Wind power plant fault determination method and system based on a rail-mounted robot

Abstract

The application discloses a wind power plant fault judgment method and system based on a hanging rail robot, which comprises the following steps: judging the fault occurrence position of the wind power plant according to the current power distribution map, ensuring the accuracy of the fault occurrence position of the wind power plant, further determining a plurality of fault parameters according to the reexamination of the fault occurrence position by the hanging rail robot, determining a fault event according to the plurality of fault parameters, the reexamination position of the plurality of fault parameters relative to the fault occurrence position and the service life of the wind power plant, determining the corresponding repair operation track according to the fault event, the judgment module of the hanging rail robot and the repair tool possessed by the hanging rail robot, and ensuring the accuracy of the repair operation track.

Description

Technical Field

[0001] This invention relates to the technical field of wind power plants, and in particular to a method and system for fault diagnosis in wind power plants based on a rail-mounted robot. Background Technology

[0002] With the development of technology, wind power plants are gradually being applied to daily life and generate electricity through wind. In a wind power plant, there are multiple conductors, all of which are used for conducting electricity and forming conductor paths. In the current technology, the current power distribution map is introduced, and the corresponding fault location is output based solely on the judgment of the current power distribution map. However, the track-mounted robot is not fully utilized to re-inspect the fault location, and it is impossible to achieve accurate control of fault events. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art. This invention provides a method and system for fault diagnosis in wind power plants based on a rail-mounted robot.

[0004] This invention provides a method for fault diagnosis in wind power plants based on a rail-mounted robot, comprising: determining the conductor path of the wind power plant based on a line model of the wind power plant; determining the corresponding rail-mounted robot based on the conductor path and conductor diameter; collecting multiple current parameters based on the dynamic detection of the conductor by the rail-mounted robot, and constructing a current power distribution map based on the multiple current parameters and the detection position of the conductor; determining the location of the fault in the wind power plant based on the current power distribution map; determining multiple fault parameters based on the re-inspection of the fault location by the rail-mounted robot, determining a fault event based on the multiple fault parameters, the re-inspection position of the multiple fault parameters relative to the fault location, and the service life of the wind power plant; and determining the corresponding repair operation trajectory based on the fault event, the judgment module of the rail-mounted robot, and the repair tools possessed by the rail-mounted robot.

[0005] This invention provides a wind power plant fault diagnosis system based on a rail-mounted robot. The system applies the aforementioned wind power plant fault diagnosis method based on a rail-mounted robot. The system includes:

[0006] The conductor path module is used to determine the conductor path of a wind farm based on the line model of the wind farm.

[0007] The robot module is used to determine the corresponding rail-mounted robot based on the conductor path and conductor diameter of the wind power plant.

[0008] The power distribution map module is used to collect multiple current parameters based on the dynamic detection of the conductor by the rail-mounted robot, and to construct the current power distribution map based on the multiple current parameters and the detection position of the conductor.

[0009] The fault location module is used to determine the location of a fault in a wind power plant based on the current power distribution map.

[0010] The fault event module is used to determine multiple fault parameters based on the re-inspection of the fault location by the rail-mounted robot, and to determine the fault event based on the multiple fault parameters, the re-inspection location of the multiple fault parameters relative to the fault location, and the service life of the wind power plant.

[0011] The repair module is used to determine the corresponding repair operation trajectory based on the fault event, the judgment module of the rail-mounted robot, and the repair tools possessed by the rail-mounted robot.

[0012] Compared with the prior art, the beneficial effects of the present invention are:

[0013] In this embodiment of the invention, the method is used to determine the conductor path of the wind power plant based on the line model of the wind power plant; determine the corresponding rail-mounting robot based on the conductor path and conductor diameter; collect multiple current parameters based on the dynamic detection of the conductor by the rail-mounting robot; construct the current power distribution map based on the multiple current parameters and the detection position of the conductor; and determine the location of the fault in the wind power plant based on the current power distribution map, thus ensuring the accuracy of the location of the fault in the wind power plant.

[0014] Furthermore, multiple fault parameters are determined based on the re-inspection of the fault location by the rail-mounted robot. The fault event is determined based on the multiple fault parameters, the re-inspection location of the multiple fault parameters relative to the fault location, and the service life of the wind power plant. Multiple interactions of multiple fault parameters, the re-inspection location of the multiple fault parameters relative to the fault location, and the service life of the wind power plant are introduced to achieve precise control of the fault event.

[0015] Therefore, by determining the corresponding repair operation trajectory based on the fault event, the judgment module of the rail-mounted robot, and the repair tools possessed by the rail-mounted robot, the overall consideration of the fault event, the judgment module of the rail-mounted robot, and the repair tools possessed by the rail-mounted robot is realized, ensuring the accuracy of the repair operation trajectory. Attached Figure Description

[0016] Figure 1 This is a schematic diagram illustrating the application scenario of the wind power plant fault diagnosis method based on a rail-mounted robot in this invention.

[0017] Figure 2This is a flowchart illustrating the fault diagnosis method for wind power plants based on a rail-mounted robot in this invention.

[0018] Figure 3 This is a schematic diagram of the structural composition of the wind power plant fault diagnosis system based on a rail-mounted robot in this invention. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0020] Example 1

[0021] This application provides a fault diagnosis method for wind power plants based on a rail-mounted robot, applicable to applications such as... Figure 1 In the application environment shown, computer 102 communicates with server 104 via a network. Terminal 102 is not limited to various personal computers, servers, or wind power plants, and server 104 is implemented using a standalone server or a server cluster.

[0022] Please see Figures 1 to 3 The fault diagnosis method for wind power plants based on a rail-mounted robot includes:

[0023] Step S11: Determine the conductor path of the wind power plant based on the line model of the wind power plant;

[0024] Step S12: Determine the corresponding rail-mounted robot based on the conductor path and conductor diameter of the wind power plant.

[0025] Step S13: Collect multiple current parameters based on the dynamic detection of the conductor by the rail-mounted robot, and construct the current power distribution map based on the multiple current parameters and the detection position of the conductor;

[0026] Step S14: Determine the location of the wind power plant failure based on the current power distribution map;

[0027] Step S15: Determine multiple fault parameters based on the re-inspection of the fault location by the rail-mounted robot, and determine the fault event based on the multiple fault parameters, the re-inspection location of the multiple fault parameters relative to the fault location, and the service life of the wind power plant.

[0028] Step S16: Determine the corresponding repair operation trajectory based on the fault event, the judgment module of the rail-mounted robot, and the repair tools available to the rail-mounted robot.

[0029] Specifically, in step S11, the conductor path of the wind power plant is determined based on the line model of the wind power plant. The specific steps are as follows:

[0030] S111: Collect the line model of the wind power plant;

[0031] S112: Multiple functional areas are determined based on the division of the line model of the wind power plant;

[0032] S113: Multiple real-time images are collected based on the circumferential aerial survey of the wind power plant by the UAV. The multiple real-time images show the real-time status of the wind power plant at different locations.

[0033] S114: Determine multiple sub-trajectory paths based on the matching of multiple functional areas and multiple real-time images;

[0034] S115: Determine the conductor path of the wind power plant based on multiple sub-conductor paths, multiple functional areas, and the line model of the wind power plant.

[0035] At this point, a line model of the wind power plant is collected, introducing a pre-set line model. This data is usually stored in the wind power plant's archives or engineering department. Based on the division of the wind power plant's line model, multiple functional areas are determined. At the same time, after obtaining the line model, the functions of the wind power plant are analyzed. This includes identifying key functional areas such as power generation, transmission, distribution, and control. Based on the results of the functional analysis, the wind power plant is divided into multiple functional areas. These areas are divided according to the type, layout, and operating logic of the equipment. Each functional area is marked and recorded in the line model for easy reference in subsequent steps.

[0036] Furthermore, multiple real-time images are collected based on the circumferential aerial survey of the wind power plant by UAVs, and these images respectively show the real-time status of the wind power plant at different locations. Multiple sub-traverse paths are determined by matching multiple functional areas and multiple real-time images. Multiple functional areas and multiple real-time images are introduced and matched to ensure the accuracy of multiple sub-traverse paths.

[0037] At this point, a drone equipped with a high-resolution camera and a stable flight control system is selected to ensure that clear and stable images are captured during flight. A detailed flight plan is formulated based on the layout and scale of the wind power plant, including flight altitude, speed, and route, to ensure that the drone fully covers all functional areas of the wind power plant. During the drone's flight, real-time images of the wind power plant are collected through its onboard camera. These real-time images clearly show the real-time status of the wind power plant at different locations, including equipment operating status, wiring layout, and environmental conditions. The collected images undergo preliminary processing, such as noise reduction and contrast enhancement, and are then stored in a designated data storage device for subsequent analysis and use.

[0038] Based on the aforementioned functional areas (such as power generation area, transmission area, and distribution area), the location and extent of these areas are identified in the real-time images. The acquired real-time images are matched and aligned with the line model of the wind power plant to ensure that the equipment in the images corresponds one-to-one with the equipment in the model. Based on the matched images and model, the sub-conductor paths within each functional area are determined. These sub-conductor paths clearly show the line layout and connection relationships within each functional area.

[0039] Therefore, the conductor path of the wind power plant is determined based on multiple sub-conductor paths, multiple functional areas, and the line model of the wind power plant. Multiple interactions are carried out on the multiple sub-conductor paths, multiple functional areas, and the line model of the wind power plant to ensure the accuracy of the conductor path of the wind power plant.

[0040] At this point, the sub-conductor paths within each functional area are integrated to form a complete power plant conductor path diagram. The integrated conductor path diagram is then compared and updated with the original power plant line model to ensure that the line layout in the model is consistent with the actual situation. The updated line model is then verified and checked to ensure its accuracy and reliability. This process is completed in conjunction with on-site inspections of the actual power plant.

[0041] In another embodiment of this application, the conductor path matching table is a method for matching multiple sub-conductor paths, functional areas, and elements in the line model. The following is a conductor path matching table (Table 1):

[0042] Table 1. Conductor Path Matching Table

[0043] Serial Number Sub-wire path Functional Area Line model elements Matching results 1 ABC Power generation area Generator 1-Transformer 1-Line 1 match 2 DEF Transmission area Line 2 - Switching Station - Line 3 match 3 GH power distribution area Line 4 - Distribution Panel Not matched (needs adjustment)

[0044] Table 1 lists multiple sub-conductor paths and their corresponding functional areas and elements in the line model. By comparison and analysis, the matching relationship between the sub-conductor paths and the elements in the line model is determined, and then the final conductor path is determined. For example, in this embodiment, ABCD (i.e., power generation area - transmission area - distribution area - external power grid) can be used as the final conductor path.

[0045] Furthermore, in step S12, the corresponding rail-mounted robot is determined based on the wind power plant's conductor path and conductor diameter. The specific steps are as follows:

[0046] S121: Obtain the conductor path of the wind power plant;

[0047] S122: Collect the distribution location of each conductor based on the traversal of the conductor path of the wind power plant;

[0048] S123: Determine the distribution style of the conductors based on their location, their parallel arrangement, and their surrounding environment.

[0049] S124: Collect multiple wire diameters based on the detection of each wire, and determine the wire diameter of the wire according to the multiple wire diameters and the wire model information;

[0050] S125: Correlate the conductor paths, conductor distribution style, and conductor diameter of wind power plants;

[0051] S126: Determine the first robot parameters based on the conductor path and conductor distribution style of the wind power plant, and determine the second robot parameters based on the conductor path and conductor diameter of the wind power plant.

[0052] S127: Determine the corresponding track-mounted robot based on the first robot parameters, the second robot parameters, and previous robot data from the wind power plant.

[0053] In this embodiment, the conductor paths of the wind power plant are obtained; the distribution positions of each conductor are collected based on the traversal of the conductor paths of the wind power plant; the distribution style of the conductors is determined according to the distribution positions of each conductor, the parallel state of each conductor, and the surrounding environment of each conductor. This allows for a holistic consideration of the distribution positions of each conductor, the parallel state of each conductor, and the surrounding environment of each conductor, achieving multi-dimensional control of the distribution positions of each conductor, the parallel state of each conductor, and the surrounding environment of each conductor, and ensuring the accuracy of the conductor distribution style.

[0054] At this point, the conductor paths of the wind power plant are obtained and further controlled. All conductor paths within the wind power plant are traversed, including the collection lines from the substation to each wind turbine, as well as the grid connection lines. During the traversal, key information such as the direction, length, and height of each line is recorded in detail. The collected conductor distribution data is organized and stored in a dedicated database for subsequent analysis and use.

[0055] Based on the collected data on conductor distribution locations, the overall layout and direction of the conductors are analyzed. This helps to understand the spatial distribution characteristics of the conductors and provides a basis for determining the conductor distribution style. The parallel arrangement of conductors is considered, including the distance and angle between them. The parallel arrangement will affect the electromagnetic interference and heat dissipation of the conductors, and this will be taken into account when determining the conductor distribution style. The natural environment (such as terrain, landform, climate, etc.) and man-made environment (such as buildings, roads, farmland, etc.) around the conductors are evaluated. These environmental factors will affect the safe operation and maintenance costs of the conductors, and therefore will be fully considered when determining the conductor distribution style. The final conductor distribution style includes the direction, arrangement, and height setting of the conductors. When determining the distribution style, the principles of safety, economy, and aesthetics are followed to ensure the normal operation and maintenance of the conductors.

[0056] Furthermore, multiple wire diameters are collected based on the detection of each wire, and the wire diameter is determined according to the multiple wire diameters and the wire model information, realizing the interaction of multiple wire diameters and wire model information, thereby enabling precise control of the wire diameter.

[0057] Therefore, the conductor paths, conductor distribution patterns, and conductor diameters of the wind power plant are correlated; the first robot parameters are determined based on the conductor paths and conductor distribution patterns of the wind power plant, and the second robot parameters are determined based on the conductor paths and conductor diameters of the wind power plant; the corresponding track-mounting robot is determined based on the first robot parameters, the second robot parameters, and the previous robot data of the wind power plant. This overall consideration of the first robot parameters, the second robot parameters, and the previous robot data of the wind power plant ensures the matching accuracy of the track-mounting robot.

[0058] At this point, the conductor path, conductor distribution style, and conductor diameter data collected in the previous steps are integrated to ensure that each conductor has corresponding path information, distribution style description, and diameter data. Furthermore, the data is integrated and associated through an association model, which can be in the form of a database, where each conductor is treated as a record, containing its path coordinates, distribution style classification (such as dense, sparse, etc.), and diameter value.

[0059] The first robot parameters are mainly based on the guide path and guide distribution style. For example, if the guide is distributed along complex terrain, the robot needs higher mobility and obstacle avoidance capabilities. If the guide is densely arranged, the robot needs more accurate navigation and positioning capabilities.

[0060] Regarding the second robot parameters, these are primarily based on the wire path and wire diameter. For example, wires with larger diameters require stronger clamping forces or larger contact areas to ensure safe operation. The complexity and length of the wire path also affect the robot's power requirements and runtime.

[0061] Meanwhile, when determining the parameters, the robot's physical limitations (such as size and weight) and operational requirements (such as speed and accuracy) are also considered to ensure that the selected parameters meet the actual needs and are within the robot's capabilities. Optionally, the technical specifications for the first robot parameters and the second robot parameters, the first robot parameter matching table (Table 2), and the second robot parameter matching table (Table 3) are as follows:

[0062] Table 2 Robot Parameter Matching Table

[0063]

[0064]

[0065] Table 3 Robot Parameter Matching Table

[0066] Conductor path characteristics conductor diameter range Robot parameters Linear, barrier-free Small (<2mm) Lightweight gripper for precise control Twisted shape, with a few obstacles Medium (2-5mm) Medium strength gripper with good flexibility Complex terrain, multiple obstacles Large (>5mm) Heavy-duty gripper, high clamping force, wear-resistant design Any path, long distance No limit Long battery life, automatic charging system, fault warning

[0067] Furthermore, a database of wind power plants is acquired, and historical robot data from wind power plants is collected by traversing this database. This historical robot data includes information such as robot performance, maintenance records, and failure rates. This information helps to assess the potential of different robot models in meeting current needs. By combining the first robot parameters, the second robot parameters, and historical data, available rail-mounted robot models on the market are evaluated and compared, and the most suitable robot model for current needs is selected.

[0068] Specifically, suppose there are several conductors in a wind power plant, each of which has been assigned a unique identifier. Now, the path coordinates (such as GPS coordinates), distribution style (such as "distributed along the ridge", "densely arranged", etc.) and wire diameter data (such as 2.5mm, 4.0mm, etc.) of each conductor are integrated into a database. Each conductor is recorded in the database, and all its related information is linked together.

[0069] Assuming a set of wires is distributed along a steep hillside with a large wire diameter, based on this information, the following first robot parameters are determined for the rail-mounted robot: a highly mobile chassis to adapt to complex terrain, and an advanced obstacle avoidance system to ensure safe operation. Meanwhile, due to the large wire diameter, the following second robot parameters are determined: an enhanced gripping mechanism to provide sufficient gripping force, and a large-capacity battery to support long-term continuous operation.

[0070] Based on the previously determined first and second robot parameters, as well as the robot data used in the past, multiple rail-mounted robot models were evaluated and compared, and the final robot model was determined. After the model was selected, on-site testing was carried out to ensure that the selected robot could operate effectively in the wind power plant environment.

[0071] In step S13, multiple current parameters are collected based on the dynamic detection of the conductor by the rail-mounted robot. A current power distribution map is constructed based on these current parameters and the detected position of the conductor. The specific steps are as follows:

[0072] S131: Interact with the wind power plant's conductor path and the location of the track-mounted robot;

[0073] S132: Determine the inspection route of the rail-mounted robot based on the interaction between the conductor path of the wind power plant and the location of the rail-mounted robot.

[0074] S133: The rail-mounted robot dynamically travels along the inspection route on the guide wire and triggers the rail-mounted robot to dynamically detect the guide wire;

[0075] S134: Multiple current parameters are collected based on the dynamic detection of the conductor by the rail-mounted robot;

[0076] S135: Multiple current parameters are matched to the detection position of each conductor;

[0077] S136: Construct the current power distribution map based on multiple current parameters and the detection location of the conductor.

[0078] This involves interaction between the wind power plant's conductor path and the location of the rail-mounted robot; the inspection route of the rail-mounted robot is determined based on this interaction, thus ensuring the accuracy of the robot's inspection route.

[0079] Furthermore, the location of the wind power plant's conductor path and the location of the track-mounting robot are obtained. In S131, the conductor path of the wind power plant is usually obtained through CAD drawings, GIS systems, or on-site surveys. The conductor path information includes parameters such as the start point, end point, turning point, height, and type of the conductor. The track-mounting robot is usually equipped with a GPS positioning module or other positioning technologies (such as RFID, visual positioning, etc.) to obtain its own location information on the conductor path in real time.

[0080] Furthermore, based on the characteristics of the wind power plant's conductor path and the performance parameters of the rail-mounted robot, and through further interaction between these two parameters, one or more inspection routes can be generated based on the conductor path information and the real-time position information of the rail-mounted robot. The inspection route includes information such as the robot's starting position, inspection point (i.e., the position of the conductor being inspected), and ending position.

[0081] Furthermore, S133 and S134 ensure the dynamic acquisition of multiple current parameters and the corresponding acquisition accuracy. The rail-mounting robot is equipped with sensors and detection equipment to monitor the status of the conductor in real time. When the rail-mounting robot travels to the preset inspection point, it automatically triggers the corresponding detection equipment to perform real-time data transmission inspection of the conductor.

[0082] Imagine a large wind power plant where a track-mounted robot travels along a pre-set inspection route on a power line. When the robot reaches an inspection point on a critical power line, it automatically triggers devices such as an infrared thermal imager and a current sensor. The infrared thermal imager monitors the temperature distribution of the power line to detect potential overheating problems, while the current sensor monitors the current in the power line in real time to determine if there are any abnormalities such as overload or short circuit. The robot transmits the detection data to the ground monitoring center in real time for operators to analyze and process.

[0083] For example, when the rail-mounted robot performs dynamic detection on a critical conductor, it can collect multiple current parameters. After processing and analysis by the ground monitoring center, if it finds that the current of the conductor is fluctuating abnormally and the harmonic content exceeds the preset threshold, the system will issue an early warning signal to prompt the maintenance personnel to conduct further inspection and treatment of the conductor.

[0084] Therefore, multiple current parameters are matched with the detection positions of each conductor; the current power distribution map is constructed based on multiple current parameters and the detection positions of the conductors, which allows for a holistic consideration of multiple current parameters and the detection positions of the conductors, thus ensuring the accuracy of the current power distribution map construction.

[0085] At this time, a preset data matching mechanism is used to ensure that the collected current parameters can be accurately matched with the corresponding detection location information. The detection location information includes timestamps, location markers, wire numbers, and other types. After the data matching is completed, the system also verifies the data to ensure the accuracy and completeness of the matching.

[0086] Meanwhile, based on the collected current parameters and conductor detection location information, the system provides an intuitive way to display the power distribution map. The power distribution map includes the distribution of various parameters such as conductor current magnitude, phase angle, and harmonic content, as well as the connection relationship between conductors and the direction of power flow. In addition, to achieve the visualization of the power distribution map, the system adopts advanced data visualization technology, including two-dimensional or three-dimensional graphic display, dynamic data refresh, color coding, and other methods, so that operators can intuitively understand the operating status of the power system.

[0087] In step S14, the location of the wind power plant fault is determined based on the current power distribution map. The specific steps are as follows:

[0088] S141: Obtain the current power distribution map;

[0089] S142: Determine abnormal power ranges based on autonomous detection of the current power distribution map;

[0090] S143: Collect multiple operating parameters of the wind power plant, and determine the current power generation mode of the wind power plant based on the multiple operating parameters of the wind power plant and the power generation mode recorded by the wind power plant.

[0091] S144: Determine the location of wind power plant losses based on the current power distribution map and the loss markers of the wind power plant;

[0092] S145: Abnormal range of associated power, current power generation mode of wind power plant, and location of wind power plant loss;

[0093] S146: Determine the first fault range based on the abnormal power range and the current power generation mode of the wind power plant; determine the second fault range based on the abnormal power range and the location of the loss at the wind power plant; and determine the location of the fault at the wind power plant based on the first fault range, the second fault range, and the line model of the wind power plant.

[0094] Therefore, by acquiring the current power distribution map and determining the abnormal power range based on the autonomous detection of the current power distribution map, the autonomous detection of the current power distribution map is realized, ensuring the screening of abnormal power ranges, so as to facilitate subsequent management and control of abnormal power ranges.

[0095] The current power distribution map is input into a preset anomaly detection model. By calculating the anomaly score or probability of each data point, potential anomaly intervals are identified. Based on the output of the anomaly detection algorithm and combined with preset thresholds or rules, the anomaly intervals in the power distribution map are determined. These anomaly intervals represent areas where parameters such as current and voltage deviate significantly from the normal range. Optionally, the determined anomaly intervals are verified by comparing them with historical data, conducting on-site inspections, or making expert judgments.

[0096] Furthermore, in S143, by introducing multiple operating parameters of the wind power plant and the power generation mode recorded by the wind power plant, the interaction between the multiple operating parameters of the wind power plant and the power generation mode recorded by the wind power plant is realized, thereby ensuring the accuracy of the current power generation mode of the wind power plant.

[0097] At this time, key operating parameters of the wind power plant are acquired in real time through sensors, monitoring systems or other data acquisition methods, including but not limited to wind speed, wind direction, generator speed, output power, gearbox temperature, generator winding temperature, etc., so as to reflect the operating status and performance of the wind turbine through the key operating parameters.

[0098] Furthermore, based on the design and operation records of wind power plants, the power generation modes of generators under different wind speed conditions can be understood, such as rated power output mode, low wind speed operation mode, and variable speed constant frequency mode. The collected operating parameters are compared with known power generation modes to analyze whether the current operating parameters conform to a certain specific power generation mode. If the operating parameters have the highest matching degree with a certain power generation mode, it is considered that the current generator is operating in that mode.

[0099] Furthermore, based on the current power distribution map and the loss markers of the wind power plant, the location of the wind power plant loss is determined, so as to conduct overall management and control of the current power distribution map and the loss markers of the wind power plant.

[0100] At this point, by examining the power distribution map, we can understand the parameters such as current, voltage, and power factor of each conductor in the power system, as well as any abnormal changes in these parameters. Based on these abnormal changes, we can determine the losses or faults in the power transmission process. We can also identify loss or fault markers within the wind power plant on the power distribution map, such as resistance losses, inductance losses, and transformer faults. These loss or fault markers are usually marked based on the power plant's operating data and historical experience, reflecting the loss or fault situation in various parts of the power plant. Thus, by combining the power distribution map and the loss or fault markers, we can analyze the specific location of the loss or fault. If the current in a certain area increases abnormally or the voltage decreases abnormally, and there are loss or fault markers in that area, then that area is identified as the location of the loss or fault.

[0101] Suppose that an abnormally high current is observed on a power distribution map of a transmission line from a wind turbine to a substation. At the same time, a large resistance loss marker is found near this transmission line in the power plant's internal loss or fault markers. Combining these two pieces of information, it is analyzed that the increased loss in the transmission line is due to excessive resistance or poor contact. Therefore, this transmission line is identified as the location of the loss, and it is recommended to repair or replace it to reduce the loss.

[0102] Therefore, the system considers the abnormal power range, the current power generation mode of the wind power plant, and the location of the wind power plant's losses. Based on the abnormal power range and the current power generation mode of the wind power plant, the system determines the first fault range. Based on the abnormal power range and the location of the wind power plant's losses, the system determines the second fault range. Based on the first fault range, the second fault range, and the line model of the wind power plant, the system determines the location of the wind power plant's fault. This comprehensive approach considers the first fault range, the second fault range, and the line model of the wind power plant, achieving multi-dimensional judgment of the location of the wind power plant's fault and improving the accuracy of the fault location determination.

[0103] Furthermore, based on the abnormal power range and the current power generation mode of the wind power plant, the first fault range is determined. At this time, the relationship between the abnormal power range and the current power generation mode is analyzed to determine the cause or component that caused the abnormality. Based on the cause or component, the preliminary fault range, i.e. the first fault range, is delineated in the power system.

[0104] Furthermore, the second fault range is determined based on the abnormal power range and the location of the wind power plant loss. At this point, the fault range is further narrowed by combining the information of the abnormal power range and the location of the loss. If the location of the loss and the abnormal power range are spatially close or temporally related, these areas are included in the second fault range.

[0105] Furthermore, the line model of the wind power plant is used to cross-validate the first and second fault ranges. The line model typically includes information such as the physical layout of the power system, component parameters, and operating characteristics to determine the location of the fault.

[0106] The first fault range includes: transmission lines and their nearby components where the transmission line current increases abnormally under rated power output mode;

[0107] The second fault range includes: lines and their connection points where the loss and abnormality coincide in time;

[0108] The location of the fault includes the connection point of the transmission line where an abnormal increase in current occurs and where there is resistive loss.

[0109] In step S15, multiple fault parameters are determined based on the re-inspection of the fault location by the rail-mounted robot. The fault event is then determined based on these multiple fault parameters, the re-inspection location of the multiple fault parameters relative to the fault location, and the service life of the wind power plant. The specific steps are as follows:

[0110] S151: Obtain the location of the fault in the wind power plant;

[0111] S152: The movement of the rail-mounted robot is triggered by the location of the fault in the wind power plant, and the rail-mounted robot re-inspects the location of the fault.

[0112] S153: Based on the re-inspection of the fault location by the rail-mounted robot, multiple fault parameters are determined. At this time, the corresponding re-inspection mode is determined according to the actual image of the fault location, the re-inspection range of the fault location by the rail-mounted robot, and the importance coefficient of the fault location relative to the wind power plant.

[0113] S154: Obtain multiple fault parameters and sort them according to their size to form a fault parameter set;

[0114] S155: Associated fault parameter set, the location of multiple fault parameters relative to the fault location, and the service life of the wind power plant.

[0115] S156: Perform multiple interactions on the fault parameter set, the re-inspection location of multiple fault parameters relative to the fault location, and the service life of the wind power plant, and determine the fault event based on the multiple interactions on the fault parameter set, the re-inspection location of multiple fault parameters relative to the fault location, and the service life of the wind power plant.

[0116] Therefore, once the location of the fault in the wind power plant is determined, the system will immediately dispatch a rail-mounted robot to that location. During the robot's movement, multiple fault parameters are further determined based on the rail-mounted robot's re-inspection of the fault location. At this point, a corresponding re-inspection mode is determined based on the actual image of the fault location, the re-inspection range of the rail-mounted robot, and the importance coefficient of the fault location relative to the wind power plant. The rail-mounted robot then conducts a targeted re-inspection of the fault location along this re-inspection mode.

[0117] The fault parameters include wear level, temperature anomaly, vibration level, electrical performance, etc. After determining the fault parameters, the system will determine the corresponding inspection mode based on the actual image of the fault location, the inspection range, and the importance coefficient of the location relative to the wind power plant. The inspection modes include detailed inspection, real-time monitoring, and periodic inspection. Once the inspection mode is determined, the rail-mounted robot will conduct a comprehensive physical and electrical inspection of the fault location according to the inspection mode.

[0118] Therefore, multiple fault parameters are obtained and sorted by size to form a fault parameter set; the fault parameter set, the re-inspection location of multiple fault parameters relative to the fault occurrence location, and the service life of the wind power plant are associated to achieve multiple interactions between the fault parameter set, the re-inspection location of multiple fault parameters relative to the fault occurrence location, and the service life of the wind power plant.

[0119] At this point, the rail-mounted robot or other testing equipment has completed a detailed review of the fault location and collected multiple fault-related parameters, including the degree of physical wear, abnormal temperature values, vibration amplitude, and degree of electrical performance degradation. Each fault parameter collected is evaluated to determine its severity and impact on the operation of the wind power plant.

[0120] Furthermore, in this embodiment, the fault parameter set is also associated with the re-inspection location of the fault occurrence, and the fault parameter set is also associated with the service life of the wind power plant, in order to assess whether the fault is related to factors such as equipment aging and insufficient maintenance, thereby formulating a more effective maintenance and repair plan.

[0121] Furthermore, multiple interactions are performed on the fault parameter set, the re-inspection location of multiple fault parameters relative to the fault location, and the service life of the wind power plant. The fault event is determined based on the multiple interactions of the fault parameter set, the re-inspection location of multiple fault parameters relative to the fault location, and the service life of the wind power plant. This comprehensive consideration of the fault parameter set, the re-inspection location of multiple fault parameters relative to the fault location, and the service life of the wind power plant ensures the accuracy of the fault event.

[0122] Specifically, suppose that in a wind power plant, a track-mounted robot discovers a gearbox fault during inspection and collects the following set of fault parameters:

[0123] Gear wear degree: severe (wear amount exceeds 70% of normal value); abnormal oil temperature: extremely high (40°C higher than normal oil temperature); vibration amplitude: significant (amplitude far exceeds normal range); the re-inspection location shows the high-speed shaft area of ​​the gearbox, and the wind power plant has been operating for 15 years, close to its design life.

[0124] This indicates that wear leads to increased friction, which in turn generates more heat. A significant increase in vibration amplitude is related to gear imbalance or damage, which may be caused by long-term wear, and the gearbox is nearing or has reached its design life, making it more prone to failure.

[0125] Based on these analyses, the system will identify the following failure events:

[0126] Fault event: Severe wear in the high-speed shaft area of ​​the gearbox led to abnormally high oil temperature and significantly increased vibration amplitude;

[0127] Cause: Gear wear due to long-term operation, coupled with equipment aging nearing its design life;

[0128] Impact: If not handled in time, it can lead to complete gearbox failure, which in turn affects the operation of the entire wind turbine.

[0129] Furthermore, the relationship between fault parameter sets, re-inspection locations, service life, and fault events can be displayed through a fault event matching table (Table 4).

[0130] Table 4 Fault Event Matching Table

[0131]

[0132] Therefore, by using the above fault event matching table, each fault event and its corresponding fault parameters, inspection location, and service life can be intuitively determined.

[0133] In step S16, the corresponding repair operation trajectory is determined based on the fault event, the judgment module of the rail-mounted robot, and the repair tools possessed by the rail-mounted robot. The specific steps are as follows:

[0134] S161: Obtain the fault event;

[0135] S162: Construct the first training set based on the previous detection data and corresponding judgment results of the rail-mounted robot;

[0136] S163: A judgment module for the rail-mounted robot formed based on the autonomous training of the first training set under various fault dimensions;

[0137] S164: Based on the fault event and the judgment module of the rail-mounted robot, several repair solutions are determined;

[0138] S165: Determine the optimal repair solution based on multiple repair options, the corresponding difficulty coefficients of the repairs, and the repair tools available to the rail-mounted robot.

[0139] S166: Based on the optimal repair plan, the fault event, and the rail-mounted robot, determine the corresponding repair node and mark the corresponding repair operation trajectory at the repair node.

[0140] In the embodiments of this application, the fault event is obtained; a first training set is constructed based on the previous detection data of the rail-mounted robot and the corresponding judgment results; a judgment module of the rail-mounted robot is formed based on the autonomous training of the first training set under each fault dimension, which ensures the training of the judgment module of the rail-mounted robot and improves the accuracy of the judgment module of the rail-mounted robot.

[0141] The input data mainly refers to the set of fault parameters detected by the rail-mounted robot at different times and locations. The set of fault parameters includes severe wear, high oil temperature, abnormal vibration, and deterioration of electrical performance. Furthermore, the input data can be judged manually to determine the fault situation, or the input data can be input into a trained model, and the model can output fault prediction results, including the type, location, and severity of the fault.

[0142] The training of the model includes: training a judgment module for the rail-mounted robot using a first training set. This judgment module is a machine learning or deep learning algorithm that can extract useful features from the input data and make accurate fault judgments based on these features. At this time, different machine learning or deep learning algorithms, such as neural networks, are selected to train the judgment module according to the complexity of the problem and the availability of data. During the training process, the algorithm automatically extracts useful features from the input data. These features are certain statistics of the original data (such as mean, variance, peak value, etc.) and are also obtained through more complex transformations or combinations. In order to improve the accuracy of the judgment module, the model is usually optimized. This includes adjusting the hyperparameters of the algorithm (such as learning rate, number of iterations, etc.), selecting the best feature combination, and using techniques such as cross-validation to evaluate the performance of the model. After training is completed, the model is evaluated to verify its performance.

[0143] Specifically, suppose that the rail-mounted robot has been regularly inspecting gearboxes in wind power plants over the past year. Each inspection records the gearbox's vibration, temperature, and sound data. After each inspection, technical experts use this data to determine if the gearbox has any faults and provide the specific type and location of the fault.

[0144] Now, these detection data and judgment results are organized into a training set. First, all the detection data (vibration data, temperature data, sound data) are used as input data. Then, the judgment results of technical experts (fault type, location) are used as output data. Before the data is input into the training algorithm, the raw data is also cleaned and preprocessed, such as removing abnormal data points caused by equipment failure or environmental interference.

[0145] A training set was constructed, containing vibration data, temperature data, and sound data as input data, and the judgment results of technical experts as output data. This training set will now be used to train the judgment module of the rail-mounted robot. A deep learning algorithm (such as a convolutional neural network CNN) was selected to train the judgment module. During the training process, the algorithm extracts useful features from the input data, such as vibration frequency, temperature gradient, and sound spectrum. Then, the algorithm uses these features to train a classifier to distinguish different types of faults (such as gear wear, bearing failure, etc.).

[0146] After training, the model's performance was evaluated using an independent test dataset. The results showed that the model had high accuracy and recall in gearbox fault diagnosis. Therefore, this trained judgment module was integrated into the control system of the rail-mounted robot, enabling it to autonomously detect and diagnose gearbox faults in wind power plants.

[0147] Furthermore, based on the fault event and the judgment module of the rail-mounted robot, multiple repair solutions are determined, which are compatible with the overall considerations of the fault event and the judgment module of the rail-mounted robot. This enables the judgment module of the rail-mounted robot to judge the fault event and output multiple repair solutions, thus ensuring the rationality of multiple repair solutions.

[0148] During the inspection process, the built-in self-diagnostic system of the rail-mounted robot will monitor the operating status of each component and functional module in real time. Once an abnormality occurs in a critical part, such as fluctuations in the current of the walking drive motor, deviations in sensor data, or abnormal charging and discharging of the battery, the rail-mounted robot can quickly detect and accurately locate the fault point.

[0149] Once a fault is identified, the judgment module of the rail-mounted robot will be activated immediately. The judgment module of the rail-mounted robot will quickly analyze the fault type through the built-in algorithm, and make a preliminary judgment on whether it is caused by hardware aging and wear, software program error, or interference from complex environment. Based on the fault type and judgment result, the judgment module will generate multiple repair solutions.

[0150] Suppose that during an underground inspection, a rail-mounted robot suddenly experiences abnormal fluctuations in the current of its walking drive motor, causing the robot to stall. The following is the process for determining a repair solution for this fault event:

[0151] The robot's self-diagnostic system detected fluctuations in the current of the walking drive motor, located the fault point to the drive motor, and analyzed the current fluctuation data. The initial judgment was that the fault was caused by motor aging, obstructed signal transmission in the control system, or foreign objects obstructing the track.

[0152] Option 1: If the problem is determined to be motor aging, replace the drive motor with a new one. In this case, the robot will automatically send a fault alarm and detailed information to the ground control center, requesting that maintenance personnel be dispatched to replace the motor with a new one.

[0153] Option 2: If the signal transmission of the control system is found to be blocked, it may be due to an error in the control system software program or a fault in the signal line. In this case, the ground control center personnel can remotely debug or update the software, or command the robot to restart the control system function module. If remote control is ineffective, maintenance personnel will be dispatched to the site to check the signal line and repair it.

[0154] Option 3: If it is determined that there is a foreign object obstructing the track, the robot will adjust its position to avoid the foreign object. If it cannot avoid it, it will activate the braking device to fix itself on the track and continuously send out distress signals to wait for rescue. After receiving the alarm, the ground control center will dispatch maintenance personnel to the scene to remove the foreign object.

[0155] Therefore, the rail-mounted robot can quickly identify fault points and generate corresponding repair solutions through its built-in self-diagnosis system and judgment module, so that the ground control center can perform remote operation or on-site maintenance.

[0156] Furthermore, the optimal repair scheme is determined based on multiple repair schemes, their corresponding repair difficulty coefficients, and the repair tools available to the rail-mounted robot. This comprehensive consideration of multiple repair schemes, their corresponding repair difficulty coefficients, and the repair tools available to the rail-mounted robot ensures the accuracy of the optimal repair scheme.

[0157] Furthermore, the repair plan is determined based on factors influencing repair, including the feasibility of the repair plan, the difficulty level of the repair, and the repair capability of the rail-mounted robot itself. Specifically, assuming the rail-mounted robot discovers severe wear on a section of the track during inspection, the system generates the following two repair plans based on the fault event and the judgment module:

[0158] Option 1: Use the simple grinding tools carried by the robot to grind the worn parts. This option is simple and easy to implement, but it requires multiple rounds of grinding to achieve the desired repair effect, and it also has high energy and time costs for the robot.

[0159] Option 2: Dispatch professional maintenance personnel with specialized welding equipment to the site for welding repair. This option offers better repair results, but requires waiting for maintenance personnel to arrive, and the welding operation has high requirements for the environment and skills.

[0160] Considering factors such as repair effectiveness, time cost, robot energy consumption, and availability of maintenance personnel, the system ultimately selected Option 2 as the best repair solution.

[0161] Therefore, based on the optimal repair plan, the fault event, and the rail-mounted robot, the corresponding repair node is determined, and the corresponding repair operation trajectory is marked at the repair node. The repair node and the corresponding repair operation trajectory are controlled so that the rail-mounted robot can trigger the corresponding repair for the fault event.

[0162] At this point, the specific repair location and time point, i.e. the repair node, are determined based on the best repair plan. The repair node is usually located at the location where the fault occurred. After the repair node is determined, the system marks the corresponding repair operation trajectory in the navigation system of the rail-mounted robot. This trajectory is the optimal path for the robot to reach the repair node from the current position. The purpose of determining and marking the repair operation trajectory is to ensure that the robot can accurately reach the repair location and perform repairs according to the predetermined operation steps.

[0163] For example, the system selects the second option above as the best repair option, and after determining the specific location of the worn part on the track as the repair node, the system marks the optimal path from the robot's current position to the repair node in the navigation system of the track-mounted robot, and the robot reaches the repair position according to the optimal path.

[0164] After the robot arrives at the repair location, the system will guide the robot to perform welding repairs according to the predetermined operating steps in order to accurately complete the repair task.

[0165] Example 2:

[0166] In another embodiment of this application, the location of the fault is determined by comparing key information in the line model of the wind power plant with the first fault range, the second fault range, and the fault location. The fault location matching table typically contains the correspondence between fault characteristics, causes, and fault locations.

[0167] Assume the following fault location matching table (Table 5):

[0168] Table 5 Fault Location Matching Table

[0169]

[0170] The first fault range is the abnormally increased current in the transmission line and its nearby components. The second fault range is the transmission line with resistance loss markings and its connection points. Comparing this information with the matching table reveals that: the abnormally increased transmission line current corresponds to causes such as "overload, poor contact, and insulation damage," with the fault location being "transmission line, connection point, and insulator"; the increased resistance loss corresponds to causes such as "resistor aging, poor contact, and corrosion," with the fault location being "resistor, connection point, and line sheath." Considering both the first and second fault ranges, the fault is most likely located in...

[0171] "The connection point of the transmission line," because this is the location involved in both fault ranges.

[0172] Example 3:

[0173] In another embodiment of this application, in order to determine the optimal repair solution more scientifically, a weighted and scored method can be used to determine the optimal repair solution, for example:

[0174] Three evaluation metrics were set: repair effectiveness (weight 0.5), repair cost (weight 0.3), and repair time (weight 0.2).

[0175] Option 1: Repair effectiveness (7 points, due to limited effectiveness), repair cost (9 points, due to low cost), repair time (8 points, due to multiple trips but short time per trip). Option 2: Repair effectiveness (9 points, due to good effectiveness), repair cost (7 points, due to external support but low equipment cost), repair time (7 points, due to waiting for maintenance personnel to arrive but fast repair process). Option 3: Repair effectiveness (10 points, due to complete resolution), repair cost (5 points, due to high cost), repair time (5 points, due to long cycle).

[0176] Option 1: Weighted score = 0.57 + 0.39 + 0.2 * 8 = 3.5 + 2.7 + 1.6 = 7.8 points; Option 2: Weighted score = 0.59 + 0.37 + 0.2 * 7 = 4.5 + 2.1 + 1.4 = 8.0 points; Option 3: Weighted score = 0.510 + 0.35 + 0.2 * 5 = 5.0 + 1.5 + 1.0 = 7.5 points; Based on the weighted scores, Option 2 has the highest score (8.0 points) and is therefore determined to be the best repair option.

[0177] Example 3

[0178] Please see Figure 3 , Figure 3 This is a schematic diagram of the structural composition of the wind power plant fault diagnosis system based on a rail-mounted robot provided in this embodiment. It can implement the wind power plant fault diagnosis method described in Embodiment 1 or 2. The wind power plant fault diagnosis system based on a rail-mounted robot includes:

[0179] The conductor path module 21 is used to determine the conductor path of the wind power plant based on the line model of the wind power plant.

[0180] Robot module 22 is used to determine the corresponding rail-mounted robot based on the conductor path and conductor diameter of the wind power plant.

[0181] The power distribution map module 23 is used to collect multiple current parameters based on the dynamic detection of the conductor by the rail-mounted robot, and to construct the current power distribution map based on the multiple current parameters and the detection position of the conductor.

[0182] The fault location module 24 is used to determine the fault location of the wind power plant based on the current power distribution map;

[0183] The fault event module 25 is used to determine multiple fault parameters based on the re-inspection of the fault location by the rail-mounted robot, and to determine the fault event based on the multiple fault parameters, the re-inspection location of the multiple fault parameters relative to the fault location, and the service life of the wind power plant.

[0184] Repair module 26 is used to determine the corresponding repair operation trajectory based on the fault event, the judgment module of the rail-mounted robot, and the repair tools possessed by the rail-mounted robot.

[0185] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A method for fault diagnosis in wind power plants based on a rail-mounted robot, characterized in that, include: Determine the conductor path of the wind power plant based on the line model of the wind power plant; The appropriate rail-mounted robot is determined based on the conductor path and conductor diameter of the wind power plant. Multiple current parameters are collected based on the dynamic detection of the conductor by the rail-mounted robot, and the current power distribution map is constructed based on the multiple current parameters and the detection position of the conductor. Determine the location of the wind power plant failure based on the current power distribution map; Multiple fault parameters were determined based on the re-inspection of the fault location by the rail-mounted robot. The fault event was determined based on the multiple fault parameters, the re-inspection location of the multiple fault parameters relative to the fault location, and the service life of the wind power plant. The corresponding repair operation trajectory is determined based on the fault event, the judgment module of the rail-mounted robot, and the repair tools possessed by the rail-mounted robot. The determination of the conductor path for the wind power plant based on the line model includes: Collect wiring models of wind power plants; Multiple functional areas are determined based on the division of the wind power plant's line model; Multiple real-time images were collected based on the circumferential aerial survey of the wind power plant by the UAV, and the multiple real-time images show the real-time status of the wind power plant at different locations. Multiple sub-guide paths are determined based on the matching of multiple functional areas and multiple real-time images; The conductor path of the wind power plant is determined based on multiple sub-conductor paths, multiple functional areas, and the line model of the wind power plant. The process of determining the corresponding rail-mounted robot based on the wind power plant's conductor path and conductor diameter includes: Obtain the conductor path of the wind power plant; The distribution location of each conductor is collected by traversing the conductor paths of the wind power plant. The distribution style of the conductors is determined based on their location, their parallel arrangement, and their surrounding environment. Multiple wire diameters are collected based on the detection of each wire, and the wire diameter is determined according to the multiple wire diameters and the wire model information. The conductor paths, conductor distribution patterns, and conductor diameters of wind power plants are correlated. The first robot parameters are determined based on the conductor path and conductor distribution style of the wind power plant, and the second robot parameters are determined based on the conductor path and conductor diameter of the wind power plant. The corresponding track-mounted robot is determined based on the first robot parameters, the second robot parameters, and previous robot data from the wind power plant.

2. The method for diagnosing faults in wind power plants according to claim 1, characterized in that, The method involves collecting multiple current parameters based on the dynamic detection of the conductor by the rail-mounted robot, and constructing a current power distribution map based on these current parameters and the detected position of the conductor, including: Interact with the wind power plant's power transmission line path and the location of the track-mounted robot; The inspection route of the rail-mounted robot is determined based on the interaction between the conductor path of the wind power plant and the location of the rail-mounted robot. The rail-mounted robot dynamically travels along the inspection route on the conductor, triggering dynamic detection of the conductor by the rail-mounted robot. Multiple current parameters are collected based on the dynamic detection of the conductor by the rail-mounted robot; Multiple current parameters are matched to the detection positions of each conductor; The current power distribution map is constructed based on multiple current parameters and the detection location of the conductor.

3. The method for determining faults in wind power plants according to any one of claims 1 to 2, characterized in that, The method of determining the location of a wind power plant fault based on the current power distribution map includes: Obtain the current power distribution map; The abnormal power range is determined by autonomous detection based on the current power distribution map; Collect multiple operating parameters of the wind power plant, and determine the current power generation mode of the wind power plant based on the multiple operating parameters of the wind power plant and the power generation mode recorded by the wind power plant. The location of wind power plant losses is determined based on the current power distribution map and the loss markers of the wind power plant. Abnormal ranges in related power, the current power generation mode of wind power plants, and the location of losses at wind power plants; The first fault range is determined based on the abnormal power range and the current power generation mode of the wind power plant. The second fault range is determined based on the abnormal power range and the location of the loss in the wind power plant. The location of the fault in the wind power plant is determined based on the first fault range, the second fault range, and the line model of the wind power plant.

4. The method for diagnosing faults in wind power plants according to claim 3, characterized in that, The process involves determining multiple fault parameters based on the re-inspection of the fault location by the rail-mounted robot, and determining the fault event based on these multiple fault parameters, the re-inspection location of the multiple fault parameters relative to the fault location, and the service life of the wind power plant. This includes: Locate the location of the wind power plant failure; The location of the fault in the wind power plant triggers the movement of the rail-mounted robot, which then re-inspects the fault location. Based on the re-inspection of the fault location by the rail-mounted robot, multiple fault parameters are determined. At this time, the corresponding re-inspection mode is determined according to the actual image of the fault location, the re-inspection range of the fault location by the rail-mounted robot, and the importance coefficient of the fault location relative to the wind power plant.

5. The method for diagnosing faults in wind power plants according to claim 4, characterized in that, The method of determining multiple fault parameters based on the re-inspection of the fault location by the rail-mounted robot, and determining the fault event based on the multiple fault parameters, the re-inspection location of the multiple fault parameters relative to the fault location, and the service life of the wind power plant, also includes: Obtain multiple fault parameters and sort them by size to form a fault parameter set; The associated fault parameter set, the re-inspection location of multiple fault parameters relative to the fault location, and the service life of the wind power plant; The system performs multiple interactions on the fault parameter set, the re-inspection location of multiple fault parameters relative to the fault location, and the service life of the wind power plant, and determines the fault event based on these multiple interactions.

6. The method for diagnosing faults in wind power plants according to claim 5, characterized in that, The step of determining the corresponding repair operation trajectory based on the fault event, the judgment module of the rail-mounted robot, and the repair tools possessed by the rail-mounted robot includes: Obtain the fault event; The first training set is constructed based on the previous detection data and corresponding judgment results of the rail-mounted robot; The judgment module of the rail-mounted robot is formed based on the autonomous training of the first training set under various fault dimensions.

7. The method for diagnosing faults in wind power plants according to claim 6, characterized in that, The step of determining the corresponding repair operation trajectory based on the fault event, the judgment module of the rail-mounted robot, and the repair tools possessed by the rail-mounted robot also includes: Based on the fault event and the judgment module of the rail-mounted robot, several repair solutions were determined; The optimal repair solution is determined based on multiple repair options, the corresponding difficulty levels of the repairs, and the repair tools available to the rail-mounted robot. Based on the optimal repair plan, the fault event, and the rail-mounted robot, the corresponding repair node is determined, and the corresponding repair operation trajectory is marked at the repair node.

8. A fault diagnosis system for wind power plants based on a rail-mounted robot, characterized in that, The wind power plant fault diagnosis system based on a rail-mounted robot is applied to the wind power plant fault diagnosis method based on a rail-mounted robot as described in any one of claims 1-7. The wind power plant fault diagnosis system based on a rail-mounted robot includes: The conductor path module is used to determine the conductor path of a wind farm based on the line model of the wind farm. The robot module is used to determine the corresponding rail-mounted robot based on the conductor path and conductor diameter of the wind power plant. The power distribution map module is used to collect multiple current parameters based on the dynamic detection of the conductor by the rail-mounted robot, and to construct the current power distribution map based on the multiple current parameters and the detection position of the conductor. The fault location module is used to determine the location of a fault in a wind power plant based on the current power distribution map. The fault event module is used to determine multiple fault parameters based on the re-inspection of the fault location by the rail-mounted robot, and to determine the fault event based on the multiple fault parameters, the re-inspection location of the multiple fault parameters relative to the fault location, and the service life of the wind power plant. The repair module is used to determine the corresponding repair operation trajectory based on the fault event, the judgment module of the rail-mounted robot, and the repair tools possessed by the rail-mounted robot.

Images