Methods and systems for operational inspections and trend analysis of power plant equipment.

Abstract

The present application relates to the technical field of data processing, and provides a method and a system for the operation patrol inspection and trend analysis of power plant facilities. In the process of the patrol inspection camera's patrol, the video capture device is activated to obtain the real-time video information of the power plant facilities, which is input into the image processing recognition model to obtain the equipment status recognition information. Connected to the target power plant facilities, upload data is obtained for equipment status prediction, and the equipment status prediction result is obtained. A patrol inspection task is set, and the patrol inspection camera is made to perform an automatic patrol inspection on the target power plant facilities according to the patrol inspection task, thereby solving the technical problems that the efficiency of the operation patrol inspection of the power plant facilities is low and the accuracy is not high, and the reliability of the operation patrol inspection information of the power plant facilities is low. A recognition model is constructed to improve the accuracy, the patrol inspection route in the patrol inspection task of the patrol inspection camera is optimized by using the equipment status prediction, and an automatic patrol inspection is performed according to the patrol inspection task, so that the technical effect of improving the operation patrol inspection efficiency and maintaining the reliability of the operation patrol inspection information can be realized.

Description

Technical Field

[0006] , , , , ,

[0005]

[0001] This application relates to the technical field related to data processing, and specifically relates to a method and system for the operation patrol inspection and trend analysis of power plant equipment.

Background Art

[0002] In recent years, with the large-scale popularization of unmanned power plants in the power grid, smart patrol cameras have gradually replaced manual labor to conduct patrol inspections of power plant equipment. Most of the instrument equipment to be patrolled is pointer-type instrument equipment, such as air pressure gauges, thermometers, oil temperature gauges, lightning arrester instruments, etc. This type of instrument has a simple structure, low manufacturing cost, convenient maintenance, strong electromagnetic interference resistance, high reliability, and has the characteristics of dust-proof, waterproof, and freeze-proof, and is widely applied in the power system.

[0003] Since most of the instrument equipment is pointer-type instrument equipment, the output result is not a digital signal and cannot be directly input into the computer system. The recognition of the conventional instrument indication is mostly by manual discrimination. The recognition by human eyes is easily affected by many human factors, with low detection efficiency, low accuracy, and there are also cases where manual reading cannot be performed, such as those with high radiation or high danger.

[0004] From the above, there are technical problems in the prior art that the operation patrol inspection efficiency of power plant equipment is low and the accuracy is not high, and the reliability of the information on the operation patrol inspection of power plant equipment is low.

Summary of the Invention

Problems to be Solved by the Invention

[0005] In order to solve the technical problems in the prior art that the operation patrol inspection efficiency of power plant equipment is low and the accuracy is not high, and the reliability of the information on the operation patrol inspection of power plant equipment is low, this application provides a method and system for the operation patrol inspection and trend analysis of power plant equipment.

Means for Solving the Problems

[0006] In view of the above issues, the embodiment of the present invention provides a method and system for operational inspection and trend analysis of power plant equipment.

[0007] One aspect of the present disclosure provides a method for operational patrol inspection and trend analysis of power plant equipment, the method being applied to a system for operational patrol inspection and trend analysis, the system for operational patrol inspection and trend analysis and a patrol inspection camera being interconnected by communication, the method including activating a video capture device and acquiring real-time video information of the power plant equipment during the process of the patrol inspection camera making its rounds, inputting the real-time video information of the power plant equipment into an image processing recognition model and acquiring equipment state recognition information, connecting to the target power plant equipment and acquiring uploaded data, performing equipment state prediction using the uploaded data based on the equipment state recognition information and acquiring equipment state prediction results, setting a patrol inspection task based on the equipment state prediction results, and causing the patrol inspection camera to perform an automatic patrol inspection of the target power plant equipment in accordance with the patrol inspection task.

[0008] Another aspect of this disclosure provides a system for operational patrol inspection and trend analysis of power plant equipment, the system comprising: a video information acquisition module for activating video capture equipment and acquiring real-time video information of power plant equipment during the process of a patrol inspection camera patrolling; an equipment state recognition module for inputting the real-time video information of the power plant equipment into an image processing recognition model and acquiring equipment state recognition information; an upload data acquisition module for connecting to the target power plant equipment and acquiring upload data; an equipment state prediction module for performing equipment state prediction based on the equipment state recognition information and the upload data and acquiring equipment state prediction results; a patrol inspection task setting module for setting a patrol inspection task based on the equipment state prediction results; and an automatic patrol inspection module for causing the patrol inspection camera to perform an automatic patrol inspection of the target power plant equipment in accordance with the patrol inspection task. [Effects of the Invention]

[0009] One or more technical means provided in this application have at least the following technical effects or advantages:

[0010] This invention provides a technical effect in which, during the process of a mobile inspection camera making its rounds, a video capture device is activated to acquire real-time video information of the power plant equipment, the real-time video information of the power plant equipment is input into an image processing recognition model to acquire equipment status recognition information, a connection is made to the target power plant equipment, uploaded data is acquired, equipment status prediction is performed based on the equipment status recognition information and the uploaded data is used, the equipment status prediction results are acquired, a mobile inspection task is set, and the mobile inspection camera is made to perform an automatic mobile inspection of the target power plant equipment according to the mobile inspection task. This builds a recognition model and improves its accuracy, optimizes the mobile inspection route within the mobile inspection task of the mobile inspection camera using the equipment status prediction, performs an automatic mobile inspection according to the mobile inspection task, improves the efficiency of operational mobile inspections, and maintains the reliability of operational mobile inspection information.

[0011] The above description is merely an overview of the technical means of this application. In order to better understand the technical means of this application, to implement them in accordance with the specifications, and to better understand the above and other purposes, features, and advantages of this application, specific embodiments of this application are listed below. [Brief explanation of the drawing]

[0012] [Figure 1] This is a possible flowchart of the method for operational inspection and trend analysis of power plant equipment provided in the embodiment of the present application. [Figure 2] This flowchart allows for updating the inspection route in the method for operational inspection and trend analysis of power plant equipment provided in the embodiment of the present invention. [Figure 3] This is a possible schematic diagram of a system for operational inspection and trend analysis of power plant equipment provided in an embodiment of the present application. [Modes for carrying out the invention]

[0013] The embodiments of this application provide a method and system for operational patrol inspection and trend analysis of power plant equipment, thereby solving the technical problems of low efficiency and low accuracy in operational patrol inspections of power plant equipment, and low reliability of operational patrol inspection information for power plant equipment. By constructing a recognition model to improve accuracy, optimizing the patrol inspection route in the patrol inspection task of the patrol inspection camera using equipment state prediction, and performing automated patrol inspections according to the patrol inspection task, the invention can achieve technical effects that improve the efficiency of operational patrol inspections and maintain the reliability of operational patrol inspection information.

[0014] After introducing the basic principles of this invention, various non-limiting embodiments of this invention will be specifically described below, in conjunction with the attached drawings.

[0015] Example 1

[0016] As shown in Figure 1, an embodiment of the present invention provides a method for operational inspection and trend analysis of power plant equipment, the method being applied to an operational inspection and trend analysis system, the operational inspection and trend analysis system and an inspection camera being interconnected via communication, and the method comprising the following:

[0017] S10: During the process of the patrol inspection camera making its rounds, the video capture equipment is activated to acquire real-time video information of the power plant equipment.

[0018] S20: Real-time video information of the power plant equipment is input to the image processing and recognition model to obtain equipment status recognition information.

[0019] Step S20 includes the following steps:

[0020] S21: In the past, when the target power plant equipment is in a different equipment state, multiple sample image information is extracted in a synchronized manner.

[0021] S22: Based on the above-mentioned multiple sample image information, a corresponding set of multiple sample equipment state recognition information is identified.

[0022] S23: Employ the plurality of sample image information and the plurality of sample equipment status recognition information to construct the image processing recognition model.

[0023] S24: Input the real-time video information of the power plant equipment into the image processing recognition model to obtain the equipment status recognition information.

[0024] The operation patrol inspection and trend analysis system and the patrol inspection camera are communicatively interconnected. Specifically, the communicative interconnection is, simply put, to communicate through signal transmission, and a communication network is configured between the operation patrol inspection and trend analysis system and the patrol inspection camera. During the process of the patrol inspection camera's patrol (various sensors such as a visible light sensor, an infrared sensor, and an audio sensor are attached to the patrol inspection camera), the video capture equipment is activated, and the audio and image information of the target power plant equipment is captured using the video capture equipment to obtain the real-time video information of the power plant equipment, providing data support for subsequent processing.

[0025] Inputting the real-time video information of the power plant equipment into the image processing recognition model to obtain the equipment status recognition information specifically includes: based on the operation patrol inspection and trend analysis system, when the target power plant equipment was in different equipment statuses at a previous time, synchronously extracting a plurality of sample image information (the plurality of sample image information corresponds one-to-one with different equipment statuses); based on the plurality of sample image information, identifying the corresponding plurality of sample equipment status recognition information from the operation log of the target power plant equipment; employing the plurality of sample image information and the plurality of sample equipment status recognition information to construct the image processing recognition model; inputting the real-time video information of the power plant equipment into the image processing recognition model as input data, and outputting the equipment status recognition information (the equipment status recognition information includes notation indication information and the display status of the enclosure), providing a reference for subsequent substitution into the image processing recognition model.

[0026] Step S23 includes the following steps.

[0027] S231: The image processing and recognition model is constructed using a convolutional neural network as the basis of the model. The input data for the image processing and recognition model is real-time video information of the power plant equipment, and the output data is equipment status recognition information.

[0028] S232: The above-mentioned sample image information and sample equipment state recognition information are taken and decomposed to obtain a training information set and a verification information set.

[0029] S233: Set the first model accuracy threshold.

[0030] S234: Using the training information set and the validation information set, supervised learning and validation are performed on the image processing recognition model until the accuracy of the image processing recognition model satisfies the model accuracy threshold, thereby acquiring the image processing recognition model.

[0031] Specifically, constructing the image processing and recognition model by employing the aforementioned multiple sample image information and multiple sample equipment state recognition information involves constructing the image processing and recognition model using a convolutional neural network as the basis of the model (the input data at the input end of the image processing and recognition model is real-time video information of the power plant equipment, and the output data at the output end of the image processing and recognition model is equipment state recognition information), dividing and decomposing the multiple sample image information and multiple sample equipment state recognition information in an 8:2 ratio to obtain a training information set and a validation information set, setting a first model accuracy threshold, and inputting the multiple sample image information and multiple sample equipment state recognition information in the training information set into the convolutional neural network as input training data, and each time training results and predicted results (predicted results: multiple) are obtained. The process includes performing error analysis from a number of sample equipment state recognition information, further modifying the weight values ​​and thresholds (by further modifying the weight values ​​and thresholds, the convolutional neural network can be trained until it can be applied to equipment state recognition), performing supervised learning step by step to obtain a model that can output results that match the expected results, and after the model output has stabilized (model stabilization: output matches expected results), performing validation using multiple sample image information and multiple sample equipment state recognition information from the validation information set as inputs to calculate the image processing recognition model accuracy (image processing recognition model accuracy = number of validation passes / total number of validations × 100%), and if the image processing recognition model accuracy satisfies the first model accuracy threshold, the validation is passed, and the image processing recognition model is determined and the model is provided to support equipment state recognition.

[0032] Step S231 further includes the following steps.

[0033] S231-1: A convolutional neural network is used as the basis for the model to implement a notation level recognition algorithm and an indicator light recognition algorithm.

[0034] S231-2: The above-mentioned display indicator recognition algorithm is used to recognize the display indicator information.

[0035] S231-3: The indicator light recognition algorithm described above is used to recognize the color of the indicator light on the enclosure.

[0036] S231-4: The equipment status recognition information is recognized and output based on the indicated information and the color of the indicator light on the enclosure.

[0037] Specifically, the inspection cameras capture images of various instruments (voltmeters, ammeters, digital display instruments, and knife-gate instruments, etc.) from within the target power plant. In the analysis process of the instrument images, instrument digital recognition belongs to scene character recognition. Scene character recognition is performed in natural scenes and has characteristics such as relatively complex backgrounds, diverse fonts, variable size, variable shape, and irregular orientation, making recognition difficult.

[0038] In establishing the indicator light recognition algorithm based on a convolutional neural network model, the color of the enclosure's indicator light corresponds one-to-one with the enclosure's display state. Following the main equipment usage instructions for the enclosure's indicator light, the display state of the enclosure is confirmed by querying the color of the enclosure's indicator light. The doubleparam2 parameter (param2 is a variable within the function, double: double-precision floating-point type) in the HoughCircles() function (HoughCircles is the function name) is modified to reduce the recognition roundness criterion, and the indicator light area in question is processed independently. Using an iterative approach, the magnitude of the parameter value is gradually reduced until a circle is recognized. Setting the param2 value too low from the beginning would result in the detection of many circles that may not exist, causing misrecognition and outputting indicator light states of different colors, so this should not be done.

[0039] In setting up a notation recognition algorithm based on a convolutional neural network model, noise is removed from the input data of the image processing recognition model, providing a clear image for accurate extraction of the center line of the subsequent pointer, a smoothing window is set (if the smoothing window is too large or too small, details of the instrument image, such as boundary contours and lines, will become blurred or the boundary contours will break. In this embodiment, a 5x5 square smoothing window is used), and image noise is removed by a median filter of the smoothing window. This achieves the purpose of noise reduction while also preserving detailed information in the image, thereby enabling the positioning of the pointer. Generally, the pointer in an instrument image has the characteristics of being thin at the top and thick at the bottom, with the gray gradation being symmetrical with respect to the center line, and the center line of the pointer always passes through the axis of rotation. The positioning of the pointer can be achieved by extracting the center line of the pointer that passes through the axis of rotation of the dial, and a linear extraction method can be used to extract the center line of the pointer. The recognition of the indication of a pointer instrument is achieved by using the relationship between the angle of the center line of the pointer and the initial scale of the instrument's measurement range and the measurement range to recognize the instrument's indication.

[0040] Taking a pressure gauge dial as an example, its measurement range is (Min, Max), and to calculate the dial's reading, a Cartesian coordinate system XY and an image space coordinate system UV are constructed with the dial's rotation axis as the origin. The image pixel coordinate system UV is converted to the dial coordinate system XY, where the coordinate origin O is located at the pixel point (m,n) of the dial's rotation axis, OF is the half-line where the dial's pointer is located, OI is the half-line from the dial's rotation axis to the dial's initial scale, OA is the half-line from the dial's rotation axis to the dial's maximum scale, and the range of angle values ​​obtained by rotating clockwise from the negative X-axis direction to the half-lines OA, OI, and OF is (-180°, 180°), where clockwise rotation is a positive angle and counterclockwise rotation is a negative angle.

[0041] In recognizing the indication information by employing the aforementioned indication recognition algorithm, the indication information includes character state, color state, and pointer direction. The indicator light recognition algorithm is used to recognize the color of the indicator light of the enclosure. Based on the indication information and the color of the indicator light of the enclosure, the device status is recognized and output according to the usage instructions for the main equipment of the indicator light of the enclosure, and technical support is provided to ensure the accuracy of the equipment status recognition.

[0042] S30: Connect to the target power plant equipment and retrieve the uploaded data.

[0043] S40: Based on the equipment status recognition information, equipment status prediction is performed using the uploaded data, and the equipment status prediction result is obtained.

[0044] Step S40 includes the following steps:

[0045] S41: Construct a chain for predicting equipment condition.

[0046] S42: The equipment status recognition information is used as the current status, the uploaded data is input into the equipment status prediction chain, and equipment status prediction is performed.

[0047] S43: The equipment condition prediction results are predicted and obtained using the equipment condition prediction chain described above.

[0048] Specifically, the system connects to the target power plant equipment using a communication interconnection method to acquire uploaded data, the uploaded data includes unanswered control commands of the system (unanswered control commands are stored in registers while the response to multiple parallel control commands is being executed), and based on the equipment state recognition information, the system performs equipment state prediction using the uploaded data and acquires the equipment state prediction result. Specifically, this includes constructing an equipment state prediction chain, using a Markov chain as the basis of the model, using the equipment state recognition information as the current state, inputting the uploaded data into the equipment state prediction chain, predicting the next equipment state (the next equipment state is the equipment state following the current state), acquiring the equipment state prediction result, and providing a reference for substitution into subsequent equipment state prediction chains.

[0049] Step S41 includes the following steps.

[0050] S411: Construct the equipment state prediction chain based on a Markov chain as the model.

[0051] S412: Multiple sample upload data are used as reward information, and multiple sample equipment state recognition information is set as state information to construct an initialization equipment state prediction chain.

[0052] S413: The aforementioned multiple Sample Upload Data Then, data is randomly extracted from multiple sample equipment status recognition information to construct a set of verification information.

[0053] S414: Set the second model accuracy threshold.

[0054] S415: The initial equipment state prediction chain is verified using the verification information set until the initial equipment state prediction chain accuracy satisfies the model accuracy threshold, and the equipment state prediction chain is obtained.

[0055] Building an equipment state prediction chain based on a Markov chain specifically involves using a Markov chain as the basis of the model, setting multiple sample equipment state recognition information as state information, and constructing an initialization equipment state prediction chain, with a 10% ratio of the above multiple Sample Upload Data The process involves randomly extracting data from multiple sample equipment state recognition information (random extraction is a conventional technique) to construct a verification information set, setting a second model accuracy threshold, and multiple verification information sets. Sample Upload Data This includes verifying the initial equipment state prediction chain using multiple sample equipment state recognition information until the initial equipment state prediction chain accuracy (prediction chain accuracy = number of successful verifications / total number of verifications × 100%) satisfies the model accuracy threshold, obtaining the equipment state prediction chain, and providing model support for equipment state prediction.

[0056] S50: Based on the equipment condition prediction results, set up a routine inspection task.

[0057] S60: The patrol inspection camera is instructed to perform an automated patrol inspection of the target power plant equipment in accordance with the patrol inspection task.

[0058] As shown in Figure 2, step S50 includes the following steps.

[0059] S51: Based on the equipment condition prediction results, a route plan for routine inspections is obtained.

[0060] S52: During the process of the mobile inspection camera making its rounds, an initial mobile inspection route plan is acquired.

[0061] S53: The initial patrol inspection route plan is modified based on the patrol inspection route plan, and the modified patrol inspection route plan is obtained.

[0062] S54: Add the revised inspection route plan to the inspection task and update the inspection route.

[0063] Setting up a patrol inspection task based on the equipment condition prediction results specifically involves obtaining a patrol inspection route plan based on the equipment condition prediction results (indicator lights and pointer instruments of the same equipment are preferentially packaged, and patrol inspections are preferentially performed on equipment with large discrepancies between the equipment condition prediction results and equipment condition recognition information, and a large discrepancy between the equipment condition prediction results and equipment condition recognition information means that there is a large jump in the data and there is a possibility that there is a risk of operational abnormality of the power plant equipment), obtaining an initial patrol inspection route plan during the process of the patrol inspection camera patrolling (the initial patrol inspection route plan is set in accordance with the optimization of the patrol inspection route, so that the patrol inspection route of the initial patrol inspection route plan is the shortest), and setting up a patrol inspection route plan based on the previous This includes modifying the initial patrol inspection route plan (packaging operational patrol inspections of the same equipment and prioritizing patrol inspections of equipment within the target power plant equipment that may have a risk of operational abnormalities), obtaining a revised patrol inspection route plan (simply put, the coordinates where there may be a risk of operational abnormalities of power plant equipment are (5,8), and the starting point is generally (1,1) by default, but the patrol inspection starting point may be modified to (5,8), and the revised patrol inspection route plan may start the patrol inspection from (5,8) and finally inspect the power plant equipment between (1,1) and (5,8)), and updating the patrol inspection route by adding the revised patrol inspection route plan to the patrol inspection task in order to ensure the rationality of the patrol route of the patrol inspection camera.

[0064] As described above, the method and system for operational inspection and trend analysis of power plant equipment provided in the embodiment of the present application have the following technical effects.

[0065] 1. This invention provides a method and system for operational patrol inspection and trend analysis of power plant equipment, which involves activating video capture equipment during the process of a patrol inspection camera making its rounds, acquiring real-time video information of the power plant equipment, inputting the real-time video information of the power plant equipment into an image processing recognition model, acquiring equipment status recognition information, connecting to the target power plant equipment, acquiring uploaded data, performing equipment status prediction based on the equipment status recognition information and the uploaded data, acquiring the equipment status prediction results, setting a patrol inspection task, and having the patrol inspection camera perform an automated patrol inspection of the target power plant equipment according to the patrol inspection task. By providing this method and system, it is possible to construct a recognition model and improve its accuracy, optimize the patrol inspection route within the patrol inspection task of the patrol inspection camera using equipment status prediction, perform an automated patrol inspection according to the patrol inspection task, improve the efficiency of operational patrol inspections, and maintain the reliability of operational patrol inspection information.

[0066] 2. Based on the equipment condition prediction results, a patrol inspection route plan is obtained. During the process of the patrol inspection camera's patrol, an initial patrol inspection route plan is obtained, the initial patrol inspection route plan is modified to obtain a revised patrol inspection route plan, and this is added to the patrol inspection task, thereby updating the patrol inspection route. This ensures the rationality of the patrol camera's patrol route.

[0067] Example 2

[0068] Based on the same inventive concept as the method for operational inspection and trend analysis of power plant equipment described in the above-described embodiment, the embodiment of the present application provides a system for operational inspection and trend analysis of power plant equipment, as shown in Figure 3. The aforementioned system,

[0069] During the process of the patrol inspection camera's patrol, a video information acquisition module 100 is activated to acquire real-time video information of the power plant equipment,

[0070] A facility status recognition module 200 inputs real-time video information of the aforementioned power plant equipment into an image processing recognition model and acquires facility status recognition information,

[0071] An upload data acquisition module 300 for connecting to the target power plant equipment and acquiring upload data,

[0072] Based on the aforementioned equipment status recognition information, an equipment status prediction module 400 is used to predict the equipment status using the uploaded data and to obtain the equipment status prediction result.

[0073] Based on the equipment status prediction results, a patrol inspection task setting module 500 is provided for setting patrol inspection tasks,

[0074] The patrol inspection camera is equipped with an automatic patrol inspection module 600 for performing automatic patrol inspections on the target power plant equipment in accordance with the patrol inspection task.

[0075] The aforementioned system further,

[0076] In a previous instance, when the target power plant equipment was in a different equipment state, a sample image information extraction module for synchronously extracting multiple sample image information,

[0077] A sample equipment state recognition information identification module for identifying corresponding sample equipment state recognition information based on the aforementioned plurality of sample image information,

[0078] A processing recognition model construction module for constructing the image processing recognition model by employing the aforementioned multiple sample image information and multiple sample equipment state recognition information,

[0079] The system includes a module for acquiring equipment status recognition information, which inputs real-time video information of the power plant equipment into an image processing and recognition model and acquires the equipment status recognition information.

[0080] The aforementioned system further,

[0081] An image processing and recognition model construction module for constructing the image processing and recognition model based on a convolutional neural network, wherein the input data of the image processing and recognition model is real-time video information of power plant equipment, and the output data is equipment state recognition information.

[0082] An information set decomposition module for acquiring a training information set and a verification information set by taking the above-mentioned multiple sample image information and multiple sample equipment state recognition information and decomposing them,

[0083] A first model accuracy threshold setting module for setting the first model accuracy threshold,

[0084] The system includes an image processing recognition model acquisition module that performs supervised learning and validation on the image processing recognition model using the training information set and validation information set until the accuracy of the image processing recognition model satisfies the model accuracy threshold, thereby acquiring the image processing recognition model.

[0085] The aforementioned system further,

[0086] A recognition algorithm setting module for setting up a notation indication recognition algorithm and an indicator light recognition algorithm, based on a convolutional neural network model,

[0087] A notation indication information recognition module for recognizing notation indication information by employing the aforementioned notation indication recognition algorithm,

[0088] An enclosure indicator light color recognition module for recognizing the color of the enclosure's indicator light by employing the aforementioned indicator light recognition algorithm,

[0089] The system includes an equipment status recognition information output module for recognizing and outputting equipment status recognition information based on the aforementioned indication information and the color of the indicator light of the enclosure.

[0090] The aforementioned system further,

[0091] A prediction chain construction module for building an equipment status prediction chain,

[0092] A device state prediction module for performing device state prediction by inputting the uploaded data into the device state prediction chain, using the aforementioned device state recognition information as the current state,

[0093] The system includes an equipment status prediction result acquisition module for predicting and acquiring equipment status prediction results based on the equipment status prediction chain.

[0094] The aforementioned system further,

[0095] A Markov chain is used as the basis for the model, and a module for constructing the aforementioned equipment state prediction chain is provided.

[0096] An initialization equipment state prediction chain construction module for building an initialization equipment state prediction chain by setting multiple sample equipment state recognition information as state information using multiple sample upload data as reward information, and

[0097] The aforementioned multiple Sample Upload Data A data random extraction module for constructing a set of verification information by randomly extracting data from multiple sample equipment state recognition information,

[0098] A second model accuracy threshold setting module for setting the second model accuracy threshold,

[0099] The system includes an equipment state prediction chain acquisition module for obtaining the equipment state prediction chain, which verifies the initial equipment state prediction chain using the verification information set until the initial equipment state prediction chain accuracy satisfies the model accuracy threshold, and for obtaining the equipment state prediction chain.

[0100] The aforementioned system further,

[0101] Based on the equipment condition prediction results, a patrol inspection route plan acquisition module is provided to acquire a patrol inspection route plan,

[0102] During the process of the aforementioned mobile inspection camera making its rounds, an initialization mobile inspection route plan acquisition module is provided to acquire an initialization mobile inspection route plan,

[0103] A modified patrol route plan acquisition module for obtaining a modified patrol route plan by modifying the initial patrol route plan based on the aforementioned patrol route plan,

[0104] The system includes a patrol inspection route update module for updating the patrol inspection route by adding the modified patrol inspection route plan to the patrol inspection task.

[0105] Any step of the above method may be stored in unrestricted computer memory as a computer instruction or program, and may be called and recognized by an unrestricted computer processor to implement any of the embodiments of the present application, without further restrictions.

[0106] Furthermore, the first or second mentioned above may not only represent sequential relationships but also represent certain concepts and / or indicate that individual or all of the elements may be selected. Clearly, a person skilled in the art can make various modifications and changes to this application without departing from its scope. Therefore, if these modifications and changes to this application fall within the scope of this application and its equivalent art, this application intends to include these modifications and changes. [Explanation of Symbols]

[0107] Video information acquisition module 100, equipment status recognition module 200, upload data acquisition module 300, equipment status prediction module 400, patrol inspection task setting module 500, automatic patrol inspection module 600.

Claims

1. A method for operational inspection and trend analysis of power plant equipment, wherein the method is applied to an operational inspection and trend analysis system, the operational inspection and trend analysis system and an inspection camera are interconnected via communication, and the method is During the process of the aforementioned patrol camera making its rounds, the video capture equipment is activated to acquire real-time video information of the power plant equipment, The process involves inputting real-time video information of the aforementioned power plant equipment into an image processing and recognition model to obtain equipment status recognition information, and Connecting to the target power plant equipment and obtaining the uploaded data, Based on the aforementioned equipment status recognition information, the equipment status is predicted using the uploaded data, and the equipment status prediction result is obtained. Based on the equipment condition prediction results, a routine inspection task is set, The aforementioned patrol inspection camera is instructed to perform an automatic patrol inspection of the target power plant equipment in accordance with the aforementioned patrol inspection task. Includes, Inputting real-time video information of the aforementioned power plant equipment into an image processing and recognition model to acquire equipment status recognition information is, In a previous instance, when the aforementioned power plant equipment was in a different equipment state, multiple sample image information was extracted in a synchronized manner. Based on the aforementioned multiple sample image information, the system identifies corresponding multiple sample equipment state recognition information, The image processing and recognition model is constructed by employing the aforementioned multiple sample image information and multiple sample equipment state recognition information. The process involves inputting real-time video information of the power plant equipment into an image processing and recognition model to acquire the equipment status recognition information, Includes, By employing the aforementioned multiple sample image information and multiple sample equipment state recognition information, constructing the image processing recognition model is possible. The image processing and recognition model is constructed using a convolutional neural network as the basis of the model, wherein the input data of the image processing and recognition model is real-time video information of the power plant equipment, and the output data is equipment state recognition information. The process involves selecting and decomposing the aforementioned multiple sample image information and multiple sample equipment state recognition information to obtain a training information set and a verification information set, First, set the model accuracy threshold, Using the aforementioned training information set and validation information set, supervised learning and validation are performed on the image processing recognition model until the accuracy of the image processing recognition model satisfies the aforementioned model accuracy threshold, thereby obtaining the image processing recognition model. Includes, Building the aforementioned image processing and recognition model using a convolutional neural network as the basis of the model further involves... Based on a convolutional neural network, we will implement a display indicator recognition algorithm and an indicator light recognition algorithm. The aforementioned notation indication recognition algorithm is used to recognize notation indication information, The color of the indicator light on the enclosure is recognized by employing the aforementioned indicator light recognition algorithm, The system recognizes and outputs the equipment status recognition information based on the aforementioned indication information and the color of the indicator light on the enclosure. A method for operational inspection and trend analysis of power plant equipment, characterized by including the following.

2. Based on the aforementioned equipment status recognition information, equipment status prediction is performed using the uploaded data, and the equipment status prediction results are obtained. To construct an equipment condition prediction chain, The equipment status recognition information is used as the current status, the uploaded data is input into the equipment status prediction chain, and equipment status prediction is performed. The aforementioned equipment condition prediction chain predicts and obtains equipment condition prediction results, The method according to claim 1, characterized by including the following:

3. Constructing the aforementioned equipment status prediction chain means Using a Markov chain as the basis of the model, the equipment state prediction chain is constructed, This involves using multiple sample upload data as reward information, setting multiple sample equipment state recognition information as state information, and constructing an initialization equipment state prediction chain. The process involves randomly extracting data from the aforementioned multiple sample upload data and multiple sample equipment state recognition information to construct a set of verification information, and Setting a second model accuracy threshold, The verification information set is used to verify the initial equipment state prediction chain until the initial equipment state prediction chain accuracy satisfies the model accuracy threshold, and the equipment state prediction chain is obtained. The method according to the second invention, characterized by including the following:

4. Setting up routine inspection tasks based on the aforementioned equipment condition prediction results is possible. Based on the aforementioned equipment condition prediction results, a route plan for routine inspections is obtained, During the process of the aforementioned mobile inspection camera making its rounds, an initial mobile inspection route plan is acquired, The initial patrol inspection route plan is modified using the aforementioned patrol inspection route plan, and a modified patrol inspection route plan is obtained. The revised inspection route plan is added to the inspection task, and the inspection route is updated. The method according to claim 1, characterized by including the following:

5. A system for operational inspection and trend analysis of power plant equipment, used to carry out the method for operational inspection and trend analysis of power plant equipment described in any one of claims 1 to 4, During the process of the patrol inspection camera's patrol, a video capture device is activated, and a video information acquisition module is used to acquire real-time video information of the power plant equipment. A power plant equipment equipment real-time video information is input to an image processing and recognition model, and an equipment status recognition module is used to acquire equipment status recognition information. An upload data acquisition module for connecting to the target power plant equipment and acquiring the uploaded data, Based on the aforementioned equipment status recognition information, an equipment status prediction module is provided to perform equipment status prediction using the uploaded data and to obtain the equipment status prediction result. Based on the equipment status prediction results, a patrol inspection task setting module is provided for setting patrol inspection tasks, The aforementioned patrol inspection camera is equipped with an automatic patrol inspection module for performing automatic patrol inspections on the target power plant equipment in accordance with the patrol inspection task, A system for operational inspection and trend analysis of power plant equipment, characterized by comprising the following features.

Images