Power transmission tower structure hidden danger detection method and system based on artificial intelligence

Abstract

The invention belongs to the technical field of structure hidden danger detection, and discloses a power transmission tower structure hidden danger detection method and system based on artificial intelligence. The system comprises a dynamic risk prediction and evaluation module, an intelligent inspection strategy generation module, an inspection strategy instruction generation and issuing module, a hidden danger consequence evaluation and analysis module and a resource scheduling and sorting module, a dynamic risk prediction report is obtained, analysis is carried out based on the dynamic risk prediction report, and an intelligent inspection task report is obtained. The intelligent inspection task report is converted into an executable task instruction set, the execution state of an instruction is monitored, a maintenance strategy suggestion report is generated, analysis is performed based on the maintenance strategy suggestion report, and a maintenance work order sorting table and a resource scheduling instruction set are obtained. The method has the remarkable advantages of being good in defect detection active effect, large in polling efficiency improvement effect and high in overall maintenance decision-making capability.

Description

Technical Field

[0001] This invention relates to the field of structural hazard detection technology, and more specifically, to a method and system for detecting structural hazards in power transmission towers based on artificial intelligence. Background Technology

[0002] The power transmission tower structural hazard detection system is a system that uses technical means to monitor, identify and assess the health status of overhead power transmission line tower structures. It is mainly used to promptly detect structural defects such as bent support components, loose bolts, foundation settlement and tower corrosion, and is an important link in ensuring the safe and stable operation of the power grid.

[0003] However, traditional power transmission tower structural hazard detection systems suffer from several shortcomings in practice. First, traditional systems largely rely on periodic manual inspections or drone-based image-based inspections. Existing AI models can only identify surface defects in images, making it difficult to capture changes in the internal mechanical state of the tower structure, let alone predict defect trends under extreme weather conditions. This results in maintenance work being largely reactive. Second, traditional systems typically conduct inspections based on fixed time cycles and routes, leading to a lack of targeted inspections. Consequently, many healthy towers are repeatedly inspected, while a few with potential hazards are inspected. Firstly, defects are easily overlooked, leading to wasted maintenance costs. Secondly, traditional systems, after identifying defects in transmission towers, mostly formulate maintenance plans based solely on the severity of the defects, neglecting the importance of the tower itself. For example, a minor defect in a high-importance tower should be greater than a major defect in a redundant tower, making it difficult for the system to truly coordinate maintenance based on importance. This further leads to a lack of authenticity in the detection and assessment of potential tower hazards. In general, how to effectively solve the problems of poor defect prediction capabilities, insufficient inspection efficiency, and lack of overall maintenance decision-making in traditional systems has become a challenge that current transmission tower structural hazard detection systems need to address. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution, including:

[0005] The multi-source data acquisition and fusion module is used to acquire raw data in real time based on sensor networks and preprocess the raw data to obtain the raw fused dataset;

[0006] Furthermore, the multi-source data acquisition and fusion module is specifically used for:

[0007] The system continuously receives raw data pushed by the sensor network through a standard API interface to obtain a raw dataset. Based on data types, all data items in the raw dataset are preprocessed to obtain a preprocessed dataset. Data types include image data, point cloud data, and environmental data. Specifically: Denoising and correction are performed on the image data in the raw dataset to complete the preprocessing of image data; denoising, registration, and downsampling are performed on the point cloud data in the raw dataset to complete the preprocessing of point cloud data; and unit unification and validity checks are performed on the environmental data in the raw dataset to complete the preprocessing of environmental data.

[0008] Using the same time and the same coordinate system as a reference, all data items in the preprocessed dataset are unified, and image data and point cloud data in the preprocessed dataset are mapped accordingly to obtain the original fused dataset.

[0009] The original fused dataset is output to the structural health assessment quantification module;

[0010] The structural health assessment quantification module is used to extract defect feature sets based on the original fusion dataset and perform health assessments according to the tower structure assessment model to obtain a structural health assessment report.

[0011] Furthermore, the structural health assessment quantitative module is specifically used for:

[0012] Based on the original fusion dataset, and using AI models and image processing algorithms, the defects of the iron tower are identified and quantified to obtain a defect feature set;

[0013] The three-dimensional model of the iron tower is retrieved from the database, and the defect feature set is used as a parameter to update the three-dimensional model of the iron tower, resulting in a new three-dimensional model of the iron tower.

[0014] The environmental type data and defect feature set in the original fusion dataset are input into the new 3D model of the tower, and the tower status report is obtained by the calculation engine in the new 3D model of the tower.

[0015] The tower health score is calculated based on the tower status report, and then integrated into the tower status report to obtain the structural health assessment report.

[0016] Output the structural health assessment report to the dynamic risk prediction and assessment module;

[0017] The dynamic risk prediction and assessment module is used to predict the future stability of the tower based on future environmental data and the tower structure assessment model, and to obtain a dynamic risk prediction report.

[0018] Furthermore, the dynamic risk prediction and assessment module is specifically used for:

[0019] Weather data within a specified area for a future preset time window is obtained from a meteorological service platform based on a sensor network, and the data is arranged in chronological order based on timestamps to obtain a weather dataset.

[0020] Based on the weather dataset, the weather data for all time points in the weather dataset is output to the new 3D model of the iron tower, and the risk probability sequence of the iron tower is obtained.

[0021] Risk level assessment is conducted based on the risk probability sequence of towers to obtain a risk level report, which includes low risk, medium risk and high risk.

[0022] By packaging the risk probability sequence, risk level report, and risk causes of the iron tower, a dynamic risk prediction report is obtained.

[0023] The dynamic risk prediction report is output to the intelligent inspection strategy generation module and the hidden danger consequence assessment and analysis module, respectively.

[0024] The intelligent inspection strategy generation module is used to analyze dynamic risk prediction reports and obtain intelligent inspection task reports.

[0025] Furthermore, the intelligent inspection strategy generation module is specifically used for:

[0026] Read the dynamic risk prediction report and sort the risk levels and tower risk probabilities from largest to smallest to obtain a risk priority report;

[0027] Based on the risk priority report, a key inspection task report is generated for tower IDs with medium and high risk levels, and a normal inspection task report is generated for tower IDs with low risk levels by retrieving the standard inspection cycle from the database. The key inspection task report and the normal inspection task report are packaged together to obtain a comprehensive inspection task report.

[0028] Based on the risk causes in the dynamic risk prediction report and all key inspection task reports in the comprehensive inspection task report, an inspection task content report is generated.

[0029] Based on the tower ID, the inspection task content report is integrated into the comprehensive inspection task report to obtain an intelligent inspection task report;

[0030] The intelligent inspection task report is output to the inspection strategy instruction generation and distribution module.

[0031] The inspection strategy instruction generation and distribution module is used to convert intelligent inspection task reports into executable task instruction sets and monitor the execution status of the instructions.

[0032] Furthermore, the inspection strategy instruction generation and distribution module is specifically used for:

[0033] Based on the intelligent inspection task report, and by retrieving the standard instruction template from the database, the intelligent inspection task report is input into the standard instruction template, and the task execution instruction set is output.

[0034] The execution type of the task execution instruction set will be output to the corresponding receiving end;

[0035] Monitor the execution status of instructions and generate corresponding execution result reports based on the monitoring results. The monitoring results include received, executing, completed, received failed, and executed failed.

[0036] All execution result reports are output to the administrator's receiving end.

[0037] The hazard consequence assessment and analysis module is used to analyze dynamic risk prediction reports, simulate the severity of consequences based on the analysis results, and generate maintenance strategy recommendation reports in combination with risk levels.

[0038] Furthermore, the hazard consequence assessment and analysis module is specifically used for:

[0039] Read the dynamic risk prediction report, identify the tower IDs with high risk in the dynamic risk prediction report, and retrieve the power grid information of the tower IDs with high risk from the database to obtain the power grid information report.

[0040] Based on power grid information reports, a shutdown consequence assessment is conducted on all high-risk tower IDs to obtain tower assessment values. The specific calculation formula for the assessment is as follows:

[0041] ;

[0042] Obtain the tower's evaluation value ,in, For the number of users, For the first The load value of a user who is unable to receive power due to tower shutdown. To estimate the downtime, Economic loss per unit of electricity;

[0043] The tower assessment values ​​are classified based on the assessment threshold range (Q1, Q2). When the tower assessment value is less than Q1, a low consequence level is generated; when the tower assessment value is greater than or equal to Q1 and less than Q2, a medium consequence level is generated; and when the tower assessment value is greater than Q2, a high consequence level is generated.

[0044] Package low-consequence, medium-consequence, and high-consequence levels to obtain a consequences severity report;

[0045] The decision rule table is retrieved from the database, and the risk level in the dynamic risk prediction report and the consequence level in the consequence severity report are input into the decision rule table to output the maintenance strategy report.

[0046] Based on the maintenance strategy report, corresponding recommended maintenance measures and estimated economic losses are generated and integrated into the maintenance strategy report to obtain a maintenance strategy recommendation report;

[0047] Output the maintenance strategy recommendation report to the resource scheduling and sorting module;

[0048] The resource scheduling and sorting module is used to analyze the maintenance strategy recommendation report and obtain the maintenance work order sorting table and the resource scheduling instruction set, respectively.

[0049] Furthermore, the resource scheduling and sorting module is specifically used for:

[0050] Based on the maintenance strategy recommendation report, the optimization objectives are set by retrieving the maintenance resource table from the database and generating the optimization objective function. The maintenance resource table includes the daily workload of all maintenance personnel teams, the amount of spare parts stored, the number of maintenance vehicles, and the expected tower downtime window.

[0051] The genetic algorithm is used to solve the optimization objective function, and the optimal maintenance task report is output.

[0052] Based on the optimal maintenance task, a maintenance scheduling report is generated and output to the receiving end of the corresponding maintenance personnel team;

[0053] The system monitors the execution results of maintenance scheduling reports. When the execution result indicates completion, it triggers the multi-source data acquisition and fusion module to acquire new raw data and uses the maintenance results as feedback data to update the tower's 3D model.

[0054] Furthermore, the present invention also provides a method for detecting structural hazards in power transmission towers based on artificial intelligence, implemented according to the aforementioned system for detecting structural hazards in power transmission towers based on artificial intelligence, and comprising the following steps:

[0055] S1: Real-time acquisition of raw data based on sensor network, and preprocessing of raw data to obtain raw fusion dataset;

[0056] S2: Extract the defect feature set based on the original fusion dataset, and conduct a health assessment based on the tower structure assessment model to obtain a structural health assessment report;

[0057] S3: Based on future environmental data and according to the tower structure assessment model, predict the future stability of the tower and obtain a dynamic risk prediction report;

[0058] S4: Based on the analysis of the dynamic risk prediction report, an intelligent inspection task report is obtained;

[0059] S5: Converts intelligent inspection task reports into a set of executable task instructions and monitors the execution status of the instructions;

[0060] S6: Analyze the dynamic risk prediction report, simulate the severity of consequences based on the analysis results, and generate a maintenance strategy recommendation report in combination with the risk level;

[0061] S7: Based on the maintenance strategy recommendation report, an analysis is performed to obtain a maintenance work order sorting table and a resource scheduling instruction set.

[0062] This invention also provides:

[0063] An artificial intelligence-based device for detecting structural hazards in power transmission towers includes a memory and a processor. The memory stores a computer program, which is executed by the processor when loaded.

[0064] A readable storage medium storing a computer program, the computer program being adapted to execute the above-described artificial intelligence-based method and system for detecting structural hazards in power transmission towers when loaded by a processor.

[0065] The technical effects and advantages of this invention, based on an artificial intelligence-based method and system for detecting structural hazards in power transmission towers, are as follows:

[0066] This invention acquires raw data in real time using a sensor network, preprocesses the raw data to obtain a raw fused dataset, extracts a defect feature set from the raw fused dataset, performs a health assessment based on a tower structure evaluation model, and obtains a structural health assessment report. Based on future environmental data and the tower structure evaluation model, it predicts the future stability of the tower, resulting in a dynamic risk prediction report. Based on the dynamic risk prediction report, it generates an intelligent inspection task report, converts the intelligent inspection task report into an executable task instruction set, monitors the execution status of the instructions, analyzes the dynamic risk prediction report, simulates the severity of consequences based on the analysis results, and generates a maintenance strategy recommendation report based on the risk level. Based on the analysis of the maintenance strategy recommendation report, it obtains a maintenance work order sorting table and a resource scheduling instruction set. This allows the system to not only identify surface defects of the tower but also acquire real-time information about the internal structure of the tower through the collaborative operation of a multi-source data acquisition and fusion module, a structural health assessment quantification module, and multiple dynamic risk prediction and assessment modules. By combining health status with environmental meteorological data, the system can further predict the structural safety of towers under extreme weather conditions, thereby maximizing the safety and reliability of towers. Furthermore, through the collaborative operation of an intelligent inspection strategy generation module and an inspection strategy instruction generation and distribution module, the invention effectively optimizes the fixed inspection scheme of traditional systems, transforming the traditional "flood irrigation" inspection method into a low-cost "on-demand" inspection method. This minimizes inspection costs while ensuring inspection quality. Finally, the establishment of a hazard consequence assessment and analysis module and a resource scheduling and ranking module allows the system to effectively consider the importance of individual towers and comprehensively analyze the impact of tower downtime on the overall power grid. Simultaneously, it combines maintenance resources and various costs to formulate maintenance plans, ensuring that limited maintenance resources are used on the most important towers, thereby maximizing the overall efficiency of the power grid. Overall, this invention has significant advantages in proactive defect detection, greatly improved inspection efficiency, and strong overall maintenance decision-making capabilities. Attached Figure Description

[0067] Figure 1 This is a schematic diagram of the artificial intelligence-based power transmission tower structural hazard detection system of the present invention;

[0068] Figure 2 This is a schematic diagram of the artificial intelligence-based method for detecting structural hazards in power transmission towers according to the present invention. Detailed Implementation

[0069] 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.

[0070] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” used in the embodiments of this invention are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0071] Depending on the context, the words “if” or “suppose” as used here can be interpreted as “when” or “in response to determination” or “in response to detection.” Similarly, depending on the context, the phrases “if determination” or “if detection (of the stated condition or event)” can be interpreted as “when determination” or “in response to determination” or “when detection (of the stated condition or event)” or “in response to detection (of the stated condition or event).”

[0072] Furthermore, the timing of the steps in the following method embodiments is merely an example and not a strict limitation.

[0073] In practice, the server-side equipment deployed by the AI-based power transmission tower structural hazard detection system may consist of one or more devices. This AI-based power transmission tower structural hazard detection system can be implemented as: a business instance, a virtual machine, or hardware devices. For example, this AI-based power transmission tower structural hazard detection system can be implemented as a business instance deployed on one or more devices in a cloud node. Simply put, this AI-based power transmission tower structural hazard detection system can be understood as software deployed on a cloud node, used to provide the AI-based power transmission tower structural hazard detection system to various user terminals. Alternatively, this AI-based power transmission tower structural hazard detection system can also be implemented as a virtual machine deployed on one or more devices in a cloud node. This virtual machine contains application software for managing various user terminals. Alternatively, this AI-based power transmission tower structural hazard detection system can also be implemented as a server composed of numerous identical or different types of hardware devices, with one or more hardware devices configured to provide the AI-based power transmission tower structural hazard detection system to various user terminals.

[0074] In terms of implementation, the AI-based power transmission tower structural hazard detection system and the user terminal are mutually compatible. Specifically, if the AI-based power transmission tower structural hazard detection system is implemented as an application installed on a cloud service platform, the user terminal acts as a client establishing a communication connection with that application; or if the system is implemented as a website, the user terminal acts as a webpage; or if the system is implemented as a cloud service platform, the user terminal acts as a mini-program within an instant messaging application.

[0075] like Figure 1 The figure shown is a system architecture diagram of the power transmission tower structure hidden danger detection system based on artificial intelligence provided in an embodiment of the present invention.

[0076] The AI-based power transmission tower structural hazard detection system described in this invention can be installed on a cloud server. In terms of implementation, it can be one or more service devices, or an application installed in the cloud (e.g., a mobile service operator's server or server cluster), or it can be developed into a website. Depending on the functions implemented, the AI-based power transmission tower structural hazard detection system may include a multi-source data acquisition and fusion module, a structural health assessment and quantification module, a dynamic risk prediction and assessment module, an intelligent inspection strategy generation module, an inspection strategy instruction generation and distribution module, a hazard consequence assessment and analysis module, and a resource scheduling and sorting module. The modules described in this invention can also be called units, referring to a series of computer program segments that can be executed by an electronic device's processor and perform a fixed function, stored in the electronic device's memory.

[0077] In this embodiment of the invention, in the AI-based power transmission tower structural hazard detection system, each of the above modules can be implemented independently and can call other modules. Here, "calling" can be understood as one module connecting to multiple modules of another type and providing corresponding services to those connected modules. For example, the resource scheduling and sorting module can call the same information acquisition module to obtain information collected by that module. Based on the above characteristics, in the AI-based power transmission tower structural hazard detection system provided in this embodiment of the invention, the applicable scope of the system architecture can be adjusted by adding modules and directly calling them without modifying the program code, achieving cluster-based horizontal expansion to quickly and flexibly expand the AI-based power transmission tower structural hazard detection system. In practical applications, the above modules can be set in the same device or different devices, or they can be set in virtual devices, such as service instances in a cloud server.

[0078] Example 1

[0079] Please see Figure 1As shown in this embodiment, the artificial intelligence-based power transmission tower structural hazard detection system includes:

[0080] The multi-source data acquisition and fusion module is used to acquire raw data in real time based on the sensor network and preprocess the raw data to obtain the raw fused dataset.

[0081] Furthermore, the steps of real-time acquisition of raw data based on sensor networks and preprocessing of the raw data include:

[0082] S1.1: Continuously receive raw data pushed by the sensor network through the standard API interface to obtain the raw dataset;

[0083] It should be explained that sensor networks include, but are not limited to, drones, fixed sensors, and weather service platforms; raw datasets include, but are not limited to, data acquisition timestamps, location coordinates, sensor device IDs, and data parameters based on sensor type, wherein data parameters include, but are not limited to, image data, infrared thermal imaging data, laser point cloud data, environmental wind speed data, and environmental icing thickness data;

[0084] S1.2: Preprocess all data items in the original dataset based on data type to obtain a preprocessed dataset. Data types include image data, point cloud data, and environmental data, where:

[0085] Denoising and correction processes are performed on the image data in the original dataset to complete the preprocessing of image data.

[0086] Denoising, registration, and downsampling are performed on the point cloud data in the original dataset to complete the preprocessing of point cloud data.

[0087] It should be explained that registration preprocessing merges the point cloud data obtained from multiple scans into a complete 3D model; downsampling preprocessing is used to reduce the amount of data while preserving data characteristics.

[0088] The environmental data in the original dataset are subjected to unit unification and validity checks to complete the preprocessing of environmental data.

[0089] S1.3: Using the same time and the same coordinate system as a reference, unify all data items in the preprocessed dataset, and map the data types of image data and point cloud data in the preprocessed dataset accordingly to obtain the original fused dataset;

[0090] It should be explained that the original fused dataset is structured data. For example, staff can use the database to query the U node with tower ID 001 at the current time to obtain the image data, infrared thermal imaging data and 3D data of that node.

[0091] S1.4: Output the original fused dataset to the structural health assessment quantification module;

[0092] The structural health assessment quantification module is used to extract a defect feature set based on the original fusion dataset and perform a health assessment according to the tower structure assessment model to obtain a structural health assessment report.

[0093] Furthermore, the steps of extracting a defect feature set based on the original fused dataset and conducting a health assessment according to the tower structure assessment model include:

[0094] S2.1: Based on the original fusion dataset, and using AI models and image processing algorithms, the defects of the iron tower are identified and quantified to obtain a defect feature set;

[0095] It should be explained that the defect feature set refers to, for example, identifying the presence of rust on the tower and quantifying the rust area;

[0096] S2.2: Retrieve the 3D model of the tower from the database and update the 3D model of the tower using the defect feature set as parameters to obtain a new 3D model of the tower.

[0097] S2.3: Input the environmental type data and defect feature set from the original fusion dataset into the new tower 3D model, and perform calculations through the calculation engine in the new tower 3D model to obtain the tower status report;

[0098] It should be explained that the calculation engine is a calculation method used to disperse the new 3D model of the tower into multiple elements and simulate the overall mechanical condition of the tower based on the dispersed elements. For example, when the defect feature set includes the bending of the tower support, the calculation engine will convert the bending sag of the support into an additional moment and calculate it based on the new 3D model of the tower to obtain new stress values ​​and stability assessment values. The tower status report includes, but is not limited to, the current stress, stability coefficient, and maximum displacement of the support ID.

[0099] S2.4: Calculate the tower health score based on the tower status report, and integrate the tower health score into the tower status report to obtain the structural health assessment report;

[0100] S2.5: Output the structural health assessment report to the dynamic risk prediction and assessment module;

[0101] The dynamic risk prediction and assessment module is used to predict the future stability of the tower based on future environmental data and according to the tower structure assessment model, and to obtain a dynamic risk prediction report.

[0102] Furthermore, the steps for predicting the future stability of the tower based on future environmental data and according to the tower structure assessment model include:

[0103] S3.1: Obtain weather data within a specified area within a future preset time window based on the meteorological service platform in the sensor network, and arrange the data in chronological order based on timestamps to obtain a weather dataset;

[0104] It should be noted that the preset time window is manually set and entered into the system, for example, the preset time window is 72 hours; weather data includes, but is not limited to, environmental wind speed data and environmental icing thickness data;

[0105] S3.2: Based on the weather dataset, output the weather data for all time points in the weather dataset to the new 3D model of the iron tower in step S2.2, and output the iron tower risk probability sequence;

[0106] It should be explained that the tower risk probability sequence is obtained by sorting the tower risk probabilities based on all time points in the weather dataset. The tower risk probability is obtained by calculating the stability coefficient and then calculating the tower risk probability based on the stability coefficient. For example, by subtracting one from the stability coefficient, the risk value is obtained. When the risk value is greater than or equal to 0, the tower at that time point is determined to be safe. When the risk value is less than 0, the tower at that time point is determined to be unsafe. The risk probability is calculated using the first second moment method for the risk values ​​that are determined to be unsafe.

[0107] S3.3: Conduct risk level assessment based on the tower risk probability sequence to obtain a risk level report, where the risk level includes low risk, medium risk and high risk;

[0108] It should be explained that the risk level report is a time series risk level sequence derived from the tower risk probability sequence;

[0109] S3.4: Package the risk probability sequence, risk level report and risk causes of the iron tower to obtain a dynamic risk prediction report;

[0110] It needs to be explained that the reasons for the risk include, for example, the expected wind speed of 20 meters per second at 2 pm today, and the insufficient stability of the tower head.

[0111] S3.5: Output the dynamic risk prediction report to the intelligent inspection strategy generation module and the hidden danger consequence assessment and analysis module respectively;

[0112] The intelligent inspection strategy generation module is used to analyze the dynamic risk prediction report to obtain an intelligent inspection task report.

[0113] Further steps in the analysis based on the dynamic risk prediction report include:

[0114] S4.1: Read the dynamic risk prediction report and sort the risk levels and tower risk probabilities from largest to smallest to obtain a risk priority report;

[0115] It should be explained that the risk level from highest to lowest means that high risk is greater than medium risk, which is greater than low risk, and the tower risk probability from highest to lowest means that the numerical value of the tower risk probability is lower than the numerical value.

[0116] S4.2: Based on the risk priority report, generate key inspection task reports for tower IDs with medium and high risk levels, and generate normal inspection task reports for tower IDs with low risk levels by retrieving standard inspection cycles from the database. Package the key inspection task reports and normal inspection task reports to obtain a comprehensive inspection task report.

[0117] S4.3: Based on the risk causes in the dynamic risk prediction report and all key inspection task reports in the comprehensive inspection task report, generate an inspection task content report;

[0118] It should be explained that the inspection task content report means that, for example, if the risk cause in the L1 key inspection task report is that the probability of bolt loosening at the L2 node is high, then the inspection task content report should at least include the content of taking high-definition images of the L2 node for a duration of no less than 10 seconds.

[0119] S4.4: Based on the tower ID, the inspection task content report is integrated into the comprehensive inspection task report to obtain an intelligent inspection task report;

[0120] S4.5: Output the intelligent inspection task report to the inspection strategy instruction generation and distribution module;

[0121] The inspection strategy instruction generation and distribution module is used to convert intelligent inspection task reports into executable task instruction sets and monitor the execution status of the instructions.

[0122] Furthermore, the steps of converting intelligent inspection task reports into executable task instruction sets and monitoring the execution status of these instructions include:

[0123] S5.1: Based on the intelligent inspection task report and according to the database to retrieve the standard instruction template, input the intelligent inspection task report into the standard instruction template and output the task execution instruction set;

[0124] It should be explained that the standard instruction templates include, but are not limited to, those for drone inspection, manual inspection, and sensor inspection. Taking drone inspection as an example, drone inspection instructions include, but are not limited to, flight path instructions, photo parameter instructions, photo duration instructions, and laser scanning instructions.

[0125] S5.2: Output the execution type according to the task execution instruction set to the corresponding receiving end;

[0126] It should be explained that the execution type is based on the inspection type. For example, if the task execution instruction set includes drone inspection, the receiving end is the drone nest; if the task instruction set includes sensor inspection, the receiving end is the intelligent controller; if the task instruction set includes manual inspection, the receiving end is the mobile receiving end of the inspection personnel.

[0127] S5.3: Monitor the execution status of the instructions in step S5.2, and generate corresponding execution result reports based on the monitoring results. The monitoring results include received, executing, completed, received failed, and executed failed.

[0128] S5.4: Output all execution result reports to the administrator's receiving end;

[0129] The hazard consequence assessment and analysis module is used to analyze the dynamic risk prediction report, simulate the severity of consequences based on the analysis results, and generate a maintenance strategy recommendation report in combination with the risk level.

[0130] Furthermore, the steps involved in analyzing the dynamic risk prediction report, simulating the severity of consequences based on the analysis results, and generating a maintenance strategy recommendation report in conjunction with the risk level include:

[0131] S6.1: Read the dynamic risk prediction report, identify the tower IDs with high risk in the dynamic risk prediction report, and retrieve the power grid information of the tower IDs with high risk from the database to obtain the power grid information report;

[0132] It should be explained that power grid information includes, but is not limited to, information such as the topology of lines with high-risk tower IDs and the importance of the power grid;

[0133] S6.2: Based on the power grid information report, assess the downtime consequences for all high-risk tower IDs to obtain the tower assessment value. The specific calculation formula for the assessment is as follows:

[0134] ;

[0135] Obtain the tower's evaluation value ,in, For the number of users, For the first The load value of a user who is unable to receive power due to tower shutdown. To estimate the downtime, Economic loss per unit of electricity;

[0136] S6.3: Classify the tower assessment value based on the assessment threshold range (Q1, Q2). When the tower assessment value is less than Q1, a low consequence level is generated. When the tower assessment value is greater than or equal to Q1 and less than Q2, a medium consequence level is generated. When the tower assessment value is greater than Q2, a high consequence level is generated.

[0137] Package low-consequence, medium-consequence, and high-consequence levels to obtain a consequences severity report;

[0138] S6.4: Retrieve the decision rule table from the database, and input the risk level from the dynamic risk prediction report and the consequence level from the consequence severity report into the decision rule table to output the maintenance strategy report;

[0139] It should be explained that the decision rule table refers to, for example, when the risk level is high risk and the consequence level is high consequence level, the maintenance strategy report includes instructions to ask staff to go to the tower ID immediately for maintenance;

[0140] S6.5: Based on the maintenance strategy report, generate corresponding recommended maintenance measures and estimated economic losses, and integrate them into the maintenance strategy report to obtain a maintenance strategy recommendation report;

[0141] S6.6: Output the maintenance strategy recommendation report to the resource scheduling and sorting module;

[0142] The resource scheduling and sorting module is used to analyze the maintenance strategy recommendation report to obtain a maintenance work order sorting table and a resource scheduling instruction set.

[0143] Further steps in analyzing the maintenance strategy recommendation report include:

[0144] S7.1: Based on the maintenance strategy recommendation report, the optimization target is set by retrieving the maintenance resource table from the database and generating the optimization objective function. The maintenance resource table includes the daily workload of all maintenance personnel teams, the amount of spare parts stored, the number of maintenance vehicles, and the expected tower downtime window.

[0145] It should be explained that optimization objectives include, but are not limited to, minimizing overall grid risk, minimizing total maintenance costs, or maximizing maintenance efficiency;

[0146] S7.2: Use a genetic algorithm to solve the objective function and output the optimal maintenance task report;

[0147] S7.3: Based on the optimal maintenance task, generate a maintenance scheduling report and output the maintenance scheduling report to the receiving end of the corresponding maintenance personnel team;

[0148] It needs to be explained that a maintenance scheduling report refers to, for example, a report that explains that the maintenance team of Group B is requested to go to tower ID 001 for maintenance during the estimated maintenance time interval from 09:00 to 15:00, with work order number ***-********-001;

[0149] S7.4: Monitor the execution result of the maintenance scheduling report in step S7.3. When the execution result is completed, trigger the multi-source data acquisition and fusion module to acquire new raw data and use the maintenance result as feedback data to update the three-dimensional model of the tower in step S2.2.

[0150] In this embodiment, the beneficial effects are achieved by real-time acquisition of raw data based on a sensor network, preprocessing the raw data to obtain a raw fused dataset, extracting a defect feature set from the raw fused dataset, and performing a health assessment based on a tower structure evaluation model to obtain a structural health assessment report. Based on future environmental data and the tower structure evaluation model, the future stability of the tower is predicted to obtain a dynamic risk prediction report. Analysis of the dynamic risk prediction report yields an intelligent inspection task report, which is then converted into an executable task instruction set, and the execution status of the instructions is monitored. The dynamic risk prediction report is analyzed, and the severity of consequences is simulated based on the analysis results. A maintenance strategy recommendation report is generated based on the risk level, and analysis of the maintenance strategy recommendation report yields a maintenance work order ranking table and a resource scheduling instruction set. This allows the system to not only identify surface defects of the tower but also acquire real-time internal information through the collaborative operation of the multi-source data acquisition and fusion module, the structural health assessment quantification module, and the multiple dynamic risk prediction and assessment modules. By combining structural health data with environmental meteorological data, the system can further predict the structural safety of towers under extreme weather conditions, thereby maximizing the safety and reliability of towers. Furthermore, this invention effectively optimizes the traditional fixed inspection scheme through the collaborative operation of an intelligent inspection strategy generation module and an inspection strategy instruction generation and distribution module. It transforms the traditional "flood irrigation" inspection method into a low-cost "on-demand" inspection method, thus minimizing inspection costs while ensuring inspection quality. Finally, the establishment of a hazard consequence assessment and analysis module and a resource scheduling and ranking module allows the system to effectively consider the importance of individual towers and comprehensively analyze the impact of tower downtime on the overall power grid. Simultaneously, it formulates maintenance plans based on maintenance resources and various costs, ensuring that limited maintenance resources are used on the most important towers, thereby maximizing the overall efficiency of the power grid. Overall, this invention has significant advantages such as effective proactive defect detection, substantial improvement in inspection efficiency, and strong overall maintenance decision-making capabilities.

[0151] Example 2

[0152] Please see Figure 2 As shown, the parts not described in detail in this embodiment are described in Embodiment 1. A method for detecting structural hazards of power transmission towers based on artificial intelligence is provided. The method includes: S1: Real-time acquisition of raw data based on sensor network, and preprocessing of raw data to obtain raw fusion dataset;

[0153] S2: Extract the defect feature set based on the original fusion dataset, and conduct a health assessment based on the tower structure assessment model to obtain a structural health assessment report;

[0154] S3: Based on future environmental data and according to the tower structure assessment model, predict the future stability of the tower and obtain a dynamic risk prediction report;

[0155] S4: Based on the analysis of the dynamic risk prediction report, an intelligent inspection task report is obtained;

[0156] S5: Converts intelligent inspection task reports into a set of executable task instructions and monitors the execution status of the instructions;

[0157] S6: Analyze the dynamic risk prediction report, simulate the severity of consequences based on the analysis results, and generate a maintenance strategy recommendation report in combination with the risk level;

[0158] S7: Based on the maintenance strategy recommendation report, an analysis is performed to obtain a maintenance work order sorting table and a resource scheduling instruction set.

[0159] This embodiment also provides:

[0160] An artificial intelligence-based device for detecting structural hazards in power transmission towers includes a memory and a processor. The memory stores a computer program, which is executed by the processor when loaded.

[0161] A readable storage medium storing a computer program, the computer program being adapted to execute the above-described artificial intelligence-based method and system for detecting structural hazards in power transmission towers when loaded by a processor.

[0162] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the present invention.

Claims

1. A power transmission tower structure hidden danger detection system based on artificial intelligence, characterized in that, The system comprises: a dynamic risk prediction evaluation module for predicting the future stability of the tower based on future environmental data and according to a tower structure evaluation model, to obtain a dynamic risk prediction report; an intelligent inspection strategy generation module for analyzing the dynamic risk prediction report to obtain an intelligent inspection task report; an inspection strategy instruction generation and issuing module for converting the intelligent inspection task report into an executable task instruction set and monitoring the execution status of the instructions; a hidden danger consequence evaluation and analysis module for analyzing the dynamic risk prediction report, simulating the severity of the consequences according to the analysis results, and generating a maintenance strategy recommendation report in combination with the risk level; a resource scheduling and sequencing module for analyzing the maintenance strategy recommendation report to obtain a maintenance work order sequencing table and a resource scheduling instruction set, respectively. 2.The AI-based power transmission tower structure defect detection system according to claim 1, wherein The system further comprises a multi-source data acquisition and fusion module and a structure health evaluation and quantification module, wherein: The multi-source data acquisition and fusion module is configured to acquire raw data in real time based on a sensor network and pre-process the raw data to obtain a raw fusion data set; The structure health evaluation and quantification module is configured to extract a defect feature set based on the raw fusion data set and perform health evaluation according to a tower structure evaluation model to obtain a structure health evaluation report. 3.The AI-based power transmission tower structure defect detection system of claim 2, wherein The multi-source data acquisition and fusion module is specifically configured to: continuously receive the raw data pushed by the sensor network through a standard API interface to obtain a raw data set; pre-process all data items in the raw data set based on data types to obtain a pre-processed data set, the data types including image data, point cloud data, and environmental data, wherein: the image data in the raw data set is subjected to denoising and correction processing respectively to complete the pre-processing of the data type as image data; the point cloud data in the raw data set is subjected to denoising, registration, and downsampling processing respectively to complete the pre-processing of the data type as point cloud data; the environmental data in the raw data set is subjected to unit unification and validity test processing respectively to complete the pre-processing of the data type as environmental data; unify all data items in the pre-processed data set based on the same time and the same coordinate system, and map the data types as image data and point cloud data in the pre-processed data set correspondingly to obtain a raw fusion data set. 4.The AI-based power transmission tower structure defect detection system according to claim 2, characterized in that, The structure health evaluation and quantification module is specifically configured to: identify and quantify the defects of the tower based on the raw fusion data set and using an AI model and an image processing algorithm to obtain a defect feature set; retrieve a three-dimensional model of the tower based on a database, and update the three-dimensional model of the tower by taking the defect feature set as a parameter to obtain a new three-dimensional model of the tower; input the environmental type data in the raw fusion data set and the defect feature set into the new three-dimensional model of the tower, and calculate through a calculation engine in the new three-dimensional model of the tower to obtain a tower state report; calculate based on the tower state report to obtain a tower health score, and integrate the tower health score into the tower state report to obtain a structure health evaluation report; output the structure health evaluation report to the dynamic risk prediction evaluation module. 5.The AI-based power transmission tower structure defect detection system of claim 3, wherein The dynamic risk prediction evaluation module is specifically configured to: The weather data in a future preset time window in a specified area is acquired based on a weather service platform in a sensor network, and is arranged in time sequence based on timestamps to obtain a weather data set; Based on the weather data set, the weather data of all time points in the weather data set is output to a new tower three-dimensional model to obtain a tower risk probability sequence; Based on the tower risk probability sequence, a risk level assessment is performed to obtain a risk level report, wherein the risk level includes low risk, medium risk, and high risk; The tower risk probability sequence, the risk level report, and risk causes are packaged to obtain a dynamic risk prediction report; The dynamic risk prediction report is output to an intelligent inspection strategy generation module and a hidden danger consequence assessment and analysis module. 6.The AI-based power transmission tower structure defect detection system of claim 5, wherein The intelligent inspection strategy generation module is specifically configured to: read the dynamic risk prediction report, and sort the risk levels and the tower risk probabilities in descending order to obtain a risk priority report; based on the risk priority report, generate a key inspection task report for the tower IDs with medium risk and high risk, and generate a normal inspection task report for the tower IDs with low risk according to a database call standard inspection period, and package the key inspection task report and the normal inspection task report to obtain a comprehensive inspection task report; based on the risk causes in the dynamic risk prediction report and all key inspection task reports in the comprehensive inspection task report, generate an inspection task content report; integrate the inspection task content report in the comprehensive inspection task report based on the tower ID to obtain an intelligent inspection task report; output the intelligent inspection task report to an inspection strategy instruction generation and distribution module. 7.The AI-based power transmission tower structure defect detection system of claim 6, wherein The inspection strategy instruction generation and distribution module is specifically configured to: based on the intelligent inspection task report, input the intelligent inspection task report into a standard instruction template according to a database call standard instruction template to output a task execution instruction set; output the execution type according to the task execution instruction set to the corresponding receiving end; monitor the instruction execution state, and generate corresponding execution result reports according to the monitoring results, wherein the monitoring results include received, executing, completed, failed to receive, and failed to execute; output all the execution result reports to a management personnel receiving end. 8.The AI-based power transmission tower structure defect detection system of claim 3, wherein The hidden danger consequence assessment and analysis module is specifically configured to: read the dynamic risk prediction report, identify the tower IDs with high risk in the dynamic risk prediction report, and obtain a power grid information report based on database call power grid information of the tower IDs with high risk; based on the power grid information report, perform a shutdown consequence assessment on all the tower IDs with high risk to obtain a tower assessment value; based on an assessment threshold interval (Q1, Q2), classify the tower assessment value, generate a low consequence level when the tower assessment value is less than Q1, generate a medium consequence level when the tower assessment value is greater than or equal to Q1 and less than Q2, and generate a high consequence level when the tower assessment value is greater than Q2; package the low consequence level, the medium consequence level, and the high consequence level to obtain a consequence severity report; and output the consequence severity report to a management personnel receiving end. Based on the database, the decision rule table is called, and the risk level in the dynamic risk prediction report and the consequence level in the consequence severity report are input into the decision rule table, and the maintenance strategy report is output; Based on the maintenance strategy report, corresponding recommended maintenance measures and predicted economic losses are generated and integrated into the maintenance strategy report to obtain a maintenance strategy suggestion report; The maintenance strategy suggestion report is output to the resource scheduling and sequencing module. 9.The AI-based power transmission tower structure defect detection system of claim 8, wherein The resource scheduling and sequencing module is specifically configured to: Based on the maintenance strategy suggestion report, the optimization target is set according to the database calling of the maintenance resource table, and the optimization objective function is generated, and the maintenance resource table includes the daily workload of all maintenance personnel teams, the storage amount of spare parts, the number of maintenance vehicles, and the predicted tower downtime window; The optimization objective function is solved by using a genetic algorithm, and an optimal maintenance task report is output; Based on the optimal maintenance task, a maintenance scheduling report is generated, and the maintenance scheduling report is output to the receiving end of the corresponding maintenance personnel team; The execution result of the maintenance scheduling report is monitored, and when the execution result is completed, the multi-source data acquisition and fusion module is triggered to perform new original data acquisition, and the maintenance result is used as feedback data to update the tower three-dimensional model.

10. The method for detecting hidden dangers of a power transmission tower structure based on artificial intelligence according to any one of claims 1-9, wherein the method is implemented by the system for detecting hidden dangers of a power transmission tower structure based on artificial intelligence. The following working steps are included: S1: Real-time acquisition of original data based on a sensor network, and preprocessing of the original data to obtain an original fusion data set; S2: Extracting a defect feature set based on the original fusion data set, and performing health assessment according to a tower structure assessment model to obtain a structure health assessment report; S3: Based on future environmental data, the future stability of the tower is predicted according to the tower structure assessment model to obtain a dynamic risk prediction report; S4: Based on the dynamic risk prediction report, an intelligent inspection task report is obtained; S5: The intelligent inspection task report is converted into an executable task instruction set, and the execution state of the instruction is monitored; S6: The dynamic risk prediction report is analyzed, the consequence severity simulation is performed according to the analysis result, and the maintenance strategy suggestion report is generated combined with the risk level; S7: Based on the maintenance strategy suggestion report, a repair work order sequencing table and a resource scheduling instruction set are obtained respectively.

Images