Unmanned aerial vehicle and one-key sequence control combined operation method based on knife switch identification algorithm

Abstract

The invention provides an unmanned aerial vehicle and one-key sequence control combined operation method based on a disconnecting link identification algorithm, and relates to the technical field of substation automation, the method comprises the following steps: a substation one-key sequence control host is used to receive a one-key sequence control operation instruction, and an intelligent anti-error host executes anti-error verification preview according to an operation order; knife switch image acquisition is carried out through a camera carried by the unmanned aerial vehicle; identifying the on-off state of the disconnecting link by using a disconnecting link identification algorithm; remote signaling data of the disconnecting link during the opening and closing operation are obtained in real time, state evaluation is carried out on the operation state of the disconnecting link during the opening and closing operation through a fusion set pair analysis method and an evidence theory method, and a state evaluation result is obtained; and performing bidirectional approval on the state of the disconnecting link, and executing a preset disconnecting link processing scheme according to an approval result. The intelligent level of the transformer substation is improved, and safe and stable operation of a power system is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of substation automation, and in particular to a method for combining an unmanned aerial vehicle and a one-button sequential control based on a switch recognition algorithm. Background Art

[0002] In recent years, with the full promotion of one-button sequential control technology, substation master station systems including centralized control station equipment monitoring systems have been widely used in the control operation of substation equipment. One-button sequential control is frequently used in substation equipment operation due to its high efficiency and convenience. However, in actual applications, complex model configuration work is required before one-button sequential control operation, especially the consistency configuration of equipment status names. For example, all equipment status names of a certain interval must be completely consistent between the substation system and the master station system, otherwise it will lead to operation failure or misjudgment.

[0003] The traditional method of knife switch status recognition mainly relies on single sensor data, which has obvious defects. Since a single data source is easily affected by multiple external factors such as environmental noise, equipment aging, and electromagnetic interference, data errors will gradually accumulate, resulting in deviations in the judgment of the true state of the knife switch. Especially in the complex and changeable substation environment, this recognition method that relies on a single data source is more difficult to ensure the accuracy and reliability of recognition.

[0004] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0005] In view of this, the present invention provides a combined operation method of a UAV and a one-button sequential control based on a knife switch recognition algorithm to solve the above-mentioned problems.

[0006] In order to solve the above problems, the specific technical solutions adopted by the present invention are as follows:

[0007] According to one aspect of the present invention, a method for joint operation of a drone and a one-button sequential control based on a knife switch recognition algorithm is provided, the method comprising the following steps:

[0008] S 1. Use the substation one-key sequential control host to receive the one-key sequential control operation instruction, and generate an operation ticket by analyzing the one-key sequential control operation instruction; the intelligent anti-error host performs the preview of the anti-error verification according to the operation ticket. If the verification is successful, the intelligent anti-error host transmits the operation ticket and the knife switch preview process image to the substation master station, otherwise, verify and adjust the one-key sequential control operation instruction until the verification is successful;

[0009] S2. When the substation master station receives the operation ticket, it uses the text recognition method to identify the switch operation content in the operation ticket, generates a UAV control instruction according to the switch operation content, and transmits the UAV control instruction to the pre-set UAV;

[0010] S3. The UAV autonomously flies to the target switch position according to the UAV control command, and collects the switch image through the camera carried by the UAV; based on the switch preview process image and the switch image, the switch recognition algorithm is used to identify the switch opening and closing state;

[0011] S4. Real-time acquisition of remote signal data of the knife switch during the opening and closing operation, and the state evaluation of the operation state of the knife switch during the opening and closing operation is performed by integrating the set pair analysis method and the evidence theory method to obtain the state evaluation result;

[0012] S5. Based on the switch opening and closing states identified by the switch recognition algorithm and the state evaluation results of the switch during the opening and closing operations, the state of the switch is bidirectionally verified, and a preset switch processing plan is executed according to the verification results.

[0013] Preferably, after receiving the operation ticket, the substation master station uses a text recognition method to identify the switch operation content in the operation ticket, generates a UAV control instruction according to the switch operation content, and transmits the UAV control instruction to a pre-set UAV, including the following steps:

[0014] S21, preprocessing the received operation ticket, the preprocessing includes image denoising, grayscale and binarization processing;

[0015] S22, using optical character recognition to perform text recognition on the pre-processed operation ticket, and extracting the knife switch operation content in the operation ticket, the knife switch operation content including the knife switch number, operation type and operation sequence;

[0016] S23, parsing the knife switch operation content in the operation ticket, and generating a control instruction for the UAV using a preset instruction template according to the parsing result, the control instruction for the UAV including the flight altitude, the target knife switch position and the shooting angle;

[0017] S24, using the command transmission protocol to transmit the control command of the drone to the pre-set drone.

[0018] Preferably, the UAV autonomously flies to the target switch position according to the UAV control command, and collects the switch image through the camera carried by the UAV; based on the switch preview process image and the switch image, using the switch recognition algorithm to identify the switch opening and closing state includes the following steps:

[0019] S31, the UAV flies to the target switch position according to the control command, and uses the camera to continuously collect images of the switch;

[0020] S32, using image processing technology to extract the characteristic points of the knife switch image and the knife switch action during the knife switch preview process, and to extract the three-dimensional characteristic parameters of the knife switch in the knife switch image and the knife switch preview process image;

[0021] S33. Based on a similarity calculation method, the three-dimensional characteristic parameters of the knife switch in the knife switch image and the knife switch preview process image are compared, and the opening and closing states of the knife switch are determined according to the comparison results.

[0022] Preferably, using image processing technology to extract the characteristic points of the knife switch action in the knife switch image and the knife switch preview process image respectively, and extracting the three-dimensional characteristic parameters of the knife switch in the knife switch image and the knife switch preview process includes the following steps:

[0023] S321, graying, filtering and edge detection processing are performed on the knife switch image and the knife switch preview process image respectively;

[0024] S322, extracting feature points from the knife gate image after edge detection and the knife gate preview process image respectively using a feature point detection algorithm;

[0025] S323, using a three-dimensional reconstruction algorithm to calculate the spatial coordinates of the feature points, and extracting the three-dimensional feature parameters of the knife switch in the knife switch image and the knife switch preview process image.

[0026] Preferably, using a three-dimensional reconstruction algorithm to calculate the spatial coordinates of the feature points and extracting the three-dimensional feature parameters of the knife switch in the knife switch image and the knife switch preview process image comprises the following steps:

[0027] S3231, using a three-dimensional reconstruction algorithm, mapping the coordinates of feature points of the knife switch at different positions in the knife switch image and the knife switch preview process image from the two-dimensional image to the three-dimensional space, constructing a spatial structure model of the knife switch, and calculating the three-dimensional coordinates of each feature point by using stereo vision technology;

[0028] S3232, normalizing the three-dimensional coordinates of each feature point, and determining the spatial position of the key parts of the knife switch based on the normalized three-dimensional coordinates, and calculating the relative distance and orientation between the key parts of the knife switch;

[0029] S3233. According to the movement characteristics of the knife switch, analyze the movement trajectory, speed and changes in the opening and closing angle of the characteristic point in the three-dimensional space.

[0030] Preferably, based on a similarity calculation method, the three-dimensional characteristic parameters of the knife switch in the knife switch preview process image are compared, and the opening and closing states of the knife switch are determined according to the comparison results, including the steps of:

[0031] S331, converting the three-dimensional characteristic parameters of the knife switch in the knife switch image and the knife switch preview process image into a characteristic vector by using a characteristic vector conversion method;

[0032] S332, using a similarity calculation method, calculating the similarity between the knife switch image and the feature vector of the knife switch in the knife switch preview process image;

[0033] S333. According to the similarity calculation result, determine whether the three-dimensional characteristic parameters of the knife switch in the knife switch image and the preview process image are consistent, and determine the opening and closing state of the knife switch according to the determination result.

[0034] Preferably, real-time acquisition of remote signaling data of the knife switch during the opening and closing operation, and state evaluation of the operation state of the knife switch during the opening and closing operation by integrating set pair analysis method and evidence theory method, and obtaining the state evaluation result includes the following steps:

[0035] S41, establishing an evaluation index system for the operation status of the knife switch according to the remote signal data of the knife switch during the opening and closing operation; and dividing the operation status of the knife switch into different status levels by calculating the relative degradation degree of each index according to the evaluation index system for the operation status of the knife switch;

[0036] S42, using a statistical method to determine the weight of each evaluation index of the switch operation status, and using a set pair analysis method to calculate the degree of connection between each evaluation index and the status level;

[0037] S43, calculating the basic credibility distribution of the evidence theory according to the degree of connection between each evaluation indicator and the state level, and using the evidence synthesis rule to merge the basic credibility distribution of multiple evaluation indicators to obtain the comprehensive credibility distribution of the switch operation state;

[0038] S44. According to the integrated reliability distribution after fusion, combined with the maximum membership principle and the reliability criterion, the current opening and closing operation state level of the knife switch is judged to obtain the state evaluation result of the knife switch.

[0039] Preferably, the calculation formula for calculating the connection degree between each evaluation indicator and the status level using the set pair analysis method is:

[0040]

[0041] In the formula, H t Represents the degree of connection between the evaluation index and the tth state level; N t represents the number of the lth evaluation index; w tl represents the weight of the lth evaluation indicator; a tl Indicates the degree of identity between the lth evaluation indicator and the nth level status level; b tl，n Indicates the difference between the lth evaluation indicator and the nth level state level; c tl It represents the degree of opposition between the lth evaluation index and the nth state level difference K-1 level; i represents the difference uncertainty coefficient; j represents the opposition coefficient.

[0042] Preferably, the basic credibility distribution of the evidence theory is calculated according to the degree of connection between each evaluation index and the state level, and the basic credibility distribution of multiple evaluation indicators is integrated by using the evidence synthesis rule to obtain the comprehensive credibility distribution of the knife switch operation state, which includes the following steps:

[0043] S431, determining the credibility coefficient of each evaluation indicator according to the weight of each evaluation indicator, and calculating the basic credibility distribution of the evidence theory according to the credibility coefficient of each evaluation indicator;

[0044] S432, selecting any two evaluation indicators and their corresponding arbitrary status levels as initial evidence, and preliminarily integrating the basic credibility distribution of the two evaluation indicators to obtain a new comprehensive credibility distribution;

[0045] S433: The comprehensive credibility distribution obtained by preliminary fusion is used as new evidence, and the basic credibility distribution of the next evaluation indicator is selected in turn and fused with the current comprehensive credibility distribution until the basic credibility distributions of all evaluation indicators are fused into the final comprehensive credibility distribution.

[0046] According to another aspect of the present invention, a UAV and one-key sequential control joint operation system based on a knife switch recognition algorithm is provided, the system comprising: an operation ticket generation and verification module, an operation ticket recognition module, a knife switch state recognition module, a knife switch state evaluation module and a knife switch state approval module;

[0047] The operation ticket generation and verification module is used to use the substation one-key sequential control host to receive the one-key sequential control operation instructions, and generate an operation ticket by analyzing the one-key sequential control operation instructions; the intelligent anti-error host performs a preview of the anti-error verification according to the operation ticket. If the verification is successful, the intelligent anti-error host transmits the operation ticket and the knife switch preview process image to the substation master station, otherwise, the one-key sequential control operation instruction is verified and adjusted until the verification is successful;

[0048] The operation ticket recognition module is used to use the text recognition method to identify the switch operation content in the operation ticket after the substation master station receives the operation ticket, generate the UAV control command according to the switch operation content and transmit the UAV control command to the pre-set UAV;

[0049] The switch state recognition module is used for the UAV to autonomously fly to the target switch position according to the UAV control command, and to collect the switch image through the camera carried by the UAV; based on the switch preview process image and the switch image, the switch recognition algorithm is used to identify the switch open and close state;

[0050] The switch status assessment module is used to obtain the remote signal data of the switch in real time during the opening and closing operation. By integrating the set pair analysis method and the evidence theory method, the operating status of the switch during the opening and closing operation is assessed to obtain the status assessment result;

[0051] The knife switch status approval module is used to combine the knife switch opening and closing status identified by the knife switch recognition algorithm and the status evaluation result of the knife switch during the opening and closing operation, to perform bidirectional approval on the status of the knife switch, and to execute the pre-set knife switch processing plan according to the approval result.

[0052] The beneficial effects of the present invention are:

[0053] 1. The present invention performs bidirectional verification of the switch state by combining the on / off state identified by the switch recognition algorithm and the state evaluation result based on the remote signaling data, which can effectively reduce the errors that may be caused by a single identification or evaluation method and improve the accuracy of state recognition. Through one-key sequential control, autonomous flight of unmanned aerial vehicles and image recognition and other technologies, the automation of switch state recognition is realized, which greatly improves the operation and maintenance efficiency, reduces the need for manual intervention, helps to improve the intelligence level of substations, and ensures the safe and stable operation of power systems.

[0054] 2. The present invention extracts the three-dimensional feature parameters of the knife switch through a three-dimensional reconstruction algorithm, which can more accurately reflect the actual state of the knife switch than traditional two-dimensional image recognition. The three-dimensional feature parameters are compared based on the similarity calculation method, which can more accurately judge the opening and closing state of the knife switch and reduce the misjudgment rate. The drone and image processing technology realize non-contact monitoring of the knife switch, avoiding the safety risks brought by direct contact between personnel and high-voltage equipment.

[0055] 3. The present invention establishes an evaluation index system for the operating status of the knife switch, comprehensively evaluates the status of the knife switch from multiple dimensions, avoids the one-sidedness of the single index evaluation, combines the set pair analysis method and the evidence theory method, and fully utilizes the advantages of the two methods in dealing with uncertainty and multi-source information fusion, thereby improving the accuracy of the evaluation. The evaluation method based on remote signaling data and image processing technology realizes non-contact monitoring of the knife switch, reduces the safety risks of operation and maintenance personnel, and will help to improve the intelligence level of the power system and ensure the safe and stable operation of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. 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 creative work. In the drawings:

[0057] Figure 1 is a flow chart of a method for joint operation of a UAV and one-button sequential control based on a knife switch recognition algorithm according to an embodiment of the present invention;

[0058] Figure 2It is a principle block diagram of a UAV and one-button sequential control joint operation system based on a knife switch recognition algorithm according to an embodiment of the present invention.

[0059] In the figure:

[0060] 1. Operation ticket generation and verification module; 2. Operation ticket identification module; 3. Knife switch status identification module; 4. Knife switch status assessment module; 5. Knife switch status approval module. DETAILED DESCRIPTION

[0061] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0062] According to an embodiment of the present invention, a method for jointly operating a UAV and a one-button sequential control based on a knife switch recognition algorithm is provided.

[0063] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to one embodiment of the present invention, a method for joint operation of a UAV and a one-button sequential control based on a knife switch recognition algorithm is provided, and the method comprises the following steps:

[0064] S 1. Use the substation one-key sequential control host to receive the one-key sequential control operation instruction, and generate an operation ticket by analyzing the one-key sequential control operation instruction; the intelligent anti-error host performs the preview of the anti-error verification according to the operation ticket. If the verification is successful, the intelligent anti-error host transmits the operation ticket and the knife switch preview process image to the substation master station, otherwise, verify and adjust the one-key sequential control operation instruction until the verification is successful;

[0065] It should be noted that the one-key sequential control operation instructions are generated by the operator according to the work requirements of the substation, usually including specific operation targets (such as knife switch operation, switch control, etc.) and their execution sequence. Based on the analysis results of the instructions, the one-key sequential control host will automatically generate a detailed operation ticket. The operation ticket lists in detail the specific content, sequence, expected results and safety precautions of each step of the operation.

[0066] The intelligent anti-error host receives the operation ticket generated by the one-button sequence control host. According to the steps in the operation ticket, the intelligent anti-error host will simulate the impact of each operation on the status of the substation equipment one by one. During the simulation process, the intelligent anti-error host will check whether each operation complies with the safety operating procedures of the substation and whether it will cause equipment failure or safety accidents. If all simulated operations are as expected and no potential safety risks are detected, the verification is successful.

[0067] If the verification is successful, the intelligent error prevention host will transmit the operation ticket and the image of the knife switch preview process to the substation master station. The substation master station can further review these operation tickets and images to ensure the accuracy and safety of the operation. If the verification fails, the intelligent error prevention host will feedback the verification results to the one-key sequence control host or operator, indicating the steps and reasons for the verification failure. The operator needs to adjust the one-key sequence control operation instructions or the relevant content in the operation ticket according to the feedback results, and then re-verify until the verification is successful.

[0068] S2. When the substation master station receives the operation ticket, it uses the text recognition method to identify the switch operation content in the operation ticket, generates a UAV control instruction according to the switch operation content, and transmits the UAV control instruction to the pre-set UAV;

[0069] As a preferred implementation, after the substation master station receives the operation ticket, the text recognition method is used to identify the switch operation content in the operation ticket, and the UAV control instruction is generated according to the switch operation content and the UAV control instruction is transmitted to the pre-set UAV, including the following steps:

[0070] S21, preprocessing the received operation ticket, the preprocessing includes image denoising, grayscale and binarization processing;

[0071] It should be noted that image denoising is to remove noise (such as dust marks or scanning errors) in the operation ticket image to make the text edges clearer.

[0072] Grayscale processing is to convert the operation ticket image into a grayscale image, reduce the data dimension, and highlight the text features.

[0073] Binarization processing uses adaptive threshold or Otsu algorithm to convert grayscale images into black and white images to separate text from background.

[0074] S22, using optical character recognition to perform text recognition on the pre-processed operation ticket, and extracting the knife switch operation content in the operation ticket, the knife switch operation content including the knife switch number, operation type and operation sequence;

[0075] It should be noted that optical character recognition is used to convert the image characters in the operation ticket into editable text content. By selecting a mature OCR tool (such as an open source tool for character recognition), text recognition includes:

[0076] Text detection: detect text areas in images and locate text blocks.

[0077] Text extraction: Identify characters in the detection area and convert them into editable text.

[0078] Language model optimization: Introduce a language model for post-processing to improve the accuracy of recognition results.

[0079] The extraction knife switch operation includes:

[0080] Information extraction, extracting content related to knife switch operation from the recognized and corrected text.

[0081] Among them, the knife switch number: usually a string of numbers or letters, used to uniquely identify the knife switch device.

[0082] Operation type: such as "open", "close", etc., indicating the operation instructions for the knife switch device.

[0083] Operation sequence: indicates the order of operations, which may be a numerical sequence, time stamp or other identification method.

[0084] S23, parsing the knife switch operation content in the operation ticket, and generating a control instruction for the UAV using a preset instruction template according to the parsing result, the control instruction for the UAV including the flight altitude, the target knife switch position and the shooting angle;

[0085] It should be noted that parsing the switch operation content in the operation ticket and generating the control instructions of the drone based on the parsing results is a process of converting text information into specific action instructions. Specifically, it includes the following steps:

[0086] Identify key information, switch number, operation type and operation sequence;

[0087] Analyze the instructions in the operation ticket and understand the logical relationship between each operation. For example, some operations may be conditional, that is, the next operation will be executed only after the previous operation is successfully completed.

[0088] Design the control command templates of the drone in advance. These templates contain all the parameters required for the drone to perform the mission, such as flight altitude, target position, shooting angle, etc.

[0089] Fill in the control command parameters of the drone, including:

[0090] Flight altitude: Set the flight altitude of the drone when performing operations according to the layout and safety regulations of the substation.

[0091] The target switch position is to combine the switch number extracted from the operation ticket with the substation's geographic information system (GIS) or CAD drawing to determine the precise location coordinates of each switch.

[0092] Shooting angle: According to the operation type (opening or closing) and the camera configuration of the drone, choose the appropriate shooting angle to ensure that the status of the knife switch can be clearly captured.

[0093] For each switch operation, specific drone control instructions are generated based on the analysis results and the preset instruction template. The instructions should include all necessary steps for the drone to take off, fly to the target location, take photos or videos, and return to the base.

[0094] S24, using the command transmission protocol to transmit the control command of the drone to the pre-set drone.

[0095] It should be noted that the command transmission protocol can adopt the MAVLink protocol, which is a lightweight protocol widely used in drone communications and supports multiple message types, including control commands, status updates, parameter requests, etc. The MAVLink protocol supports multiple transport layers (such as UDP, TCP, serial ports, etc.), and the appropriate transmission method can be selected according to the application scenario.

[0096] S3. The UAV autonomously flies to the target switch position according to the UAV control command, and collects the switch image through the camera carried by the UAV; based on the switch preview process image and the switch image, the switch recognition algorithm is used to identify the switch opening and closing state;

[0097] As a preferred implementation, the UAV autonomously flies to the target switch position according to the UAV control command, and collects the switch image through the camera carried by the UAV; based on the switch preview process image and the switch image, the switch recognition algorithm is used to identify the switch opening and closing state, including the following steps:

[0098] S31, the UAV flies to the target switch position according to the control command, and uses the camera to continuously collect images of the switch;

[0099] Specifically, when the drone receives the drone control command, it parses the flight parameters and plans the flight path using the path planning algorithm.

[0100] S32, using image processing technology to extract the characteristic points of the knife switch image and the knife switch action during the knife switch preview process, and to extract the three-dimensional characteristic parameters of the knife switch in the knife switch image and the knife switch preview process image;

[0101] As a preferred embodiment, the image processing technology is used to extract the characteristic points of the knife switch action in the knife switch image and the knife switch preview process image, and the three-dimensional characteristic parameters of the knife switch in the knife switch image and the knife switch preview process are extracted, including the following steps:

[0102] S321, graying, filtering and edge detection processing are performed on the knife switch image and the knife switch preview process image respectively;

[0103] It should be noted that grayscale conversion is to convert a color image into a grayscale image to simplify image information, reduce the amount of calculation, and retain sufficient structural information for subsequent processing. The specific steps include:

[0104] Read a color image (usually in RGB format).

[0105] Convert RGB images to grayscale images. Common grayscale conversion methods include weighted average method.

[0106] Filtering is to remove noise from images, improve image quality, and prepare for subsequent processing steps such as edge detection.

[0107] Edge detection is the identification of edge features in an image. These edges usually correspond to locations where the grayscale value in the image changes dramatically, and are very important for subsequent feature extraction and image recognition.

[0108] S322, extracting feature points from the knife gate image after edge detection and the knife gate preview process image respectively using a feature point detection algorithm;

[0109] Specifically, the feature point detection algorithm includes:

[0110] SIFT algorithm: It is a powerful feature point detection and description algorithm that can detect stable feature points under different scales, rotations and lighting conditions, and generate a 128-dimensional feature vector for each feature point. SIFT algorithm has good robustness to image transformation and noise, but the amount of calculation is relatively large.

[0111] ORB algorithm: It is a fast feature point detection and description algorithm that combines the FAST feature point detector and the BRIEF feature descriptor. The ORB algorithm has fast calculation speed and good real-time performance, but it may not be as stable as the SIFT algorithm in some complex scenarios.

[0112] Among them, through the feature point detection algorithm, such as SIFT algorithm or ORB algorithm, the key feature points in the knife switch image and the preview image are extracted. The detected feature points should include the significant points of the main parts of the knife switch (such as hinges, contacts, and arm rods), and the two-dimensional coordinates of these feature points are recorded.

[0113] S323, using a three-dimensional reconstruction algorithm to calculate the spatial coordinates of the feature points, and extracting the three-dimensional feature parameters of the knife switch in the knife switch image and the knife switch preview process image.

[0114] As a preferred embodiment, the spatial coordinates of the feature points are calculated using a three-dimensional reconstruction algorithm, and the three-dimensional feature parameters of the knife switch in the knife switch image and the knife switch preview process image are extracted, including the following steps:

[0115] S3231, using a three-dimensional reconstruction algorithm, mapping the coordinates of feature points of the knife switch at different positions in the knife switch image and the knife switch preview process image from the two-dimensional image to the three-dimensional space, constructing a spatial structure model of the knife switch, and calculating the three-dimensional coordinates of each feature point by using stereo vision technology;

[0116] It should be noted that for image pairs taken from different perspectives, a feature matching algorithm (such as FLANN matcher) is used to find corresponding feature point pairs. These point pairs should reflect the projection of the same point in three-dimensional space under different perspectives. Then, by calculating the difference in pixel coordinates (i.e., disparity) of the matching feature point pairs, the depth information of these points in three-dimensional space can be inferred.

[0117] Using the principles of stereo vision, matching pairs of feature points in at least two images taken from different viewpoints can be used to triangulate the locations of these points in 3D space. This usually involves solving a system of linear equations to solve for the 3D coordinates.

[0118] S3232, normalizing the three-dimensional coordinates of each feature point, and determining the spatial position of the key parts of the knife switch based on the normalized three-dimensional coordinates, and calculating the relative distance and orientation between the key parts of the knife switch;

[0119] It should be noted that normalization processing usually scales the data to a specific range, such as [0, 1] or [-1, 1]. The straight-line distance between two key parts is calculated using the Euclidean distance formula. For example, for key parts A and B, their normalized three-dimensional coordinates are (X A , Y A ,Z A ) and (X B , Y B ,Z B ), then the relative distance between them is:

[0120]

[0121] Where Dis represents the relative distance between key parts A and B.

[0122] The orientation can be determined by calculating the vector between two key parts. Vector AB can be expressed as (X B -XA , Y B -Y A , Z B -Z A ). You can then use this vector to calculate the azimuth (horizontal and elevation) or use other methods such as quaternions to represent rotation and orientation.

[0123] S3233. According to the movement characteristics of the knife switch, analyze the movement trajectory, speed and changes in the opening and closing angle of the characteristic point in the three-dimensional space.

[0124] It should be noted that the motion trajectory refers to the path of a feature point changing over time in three-dimensional space. By analyzing the position changes of feature points in consecutive frame images, the motion trajectory of the feature points can be tracked, which usually involves technologies such as image registration, feature point matching and trajectory smoothing.

[0125] Speed ​​refers to the distance that a feature point moves per unit time in three-dimensional space. The speed can be obtained by differentiating the motion trajectory, that is, calculating the ratio of the displacement of the feature point at different time points to the time.

[0126] The opening and closing angle refers to the angle change between the contacts or key components of the knife switch during the opening and closing process. The opening and closing angle can be calculated by measuring the relative position change of characteristic points (such as the contact tip) in three-dimensional space. It is usually necessary to establish a geometric model of the knife switch and use three-dimensional coordinate transformation technology to determine the angle change.

[0127] S33. Based on a similarity calculation method, the three-dimensional characteristic parameters of the knife switch in the knife switch image and the knife switch preview process image are compared, and the opening and closing states of the knife switch are determined according to the comparison results.

[0128] As a preferred implementation, based on a similarity calculation method, the three-dimensional characteristic parameters of the knife switch in the knife switch preview process image are compared, and the opening and closing states of the knife switch are judged according to the comparison results, including the following steps:

[0129] S331, converting the three-dimensional characteristic parameters of the knife switch in the knife switch image and the knife switch preview process image into a characteristic vector by using a characteristic vector conversion method;

[0130] It should be noted that for each three-dimensional feature parameter, a corresponding weight is assigned according to its nature and importance, and these weighted feature parameters are combined into a high-dimensional vector, namely, a feature vector. Each dimension of the feature vector corresponds to a feature parameter or its weighted value.

[0131] S332, using a similarity calculation method, calculating the similarity between the knife switch image and the feature vector of the knife switch in the knife switch preview process image;

[0132] It should be noted that a suitable similarity calculation method is selected, such as Euclidean distance, cosine similarity, Manhattan distance, etc., to calculate the similarity score between two feature vectors. The lower the score (for distance measurement) or the higher the score (for similarity measurement), the more similar the two feature vectors are.

[0133] S333. According to the similarity calculation result, determine whether the three-dimensional characteristic parameters of the knife switch in the knife switch image and the preview process image are consistent, and determine the opening and closing state of the knife switch according to the determination result.

[0134] It should be noted that a similarity threshold is pre-set according to the actual application scenario and accuracy requirements. If the calculated similarity score is higher (or lower, depending on the type of similarity calculation method) than the set threshold, it is considered that the three-dimensional feature parameters of the knife switch in the knife switch image and the preview process image are consistent, and the knife switch is in the expected open and closed state. Otherwise, it is considered that the knife switch state is inconsistent with the preview state, and further inspection or alarm is required. Finally, the judgment result is output, including the open and closed state of the knife switch, the similarity score, and any related warning or error information.

[0135] For example, assume that the three-dimensional feature parameters of the knife switch image and the knife switch preview process image, such as the coordinates and angles of key parts (hinge, contact, arm), etc., are assigned weights according to the importance of each feature parameter, such as the contact position has a higher weight.

[0136] Among them, the three-dimensional characteristic parameters of the knife switch image are: 1.0, 2.5, 0.8, ...; the three-dimensional characteristic parameters of the knife switch preview process image are: 1.2, 2.7, 0.9, ...;

[0137] Assume that the contact position weight w1 = 0.4, the arm length weight w2 = 0.3, and the other feature weights w3, w4, ... = 0.1.

[0138] The knife switch image feature vectors include: 0.4·1.0, 0.3·2.5, 0.1·0.8, ...;

[0139] The image characteristics of the knife switch preview process include: 0.4·1.2, 0.3·2.7, 0.1·0.9, ...;

[0140] The similarity threshold T is set to 0.9.

[0141] If the calculated similarity score is higher than the threshold T, it is considered that the three-dimensional characteristic parameters of the knife switch in the knife switch image and the preview process image are consistent, and the knife switch is in the expected open and close state. Otherwise, it is considered that the knife switch state is inconsistent with the preview state, and further inspection or alarm is required.

[0142] Then, the switch is in the closed state. The similarity score is 0.95, and there is no warning or error message.

[0143] S4. Real-time acquisition of remote signal data of the knife switch during the opening and closing operation, and the state evaluation of the operation state of the knife switch during the opening and closing operation is performed by integrating the set pair analysis method and the evidence theory method to obtain the state evaluation result;

[0144] As a preferred implementation, the remote signal data of the knife switch during the opening and closing operation is obtained in real time, and the operating state of the knife switch during the opening and closing operation is evaluated by integrating the set pair analysis method and the evidence theory method. The state evaluation result includes the following steps:

[0145] S41, establishing an evaluation index system for the operation status of the knife switch according to the remote signal data of the knife switch during the opening and closing operation; and dividing the operation status of the knife switch into different status levels by calculating the relative degradation degree of each index according to the evaluation index system for the operation status of the knife switch;

[0146] It should be noted that the remote signal data of the knife switch during the opening and closing operation reflects its operating status. By extracting and classifying these data, an evaluation index system is established. And because the dimensions of different indicators are different, the indicators need to be normalized;

[0147] According to the operating characteristics of the knife switch, the following key evaluation indicators are selected:

[0148] Operation time (unit: seconds): The time it takes to complete the opening and closing operation. If it is too long, it may indicate mechanical jam.

[0149] Contact position offset (unit: meter): The deviation of the knife switch contact position from the standard position.

[0150] Opening and closing angle deviation (unit: degree): The deviation between the actual movement angle of the knife gate lever arm and the designed angle.

[0151] Electrical parameters (such as contact resistance, unit: ohm): reflect the degree of contact quality of the contacts.

[0152] Environmental parameters (such as temperature, humidity): external conditions that affect the performance of the knife switch.

[0153] Among them, the relative degradation measures the degree of deviation of each evaluation indicator. The calculation formula of relative degradation is: P = |R norm -R ref | / R ref ; P represents the relative degradation degree, R norm Represents the normalized value of the evaluation index, R ref Indicates the threshold of the evaluation index. The degradation degree range is [0, 1], and the larger the degradation degree value, the worse the state.

[0154] According to the calculation results of the relative degradation degree, the operating status of the knife switch is divided into different levels, such as normal, attention, warning, severe, etc. The level classification of the operating status of the knife switch is shown in Table 1:

[0155] Table 1 Classification of the operating status of the knife switch

[0156]

[0157]

[0158] S42, using a statistical method to determine the weight of each evaluation index of the switch operation status, and using a set pair analysis method to calculate the degree of connection between each evaluation index and the status level;

[0159] It should be noted that the weights of various evaluation indicators are determined by using statistical methods (such as analytic hierarchy process, principal component analysis, entropy weight method, etc.) These methods can objectively and scientifically assign the weights of various indicators based on the statistical characteristics of the data.

[0160] For example, the hierarchical analysis method can determine the weights by constructing a hierarchical model and combining expert scoring and consistency testing; the entropy weight law can assign weights according to the information entropy of each indicator. The smaller the information entropy, the greater the amount of information provided by the indicator, and the weight should be increased accordingly.

[0161] Among them, set pair analysis is a systematic analysis method for dealing with uncertainty problems. It takes both certainty and uncertainty into account and comprehensively characterizes the connection between two sets by calculating the sameness, difference and opposition. Each evaluation indicator is constructed into a set pair with different status levels. For example, the "operation time" indicator is constructed into set pairs with the status levels of "normal", "attention", "warning", "serious", etc. According to the principle of set pair analysis, the connection degree of each set pair is calculated. The connection degree usually includes three components: the sameness, the difference and the opposition.

[0162] As a preferred implementation, the set pair analysis method is used to calculate the degree of connection between each evaluation indicator and the status level:

[0163]

[0164] In the formula, H t Represents the degree of connection between the evaluation index and the tth state level; N t represents the number of the lth evaluation index; w tl represents the weight of the lth evaluation indicator; a tl Indicates the degree of identity between the lth evaluation indicator and the nth level status level; b tl，n Indicates the difference between the lth evaluation indicator and the nth level state level; c tl It represents the degree of opposition between the lth evaluation index and the nth state level difference K-1 level; i represents the difference uncertainty coefficient; j represents the opposition coefficient.

[0165] S43, calculating the basic credibility distribution of the evidence theory according to the degree of connection between each evaluation indicator and the state level, and using the evidence synthesis rule to merge the basic credibility distribution of multiple evaluation indicators to obtain the comprehensive credibility distribution of the switch operation state;

[0166] As a preferred implementation, the basic credibility distribution of the evidence theory is calculated according to the degree of connection between each evaluation index and the state level, and the basic credibility distribution of multiple evaluation indicators is integrated by using the evidence synthesis rule to obtain the comprehensive credibility distribution of the knife switch operation state, which includes the following steps:

[0167] S431, determining the credibility coefficient of each evaluation indicator according to the weight of each evaluation indicator, and calculating the basic credibility distribution of the evidence theory according to the credibility coefficient of each evaluation indicator;

[0168] It should be noted that the credibility coefficient reflects the reliability of each evaluation indicator for state evaluation, which is determined by the indicator weight and the support rate of historical data. The credibility coefficient calculation formula is:

[0169] C tl =w tl / w max ;

[0170] C tl represents the credibility coefficient of the lth evaluation indicator, w tl represents the weight of the lth evaluation indicator; w max Represents the maximum value among the weights of the evaluation indicators.

[0171] For each evaluation indicator and status level, the connection degree and the credibility coefficient are multiplied to obtain the basic credibility distribution.

[0172] S432, selecting any two evaluation indicators and their corresponding arbitrary status levels as initial evidence, and preliminarily integrating the basic credibility distribution of the two evaluation indicators to obtain a new comprehensive credibility distribution;

[0173] S433: The comprehensive credibility distribution obtained by preliminary fusion is used as new evidence, and the basic credibility distribution of the next evaluation indicator is selected in turn and fused with the current comprehensive credibility distribution until the basic credibility distributions of all evaluation indicators are fused into the final comprehensive credibility distribution.

[0174] Specifically, through the aforementioned steps, a preliminary comprehensive credibility assignment is obtained, which will become the new evidence base. The next indicator is selected from the remaining evaluation indicators, which has its basic credibility assignment, and the new basic credibility assignment is fused with the current preliminary comprehensive credibility assignment using the Dempster-Shafer combination rule.

[0175] According to the Dempster-Shafer combination rule, the new comprehensive reliability distribution m new It can be calculated by the following formula:

[0176]

[0177] In the formula, m new represents the new comprehensive credibility distribution, Q represents the subset of the status level, u represents the conflict coefficient, which represents the degree of conflict between the evidences, and m current represents the current preliminary comprehensive credit assignment, m l represents the basic confidence distribution of the new evaluation indicator, and W and E represent other subsets of the status level.

[0178] S44. According to the integrated reliability distribution after fusion, combined with the maximum membership principle and the reliability criterion, the current opening and closing operation state level of the knife switch is judged to obtain the state evaluation result of the knife switch.

[0179] It should be noted that this is the result obtained by integrating the basic reliability distribution of all evaluation indicators through the evidence synthesis rule, which reflects the reliability distribution of the knife switch at each possible state level.

[0180] In fuzzy set theory, membership indicates the degree to which an element belongs to a fuzzy set. In state assessment, each state level can be regarded as a fuzzy set, and the current state of the knife switch is a fuzzy element. The maximum membership principle means that an element is most likely to belong to the fuzzy set with the largest membership. Find the state level with the largest membership (i.e., credibility value) from the comprehensive credibility distribution. Assume that the membership of this state level is Bel(Fc), where Fc represents the state level with the largest membership.

[0181] In order to ensure the reliability of the judgment result, a confidence threshold γ needs to be set. Only when the maximum membership exceeds this threshold can the judgment result be considered reliable.

[0182] Compare Bel(Fc) with the set confidence threshold γ.

[0183] If Bel(Fc)>γ, the knife switch is considered to be in state level Fc, and the state is output as the evaluation result. If Bel(Fc)≤γ, the current comprehensive confidence distribution is considered insufficient to make a reliable judgment, and further data collection or re-evaluation is required.

[0184] S5. Based on the switch opening and closing states identified by the switch recognition algorithm and the state evaluation results of the switch during the opening and closing operations, the state of the switch is bidirectionally verified, and a preset switch processing plan is executed according to the verification results.

[0185] Specifically, by comparing and verifying the status information from two different sources, the accuracy and reliability of the switch status judgment can be improved. Specifically, the following steps are included:

[0186] Compare the on / off state identified by the switch identification algorithm with the state assessment result.

[0187] If the two are consistent, the current state of the knife switch can be confirmed and the corresponding processing plan can be executed according to the state.

[0188] If the two are inconsistent, further analysis is required, which may include image recognition errors, inaccurate state assessment models, or data source issues.

[0189] Depending on the analysis results, the image recognition algorithm needs to be adjusted, or the status assessment model needs to be recalibrated, or even manual intervention for on-site inspection is required.

[0190] The treatment plan is usually pre-made according to the different states of the knife switch, for example:

[0191] If the knife switch is in the normal open or closed state, you only need to record the status information and continue monitoring.

[0192] If the knife switch is in an abnormal state, an alarm needs to be triggered immediately to notify the operation and maintenance personnel to conduct inspection and repair.

[0193] Once the accurate status of the knife switch is determined, the corresponding treatment plan should be triggered automatically or manually, and the execution process should be recorded and tracked.

[0194] like Figure 2 As shown, according to another embodiment of the present invention, a UAV and one-key sequential control joint operation system based on a knife switch recognition algorithm is provided, the system comprising: an operation ticket generation and verification module 1, an operation ticket recognition module 2, a knife switch state recognition module 3, a knife switch state evaluation module 4 and a knife switch state approval module 5;

[0195] The operation ticket generation and verification module 1 is used to use the substation one-key sequential control host to receive the one-key sequential control operation instruction, and generate an operation ticket by analyzing the one-key sequential control operation instruction; the intelligent anti-error host performs a preview of the anti-error verification according to the operation ticket. If the verification is successful, the intelligent anti-error host transmits the operation ticket and the knife switch preview process image to the substation master station, otherwise, the one-key sequential control operation instruction is verified and adjusted until the verification is successful;

[0196] The operation ticket recognition module 2 is used to use the text recognition method to recognize the switch operation content in the operation ticket after the substation master station receives the operation ticket, generate the UAV control command according to the switch operation content and transmit the UAV control command to the pre-set UAV;

[0197] The switch state recognition module 3 is used for the UAV to autonomously fly to the target switch position according to the UAV control command, and to collect the switch image through the camera carried by the UAV; based on the switch preview process image and the switch image, the switch recognition algorithm is used to identify the switch opening and closing state;

[0198] The switch state evaluation module 4 is used to obtain the remote signal data of the switch during the opening and closing operation in real time, and to evaluate the operation state of the switch during the opening and closing operation by integrating the set pair analysis method and the evidence theory method to obtain the state evaluation result;

[0199] The switch state verification module 5 is used to perform bidirectional verification on the switch state by combining the switch opening and closing state identified by the switch recognition algorithm and the switch state evaluation result during the switch opening and closing operation, and execute a preset switch processing plan according to the verification result.

[0200] In summary, with the help of the above technical solutions of the present invention, the present invention performs bidirectional verification of the switch state by combining the on / off state identified by the switch recognition algorithm and the state evaluation result based on the telesignaling data, which can effectively reduce the error that may be caused by a single recognition or evaluation method, improve the accuracy of state recognition, and realize the automation of switch state recognition through one-key sequential control, autonomous flight of drones and image recognition and other technologies, greatly improve the operation and maintenance efficiency, reduce the need for manual intervention, help improve the intelligence level of substations, and ensure the safe and stable operation of power systems. The present invention extracts the three-dimensional feature parameters of the switch through a three-dimensional reconstruction algorithm, which can more accurately reflect the actual state of the switch than traditional two-dimensional image recognition. The three-dimensional feature parameters are compared based on the similarity calculation method, which can more accurately judge the on / off state of the switch and reduce the misjudgment rate. The drone and image processing technology realize non-contact monitoring of the switch, avoiding the safety risks caused by direct contact between personnel and high-voltage equipment. The present invention establishes an evaluation index system for the operating status of the knife switch, comprehensively evaluates the status of the knife switch from multiple dimensions, avoids the one-sidedness of single index evaluation, combines set pair analysis and evidence theory, and fully utilizes the advantages of the two methods in dealing with uncertainty and multi-source information fusion, thereby improving the accuracy of the evaluation. The evaluation method based on remote signaling data and image processing technology realizes non-contact monitoring of the knife switch, reduces the safety risks of operation and maintenance personnel, and will help to improve the intelligence level of the power system and ensure the safe and stable operation of the power grid.

[0201] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, optical storage, etc.) containing computer-usable program code.

[0202] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. The combined operation method of UAV and one-button sequential control based on knife switch recognition algorithm is characterized in that: The method comprises the following steps: S1. Use the substation one-key sequential control host to receive the one-key sequential control operation instruction, and generate an operation ticket by analyzing the one-key sequential control operation instruction; the intelligent anti-error host performs a preview of the anti-error verification according to the operation ticket. If the verification is successful, the intelligent anti-error host transmits the operation ticket and the knife switch preview process image to the substation master station. Otherwise, verify and adjust the one-key sequential control operation instruction until the verification is successful; S2. When the substation master station receives the operation ticket, it uses the text recognition method to identify the switch operation content in the operation ticket, generates a UAV control instruction according to the switch operation content, and transmits the UAV control instruction to the pre-set UAV; S3. The UAV autonomously flies to the target switch position according to the UAV control command, and collects the switch image through the camera carried by the UAV; based on the switch preview process image and the switch image, the switch recognition algorithm is used to identify the switch opening and closing state; S4. Real-time acquisition of remote signal data of the knife switch during the opening and closing operation, and the state evaluation of the operation state of the knife switch during the opening and closing operation is performed by integrating the set pair analysis method and the evidence theory method to obtain the state evaluation result; S5. Based on the switch opening and closing states identified by the switch recognition algorithm and the state evaluation results of the switch during the opening and closing operations, the state of the switch is bidirectionally verified, and a preset switch processing plan is executed according to the verification results.

2. The combined operation method of UAV and one-button sequential control based on knife switch recognition algorithm according to claim 1 is characterized in that: The substation master station receives the operation ticket, identifies the switch operation content in the operation ticket by using a text recognition method, generates a UAV control instruction according to the switch operation content, and transmits the UAV control instruction to a preset UAV, including the following steps: S21, preprocessing the received operation ticket, wherein the preprocessing includes image denoising, grayscale and binarization processing; S22, using optical character recognition to perform text recognition on the pre-processed operation ticket, and extracting the knife switch operation content in the operation ticket, wherein the knife switch operation content includes the knife switch number, operation type and operation sequence; S23, parsing the knife switch operation content in the operation ticket, and generating a control instruction of the drone using a preset instruction template according to the parsing result, wherein the control instruction of the drone includes a flight altitude, a target knife switch position and a shooting angle; S24, using the command transmission protocol to transmit the control command of the drone to the pre-set drone.

3. The combined operation method of UAV and one-button sequential control based on knife switch recognition algorithm according to claim 1 is characterized in that: The UAV autonomously flies to the target switch position according to the UAV control command, and collects the switch image through the camera carried by the UAV; based on the switch preview process image and the switch image, the switch recognition algorithm is used to identify the switch opening and closing state, including the following steps: S31, the UAV flies to the target switch position according to the control command, and uses the camera to continuously collect images of the switch; S32, using image processing technology to extract the characteristic points of the knife switch image and the knife switch action during the knife switch preview process, and to extract the three-dimensional characteristic parameters of the knife switch in the knife switch image and the knife switch preview process image; S33. Based on a similarity calculation method, the three-dimensional characteristic parameters of the knife switch in the knife switch image and the knife switch preview process image are compared, and the opening and closing states of the knife switch are determined according to the comparison results.

4. The combined operation method of UAV and one-button sequential control based on knife switch recognition algorithm according to claim 3 is characterized in that: The method of using image processing technology to respectively extract the characteristic points of the knife switch action in the knife switch image and the knife switch preview process image, and to extract the three-dimensional characteristic parameters of the knife switch in the knife switch image and the knife switch preview process comprises the following steps: S321, graying, filtering and edge detection processing are performed on the knife switch image and the knife switch preview process image respectively; S322, extracting feature points from the knife gate image after edge detection and the knife gate preview process image respectively using a feature point detection algorithm; S323, using a three-dimensional reconstruction algorithm to calculate the spatial coordinates of the feature points, and extracting the three-dimensional feature parameters of the knife switch in the knife switch image and the knife switch preview process image.

5. The combined operation method of UAV and one-button sequential control based on knife switch recognition algorithm according to claim 4 is characterized in that: The method of calculating the spatial coordinates of the feature points by using a three-dimensional reconstruction algorithm and extracting the three-dimensional feature parameters of the knife switch in the knife switch image and the knife switch preview process image comprises the following steps: S3231, using a three-dimensional reconstruction algorithm, mapping the coordinates of feature points of the knife switch at different positions in the knife switch image and the knife switch preview process image from the two-dimensional image to the three-dimensional space, constructing a spatial structure model of the knife switch, and calculating the three-dimensional coordinates of each feature point by using stereo vision technology; S3232, normalizing the three-dimensional coordinates of each feature point, and determining the spatial position of the key parts of the knife switch based on the normalized three-dimensional coordinates, and calculating the relative distance and orientation between the key parts of the knife switch; S3233. According to the movement characteristics of the knife switch, analyze the movement trajectory, speed and changes in the opening and closing angle of the characteristic point in the three-dimensional space.

6. The combined operation method of UAV and one-button sequential control based on knife switch recognition algorithm according to claim 3 is characterized in that: The method of comparing the three-dimensional characteristic parameters of the knife switch in the knife switch preview process image with the knife switch image based on the similarity calculation method, and judging the opening and closing state of the knife switch according to the comparison result includes the following steps: S331, converting the three-dimensional characteristic parameters of the knife switch in the knife switch image and the knife switch preview process image into a characteristic vector by using a characteristic vector conversion method; S332, using a similarity calculation method, calculating the similarity between the knife switch image and the feature vector of the knife switch in the knife switch preview process image; S333. According to the similarity calculation result, determine whether the three-dimensional characteristic parameters of the knife switch in the knife switch image and the preview process image are consistent, and determine the opening and closing state of the knife switch according to the determination result.

7. The combined operation method of UAV and one-button sequential control based on knife switch recognition algorithm according to claim 1 is characterized in that: The real-time acquisition of remote signaling data of the knife switch during the opening and closing operation, and the state evaluation of the operation state of the knife switch during the opening and closing operation by integrating the set pair analysis method and the evidence theory method to obtain the state evaluation result include the following steps: S41, establishing an evaluation index system for the operation status of the knife switch according to the remote signal data of the knife switch during the opening and closing operation; and dividing the operation status of the knife switch into different status levels by calculating the relative degradation degree of each index according to the evaluation index system for the operation status of the knife switch; S42, using a statistical method to determine the weight of each evaluation index of the switch operation status, and using a set pair analysis method to calculate the degree of connection between each evaluation index and the status level; S43, calculating the basic credibility distribution of the evidence theory according to the degree of connection between each evaluation indicator and the state level, and using the evidence synthesis rule to merge the basic credibility distribution of multiple evaluation indicators to obtain the comprehensive credibility distribution of the switch operation state; S44. According to the integrated reliability distribution after fusion, combined with the maximum membership principle and the reliability criterion, the current opening and closing operation state level of the knife switch is judged to obtain the state evaluation result of the knife switch.

8. The combined operation method of UAV and one-button sequential control based on knife switch recognition algorithm according to claim 7 is characterized in that: The formula for calculating the degree of connection between each evaluation indicator and the status level using the set pair analysis method is: In the formula, H t Represents the degree of connection between the evaluation index and the t-th state level; N t represents the number of the lth evaluation indicator; w tl represents the weight of the lth evaluation indicator; a tl It indicates the identity between the lth evaluation index and the nth level status level; b tl，n It indicates the difference between the lth evaluation indicator and the nth state level; c tl It indicates the degree of opposition between the lth evaluation indicator and the nth state level difference K-1 level; i represents the uncertainty coefficient of difference; j represents the opposition coefficient.

9. The combined operation method of UAV and one-button sequential control based on knife switch recognition algorithm according to claim 7 is characterized in that: The method of calculating the basic credibility distribution of the evidence theory according to the degree of connection between each evaluation index and the state level, and fusing the basic credibility distribution of multiple evaluation indexes by using the evidence synthesis rule to obtain the comprehensive credibility distribution of the knife switch operation state includes the following steps: S431, determining the credibility coefficient of each evaluation indicator according to the weight of each evaluation indicator, and calculating the basic credibility distribution of the evidence theory according to the credibility coefficient of each evaluation indicator; S432, selecting any two evaluation indicators and their corresponding arbitrary status levels as initial evidence, and preliminarily integrating the basic credibility distribution of the two evaluation indicators to obtain a new comprehensive credibility distribution; S433: The comprehensive credibility distribution obtained by preliminary fusion is used as new evidence, and the basic credibility distribution of the next evaluation indicator is selected in turn and fused with the current comprehensive credibility distribution until the basic credibility distributions of all evaluation indicators are fused into the final comprehensive credibility distribution.

10. The UAV and one-button sequential control joint operation system based on the knife switch recognition algorithm is characterized by: Used to implement the UAV and one-button sequential control joint operation method based on the knife switch recognition algorithm described in any one of claims 1-9, the system includes: an operation ticket generation and verification module, an operation ticket recognition module, a knife switch state recognition module, a knife switch state evaluation module and a knife switch state approval module; The operation ticket generation and verification module is used to use the substation one-key sequential control host to receive the one-key sequential control operation instruction, and generate an operation ticket by analyzing the one-key sequential control operation instruction; the intelligent anti-error host performs a preview of the anti-error verification according to the operation ticket. If the verification is successful, the intelligent anti-error host transmits the operation ticket and the knife switch preview process image to the substation master station, otherwise, the one-key sequential control operation instruction is verified and adjusted until the verification is successful; The operation ticket recognition module is used to use a text recognition method to recognize the switch operation content in the operation ticket after the substation master station receives the operation ticket, generate a UAV control instruction according to the switch operation content, and transmit the UAV control instruction to a pre-set UAV; The switch state recognition module is used for the UAV to autonomously fly to the target switch position according to the UAV control command, and to collect the switch image through the camera carried by the UAV; based on the switch preview process image and the switch image, the switch recognition algorithm is used to identify the switch open and close state; The switch state evaluation module is used to obtain the remote signal data of the switch during the opening and closing operation in real time, and to evaluate the operation state of the switch during the opening and closing operation by integrating the set pair analysis method and the evidence theory method to obtain the state evaluation result; The switch state verification module is used to perform bidirectional verification on the switch state by combining the switch open / close state identified by the switch recognition algorithm and the switch state evaluation result during the switch open / close operation, and execute a preset switch processing plan according to the verification result.

Images