Intelligent early warning method for fan robot inspection system

Abstract

The invention provides an intelligent early warning method for a fan robot inspection system, and relates to the technical field of intelligent control, and the method comprises the steps: determining a robot inspection route, and obtaining fan equipment and environment state information through a robot; preprocessing the fan equipment and environment state information to obtain the processed fan equipment and environment state information; analyzing the processed fan equipment and environment state information to extract key information and feature data; identifying the abnormal state of the equipment according to the key information and the feature data; historical operation data, including occurrence frequencies, duration and influence ranges of various abnormal events, of the fan equipment is obtained, and an adjustment factor is calculated according to the historical operation data of the fan equipment. Through intelligent early warning of the fan robot inspection system, real-time monitoring, fault prediction and intelligent early warning of the running state of the fan can be realized, so that the operation and maintenance efficiency and safety of a wind power plant are effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control technology, and in particular to an intelligent early warning method for a wind turbine robot inspection system. Background Art

[0002] In the field of wind power generation, the stable operation of wind turbine equipment is crucial to ensuring energy supply and reducing maintenance costs. In order to promptly detect and deal with potential failures of wind turbine equipment, traditional inspection methods mainly rely on regular manual inspections. However, this method has some defects and may not meet the modern wind power industry's needs for high efficiency, high reliability and safety.

[0003] First, the traditional manual inspection method is inefficient. Since wind turbine equipment is usually distributed in a wide area with complex environment, manual inspection requires a lot of time and human resources. In addition, manual inspection is also limited by many factors such as weather conditions, personnel skill level and work fatigue, making it difficult to ensure the comprehensiveness and timeliness of the inspection.

[0004] Secondly, traditional inspection methods have limitations in data collection and processing. Manual inspections mainly rely on the observation and recording of inspectors, and the accuracy and consistency of data acquisition are difficult to guarantee. At the same time, due to the lack of effective data analysis and processing methods, traditional methods are difficult to extract key information and features from large amounts of data, and cannot achieve accurate assessment of equipment status. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide an intelligent early warning method for a wind turbine robot inspection system, which can realize real-time monitoring of the wind turbine operating status, fault prediction and intelligent early warning, thereby effectively improving the operation and maintenance efficiency and safety of the wind farm.

[0006] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0007] In a first aspect, an intelligent early warning method for a wind turbine robot inspection system is provided, the method comprising:

[0008] Step 11, determine the robot inspection route, and obtain wind turbine equipment and environmental status information through the robot;

[0009] Step 12, preprocessing the fan equipment and environmental status information to obtain processed fan equipment and environmental status information;

[0010] Step 13, analyzing the processed wind turbine equipment and environmental status information to extract key information and feature data; identifying abnormal equipment status based on the key information and feature data;

[0011] Step 14, obtaining historical operation data of the wind turbine equipment, including the occurrence frequency, duration and impact range of various abnormal events, and calculating the adjustment factor according to the historical operation data of the wind turbine equipment;

[0012] Step 15, performing a preliminary analysis on the abnormal state of the device to obtain a preliminary abnormal urgency; multiplying the adjustment factor by the preliminary abnormal urgency to obtain a final abnormal urgency;

[0013] Step 16, execute emergency response warning according to the abnormal urgency, including continuous monitoring and issuing alarm information.

[0014] Furthermore, the robot inspection route is determined, and the wind turbine equipment and environmental status information is obtained through the robot, including:

[0015] Generate an initial set of candidate inspection paths based on the physical layout, criticality, accessibility, and safety requirements of the wind turbine equipment;

[0016] Set the initial parameters of the genetic algorithm, including population size, crossover rate, mutation rate, and number of iterations;

[0017] Encode each inspection path as an individual of a genetic algorithm, and each individual represents a possible inspection path; generate the initial population of the genetic algorithm based on the initial inspection path candidates;

[0018] Define a fitness function to evaluate the quality of each inspection path;

[0019] According to the fitness function, the corresponding individuals are selected from the current population as the parents to generate the next generation; through the crossover operation, new inspection path individuals are generated; the newly generated individuals are randomly mutated, and the selection, crossover and mutation operations are repeated until the preset number of iterations is reached to obtain the corresponding individuals, that is, the final inspection path;

[0020] According to the final inspection route, the robot is controlled to perform inspection tasks. During the inspection process, the status information of the wind turbine equipment and the surrounding environment is collected in real time, including wind turbine equipment images, temperature, humidity, vibration, sound, wind speed, wind direction, atmospheric pressure, equipment voltage, current, oil pressure, oil temperature and oil level of bearings and gearboxes, yaw and pitch angles, stress and strain data of towers and blades, smoke inside the equipment, settlement of soil and foundation, and displacement monitoring.

[0021] Furthermore, based on the physical layout, criticality, accessibility, and safety requirements of the wind turbine equipment, an initial inspection path candidate set is generated, including:

[0022] Determine the relative positions and distances between the devices according to the physical layout of the fan equipment; regard each fan equipment as a node in the graph, and the feasible paths between the devices as edges; assign a weight value to each edge according to the actual distance, movement difficulty and terrain factors between the devices, representing the cost of passing through the corresponding path.

[0023] Identify and mark the inaccessible or dangerous areas, and select a starting node, which is the starting position of the inspection task; initialize the distance array to record the shortest distance from the starting node to each node in the graph. Initially, except for the starting node which is set to 0, the remaining nodes are set to infinity; create an empty set to save the nodes for which the shortest paths have been found; create a priority queue using a min-heap, and the elements in the queue are sorted according to the distance from the starting node to that node. Initially, only the starting node is in the queue.

[0024] Take out the node closest to the starting node from the priority queue. For each adjacent node of this node, calculate the distance from the starting node to the adjacent node through the current node. If this distance is less than the previously recorded distance, update the distance array and add the adjacent node to the priority queue.

[0025] Mark the current node as having found the shortest path and add it to the set of processed nodes. Repeat the above steps until the priority queue is empty, that is, the shortest paths of all reachable nodes have been found.

[0026] Determine the shortest path from the starting node to each fan equipment according to the distance array; evaluate the generated inspection paths according to the total length of the paths, the number of key equipment covered and the safety to obtain the evaluation results.

[0027] Select multiple corresponding inspection paths as the initial inspection path candidate set according to the evaluation results. The initial inspection path candidate set includes the information of the nodes and edges passed on the path.

[0028] Furthermore, the calculation formula of the fitness function is:

[0029]

[0030] Among them, F is the fitness value of the individual, D is the total length of the inspection path, a is the exponent of the length factor. When a > 1, the penalty for longer paths will be increased; when 0 < a < 1, the penalty for longer paths will be reduced; W D is the weight coefficient of the path length; K is the number of key equipment covered by the inspection path, K max is the total number of key equipment, g is the exponent of the key equipment coverage factor, W K is the weight coefficient of the key equipment coverage rate; S is the safety score of the inspection path, S mαxis the highest score of safety, γ is the index of safety factor, W S is the safety weight coefficient, R represents the accessibility score of the path, and W R is the weight coefficient of accessibility, C is the continuity score of the path, and W C is the continuity weight coefficient.

[0031] Furthermore, the processed wind turbine equipment and environmental status information is analyzed to extract key information and characteristic data; based on the key information and characteristic data, the abnormal status of the equipment is identified, including:

[0032] Analyze the processed fan equipment and environmental status information, calculate and summarize the basic statistics of each data point, including the mean, standard deviation, maximum and minimum values;

[0033] According to the historical data distribution, set corresponding thresholds for each state information, and compare the statistics of each data point with the set threshold one by one to identify potential outliers;

[0034] Using potential outliers, construct time series graphs to initially identify the existence of cyclical fluctuations;

[0035] Through Fourier analysis, the periodic components in the data are extracted, and the Spearman rank correlation coefficient between each data point is calculated using the data set containing periodic change information. The Spearman rank correlation coefficient is used to quantify the strength and direction of the correlation between different data points.

[0036] Extract key information and characteristic data representing the status of wind turbine equipment and environment based on abnormal values, periodic fluctuations and correlation strength and direction;

[0037] Identify abnormal status of equipment based on key information and feature data.

[0038] Furthermore, the calculation formula of the Spearman rank correlation coefficient is:

[0039]

[0040] Among them, r s represents the Spearman rank correlation coefficient; 6 represents the scaling factor; Q x( (j) and Q Y (j) represents x j and j The rank order in the respective dataset; M represents the number of data points; j represents the index variable; and m represents the total number of data points.

[0041] Furthermore, the calculation formula of the adjustment factor is:

[0042]

[0043] Among them, p i is the frequency of occurrence of the ith abnormal event, q i is the average duration of the ith abnormal event, v i is the impact range of the ith abnormal event, u i is the basic weight coefficient of the ith abnormal event, m i is the severity coefficient of the ith abnormal event, n i is the complexity score of the ith abnormal event, τ is the time decay coefficient, N is the total number of abnormal event types considered, T m is the current time, T t is the recording time of the ith abnormal event data, k, b, h are exponential parameters for duration and impact range, n max is the maximum complexity score among all abnormal events, and i is the index.

[0044] In the second aspect, an intelligent early warning system for a wind turbine robot inspection system includes:

[0045] Determine the module, determine the robot inspection route, and obtain wind turbine equipment and environmental status information through the robot;

[0046] A preprocessing module is used to preprocess the fan equipment and environmental status information to obtain processed fan equipment and environmental status information;

[0047] The analysis module analyzes the processed wind turbine equipment and environmental status information to extract key information and characteristic data; identifies abnormal equipment status based on the key information and characteristic data; obtains historical operation data of the wind turbine equipment, including the frequency, duration and impact range of various abnormal events, and calculates adjustment factors based on the historical operation data of the wind turbine equipment;

[0048] The execution module performs a preliminary analysis of the abnormal state of the equipment to obtain a preliminary abnormal urgency; multiplies the adjustment factor by the preliminary abnormal urgency to obtain a final abnormal urgency; and executes emergency response warnings based on the abnormal urgency, including continuous monitoring and issuing alarm information.

[0049] According to a third aspect, a computing device includes:

[0050] one or more processors;

[0051] The storage device is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the method described.

[0052] In a fourth aspect, a computer-readable storage medium stores a program, and when the program is executed by a processor, the method described is implemented.

[0053] The above solution of the present invention includes at least the following beneficial effects:

[0054] Through robot inspection, the status information of wind turbine equipment and environment can be automatically collected, reducing the time and cost of manual inspection. The use of genetic algorithm to optimize the inspection path can ensure that the robot traverses all key equipment in the most efficient way, thereby further improving the inspection efficiency and reducing the operation risk of the robot in a complex environment. By preprocessing and analyzing the collected data, the system can extract key information and feature data, and then identify the abnormal state of the equipment, which helps to find potential problems in time and prevent the occurrence or expansion of equipment failures. The system performs emergency response warnings according to the urgency of the abnormality, including continuous monitoring and issuing alarm information, which helps relevant personnel to respond in time and take necessary measures to solve the problem, thereby minimizing equipment downtime and production losses. By considering historical operating data and the calculation of adjustment factors, the stable operation of wind turbine equipment is maintained and the overall performance is optimized. The method and system can flexibly adapt to wind turbine equipment groups of different sizes and complexities, and have good scalability, and can be adjusted and optimized accordingly with the increase in the number of equipment or technology updates. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a flow chart of an intelligent early warning method for a wind turbine robot inspection system provided by an embodiment of the present invention.

[0056] Figure 2 It is a schematic diagram of an intelligent early warning system for a wind turbine robot inspection system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0058] like Figure 1 As shown, an embodiment of the present invention provides an intelligent early warning method for a wind turbine robot inspection system, the method comprising the following steps:

[0059] Step 11, determine the robot inspection route, and obtain wind turbine equipment and environmental status information through the robot;

[0060] Step 12, preprocessing the fan equipment and environmental status information to obtain processed fan equipment and environmental status information;

[0061] Step 13, analyzing the processed wind turbine equipment and environmental status information to extract key information and feature data; identifying abnormal equipment status based on the key information and feature data;

[0062] Step 14, obtaining historical operation data of the wind turbine equipment, including the occurrence frequency, duration and impact range of various abnormal events, and calculating the adjustment factor according to the historical operation data of the wind turbine equipment;

[0063] Step 15, performing a preliminary analysis on the abnormal state of the device to obtain a preliminary abnormal urgency; multiplying the adjustment factor by the preliminary abnormal urgency to obtain a final abnormal urgency;

[0064] Step 16, execute emergency response warning according to the abnormal urgency, including continuous monitoring and issuing alarm information.

[0065] In the embodiment of the present invention, through robot inspection, the efficiency and accuracy of data collection can be improved, the cost and safety risks of manual inspection can be reduced, and the robot can reach areas that are difficult to access manually, providing more comprehensive equipment and environmental status information. Preprocessing can clean and organize raw data, remove noise and outliers, and improve data quality. Extracting key information and feature data can simplify complex data sets and highlight key information. By analyzing key information and feature data, equipment abnormalities can be discovered in time and potential failures can be prevented. The use of historical operating data can better understand the operating conditions and abnormal modes of wind turbine equipment. The calculation of the adjustment factor takes into account the past performance of the equipment, making the assessment of abnormal urgency more accurate and personalized. The analysis of the preliminary abnormal urgency provides a preliminary judgment basis for rapid response. The application of the adjustment factor makes the assessment of abnormal urgency more refined, which helps to optimize resource allocation and response strategies. Executing emergency response warnings according to abnormal urgency can achieve rapid and targeted intervention and reduce losses caused by failures. Continuous monitoring and timely release of alarm information help improve the reliability and safety of the system.

[0066] In a preferred embodiment of the present invention, the above step 11 includes:

[0067] Step 111, generating an initial inspection path candidate set according to the physical layout, criticality, accessibility and safety requirements of the wind turbine equipment;

[0068] Step 112, setting the initial parameters of the genetic algorithm, including population size, crossover rate, mutation rate and number of iterations;

[0069] Step 113, encoding each inspection path as an individual of a genetic algorithm, each individual representing a possible inspection path; generating an initial population of the genetic algorithm based on the initial inspection path candidates;

[0070] Step 114, defining a fitness function for evaluating the quality of each inspection path;

[0071] Step 115, according to the fitness function, select the corresponding individual from the current population as the parent generation to generate the next generation; generate a new inspection path individual through a crossover operation; perform random mutation on the newly generated individuals, and repeat the selection, crossover and mutation operations until a preset number of iterations is reached to obtain the corresponding individual, that is, the final inspection path;

[0072] Step 116, according to the final inspection route, control the robot to perform the inspection task. During the inspection process, real-time status information of the wind turbine equipment and the surrounding environment is collected, including wind turbine equipment image, temperature, humidity, vibration, sound, wind speed, wind direction, atmospheric pressure, equipment voltage, current, oil pressure, oil temperature and oil level of bearings and gearboxes, yaw and pitch angles, stress and strain data of towers and blades, smoke inside the equipment, settlement of soil and foundation, and displacement monitoring.

[0073] In the embodiment of the present invention, by considering the physical layout, criticality, accessibility and safety of the wind turbine equipment, a more reasonable and efficient inspection path candidate set can be generated, and setting the initial parameters of the genetic algorithm can ensure that the algorithm searches within a suitable range, thereby improving the convergence speed and optimization ability of the algorithm. By encoding the inspection path as an individual of the genetic algorithm, the global search ability of the genetic algorithm can be used to find the optimal solution in a huge solution space. Generating an initial population increases the diversity of the inspection path, which helps the algorithm to find a better solution during the iteration process. Defining a fitness function can quantitatively evaluate the pros and cons of each inspection path, provide a clear basis for the selection operation of the algorithm, and ensure that excellent inspection paths can be retained and passed on to the next generation. Through selection, crossover and mutation operations, new inspection path individuals can be continuously generated, the diversity of the population can be increased, and the optimal solution can be gradually approached. The final inspection path obtained after reaching the preset number of iterations is the result of multiple optimizations, and has high reliability and efficiency. Carrying out robot inspections according to the final inspection path can ensure the comprehensiveness and efficiency of the inspection, and reduce the possibility of missed inspections and repeated inspections. Real-time collection of status information of wind turbine equipment and surrounding environment helps to timely identify potential problems and take corresponding maintenance measures. Diversified data collection can also help the operation and maintenance team to have a more comprehensive understanding of the operating status of wind turbine equipment.

[0074] When specifically applied, this includes:

[0075] Step 111, analyze in detail the physical layout of the wind turbine equipment, including the spatial coordinates of the equipment, the width of the passages between the equipment, the size of the equipment, and other information, understand the relative positions and distances between the various equipment, and judge and evaluate the criticality of each equipment based on factors such as the importance of the equipment, the frequency of failures, and the maintenance history, and determine which equipment must be inspected during the inspection. Analyze the accessibility of each device, considering whether the inspection personnel or robots can easily reach the equipment location, including factors such as the width, slope, and obstacles of the passage. Evaluate the safety of the inspection path and avoid areas that may pose safety risks, such as high temperature, high pressure, and fragile areas. Based on the above analysis, use graph theory or gridding methods to generate a series of initial inspection path candidates. These paths cover all critical equipment and meet accessibility and safety requirements.

[0076] Steps 112-113, determine the size of the population in the genetic algorithm, that is, the number of individuals in each generation, set the probability of the crossover operation to control the frequency of genetic information exchange between individuals, set the probability of the mutation operation to introduce new genetic mutations and increase the diversity of the population, determine the maximum number of iterations of the algorithm, that is, the maximum number of generations of the algorithm, and encode each inspection path as an individual of the genetic algorithm. Based on the initial inspection path candidate set generated in step 111, a certain number of paths are randomly selected or selected according to a certain strategy as the initial population of the genetic algorithm.

[0077] Step 114, count the number of key devices covered on the path, score the safety of the path, consider factors such as obstacles and dangerous areas on the path, and weight the above factors to form a fitness function for evaluating the quality of each path.

[0078] Step 115, according to the fitness function, select excellent individuals from the current population as parents, perform crossover operations on the selected parent individuals to generate new offspring individuals, perform random mutations on the newly generated offspring individuals to increase the diversity of the population, merge the newly generated offspring individuals with the parent individuals to form a new population. Then repeat the selection, crossover and mutation operations until the preset number of iterations is reached.

[0079] Step 116, based on the final inspection path obtained by genetic algorithm optimization, control the robot or inspection personnel to perform inspection along the path. During the inspection process, real-time status information of the fan equipment and the surrounding environment is collected, including the use of various sensors (such as temperature sensors, humidity sensors, vibration sensors, etc.) and equipment (such as cameras, infrared thermal imagers, etc.) to collect data.

[0080] In a preferred embodiment of the present invention, the above step 111 includes:

[0081] Step 1111, according to the physical layout of the wind turbine equipment, determine the relative position and distance between the equipment; regard each wind turbine equipment as a node in the graph, and the feasible paths between the equipment as edges; according to the actual distance between the equipment, the difficulty of movement and the terrain factors, assign a weight value to each edge, which represents the cost of passing through the corresponding path;

[0082] Step 1112, identify and mark the unreachable or dangerous areas, select a starting node, which is the starting position of the inspection task; initialize the distance array, record the shortest distance from the starting node to each node in the graph, initially set the starting node to 0, and the remaining nodes to infinity; create an empty set to save the nodes for which the shortest path has been found; use a minimum heap to create a priority queue, the elements in the queue are sorted according to the distance from the starting node to the node, and initially only the starting node is in the queue;

[0083] Step 1113, taking out the node closest to the starting node from the priority queue, and for each adjacent node of the node, calculating the distance from the starting node through the current node to the adjacent node, if the distance is less than the previously recorded distance, updating the distance array and adding the adjacent node to the priority queue;

[0084] Step 1114, mark the current node as having found the shortest path, and add it to the set of processed nodes, repeat the above steps until the priority queue is empty, that is, the shortest paths to all reachable nodes have been found;

[0085] Step 1115, determining the shortest path from the starting node to each wind turbine device according to the distance array; evaluating the generated inspection path according to the total length of the path, the number of key devices covered, and the safety to obtain an evaluation result;

[0086] Step 1116: select multiple corresponding inspection paths as an initial inspection path candidate set according to the evaluation result. The initial inspection path candidate set includes information about nodes and edges passed on the path.

[0087] In an embodiment of the present invention, by converting the physical layout of the wind turbine equipment into a graph theory model, complex spatial relationships can be simplified, which facilitates path planning and optimization. The setting of nodes and edges can clearly represent the relative positions and feasible paths between devices. Assigning weights to each edge according to actual distance, difficulty of movement and terrain factors can more realistically reflect the actual cost in the inspection process, and help plan an inspection path that is both efficient and safe. Identifying and marking unreachable or dangerous areas can avoid these areas during path planning to ensure the safe execution of inspection tasks, which helps to reduce safety risks in the inspection process and protect the safety of inspection personnel and equipment. By using data structures such as priority queues and minimum heaps, the shortest path from the starting node to each wind turbine device can be efficiently searched. The generated inspection path is evaluated according to the total length of the path, the number of key devices covered and the safety, and multiple factors can be comprehensively considered to select the most suitable inspection path. The initial inspection path candidate set generated by the above steps includes multiple optimized and evaluated inspection paths, which provides rich options for subsequent path selection and adjustment.

[0088] When specifically applied, this includes:

[0089] Step 1111, analyze the physical layout of the wind turbine equipment in detail, and determine the exact location of each device in space through GPS coordinates, equipment installation drawings or on-site surveys. Each wind turbine device is regarded as a node in the graph. The feasible paths between devices, that is, the channels through which personnel or robots can move, are regarded as edges in the graph. In this way, the layout of the entire wind turbine equipment is converted into a graph model. According to the actual distance, the difficulty of movement (such as slope, obstacles, etc.) and terrain factors, a weight value is assigned to each edge in the graph. The weight value represents the difficulty or cost of passing through the path. For example, a flat and short distance may have a lower weight, while a steep and long distance may have a higher weight.

[0090] Step 1112, mark those areas that are inaccessible due to safety reasons or physical limitations. These areas will be regarded as inaccessible nodes or edges in the graph. Determine the starting position of the inspection task and use it as a specific node in the graph. This node will be the starting point for the shortest path search. Create an array to record the shortest distance from the starting node to each node in the graph. Initially, except for the distance of the starting node set to 0, the distances of the remaining nodes are set to infinity. Create a priority queue using the minimum heap data structure. The elements in the queue are sorted according to the currently known shortest distance from the starting node to the node. Initially, only the starting node is in the queue and its distance is 0.

[0091] Step 1113: Take out the node closest to the starting node from the priority queue. This node is the unprocessed node that is currently known to be the closest to the starting node. For each adjacent node taken out currently, calculate the distance from the starting node to this adjacent node through the current node. If this newly calculated distance is less than the previously recorded distance, update the distance array, and add the adjacent node (if it is not already in the queue) to the priority queue, or update its distance value in the queue. Continuously take out the closest node from the priority queue and update the distances of its adjacent nodes until the queue is empty.

[0092] Step 1114: Mark the currently processed node as having found the shortest path, meaning that the shortest path from the starting node to this node has been found. Add the current node to a set of processed nodes so as to ignore it in subsequent iterations.

[0093] Step 1115: According to the distance array, the shortest paths from the starting node to each fan device can be determined. These paths are connected by a series of nodes and edges. For each generated inspection path, evaluate it according to the total length of the path, the number of key devices covered, and the security.

[0094] Step 1116: According to the evaluation results of the previous step, select multiple excellent inspection paths as the initial candidate set of inspection paths. For each selected inspection path, record in detail the information of the nodes and edges it passes through.

[0095] In a preferred embodiment of the present invention, in the above step 114, the calculation formula of the fitness function is:

[0096]

[0097] where F is the fitness value of the individual, D is the total length of the inspection path, a is the exponent of the length factor. When a > 1, it will increase the penalty for longer paths; when 0 < a < 1, it will reduce the penalty for longer paths; W D is the weight coefficient of the path length; K is the number of key devices covered by the inspection path, K max is the total number of key devices, g is the exponent of the key device coverage factor, W K is the weight coefficient of the key device coverage rate; S is the security score of the inspection path, S max is the highest score of security, γ is the exponent of the security factor, W S is the weight coefficient of security, R represents the reachability score of the path, W R is the weight coefficient of reachability, C is the continuity score of the path, W C is the weight coefficient of continuity.

[0098] In an embodiment of the present invention, the fitness function takes multiple key factors into consideration to ensure that the generated inspection path is optimized in multiple dimensions. By setting different weight coefficients, the importance of each factor in path planning can be flexibly adjusted according to actual needs to meet the inspection needs in different scenarios. When the exponent of the length factor is greater than 1, the function will increase the penalty for longer paths, which helps to generate shorter inspection paths and improve inspection efficiency. Conversely, when the index is between 0 and 1, the penalty for longer paths can be reduced to adapt to different requirements for path length in certain specific scenarios. By considering the coverage of key equipment, the function ensures that the inspection path can cover as many key equipment as possible, so that potential problems can be discovered and handled in a timely manner, and the reliability and operating efficiency of the equipment can be improved.

[0099] In a preferred embodiment of the present invention, the above step 12 includes:

[0100] Step 121, collecting real-time data from the wind turbine robot and ground monitoring equipment;

[0101] Step 122, establishing a communication connection between the wind turbine robot and the ground monitoring equipment and the inspection intelligent system;

[0102] Step 123, formatting the acquired real-time data to obtain formatted data;

[0103] Step 124 , encapsulate the formatted data into a data packet and transmit it from the robot to the inspection intelligent system in real time.

[0104] In an embodiment of the present invention, real-time transmission of real-time data collected by a wind turbine robot and ground monitoring equipment is achieved. By establishing a stable communication connection, the real-time and accuracy of the data are ensured. The acquired real-time data is formatted so that the data has a unified format and standard. The formatting process improves the quality and consistency of the data and reduces analysis errors caused by inconsistent data formats. The formatted data is encapsulated into a data packet and transmitted to the inspection intelligent system in real time through a robot or other communication methods, thereby achieving rapid data transmission and sharing. This real-time data transmission mechanism enables the inspection intelligent system to obtain the latest data in a timely manner, providing timely and accurate data support for intelligent early warning.

[0105] When specifically applied, this includes:

[0106] Step 121, the wind turbine robot and the ground monitoring equipment collect data according to the preset inspection route and monitoring points. The wind turbine robot uses the monitoring camera and infrared thermal imager to obtain real-time video images and temperature data inside the cabin; the ground monitoring equipment collects various environmental parameter data such as temperature, humidity, wind direction, and ambient gas. The collected real-time data is stored in the local memory of the device.

[0107] Step 122, establish a stable communication connection between the wind turbine robot and the ground monitoring equipment and the inspection intelligent system. The communication connection can be wired or wireless, such as wireless communication technologies such as Wi-Fi, Bluetooth, ZigBee, or wired communication technologies such as optical fiber and network cable. After the communication connection is established, the equipment can transmit the collected data to the inspection intelligent system in real time.

[0108] Step 123, formatting the acquired real-time data to make it have a unified format and standard. The formatting includes conversion of data types, unification of data units, and standardization of data formats, for example, converting temperature data from Celsius to Kelvin, converting humidity data from percentage to decimal form, converting image data to a unified pixel format, etc.

[0109] Step 124, encapsulate the formatted data into a data packet. The data packet includes a data header, a data body, and a checksum, wherein the data header contains information such as the source, type, and size of the data, the data body contains the actual data content, and the checksum is used to verify the integrity and accuracy of the data. The encapsulated data packet is transmitted to the inspection intelligent system in real time through a robot or other communication methods (such as a wireless communication network, a wired communication network, etc.).

[0110] In a preferred embodiment of the present invention, the above step 13 analyzes the processed wind turbine equipment and environmental status information to extract key information and feature data; and identifies the abnormal state of the equipment based on the key information and feature data, including:

[0111] Step 131, analyzing the processed wind turbine equipment and environmental status information, calculating and summarizing basic statistics of each data point, including mean value, standard deviation, maximum value and minimum value;

[0112] Step 132, according to the historical data distribution, set corresponding thresholds for each state information, and compare the statistics of each data point with the set thresholds one by one to identify potential outliers;

[0113] Step 133, using potential outliers, construct a time series graph to preliminarily identify existing cyclical fluctuations;

[0114] Step 134, extracting the periodic components in the data through Fourier analysis, using the data set containing periodic change information, calculating the Spearman rank correlation coefficient between each data point, and quantifying the strength and direction of the correlation between different data points through the Spearman rank correlation coefficient;

[0115] Step 135, extracting key information and characteristic data representing the status of the wind turbine equipment and the environment according to the abnormal values, periodic fluctuations and correlation strength and direction;

[0116] Step 136, identifying the abnormal state of the device based on the key information and characteristic data.

[0117] In the embodiment of the present invention, statistics such as the mean value and standard deviation are helpful to discover the central trend and dispersion of the data, the maximum value and the minimum value are helpful to identify the extreme cases of the data, setting the threshold value is helpful to establish the benchmark for data monitoring, and it is convenient to find data anomalies in time. The identification of outliers is helpful to prevent potential equipment failures. Through outlier analysis, data support can be provided for the preventive maintenance of the equipment and the maintenance cost can be reduced. The time series graph intuitively shows the trend of data changes over time, which is convenient for discovering patterns such as periodic fluctuations. By identifying periodic fluctuations, the operating law of the equipment can be better understood, providing a basis for the optimized operation of the equipment, helping to predict the state of the equipment at a future time point, and improving the predictability of equipment management. Fourier analysis can deeply explore the periodic components in the data and reveal the laws hidden behind the complex data. The Spearman rank correlation coefficient can quantify the correlation between different data points, which is helpful to discover the intrinsic connection and influencing factors between the data. Through correlation analysis, new perspectives and ideas can be provided for equipment fault diagnosis and performance optimization. Extracting key information and feature data helps to simplify complex data sets and highlight core information. These key information and feature data can be used as important indicators for equipment status monitoring and fault diagnosis. It helps to build an efficient equipment status assessment model and improve the intelligent level of equipment management. Timely identification of abnormal equipment conditions helps prevent equipment failures and improves equipment safety and reliability. By identifying abnormal conditions, equipment maintenance plans can be optimized, unnecessary downtime can be reduced, equipment utilization can be improved, equipment maintenance costs and operating costs can be reduced, and the economic benefits of the enterprise can be improved.

[0118] When specifically applied, this includes:

[0119] Step 131, for each data point, calculate the average of all its observations, calculate the standard deviation of each data point to measure the dispersion or volatility of the data, find the maximum and minimum values ​​of each data point, and organize the calculated statistics into a table or report.

[0120] Step 132, study the distribution characteristics of historical data, obtain the data range under normal conditions, and set a reasonable threshold for each data point based on the historical data distribution. The threshold is a fixed value or a dynamic value based on a statistic. The actual observed value of each data point is compared with the set threshold. Data points that exceed the threshold range are regarded as potential outliers.

[0121] Step 133, combining the potential outliers with their corresponding timestamps to form time series data, using the time series data to draw a line graph or a scatter plot with time as the horizontal axis and the data value as the vertical axis, observing the time series graph, observing the occurrence pattern of peaks and troughs, so as to preliminarily identify possible periodic fluctuations.

[0122] Step 134, applying Fourier transform, converting the time series data from the time domain to the frequency domain, identifying significant frequency components in the frequency domain, these components correspond to the periodic fluctuations in the time domain, converting the extracted periodic components back to the time domain, obtaining a data set containing periodic change information, selecting data pairs that need to be compared (such as wind speed and generator power) from the data set containing periodic change information, and calculating the Spearman rank correlation coefficient for each pair of data points, which can measure the hierarchical correlation between two variables and is not affected by the specific distribution form of the data. According to the calculated correlation coefficient, the strength and direction of the correlation between the data points (positive correlation, negative correlation or no correlation) are determined.

[0123] Step 135, organize all potential outliers identified in step 132 and their related information, extract the main periodic fluctuation characteristics, such as cycle length, fluctuation amplitude, etc. based on the Fourier analysis results in step 134, extract significantly correlated data point pairs and their correlation coefficients according to the correlation calculation results in step 134, integrate the above extracted information and features together to form a comprehensive data set that can represent the key characteristics and behaviors of the wind turbine equipment and environmental status.

[0124] Step 136, based on historical experience, set rules for identifying abnormal status of equipment, apply the set rules to the key information and feature data extracted in step 135, determine whether the equipment is in an abnormal state, and output the identified abnormal status of the equipment and its related information as a report or alarm.

[0125] In a preferred embodiment of the present invention, in the above step 134, the calculation formula of the Spearman rank correlation coefficient is:

[0126]

[0127] Among them, r s represents the Spearman rank correlation coefficient; 6 represents the scaling factor; Q x( (j) and Q Y (j) represents x j and j The rank order in the respective dataset; M represents the number of data points; j represents the index variable; and m represents the total number of data points.

[0128] In an embodiment of the present invention, an innovative Spearman-level correlation coefficient calculation formula is used in the data analysis process. By introducing a scaling factor, the linear relationship strength and direction between different data can be quantified more accurately, which not only improves the accuracy of data correlation analysis, but also provides a more reliable basis for the identification of abnormal equipment status. By calculating the Spearman-level correlation coefficient between different data, and combining time series graphs, periodic change analysis results, etc., the correlation between data can be fully and deeply explored to help operation and maintenance personnel quickly locate potential abnormal sources and fault points. This precise data correlation analysis method significantly improves the accuracy and efficiency of wind turbine inspections and reduces the risk of misjudgment and missed judgments.

[0129] In a preferred embodiment of the present invention, in the above step 14, the calculation formula of the adjustment factor is:

[0130]

[0131] Among them, p i is the frequency of occurrence of the ith abnormal event, q i is the average duration of the ith abnormal event, v i is the impact range of the ith abnormal event, u i is the basic weight coefficient of the ith abnormal event, m i is the severity coefficient of the ith abnormal event, n i is the complexity coefficient of the ith abnormal event, τ is the time decay coefficient, N is the total number of abnormal event types considered, T m is the current time, T t is the recording time of the ith abnormal event data, k, b, h are exponential parameters for duration and impact range, n max is the maximum complexity score among all abnormal events, and i is the index.

[0132] In the embodiment of the present invention, the formula comprehensively considers multiple factors, such as the frequency of occurrence, average duration, impact range, basic weight coefficient, severity coefficient and complexity score of abnormal events, so that the impact of various abnormal events can be more accurately evaluated and adjusted, which helps to more accurately identify and solve key problems and improve the overall performance of the system. The time decay coefficient is introduced in the formula so that the adjustment factor can change dynamically over time. This means that the system can automatically adjust the response to abnormal events according to actual conditions and better adapt to the changing environment and needs. By comprehensively considering multiple abnormal events and weighting them according to their respective characteristics (such as duration, impact range, etc.), a comprehensive abnormal event evaluation framework can be provided, which helps to fully understand the system status and formulate more comprehensive maintenance and optimization strategies. Parameters such as the exponential parameter (for duration and impact range) and the maximum value of the complexity score in the formula can be adjusted according to actual needs. By accurately calculating the adjustment factor, resources can be allocated more effectively to deal with various abnormal events. For example, for abnormal events with high frequency and wide impact range, more resources can be allocated for prevention and processing, thereby improving resource utilization efficiency and system stability.

[0133] In a preferred embodiment of the present invention, the above step 15 includes:

[0134] Step 151, conduct a preliminary analysis of the abnormal state of the equipment, and classify it according to various parts, equipment, and components to obtain data analysis results for each category;

[0135] Step 152, identifying the urgency of the data analysis results;

[0136] Step 153: According to the urgency, the adjustment factor is multiplied by the preliminary abnormal urgency to obtain a final abnormal urgency.

[0137] In an embodiment of the present invention, by performing a preliminary analysis of the abnormal state of the equipment and classifying it according to various parts, equipment, and components, the specific location where the fault may occur can be located more quickly. This classification analysis method helps to reduce the time for troubleshooting and improve the work efficiency of maintenance personnel. Identifying the urgency of the data analysis results, introducing an adjustment factor to multiply the preliminary abnormal urgency, makes the assessment of urgency more flexible and accurate. Through the calculation of the final abnormal urgency, through a detailed analysis of the abnormal state of the equipment and an assessment of the urgency, potential problems can be discovered and resolved in a timely manner, thereby improving the overall reliability and operating efficiency of the equipment. This helps to reduce equipment downtime and improve production efficiency.

[0138] When specifically applied, this includes:

[0139] Step 151, conduct an in-depth analysis of the abnormal state of the equipment in step 13, and classify the data according to the various parts of the wind turbine (such as the nacelle, blades, tower, etc.), equipment (such as generators, gear boxes, converters, etc.) and components (such as bearings, sensors, etc.) according to the source, type and characteristics of the data. For each type of data, extract key information and feature data, such as temperature, humidity, vibration frequency, image features, etc., to form the analysis results of this type of data.

[0140] Step 152: Based on historical experience, conduct an urgency assessment for each type of data analysis result obtained in step 151. For each urgency determination result, determine its scope of impact, including the affected equipment, components, areas, and possible consequences. According to the importance of urgency and scope of impact, set a weight for each factor (such as urgency, size of the affected area, potential losses, etc.), and the weight reflects the importance of the factor in the overall evaluation. Develop specific scoring criteria, such as using a five-level scoring system (1-5 points) or a more detailed scoring range (such as 1-10 points) to quantitatively evaluate each factor. According to the scoring criteria, assign different scores to each urgency and scope of impact factor, and record the urgency information obtained from the assessment.

[0141] Step 153, multiplying the preliminary abnormal urgency score obtained by the evaluation in step 152 by the adjustment factor, and the final abnormal urgency calculated reflects the comprehensive urgency after considering multiple factors.

[0142] In a preferred embodiment of the present invention, the above step 16 includes:

[0143] Step 161, determining the extent of the fault according to the abnormal urgency, predicting the data trend within a preset period of time in the future, and obtaining a prediction result;

[0144] Step 162, based on the degree of the fault and the prediction result, real-time monitoring is performed on the fault area, and the collected real-time data is continuously transmitted back to the inspection intelligent system;

[0145] Step 163, the control system processes and analyzes the received data to evaluate the current fault condition and issue an alarm message.

[0146] In an embodiment of the present invention, by defining a scoring standard and scoring each urgency, the present invention realizes a quantitative evaluation of abnormal phenomena, so that different abnormal phenomena can be compared. This quantitative evaluation method helps operation and maintenance personnel to quickly identify the most important abnormal phenomena and handle them with priority. The scores of each urgency and impact range are added to obtain a total score of the data results, and the results are sorted from high to low according to the total score to obtain a priority ranking result of the abnormal phenomena. This helps operation and maintenance personnel to reasonably allocate resources, give priority to abnormal phenomena with high urgency and high impact difficulty, and improve operation and maintenance efficiency and safety.

[0147] When specifically applied, this includes:

[0148] Step 161, conduct a comprehensive analysis of the urgency and impact scope of each task to determine the extent of the fault, which can be divided into multiple levels, such as minor faults, general faults, serious faults and emergency faults. Each level corresponds to a different emergency response strategy and priority. For faults of different levels, a detailed emergency response plan is formulated, including fault handling procedures, required resources, responsible personnel, etc. Use a time series analysis algorithm to analyze the processed wind turbine equipment and environmental status information, and predict data trends within a preset period of time in the future (such as a few hours, days or weeks). The prediction content can include the changing trends of key equipment parameters (such as temperature, vibration frequency, current, etc.), as well as equipment failures or abnormal conditions that may be caused by these parameters. Based on the prediction results, evaluate the development trend and potential risks of the fault.

[0149] Step 162: For the identified fault area, use inspection tools such as rail-hanging robots, ground monitoring equipment or robots to conduct real-time monitoring. The real-time monitoring content should include video images, temperature, humidity, vibration frequency, current and other data to fully reflect the status of the fault area. The collected real-time data is transmitted back to the inspection intelligent system.

[0150] Step 163, the patrol intelligent system pre-processes and analyzes the received real-time data to extract key information and feature data, and the control system of the patrol intelligent system performs in-depth processing and analysis on the received real-time data to evaluate the current fault situation. The analysis content may include the development speed of the fault, the scope of impact, potential risks, etc., as well as the impact of the fault on the overall operation of the wind turbine. According to the evaluation results, the control system automatically generates alarm information, which should include the specific location, type, urgency, predicted trend, and recommended emergency response measures of the fault. The alarm information is sent to the operation and maintenance personnel through various channels (such as SMS, email, system interface prompts, sound alarms, etc.) to ensure that they can receive the alarm in time and take corresponding measures. At the same time, the patrol intelligent system can also record the alarm information in the log for subsequent tracking and analysis.

[0151] like Figure 2 As shown, an embodiment of the present invention further provides an intelligent early warning system 20 for a wind turbine robot inspection system, comprising:

[0152] Determine module 21, determine the robot inspection route, and obtain wind turbine equipment and environmental status information through the robot;

[0153] A preprocessing module 22 is used to preprocess the fan equipment and environmental status information to obtain processed fan equipment and environmental status information;

[0154] The analysis module 23 analyzes the processed wind turbine equipment and environmental status information to extract key information and characteristic data; identifies the abnormal status of the equipment based on the key information and characteristic data; obtains the historical operation data of the wind turbine equipment, including the frequency, duration and impact range of various abnormal events, and calculates the adjustment factor based on the historical operation data of the wind turbine equipment;

[0155] Module 24 is executed to perform a preliminary analysis on the abnormal state of the equipment to obtain a preliminary abnormal urgency; the adjustment factor is multiplied by the preliminary abnormal urgency to obtain a final abnormal urgency; and emergency response warning is executed according to the abnormal urgency, including continuous monitoring and issuing alarm information.

[0156] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. An intelligent early warning method for a wind turbine robot inspection system, characterized in that: The method comprises: Step 11, determine the robot inspection route, and obtain wind turbine equipment and environmental status information through the robot; Step 12, preprocessing the fan equipment and environmental status information to obtain processed fan equipment and environmental status information; Step 13, analyzing the processed wind turbine equipment and environmental status information to extract key information and feature data, and identifying abnormal equipment status based on the key information and feature data; Step 14, obtaining historical operation data of the wind turbine equipment, including the occurrence frequency, duration and impact range of various abnormal events, and calculating the adjustment factor according to the historical operation data of the wind turbine equipment; Step 15, performing a preliminary analysis on the abnormal state of the device to obtain a preliminary abnormal urgency; multiplying the adjustment factor by the preliminary abnormal urgency to obtain a final abnormal urgency; Step 16, execute emergency response warning according to the abnormal urgency, including continuous monitoring and issuing alarm information.

2. The intelligent early warning method for a wind turbine robot inspection system according to claim 1 is characterized in that: Determine the robot inspection route and use the robot to obtain wind turbine equipment and environmental status information, including: Generate an initial set of candidate inspection paths based on the physical layout, criticality, accessibility, and safety requirements of the wind turbine equipment; Set the initial parameters of the genetic algorithm, including population size, crossover rate, mutation rate, and number of iterations; Encode each inspection path as an individual of a genetic algorithm, and each individual represents a possible inspection path; generate the initial population of the genetic algorithm based on the initial inspection path candidates; Define a fitness function to evaluate the quality of each inspection path; According to the fitness function, the corresponding individuals are selected from the current population as the parents to generate the next generation; through the crossover operation, new inspection path individuals are generated; the newly generated individuals are randomly mutated, and the selection, crossover and mutation operations are repeated until the preset number of iterations is reached to obtain the corresponding individuals, that is, the final inspection path; According to the final inspection route, the robot is controlled to perform inspection tasks. During the inspection process, the status information of the wind turbine equipment and the surrounding environment is collected in real time, including wind turbine equipment images, temperature, humidity, vibration, sound, wind speed, wind direction, atmospheric pressure, equipment voltage, current, oil pressure, oil temperature and oil level of bearings and gearboxes, yaw and pitch angles, stress and strain data of towers and blades, smoke inside the equipment, settlement of soil and foundation, and displacement monitoring.

3. The intelligent early warning method for the wind turbine robot inspection system according to claim 2 is characterized in that: Based on the physical layout, criticality, accessibility, and safety requirements of the wind turbine equipment, an initial set of candidate inspection paths is generated, including: According to the physical layout of wind turbine equipment, the relative position and distance between equipment are determined; each wind turbine equipment is regarded as a node in the graph, and the feasible paths between equipment are regarded as edges; according to the actual distance between equipment, movement difficulty and terrain factors, a weight value is assigned to each edge, representing the cost of passing the corresponding path; Identify and mark unreachable or dangerous areas, select a starting node, which is the starting position of the inspection task; initialize the distance array, record the shortest distance from the starting node to each node in the graph, initially set the starting node to 0, and the remaining nodes to infinity; create an empty set to save the nodes for which the shortest path has been found; use a minimum heap to create a priority queue, and the elements in the queue are sorted according to the distance from the starting node to the node, and initially only the starting node is in the queue; Take the node closest to the starting node from the priority queue. For each adjacent node of the node, calculate the distance from the starting node through the current node to the adjacent node. If the distance is less than the previously recorded distance, update the distance array and add the adjacent node to the priority queue. Mark the current node as having found the shortest path, and add it to the set of processed nodes, repeating the above steps until the priority queue is empty, that is, the shortest paths to all reachable nodes have been found; According to the distance array, the shortest path from the starting node to each wind turbine device is determined; according to the total length of the path, the number of key devices covered and the safety, the generated inspection path is evaluated to obtain the evaluation result; According to the evaluation results, multiple corresponding inspection paths are selected as an initial inspection path candidate set, and the initial inspection path candidate set includes information about nodes and edges passed on the path.

4. The intelligent early warning method for a wind turbine robot inspection system according to claim 3 is characterized in that: The calculation formula of the fitness function is: Among them, F is the fitness value of an individual, D is the total length of the inspection path, a is the exponent of the length factor. When a > 1, the penalty for longer paths will increase; when 0 < a < 1, the penalty for longer paths will decrease; W D is the weight coefficient of the path length; K is the number of key devices covered by the inspection path, K max is the total number of key devices, g is the exponent of the key device coverage factor, W K is the weight coefficient of the key device coverage rate; S is the safety score of the inspection path, S max is the highest score of safety, γ is the exponent of the safety factor, W S is the weight coefficient of safety, R represents the reachability score of the path, W R is the weight coefficient of reachability, C is the continuity score of the path, W C is the weight coefficient of continuity.

5. The intelligent early warning method for a wind turbine robot inspection system according to claim 4 is characterized in that: Analyze the processed wind turbine equipment and environmental status information to extract key information and feature data; Identify abnormal equipment status based on key information and feature data, including: Analyze the processed fan equipment and environmental status information, calculate and summarize the basic statistics of each data point, including the mean, standard deviation, maximum and minimum values; According to the historical data distribution, set corresponding thresholds for each state information, and compare the statistics of each data point with the set threshold one by one to identify potential outliers; Using potential outliers, construct time series graphs to initially identify the existence of cyclical fluctuations; Through Fourier analysis, the periodic components in the data are extracted, and the Spearman rank correlation coefficient between each data point is calculated using the data set containing periodic change information. The Spearman rank correlation coefficient is used to quantify the strength and direction of the correlation between different data points. Extract key information and characteristic data representing the status of wind turbine equipment and environment based on abnormal values, periodic fluctuations and correlation strength and direction; Identify abnormal status of equipment based on key information and feature data.

6. The intelligent early warning method for a wind turbine robot inspection system according to claim 5 is characterized in that: The calculation formula of Spearman rank correlation coefficient is: Among them, r s represents the Spearman rank correlation coefficient; 6 represents the scaling factor; Q x( (j) and Q Y (j) represents x j and j The ranking level in the respective dataset; M represents the number of data points; j represents the index variable; m Represents the total number of data points.

7. The intelligent early warning method for a wind turbine robot inspection system according to claim 6 is characterized in that: The adjustment factor is calculated as: Among them, p i is the frequency of occurrence of the ith abnormal event, q i is the average duration of the ith abnormal event, v i is the impact range of the ith abnormal event, u i is the basic weight coefficient of the ith abnormal event, m i is the severity coefficient of the ith abnormal event, n i is the complexity score of the ith abnormal event, τ is the time decay coefficient, N is the total number of abnormal event types considered, T m is the current time, T t is the recording time of the ith abnormal event data, k, b, h are exponential parameters for duration and impact range, n max is the maximum complexity score among all abnormal events, and i is the index.

8. An intelligent early warning system for a wind turbine robot inspection system, characterized in that: The system is used to perform the method according to any one of claims 1 to 7, comprising: Determine the module, determine the robot inspection route, and obtain wind turbine equipment and environmental status information through the robot; A preprocessing module is used to preprocess the fan equipment and environmental status information to obtain processed fan equipment and environmental status information; The analysis module analyzes the processed wind turbine equipment and environmental status information to extract key information and characteristic data; identifies abnormal equipment status based on the key information and characteristic data; obtains historical operation data of the wind turbine equipment, including the frequency, duration and impact range of various abnormal events, and calculates adjustment factors based on the historical operation data of the wind turbine equipment; The execution module performs a preliminary analysis of the abnormal state of the equipment to obtain a preliminary abnormal urgency; multiplies the adjustment factor by the preliminary abnormal urgency to obtain a final abnormal urgency; and executes emergency response warnings based on the abnormal urgency, including continuous monitoring and issuing alarm information.

9. A computing device, characterized in that include: one or more processors; A storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the method as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program, which, when executed by a processor, implements the method according to any one of claims 1 to 7.

Images