone to congestion also needs a reasonable weight.

[0086] According to the target inspection well q i Go to the target inspection well q j Real-time traffic congestion data on the road C i,j and historical average traffic data V i,j Determine the traffic condition weight w on the road between the two i,j The calculation formula is w i,j =α×C i,j +(1-α)×V i,j , where 0 < α < 1. Real-time traffic congestion data can be obtained in real time through traffic monitoring systems, while historical average traffic data is the result of a statistical analysis of traffic conditions on a particular road section over a period of time. This approach can more accurately reflect the impact of current and historical traffic conditions on route selection.

[0087] Calculate the distance from coordinate S to each target inspection well q i The comprehensive cost m s,i , the formula is m s,i =w s,i ×d s,i The comprehensive cost takes into account the impact of distance and traffic conditions, that is, the longer the distance and the worse the traffic conditions (the greater the weight), the higher the comprehensive cost.

[0088] Calculate the target manhole q i Go to the target inspection well q j The comprehensive cost m i,j, the formula is m i,j =d i,j ×w i,j This cost also takes into account distance and traffic conditions to assess the cost of moving between different target manholes.

[0089] Randomly generate P individuals as a population. Each individual k represents a target manhole sequence with coordinate S as the starting point, which can be expressed as Here, k = 1, 2, ..., P, and i = 1, 2, ..., n. These initial individuals are the starting point for path planning and are continuously optimized through subsequent genetic operations.

[0090] Calculate the total path cost L for each individual k k , which is the sum of the comprehensive costs of passing through each target inspection well in the order represented by individual k starting from coordinate S. For example, if the order of individual k is (S, q1, q2, q3), then L k =m s,1 +m 1,2 +m 2,3 .

[0091] Calculate the congestion time penalty T for each individual k k The formula is in Indicates the distance from coordinate S to the target inspection well The comprehensive cost, Indicates the inspection well from the target Go to the target inspection well The time penalty coefficient, Indicates the coordinate from coordinate S to the target inspection well The time penalty coefficient, the calculation of the congestion time penalty takes into account the additional time cost that may be caused by traffic congestion at each stage of the path. The worse the traffic conditions (the higher the overall cost), the greater the penalty.

[0092] According to the total path cost L k and congestion time penalty T k Constructing the fitness function F k , the formula is F k =γ×L k +δ×T k , where γ is the total path cost L k The weight coefficient, δ is the congestion time penalty T k The fitness function is used to measure the quality of each individual (path plan). The higher the fitness value, the more the path plan meets the optimization goal, that is, the total distance is short and the congestion time is less.

[0093] According to the fitness function Fk The selection probability of each individual is calculated by Then based on the individual's selection probability p k , a roulette wheel selection strategy is used to select parent individuals from the individuals in the population. The roulette wheel selection strategy makes individuals with high fitness values ​​have a greater probability of being selected, thereby retaining the better path plan.

[0094] Perform a crossover operation on the parent individuals to generate offspring individuals. Crossover is one of the important operations in genetic algorithms. It generates new offspring individuals by exchanging some genes (target manhole sequence) in the parent individuals, thereby increasing the diversity of the population and potentially generating a better path solution.

[0095] Mutation is performed on offspring individuals with a preset probability. This operation randomly changes certain genes in the offspring individuals, preventing the algorithm from falling into local optimal solutions. This allows for further exploration of the solution space and increases the likelihood of finding the global optimal path. The offspring individual with the highest fitness is considered the optimal individual of the current generation and is used for comparison with the optimal individuals of previous generations.

[0096] The selection, crossover, and mutation steps are repeated until the preset number of iterations is reached. During each iteration, the individuals in the population are continuously optimized through selection, crossover, and mutation, gradually bringing the path plan closer to the optimal solution. Finally, by comparing the fitness values ​​of the optimal individuals from previous generations, the offspring individual with the highest fitness value is determined as the global optimal individual. The arrival sequence of the target manholes represented by this global optimal individual becomes the desired target path. This target path guides repair personnel through each target manhole along the optimal route, improving work efficiency and reducing resource consumption.

[0097] The advantage of this path planning approach is its ability to handle complex routing problems, particularly those involving multiple targets and dynamic traffic conditions. By simulating natural selection and genetic mechanisms, the algorithm can find the optimal solution under multiple constraints, improving the efficiency of repair personnel and reducing travel costs.

[0098] In an optional embodiment, when there are at least two maintenance personnel, a genetic algorithm is used to perform multi-target manhole path planning based on the starting point location information of the maintenance personnel and the location information of each target manhole to obtain the target path, including the following steps:

[0099] S1: All restoration personnel are recorded as a set H = {h1,h2,…,h M}, M is the number of repair personnel, M ≥ 2;

[0100] S2: Using density-based spatial clustering algorithm, the target inspection wells are divided into M target inspection well groups G1, G2, ..., G M;

[0101] S3: Assign a capacity coefficient to each repair worker, and assign each target inspection well group G to each target inspection well group based on the capacity coefficient and the number of target inspection wells contained in each target inspection well group. u Assigned to the corresponding repair personnel h u ; Where u represents the serial number of the inspection well group, u=1,2,…,M; h u ∈H;

[0102] S4: For each target inspection well group G u and the corresponding operators h u , calculate the number of workers h u From the starting point to the target inspection well group G u The comprehensive cost of each target inspection well in the target inspection well group G is calculated. u The comprehensive cost among the target inspection wells;

[0103] S5: In each target inspection well group G u Within, based on the operator h u From the starting point to the target inspection well group G u The comprehensive cost of each target inspection well and the target inspection well group G u The comprehensive cost between each target inspection well is calculated by using genetic algorithm to plan the path within the group, and each target inspection well group G is obtained. u target path.

[0104] Specifically, firstly, all maintenance personnel involved in the operation are regarded as a set, which contains the information of all maintenance personnel, such as their starting points.

[0105] Next, a density-based spatial clustering algorithm was used to group the target manholes. This clustering algorithm divides the manholes into groups based on their spatial distances. Manholes within a group are closer together, while manholes in different groups are farther apart. This approach reduces the time maintenance personnel spend traveling between manholes and improves efficiency.

[0106] Each maintenance worker is then assigned a capacity factor, which can be based on their work experience, skill level, or other relevant factors. Combined with the target number of manholes within each manhole group, these manhole groups are assigned to the corresponding maintenance workers. This step ensures that each maintenance worker's workload is balanced while also taking into account their individual capabilities.

[0107] Finally, similar to the previous example, within each manhole group, the comprehensive cost of a maintenance worker traveling from their starting point to each manhole within the group, as well as the comprehensive cost between manholes within the group, must be calculated. These costs include not only distance costs but also factors such as time costs and traffic conditions. Within each manhole group, a genetic algorithm is used for intra-group path planning.

[0108] This method can effectively plan the routes of multiple maintenance personnel in manholes, improving operational efficiency and reducing costs. This method is particularly suitable for situations where manholes are widely distributed and the number of maintenance personnel is large, significantly improving the organization and management efficiency of operations.

[0109] In an optional embodiment, before generating the downhole operation instruction, the following steps are further included:

[0110] Conduct environmental assessment on the target inspection well and obtain the assessment results;

[0111] Based on the evaluation results, job instructions are generated.

[0112] Specifically, the environmental assessment can be a gas environment assessment. Use professional gas detection equipment to conduct a comprehensive test of the gas composition and concentration in the target inspection well. The types of gases detected may include methane, hydrogen sulfide, carbon monoxide, etc. Methane is a common flammable gas in sewage pipes, and its high concentration may cause explosion risks; hydrogen sulfide is highly toxic and can cause serious damage to the human respiratory and nervous systems; carbon monoxide is also toxic and may cause poisoning to people. By detecting the concentration of these gases, it is determined whether the underground gas environment is safe and whether measures such as ventilation are needed to reduce the concentration of harmful gases. For example, if the methane concentration is detected to be close to the lower explosion limit, the assessment results will clearly indicate the existence of an explosion risk.

[0113] At the same time, attention should also be paid to oxygen content. Normally, the oxygen content in air is about 21%. If the oxygen content in the manhole is too low (such as below 19.5%), it may cause suffocation due to lack of oxygen; if the oxygen content is too high (such as above 23.5%), it will increase the risk of fire and explosion.

[0114] Based on the results of the gas environment assessment, if hazardous gas concentrations exceed standards or oxygen levels are abnormal, work instructions can explicitly require workers to wear appropriate protective equipment, such as gas masks with gas filtration functions or oxygen respirators. Furthermore, they can stipulate that adequate ventilation must be carried out before entering the manhole until the gas environment reaches safety standards. For example, if hydrogen sulfide concentrations exceed standards, instructions can emphasize the use of gas masks specifically designed for hydrogen sulfide and ensure that ventilation equipment is continuously operated for a certain period of time.

[0115] In an optional embodiment, performing an environmental assessment on a target inspection well to obtain an assessment result; and generating an operation instruction based on the assessment result includes the following steps:

[0116] Based on the status data, determine whether the target inspection well that matches the location information of the repair personnel meets the preset well environment requirements;

[0117] If the conditions are met, a permission to go down the well instruction is generated; the permission to go down the well instruction is used to instruct the generation of a down-the-well operation instruction;

[0118] If not, a "no going down the well" instruction is generated; the "no going down the well" instruction is used to instruct the repair personnel to purify the underground environment of the target inspection well.

[0119] Specifically, using gas assessment as an example, the concentration data for gases such as methane, hydrogen sulfide, and carbon monoxide, as well as oxygen content, within the target manhole are accurately extracted from the status data. This data is continuously monitored and recorded by gas sensors installed within the manhole. For example, the sensors collect gas concentration information at regular intervals (e.g., 15 minutes), generating a time series.

[0120] A preset gas environment requirement for underground wells can be set, which clearly defines the safety limits of various gas concentrations. Methane concentration is usually required to be a certain percentage lower than its lower explosion limit. Assuming the lower explosion limit is 5%, the safety standard may be set to less than 0.25%. The occupational exposure limit of hydrogen sulfide is generally 10mg / m 3 The safe concentration of carbon monoxide is 30mg / m 3 The oxygen content must be within the range of 19.5%-23.5%. Compare the actual monitored gas concentration data with these standards one by one. If the methane concentration is 0.1% and the hydrogen sulfide concentration is 5mg / m 3 , carbon monoxide concentration is 20mg / m 3 If the oxygen content is 20.5%, then the gas environment is preliminarily judged to meet the requirements for going down the well; if the methane concentration reaches 0.3%, then the requirements for the environment for going down the well are not met.

[0121] When the gas environment meets the requirements, the system automatically generates a permission to go down the well command. This command triggers the subsequent process of generating instructions related to specific operations. For example, if the target inspection well fault is to check the sealing of the pipeline connection, the permission to go down the well command will prompt the system to prepare the corresponding testing tools (such as sealing testing instruments) and plan the specific operational steps for testing the sealing condition after going down the well, such as starting from the wellhead and stepping along the pipeline connection, recording the sealing status of each connection.

[0122] If the gas environment does not meet the requirements, the instruction not to go down the well will inform the repair personnel in detail of the gas purification measures to be taken.

[0123] For example, the directive may explicitly require the start of strong ventilation equipment, such as high-power exhaust fans. The operating time and ventilation volume of the ventilation equipment will be determined according to the degree of excess gas concentration. If the methane concentration is slightly higher than the safety standard, the ventilation equipment may be required to run continuously for 1 hour, and the ventilation volume may be maintained at a certain number of cubic meters per minute (such as 5 cubic meters / minute) to ensure that fresh air is fully delivered to the inspection well, reducing the concentration of harmful gases and increasing the oxygen content. During the ventilation process, repair personnel need to continuously monitor changes in gas concentrations until the concentrations of various gases reach a safe range.

[0124] If necessary, other auxiliary measures can be taken. For example, if the concentration of harmful gases exceeds the standard seriously, in addition to ventilation, other auxiliary measures may be needed. For example, in the case of excessively high hydrogen sulfide concentrations, an appropriate amount of chemical neutralizer can be sprayed near the wellhead while ventilating to accelerate the decomposition and removal of hydrogen sulfide. However, when using chemical neutralizers, they must strictly follow the operating procedures to ensure that no new safety hazards are created. During the entire gas environment purification process, repair personnel must closely monitor changes in gas concentration. Only when the gas environment fully meets the requirements for going down the well can they consider the next step of preparation for going down the well.

[0125] In an embodiment of the present application, a comprehensive inspection, management, and repair method for manholes is provided. This method utilizes sensors, inspections, and system interactions to acquire manhole status data. Feature extraction and intelligent model diagnosis determine the fault type, providing a basis for precise repair. A genetic algorithm is used to plan the optimal path based on the location of the repair personnel and the target manhole, saving time and resources. An environmental assessment of the target manhole is conducted before descending, and the decision to descend and the corresponding measures are determined based on the assessment results to ensure operational safety. Detailed operational instructions are generated based on the fault type to ensure the quality of the repair work. Overall, this method improves the accuracy and efficiency of manhole repairs, reduces operational risks, and enhances the infrastructure management level of petrochemical enterprises.

[0126] It should be understood that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence according to the instructions of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above can include multiple steps or multiple stages, and these steps or stages are not necessarily executed and completed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0127] Based on the same inventive concept, embodiments of the present application also provide a comprehensive inspection, management, and repair system for manholes, which is used to implement the aforementioned comprehensive inspection, management, and repair method for manholes. The solution provided by this system is similar to the solution described in the aforementioned method. Therefore, the specific limitations in one or more embodiments of the comprehensive inspection, management, and repair system for manholes provided below can be found in the limitations of the comprehensive inspection, management, and repair method for manholes described above, and will not be further elaborated here.

[0128] In an exemplary embodiment, Figure 2 As shown, a comprehensive inspection, management and repair system 200 for inspection wells is provided, comprising:

[0129] The data acquisition module 201 is used to obtain the target location and fault type of the target inspection well to be repaired;

[0130] The path planning module 202 is used to obtain the starting point location of the repair personnel and, based on the starting point location and the target location, use a genetic algorithm to perform multi-target inspection well path planning to obtain a target path; the target path is used to instruct the repair personnel to pass through each target inspection well along the optimal route;

[0131] The position matching module 203 is used to generate a downhole operation instruction according to the fault type when the position information of the repair personnel matches the position information of the target inspection well; the downhole operation instruction is used to instruct the repair personnel to repair the target inspection well according to the fault type.

[0132] Optionally, the data acquisition module 201 includes:

[0133] The status data acquisition unit is used to obtain the status data of all inspection wells.

[0134] The feature extraction unit is used to extract features from the state data to obtain a state feature vector.

[0135] The state diagnosis unit is used to input each of the state feature vectors into a trained intelligent state diagnosis model to obtain a state diagnosis result of each of the inspection wells; the state diagnosis result is a normal state or a fault type.

[0136] The target inspection well acquisition unit is used to take the inspection well corresponding to the state characteristic vector of the fault type of the state diagnosis result as the target inspection well, and acquire the target position of the target inspection well.

[0137] Optionally, the status diagnosis unit includes a model training subunit, which is used to perform the following operations when obtaining the training set:

[0138] Obtain the historical data of the inspection well, which includes: historical status data set A and historical fault records, where A = (a1, a2, ..., a N ), a1, a2, …, a N is the historical status data of different samples, N is the number of samples;

[0139] Extract the features of the historical state data and obtain the historical state feature vector set B=(b1,b2,…,b N );

[0140] According to the historical fault records, the normal state or fault type corresponding to each historical state feature vector is determined, the normal state is marked, and different fault types are marked differently, and the state type set Z = (z1, z2, ..., z N ); where z1, z2, …, z N For b1, b2, …, b N The corresponding status type label;

[0141] Based on the historical state feature vector set B and the state type set P, the training set E={(b1,z1),(b2,z2),…,(b N ,z N )}.

[0142] Optionally, the path planning module 202 is specifically configured to perform the following operations:

[0143] S1: Record the starting point of the repairer as coordinate S(x s ,y s ); all target inspection wells are recorded as set Q = {q1,q2,…,q n}, determine the target inspection well q according to the location information of the target inspection well i The coordinates of (x i ,y i ), i = 1, 2, ..., n, n is the number of target inspection wells.

[0144] S2: Calculate the coordinates S to each target inspection well q i distance Calculate the distance between each target inspection well Where j = 1, 2, ..., n, d i,j Indicates the target inspection well q i Go to the target inspection well q j The distance between them.

[0145] S3: Determine the distance from coordinate S to each target inspection well q i Traffic condition weight on the road w s,i ; According to the target inspection well qi Go to the target inspection well q j Real-time traffic congestion data on the road C i,j and historical average traffic data H i,j Determine the target inspection well q i Go to the target inspection well q j Traffic condition weight on the road w i,j , where w i,j =α×C i,j +(1-α)×H i,j , 0<α<1.

[0146] S4: Calculate the distance from coordinate S to each target inspection well q i The comprehensive cost m s,i =w s,i ×d s,i ; Calculate the target inspection well q i Go to the target inspection well q j The comprehensive cost m i,j =d i,j ×w i,j .

[0147] S5: Randomly generate P individuals as a population, each individual k represents a target manhole arrival sequence with coordinate S as the starting point, where

[0148] S6: Calculate the total path cost L for each individual k k and congestion time penalty T k ;in, Indicates the distance from coordinate S to the target inspection well The comprehensive cost, Indicates the inspection well from the target Go to the target inspection well The time penalty coefficient, Indicates the coordinate from coordinate S to the target inspection well The time penalty coefficient.

[0149] S7: According to the total path cost L k and congestion time penalty T k Constructing the fitness function F k , where F k =γ×L k +δ×T k , γ is the total path cost L k The weight coefficient, δ is the congestion time penalty T k The weight coefficient is , and γ+δ=1.

[0150] S8: According to the fitness function Fk The selection probability of each individual is calculated by And based on the selection probability p of each individual k , the roulette wheel selection strategy is used to select the parent individuals from the individuals of the population; where F l is the fitness value of the lth individual in the population.

[0151] S9: Perform a crossover operation on the parent individuals to generate offspring individuals, and perform a mutation operation on the offspring individuals with a preset probability μ; the offspring individual with the highest fitness value is taken as the optimal individual of this generation.

[0152] S10: Repeat S8 to S9 until the preset number of iterations is reached; by comparing the fitness values ​​of the best individuals of previous generations, the offspring individual with the highest fitness value is taken as the global optimal individual, and the arrival order of the target inspection well represented by the global optimal individual is the target path.

[0153] Optionally, when there are at least two repair personnel, the path planning module 202 is specifically configured to perform the following operations:

[0154] S1: All restoration personnel are recorded as a set H = {h1,h2,…,h M}, M is the number of repair personnel, M ≥ 2;

[0155] S2: Using density-based spatial clustering algorithm, the target inspection wells are divided into M target inspection well groups G1, G2, ..., G M ;

[0156] S3: Assign a capacity coefficient to each repair worker, and assign each target inspection well group G to each target inspection well group based on the capacity coefficient and the number of target inspection wells contained in each target inspection well group. u Assigned to the corresponding repair personnel h u ; Where u represents the serial number of the inspection well group, u=1,2,…,M; h u ∈H;

[0157] S4: For each target inspection well group G u and the corresponding operators h u , calculate the number of workers h u From the starting point to the target inspection well group G u The comprehensive cost of each target inspection well in the target inspection well group G is calculated. u The comprehensive cost among the target inspection wells;

[0158] S5: In each target inspection well group G u Within, based on the operator h u From the starting point to the target inspection well group G u The comprehensive cost of each target inspection well and the target inspection well group Gu The comprehensive cost between each target inspection well is calculated by using genetic algorithm to plan the path within the group, and each target inspection well group G is obtained. u target path.

[0159] Optionally, the location matching module includes:

[0160] The evaluation unit is used to perform an environmental evaluation on the target inspection well and obtain an evaluation result before generating a downhole operation instruction.

[0161] The instruction generation unit is used to generate a job instruction based on the evaluation result.

[0162] Optionally, the evaluation unit is specifically configured to determine, based on the status data, whether the target inspection well that matches the location information of the repair personnel meets preset downhole environment requirements.

[0163] The instruction generation unit is specifically used to:

[0164] When the preset downhole environment requirements are met, a downhole permission instruction is generated; the downhole permission instruction is used to instruct the generation of a downhole operation instruction;

[0165] When the preset downhole environment requirements are not met, a no-downhole instruction is generated; the no-downhole instruction is used to instruct the repair personnel to purify the downhole environment of the target inspection well.

[0166] An embodiment of the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0167] An embodiment of the present application further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0168] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to the partial description of the method embodiments. The device embodiments described above are merely illustrative, wherein the components described as separate parts may or may not be physically separated, and the parts displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the disclosed solution. A person of ordinary skill in the art can understand and implement it without expending creative work.

[0169] The above-described embodiments merely represent several implementation methods of the embodiments of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that a person skilled in the art may make various modifications and improvements without departing from the concept of the embodiments of the present application, and these modifications and improvements fall within the scope of protection of the embodiments of the present application.

Claims

1. A method for comprehensive inspection, management and repair of inspection wells, characterized in that: The method comprises: Obtain the target location and fault type of the target inspection well to be repaired; Obtaining the starting point location of the repair personnel, and based on the starting point location and the target location, using a genetic algorithm to perform multi-target inspection well path planning to obtain a target path; the target path is used to instruct the repair personnel to pass through each target inspection well along the optimal route; When the location information of the repair personnel matches the target location, a downhole operation instruction is generated according to the fault type; the downhole operation instruction is used to instruct the repair personnel to repair the target inspection well according to the fault type.

2. A method for comprehensive inspection, management and repair of inspection wells according to claim 1, characterized in that: The step of obtaining the target location and fault type of the target inspection well to be repaired includes: Get the status data of all inspection wells; Performing feature extraction on the state data to obtain a state feature vector; Inputting each state feature vector into a trained intelligent state diagnosis model to obtain a state diagnosis result of each manhole; the state diagnosis result is a normal state or a fault type; The inspection well corresponding to the state characteristic vector whose state diagnosis result is a fault type is used as the target inspection well, and the target position of the target inspection well is obtained.

3. A method for comprehensive inspection, management and repair of inspection wells according to claim 2, characterized in that: The training set of the intelligent state diagnosis model is obtained by the following method: Obtain the historical data of the inspection well, the historical data includes: historical status data set A and historical fault records, wherein A=(a1, a2, ..., a N ), a1, a2, …, a N is the historical status data of different samples, N is the number of samples; The feature extraction is performed on the historical state data to obtain a historical state feature vector set B = (b1, b2, ..., b N ); Determine the normal state or the fault type corresponding to each historical state feature vector according to the historical fault record, mark the normal state and mark different fault types differently, and obtain a state type set Z = (z1, z2, ..., z N ); where z1, z2, …, z N For b1, b2, …, b N The corresponding status type label; Based on the historical state feature vector set B and the state type set P, the training set E of the intelligent state diagnosis model is constructed as follows: {(b1, z1), (b2, z2), ..., (b N ,z N )}.

4. A comprehensive inspection, management and repair method for inspection wells according to claim 1, characterized in that: The method of using a genetic algorithm to perform multi-target inspection well path planning based on the starting point location information of the repair personnel and the location information of each target inspection well to obtain a target path includes: S1: Record the starting point of the repairer as coordinate S(x s ,y s ); all the target inspection wells are recorded as a set Q = {q1, q2, ..., q n }, determine the target inspection well q according to the location information of the target inspection well i The coordinates of (x i ,y i ), i=1, 2, ..., n, n is the number of the target inspection wells; S2: Calculate the coordinates S to each target inspection well q i distance Calculate the distance between each target inspection well Where j = 1, 2, ..., n, d i,j Indicates the target inspection well q i Go to the target inspection well q j the distance between them; S3: Determine the distance from the coordinate S to each target inspection well q i Traffic condition weight on the road w s,i ; According to the target inspection well q i Go to the target inspection well q j Real-time traffic congestion data on the road C i,j and historical average traffic data V i,j Determine the target inspection well q i Go to the target inspection well q j Traffic condition weight on the road w i,j , where w i,j =α×C i,j +(1-α)×V i,j , 0<α<1; S4: Calculate the distance from coordinate S to each target inspection well q i The comprehensive cost m s,i =w s,i ×d s,i ; Calculate the target inspection well q i Go to the target inspection well q j The comprehensive cost m i,j =d i,j ×w i,k ; S5: Randomly generate P individuals as a population, each individual k represents a target manhole arrival sequence with coordinate S as the starting point, where S6: Calculate the total path cost L for each individual k k and congestion time penalty T k ;in, Indicates the distance from coordinate S to the target inspection well The comprehensive cost, Indicates the inspection well from the target Go to the target inspection well The time penalty coefficient, Indicates the coordinate from coordinate S to the target inspection well Time penalty coefficient; S7: Based on the total path cost L k and the congestion time penalty T k Constructing the fitness function F k , where F k =γ×L k +δ×T k , γ is the total cost of the path L k The weight coefficient, δ is the congestion time penalty T k The weight coefficient of , and γ+δ=1; S8: According to the fitness function F k The selection probability of each individual is calculated by And based on the selection probability p of each individual k , a roulette wheel selection strategy is used to select parent individuals from the individuals of the population; where F l is the fitness value of the lth individual in the population; S9: performing a crossover operation on the parent individuals to generate offspring individuals, and performing a mutation operation on the offspring individuals with a preset probability; the offspring individual with the highest fitness value is taken as the optimal individual of this generation; S10: Repeat S8 to S9 until the preset number of iterations is reached; by comparing the fitness values ​​of the best individuals of previous generations, the offspring individual with the highest fitness value is taken as the global optimal individual, and the arrival order of the target inspection well represented by the global optimal individual is the target path.

5. The method for comprehensive inspection, management and repair of inspection wells according to claim 1, characterized in that: There are at least two maintenance personnel; Based on the starting point location information of the repair personnel and the location information of each target inspection well, a genetic algorithm is used to perform multi-target inspection well path planning to obtain a target path, including: S1: All the repair personnel are recorded as a set H = {h1,h2,…,h M }, M is the number of the repair personnel, M≥2; S2: Using density-based spatial clustering algorithm, the target inspection well is divided into M target inspection well groups G1, G2, ..., G M ; S3: assign a capability coefficient to each of the repair personnel, and assign each of the target inspection well groups G based on the capability coefficient and the number of target inspection wells included in each of the target inspection well groups. u Assigned to the corresponding repair personnel h u ; Wherein, u represents the serial number of the inspection well group, u=1,2,…,M; h u ∈H; S4: For each target inspection well group G u and the corresponding operators h u , calculate the number of workers h u The inspection well group G from the starting point to the target u The comprehensive cost of each target inspection well in the target inspection well group G is calculated. u The comprehensive cost among the target inspection wells; S5: In each target inspection well group G u Within, based on the operator h u The inspection well group G from the starting point to the target u The comprehensive cost of each target inspection well and the target inspection well group G u The comprehensive cost between each target inspection well is calculated by using genetic algorithm to plan the path within the group, and each target inspection well group G is obtained. u The target path.

6. A method for comprehensive inspection, management and repair of inspection wells according to claim 1, characterized in that: Before generating the downhole operation instruction, the method further includes: Conducting an environmental assessment on the target inspection well to obtain an assessment result; Based on the evaluation result, the work instruction is generated.

7. A method for comprehensive inspection, management and repair of inspection wells according to claim 6, characterized in that: The step of performing an environmental assessment on the target inspection well to obtain an assessment result and generating the operation instruction based on the assessment result includes: Based on the status data, determining whether the target inspection well matching the location information of the repair personnel meets the preset downhole environment requirements; If the conditions are met, a permission instruction for going down the well is generated; the permission instruction for going down the well is used to instruct the generation of the instruction for going down the well operation; If not, a no-going-well instruction is generated; the no-going-well instruction is used to instruct the repair personnel to purify the underground environment of the target inspection well.

8. A system for comprehensive inspection, management and repair of inspection wells, characterized by: The system comprises: A data acquisition module is used to obtain the target location and fault type of the target inspection well to be repaired; A path planning module is used to obtain the starting point location of the repair personnel and, based on the starting point location and the target location, use a genetic algorithm to perform multi-target inspection well path planning to obtain a target path; the target path is used to instruct the repair personnel to pass through each target inspection well along the optimal route; A position matching module is used to generate a downhole operation instruction according to the fault type when the position information of the repair personnel matches the position information of the target inspection well; the downhole operation instruction is used to instruct the repair personnel to repair the target inspection well according to the fault type.

9. A computer device for comprehensive inspection, management and repair of manholes, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A storage medium for comprehensive inspection, management and repair of manholes, on which a computer program is stored, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images