Large-scale transformer substation unmanned aerial vehicle inspection method based on intelligent operation and maintenance
By establishing panoramic coordinates and anomaly matrix analysis in large substations, identifying and positioning abnormal points, the problem of time lag in risk sources found during drone inspections is solved, the inspection efficiency and safety are improved, and the on-site risks are reduced.
Patent Information
- Application Number
- CN202410203036.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-02-23
- Publication Date
- 2024-11-26
AI Technical Summary
In large substations, drone inspections are prone to delay time discovery of risk sources. Traditional fixed path inspections are inefficient and difficult to quickly identify abnormal points, resulting in low risk response efficiency.
Create a model by obtaining substation information, setting panoramic coordinates, conducting simulation inspections and image evaluations, generating panoramic value distribution maps, determining the combination of points to be selected, selecting the point corresponding to the minimum simulation time as panoramic points, identifying abnormal points and performing positioning inspections , the diode model is used to analyze the anomaly matrix to calculate the comprehensive outlier value, mark the abnormal points, and control the drone to perform positioning inspection based on the abnormal point coordinates.
It has achieved rapid detection of abnormal points in large substation inspections, improved risk response time, improved patrol efficiency, reduced workload and time costs of operation and maintenance personnel, and reduced on-site operation risks.
Smart Images

Figure CN119024858A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of large-scale substation drone inspection, and specifically is a large-scale substation drone inspection method based on intelligent operation and maintenance. Background Art
[0002] With the rapid development of science and technology, drone technology has been widely used in various fields, especially in the field of power grids. Drone inspection technology is gradually replacing traditional manual inspection methods with its high efficiency, flexibility and safety. In large substations, drone inspection technology can realize functions such as remote equipment monitoring and automatic temperature measurement, thereby greatly reducing the risk of on-site operations, improving equipment operation and maintenance efficiency, and ensuring the safe and reliable operation of the power grid. In particular, drones can be equipped with advanced equipment such as high-definition cameras and infrared thermal imagers to inspect high-altitude equipment such as substation structures, lightning rods, and busbars, and promptly discover hidden dangers and defects such as rust, looseness, and falling off.
[0003] However, the current drone inspection has certain shortcomings, such as risk response efficiency, especially for large substations, which need to quickly discover risk sources. Inspections only along fixed routes are prone to time lags in discovering risk sources. Based on this, the present invention provides a large substation drone inspection method based on intelligent operation and maintenance. Summary of the invention
[0004] In order to solve the problems existing in the above-mentioned solutions, the present invention provides a large-scale substation drone inspection method based on intelligent operation and maintenance.
[0005] The purpose of the present invention can be achieved by the following technical solutions:
[0006] A large-scale substation drone inspection method based on intelligent operation and maintenance, the method comprising:
[0007] Step 1: Obtain substation information, establish a substation model according to the substation information, set panoramic coordinates based on the substation model; and mark the panoramic coordinates in the substation model;
[0008] Furthermore, the method for setting the panoramic coordinates includes:
[0009] Acquire drone information, and determine a flight restricted area according to the drone information; control the drone to perform a simulated inspection in the flight restricted area to obtain a simulated image of each point in the flight restricted area; evaluate the simulated image to obtain a corresponding panoramic value;
[0010] Generate a panoramic value distribution map within the flight restriction area according to the panoramic value of each point, and determine each candidate point combination according to the panoramic value distribution map; estimate the simulation time corresponding to each candidate point combination, select the point in the candidate point combination corresponding to the minimum simulation time as the panoramic point, and identify the panoramic coordinates of the panoramic point.
[0011] Further, the method for evaluating the simulated image includes:
[0012] Identify the information of each power device in the substation model, calculate the initial value of each power device according to the power device information; and mark the obtained initial value of each device in the substation model;
[0013] Identify the simulation point range corresponding to the simulation image according to the substation model, identify the equipment initial value corresponding to each power device within the simulation point range, calculate the sum of the equipment initial values of each power device within the simulation point range, and obtain the simulation range value; calculate the sum of the equipment initial values of all power devices in the substation model to obtain the regional equipment value;
[0014] The obtained simulation range value and regional device value are marked as FZ and QZ respectively;
[0015] Calculate the coverage of the simulation point range for the substation model; mark the obtained coverage as FP;
[0016] According to the formula Calculate the corresponding panoramic value; where: QUZ is the panoramic value.
[0017] Furthermore, the calculation method of the initial value of the device includes:
[0018] According to the power equipment information, the risk category corresponding to each power equipment and the probability of occurrence and economic loss of each risk category are counted; the risk category corresponding to the power equipment is marked as c, z=1, 2, ..., v, v is a positive integer; the corresponding probability of occurrence and economic loss are marked as PLc and SLc respectively;
[0019] According to the formula Calculate the initial value of each power equipment; where: CL is the initial value of the equipment; e is a constant.
[0020] Step 2: Control the drone to collect data at the panoramic coordinates to obtain a panoramic image of the substation; analyze the panoramic image to obtain corresponding abnormal points;
[0021] Furthermore, the method for analyzing the panoramic image includes:
[0022] Obtain a substation model, identify the initial value of each power device in the substation model; set reference characteristics of each power device; obtain the operating data of each power device in the substation in real time, extract the operating characteristics of the power device according to the operating data; compare and analyze the reference characteristics with the operating characteristics to obtain the corresponding adjustment coefficient;
[0023] Acquire a panoramic image, segment the panoramic image according to each power device, and obtain a single image corresponding to each power device;
[0024] Perform grayscale processing on a single image to obtain a single grayscale image, mark each pixel in the single grayscale image as q, q = 1, 2, ..., p, and p is a positive integer; determine the positioning attribute and grayscale value of each pixel; mark the grayscale value and positioning attribute corresponding to the pixel as Hq and Wq respectively;
[0025] A bipolar model is established, and each pixel is analyzed by the bipolar model to obtain the pixel abnormal value corresponding to each pixel;
[0026] The abnormal matrix R is established according to the position of each pixel and the abnormal value of the pixel. rij represents the corresponding element in the abnormal matrix R, i = 1, 2, ..., n, n is a positive integer; j = 1, 2, ..., m, m is a positive integer;
[0027] According to the formula Calculate the corresponding comprehensive outlier value;
[0028] Where: YCF is the comprehensive outlier value; ||R|| F represents the Frobenius norm of the abnormal matrix R; CL is the initial value of the equipment; τ is the adjustment coefficient;
[0029] The power equipment whose comprehensive abnormal value is greater than the threshold value X1 is marked as an abnormal point.
[0030] Furthermore, the expression of the two-pole model is Where: x is the input data, x = (Hq, Wq); the output data is the pixel outlier value.
[0031] Step 3: Obtain the coordinates of the abnormal point, and control the drone to perform positioning inspection on the abnormal point according to the coordinates of the abnormal point; and obtain corresponding abnormal inspection data;
[0032] When there are no abnormal points, go to step 5;
[0033] Furthermore, when there are multiple abnormal points, the corresponding abnormal points are located and inspected in order of the comprehensive abnormal values from high to low.
[0034] Step 4: Analyze the abnormal inspection data to determine whether there is a security risk;
[0035] When it is determined that there is no security risk, proceed to step 5;
[0036] When it is determined that there is a security risk, appropriate security processing is carried out;
[0037] Step 5: Obtain the original inspection plan and control the drone to conduct inspection according to the original inspection plan.
[0038] Compared with the prior art, the present invention has the following beneficial effects:
[0039] By using the present invention, during the inspection of large-scale substations, rapid inspection of large-scale substations can be achieved in the early stage of the inspection, so that existing abnormal points can be discovered in time, and positioning inspections can be carried out on the abnormal points, which greatly improves the risk response time; substations can be inspected quickly by drones, which greatly improves the inspection efficiency and reduces the workload and time cost of operation and maintenance personnel compared to traditional manual inspection methods; at the same time, drone inspections can prevent operation and maintenance personnel from entering high-risk areas, reducing the risks of on-site operations. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0041] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION
[0042] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0043] like Figure 1 As shown, a large-scale substation drone inspection method based on intelligent operation and maintenance includes:
[0044] Step 1: Obtain substation information, including relevant data such as the type, model, distribution location, coordinate information of each device, establish a corresponding substation model based on the substation information, and set the panoramic coordinates based on the substation model;
[0045] The substation model is a three-dimensional data model built based on existing modeling technology, which is used to intuitively display substation information.
[0046] The methods for setting panorama coordinates include:
[0047] Obtain the information of the drone used for inspection, control the drone to perform simulated inspection, and obtain the space area that meets the requirements of drone inspection, that is, determine the maximum and minimum flight altitudes of the drone according to the drone performance and inspection requirements, and then adjust them according to the actual flight control situation to obtain the space area where the drone can fly for the current substation; mark the corresponding space area as a flight restricted area;
[0048] Control the UAV to fly in the flight restriction area, obtain the simulated image of each point, evaluate the simulated image, and obtain the corresponding panoramic value; generate a panoramic value distribution map in the flight restriction area according to the panoramic value of each point, and identify the comprehensive panoramic value of each point combination. 0.7 The combination is marked as the candidate point combination. The comprehensive panoramic value is calculated according to the comprehensive image corresponding to the combination point. Multiple simulated images are combined to calculate the corresponding panoramic value as the comprehensive panoramic value of the combination. An independent point satisfies e 0.7 The requirement can also be used as a combination of candidate points;
[0049] Estimate the collection time of controlling the drone according to the candidate point combination, that is, the time from takeoff to completion of the collection of the corresponding position of the candidate point combination, marked as simulation time; select the point in the candidate point combination corresponding to the minimum simulation time as the panoramic point, and identify the panoramic coordinates of the panoramic point.
[0050] Among them, the evaluation methods of panoramic value include:
[0051] Identify the information of each power equipment in the substation model, and count the risk categories corresponding to each power equipment, the probability of occurrence of each risk category, and the economic loss after occurrence according to the equipment information, and take the average value; mark the risk category corresponding to the power equipment as c, c = 1, 2, ..., v, v is a positive integer; mark the corresponding probability of occurrence and economic loss as PLc and SLc respectively;
[0052] According to the formula Calculate the initial value of each power equipment; where: CL is the initial value of the equipment; e is a constant;
[0053] Mark the obtained initial values of each device in the substation model accordingly;
[0054] Identify the acquisition range of the simulation image for the substation model, that is, the shooting range that can be collected and meets the inspection requirements, mark the corresponding acquisition range as the simulation point range, and take the maximum simulation point range shot at each angle of the point as the standard; identify the equipment initial value corresponding to each power equipment within the simulation point range, calculate the sum of the equipment initial values of each power equipment within the simulation point range, and obtain the simulation range value; calculate the sum of the equipment initial values of all power equipment in the substation model to obtain the regional equipment value;
[0055] The obtained simulation range value and regional device value are marked as FZ and QZ respectively;
[0056] Calculate the coverage of the simulation point range for the substation model; mark the obtained coverage as FP;
[0057] According to the formula Calculate the corresponding panoramic value; where: QUZ is the panoramic value.
[0058] Step 2: When inspection is required, control the drone to collect data at the panoramic point to obtain a panoramic image of the substation; analyze the obtained panoramic image to obtain the corresponding abnormal points;
[0059] Methods for analyzing panoramic images include:
[0060] Obtain the substation model and identify the initial value of each power device in the substation model; set the reference characteristics of each power device, which is composed of the characteristic item data corresponding to each characteristic item. The characteristic item is the data classification related to its safety, such as power, current, voltage, etc.; specifically, the reference characteristics are set according to the corresponding data when the initial value of the calculated device corresponds to each risk category, which is equivalent to the critical condition for the occurrence of the risk; in actual applications, the reference characteristics of each power device are generally set directly by professionals in the substation;
[0061] Obtain the operating data of each power device in the current substation, mainly referring to the relevant data related to its operating safety, and extract the operating characteristics of each power device in real time according to the operating data, that is, extract data according to each feature item corresponding to the reference feature;
[0062] Compare and analyze the reference characteristics with the operating characteristics to obtain the corresponding adjustment coefficient, that is, make a comparison and analyze whether the risk probability has changed. If it is within the normal range, it will not change. Otherwise, it will be adjusted according to the situation, and the adjustment coefficient is greater than or equal to 1; such as high-load operation, long-term operation, etc.; specifically, establish a corresponding adjustment analysis model based on a neural network such as a CNN network or a DNN network, and manually establish a corresponding training set for training. The training set includes input data and output data. The input data includes power equipment information, reference characteristics and operating characteristics, and the output data is the adjustment coefficient; analyze the adjustment analysis model after successful training to obtain the corresponding adjustment coefficient of each power equipment.
[0063] Obtain a panoramic image, segment the panoramic image according to each power device, obtain a panoramic image corresponding to each power device, and mark it as a single image; for example, the image at the location of the power device can be directly cut;
[0064] Perform grayscale processing on a single image to obtain a corresponding grayscale image, which is marked as a single grayscale image. Mark each pixel in the single grayscale image as q, where q=1, 2, ..., p, and p is a positive integer; determine the positioning attribute of each pixel, which is the position on the power device to which the pixel position is equivalent. Therefore, it is necessary to first perform corresponding positioning based on the existing technology, and then determine the corresponding positioning attribute. Because the position corresponding to the pixel is extremely small, it is sufficient to accurately determine the specific component or structure, and further positioning is also possible; identify the grayscale value of each pixel; mark the grayscale value and positioning attribute corresponding to the pixel as Hq and Wq respectively;
[0065] A two-pole model is established. The two-pole model is based on the isolation forest algorithm and is trained using a large amount of historical risk data. The location attributes outside the power equipment are directly regarded as always normal. The location attributes on the power equipment are judged to be abnormal based on the gray value. The expression is: Where: x is the input data, x = (Hq, Wq); the output data is the abnormal value, that is, 1 or 0;
[0066] Each pixel is analyzed by a bipolar model to obtain the pixel abnormal value corresponding to each pixel;
[0067] According to the position of each pixel and the pixel abnormal value, the corresponding abnormal matrix R is established. That is, each pixel corresponds to an element in the matrix, and the corresponding pixel abnormal value is substituted; rij represents the corresponding element in the abnormal matrix R, i = 1, 2, ..., n, n is a positive integer; j = 1, 2, ..., m, m is a positive integer;
[0068] According to the formula Calculate the corresponding comprehensive outlier value;
[0069] Where: YCF is the comprehensive outlier value; ||R|| F represents the Frobenius norm of the abnormal matrix R; CL is the initial value of the equipment; τ is the adjustment coefficient;
[0070] The power equipment whose comprehensive abnormal value is greater than the threshold value X1 is marked as an abnormal point.
[0071] Step 3: Obtain the coordinates of the abnormal point, and control the drone to locate and inspect the abnormal point according to the coordinates of the abnormal point, that is, conduct a direct, comprehensive and detailed inspection of the abnormal point; obtain the corresponding abnormal inspection data;
[0072] When there are no abnormal points, go to step 5;
[0073] Step 4: Analyze the abnormal inspection data to determine whether there is a security risk;
[0074] When it is determined that there is no security risk, proceed to step 5;
[0075] When it is determined that there is a security risk, corresponding security processing is carried out; that is, it is handled according to the preset security risk processing measures.
[0076] The abnormal inspection data is analyzed and evaluated using existing risk assessment methods. Since there are many mature inspection data risk assessment technologies, no corresponding additional explanation is given.
[0077] Step 5: Get the original inspection plan, and control the drone to perform inspection according to the original inspection plan. The original inspection plan is the previously applied drone inspection control method, or it can be any existing inspection control method.
[0078] The above formulas are all calculated by removing dimensions and taking numerical values. The formula is a formula that is closest to the actual situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained by simulating a large amount of data.
[0079] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.
Claims
1. A large-scale substation drone inspection method based on intelligent operation and maintenance, characterized in that: Methods include: Step 1: Obtain substation information, establish a substation model according to the substation information, set panoramic coordinates based on the substation model; and mark the panoramic coordinates in the substation model; Step 2: Control the drone to collect data at the panoramic coordinates to obtain a panoramic image of the substation; analyze the panoramic image to obtain corresponding abnormal points; Step 3: Obtain the coordinates of the abnormal point, and control the drone to perform positioning inspection on the abnormal point according to the coordinates of the abnormal point; and obtain corresponding abnormal inspection data; When there are no abnormal points, go to step 5; Step 4: Analyze the abnormal inspection data to determine whether there is a security risk; When it is determined that there is no security risk, proceed to step 5; When it is determined that there is a security risk, appropriate security processing is carried out; Step 5: Obtain the original inspection plan and control the drone to conduct inspection according to the original inspection plan.
2. A large-scale substation drone inspection method based on intelligent operation and maintenance according to claim 1, characterized in that: The methods for setting panorama coordinates include: Acquire drone information, and determine a flight restricted area according to the drone information; control the drone to perform a simulated inspection in the flight restricted area to obtain a simulated image of each point in the flight restricted area; evaluate the simulated image to obtain a corresponding panoramic value; Generate a panoramic value distribution map within the flight restriction area according to the panoramic value of each point, and determine each candidate point combination according to the panoramic value distribution map; estimate the simulation time corresponding to each candidate point combination, select the point in the candidate point combination corresponding to the minimum simulation time as the panoramic point, and identify the panoramic coordinates of the panoramic point.
3. A large-scale substation drone inspection method based on intelligent operation and maintenance according to claim 2, characterized in that: Methods for evaluating simulated images include: Identify the information of each power device in the substation model, calculate the initial value of each power device according to the power device information; and mark the obtained initial value of each device in the substation model; Identify the simulation point range corresponding to the simulation image according to the substation model, identify the equipment initial value corresponding to each power device within the simulation point range, calculate the sum of the equipment initial values of each power device within the simulation point range, and obtain the simulation range value; calculate the sum of the equipment initial values of all power devices in the substation model to obtain the regional equipment value; The obtained simulation range value and regional device value are marked as FZ and QZ respectively; Calculate the coverage of the simulation point range for the substation model; mark the obtained coverage as FP; According to the formula Calculate the corresponding panoramic value; where: QUZ is the panoramic value.
4. A large-scale substation drone inspection method based on intelligent operation and maintenance according to claim 3, characterized in that: The calculation method of the initial value of the device includes: According to the power equipment information, the risk category corresponding to each power equipment and the probability of occurrence and economic loss of each risk category are counted; the risk category corresponding to the power equipment is marked as c, z=1, 2, ..., v, v is a positive integer; the corresponding probability of occurrence and economic loss are marked as PLc and SLc respectively; According to the formula Calculate the initial value of each power equipment; where: CL is the initial value of the equipment; e is a constant.
5. A large-scale substation drone inspection method based on intelligent operation and maintenance according to claim 1, characterized in that: Methods for analyzing panoramic images include: Obtain a substation model, identify the initial value of each power device in the substation model; set reference characteristics of each power device; obtain the operating data of each power device in the substation in real time, extract the operating characteristics of the power device according to the operating data; compare and analyze the reference characteristics with the operating characteristics to obtain the corresponding adjustment coefficient; Acquire a panoramic image, segment the panoramic image according to each power device, and obtain a single image corresponding to each power device; Perform grayscale processing on a single image to obtain a single grayscale image, mark each pixel in the single grayscale image as q, q = 1, 2, ..., p, and p is a positive integer; determine the positioning attribute and grayscale value of each pixel; mark the grayscale value and positioning attribute corresponding to the pixel as Hq and Wq respectively; A bipolar model is established, and each pixel is analyzed by the bipolar model to obtain the pixel abnormal value corresponding to each pixel; The abnormal matrix R is established according to the position of each pixel and the abnormal value of the pixel. rij represents the corresponding element in the abnormal matrix R, i = 1, 2, ..., n, n is a positive integer; j = 1, 2, ..., m, m is a positive integer; According to the formula Calculate the corresponding comprehensive outlier value; Where: YCF is the comprehensive outlier value; ||R|| F represents the Frobenius norm of the abnormal matrix R; CL is the initial value of the equipment; τ is the adjustment coefficient; The power equipment whose comprehensive abnormal value is greater than the threshold value X1 is marked as an abnormal point.
6. A large-scale substation drone inspection method based on intelligent operation and maintenance according to claim 5, characterized in that: The expression of the two-pole model is Where: x is the input data, x = (Hq, Wq); the output data is the pixel outlier value.
7. A large-scale substation drone inspection method based on intelligent operation and maintenance according to claim 1, characterized in that: In step three, when there are multiple abnormal points, the corresponding abnormal points are located and inspected in order of the comprehensive abnormal values from high to low.
Citation Information
Cited By
Transformer abnormal state intelligent identification method, system and device
CN120635756A
