All-fiber current transformer online monitoring method and system, and storage medium

Abstract

The embodiment of the invention provides an on-line monitoring method and system for an all-fiber current transformer and a storage medium, and belongs to the technical field of power system monitoring. The on-line monitoring method comprises the following steps: acquiring a key state quantity and a preset ratio difference set of the all-fiber current transformer at the current moment; acquiring the state quantity of the next moment according to the key state quantity; establishing a pre-estimation model according to the key state quantity set and the state quantity at the next moment and the preset ratio difference set; obtaining a first ratio difference value at the current moment and a second ratio difference value at the next moment according to the pre-estimation model; judging whether the difference between the first ratio difference value and the second ratio difference value is greater than a preset value; when it is judged that the difference between the first ratio difference value and the second ratio difference value is larger than a preset value, it is determined that the all-fiber current transformer at the next moment is abnormal; and when it is judged that the difference between the first ratio difference value and the second ratio difference value is smaller than or equal to the preset value, it is determined that the all-fiber current transformer operates normally at the next moment.

Description

Technical Field

[0001] This invention relates to the field of power system monitoring technology, and specifically to an online monitoring method, system, and storage medium for an all-fiber optic current transformer. Background Technology

[0002] All-fiber current transformers play an important role in power systems, but in actual operation, their performance and reliability need to be monitored in real time to ensure the safe and stable operation of the power grid.

[0003] Currently, existing monitoring technologies have certain limitations. For example, some monitoring systems are complex in design and have low reliability, while others cannot comprehensively and accurately obtain status information of key operating parameters.

[0004] Therefore, there is an urgent need to provide an efficient, reliable, and comprehensive online monitoring system for all-fiber current transformers to solve the above-mentioned technical problems. Summary of the Invention

[0005] The purpose of this invention is to provide an online monitoring method, system, and storage medium for all-fiber optic current transformers. This method and equipment can monitor the operating status of all-fiber optic current transformers in real time, comprehensively, and reliably, thereby improving the safety and stability of power systems.

[0006] To achieve the above objectives, embodiments of the present invention provide an online monitoring method for an all-fiber optic current transformer, the online monitoring method comprising: Obtain the key state variables and preset ratio difference set of the all-fiber current transformer at the current moment; The state quantity at the next moment is obtained based on the key state quantity; A prediction model is established based on the set of key state quantities and the state quantities at the next moment compared with the preset set of ratios. The first ratio difference at the current time and the second ratio difference at the next time time are obtained based on the prediction model. Determine whether the difference between the first ratio difference and the second ratio difference is greater than a preset value; If the difference between the first ratio difference and the second ratio difference is greater than a preset value, it is determined that the all-fiber current transformer has an abnormal situation at the next moment. If the difference between the first ratio difference and the second ratio difference is less than or equal to a preset value, the all-fiber current transformer is determined to be operating normally at the next moment.

[0007] Optionally, establishing a prediction model based on the set of key state variables and the state variables at the next time step compared with the preset set of ratios includes: Establish an initial state variable prediction model and an initial ratio difference prediction model; Predict the state variables at the next moment based on the state variable prediction model; The state quantity prediction model and the ratio difference prediction model are updated based on the predicted state quantity at the next time step. Determine whether the number of training iterations for the updated state variable prediction model and the ratio difference prediction model meets the preset number of training iterations; If the number of training iterations of the updated state variable prediction model and the ratio difference prediction model meets the preset number of training iterations, the final state variable prediction model and the ratio difference prediction model are established. If it is determined that the number of training iterations of the updated state quantity prediction model and the ratio difference prediction model does not meet the preset number of training iterations, the process returns to the step of updating the state quantity prediction model and the ratio difference prediction model based on the predicted state quantity at the next moment.

[0008] Optionally, establishing the initial state variable prediction model and the initial ratio difference prediction model includes: The initial state variable prediction model is obtained according to formula (1), and the initial ratio difference prediction model is obtained according to formula (2). (1) (2) in, , These are the initialization state variable prediction model and the initialization ratio difference prediction model, respectively. and For the true value, and For predicted values, and For loss function, For the number of samples, , and Both are integer codes.

[0009] Optionally, updating the state quantity prediction model and the ratio difference prediction model based on the predicted state quantity at the next time step includes: Based on the current state quantity and the state quantity prediction model, predict the predicted value of the state quantity at the next time step. Obtain the actual state value at the next moment, and update the state value prediction model based on the predicted value and the actual state value.

[0010] Optionally, obtaining the actual state value at the next time step and updating the state value prediction model based on the predicted value and the actual state value includes: Update the parameters of the state variable prediction model to obtain the updated next state prediction model; The updated parameters are obtained according to formula (3). (3) in, For the updated number One parameter, For the first One parameter, This represents the actual state value at the next moment. This is the predicted value for the next moment. For loss function, For the first The gradient of each parameter, For learning efficiency, It is the minimum of the true state value at the next time step and the true state value at the next time step.

[0011] Optionally, updating the state quantity prediction model and the ratio difference prediction model based on the predicted state quantity at the next time step further includes: Obtain the set of state variables and the set of ratios at a preset time. A ratio difference prediction model is established based on the set of state variables and the ratio difference set to obtain the ratio difference value; Determine whether the current training count for the difference meets the preset training count; If the number of training iterations for the current ratio difference meets the preset number of training iterations, establish a prediction ratio difference model; If the current training count for the difference does not meet the preset training count, the training model operation is repeatedly executed until the current training count for the difference meets the preset training count.

[0012] Optionally, if the number of training iterations of the updated state variable prediction model and the ratio difference prediction model meets the preset number of training iterations, establishing the final state variable prediction model and the ratio difference prediction model includes: A final state quantity prediction model is established based on formula (4). (4) in, To initialize the state variable prediction model, express Each region The maximum number of iterations, The minimum loss function value for each region. For specific input-related values, For input variables, It is a set of regions and iteration counts.

[0013] Optionally, if the number of training iterations of the updated state variable prediction model and the ratio difference prediction model meets the preset number of training iterations, establishing the final state variable prediction model and the ratio difference prediction model includes: The final ratio difference prediction model is established based on formula (5). (5) in, For the final ratio difference prediction model, To initialize the ratio difference prediction model, For the number of iterations, To regress the number of leaf nodes, For the best fit value, For the first The partitioning in the next iteration Each region For input variables.

[0014] On the other hand, the present invention also provides an online monitoring system for an all-fiber optic current transformer, the system being used to execute any of the online monitoring methods described above, the system comprising: The hardware parameter acquisition module is used to acquire various hardware parameters of the all-fiber current transformer. The status monitoring module is used to process and analyze the hardware parameters collected by the hardware parameter acquisition module. The host computer is connected to the monitoring device and is used to display the local monitoring interface and the remote monitoring interface. It is also used to receive status data sent by the monitoring device and to store, analyze and evaluate the data.

[0015] On the other hand, the present invention provides a machine-readable storage medium storing instructions that cause a machine to perform any of the online monitoring methods described above.

[0016] The above technical solution first obtains the key state variables of the all-fiber optic current transformer at the current moment and a preset set of ratio differences; then, it obtains the state variables for the next moment based on the key state variables; next, it establishes a prediction model based on the set of key state variables, the state variables for the next moment, and the preset set of ratio differences; then, it obtains the first ratio difference value for the current moment and the second ratio difference value for the next moment based on the prediction model; finally, it determines whether the difference between the first and second ratio differences is greater than a preset value. If the difference is greater than the preset value, it is determined that the all-fiber optic current transformer has an abnormal situation at the next moment; if the difference is less than or equal to the preset value, it is determined that the all-fiber optic current transformer is operating normally at the next moment. This method can monitor the operating status of the all-fiber optic current transformer in real time, comprehensively, and reliably, improving the safety and stability of the power system. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of an online monitoring method for an all-fiber current transformer according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating the process of establishing a prediction model based on a set of key state quantities and a set of ratio differences between the state quantities at the next moment and a preset set of ratio differences in an online monitoring method for an all-fiber current transformer according to an embodiment of the present invention. Figure 3 This is a flowchart illustrating the method of updating the state quantity prediction model based on the predicted state quantity at the next moment in an online monitoring method for an all-fiber current transformer according to an embodiment of the present invention. Figure 4 This is a flowchart illustrating the method of updating the ratio difference prediction model based on the predicted state quantity at the next moment in an online monitoring method for an all-fiber current transformer according to an embodiment of the present invention. Figure 5 This is a schematic diagram of the processing flow of an all-fiber optic current transformer online monitoring system according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the monitoring and prediction of an all-fiber optic current transformer online monitoring system according to one embodiment of the present invention. Detailed Implementation

[0018] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0019] It should be noted that the acquisition, transmission, storage, use, and processing of data in the technical solution of this application all comply with the relevant provisions of national laws and regulations. In the embodiments of this application, certain existing industry solutions such as software, components, and models may be mentioned. These should be considered exemplary, intended only to illustrate the feasibility of implementing the technical solution of this application, and do not imply that the applicant has already used or necessarily used such solutions.

[0020] like Figure 1 As shown, Figure 1 This is a flowchart of an online monitoring method for an all-fiber optic current transformer according to an embodiment of the present invention. Figure 1 The method includes the following steps: In step S01, the key state variables of the all-fiber optic current transformer at the current moment and the preset ratio difference set are obtained. The key state variables of the all-fiber optic current transformer at the current moment are as follows: (LED Current) (Light SourceDrive) (Average Deviation) (Second Harmonic) and (Fourth Harmonic), the preset set of ratios is .

[0021] In step S02, the state variables for the next time step are obtained based on the key state variables. The key state variables are... As input, it is fed into the state variable prediction model module. In the middle, the prediction was obtained state quantity at time 1 .

[0022] In step S03, a prediction model is established based on the set of key state variables and the state variables at the next time step compared with a preset set of ratios. After obtaining... State set at time and the preset set of comparisons ,back The set of state variables and the set of ratio differences at each time step are used as the ratio difference prediction module. The training set, and it is used for After several iterations of training, the final prediction model is obtained.

[0023] In step S04, the first ratio difference value at the current time and the second ratio difference value at the next time are obtained according to the prediction model. The first ratio difference value at the current time is... The second ratio difference at the next moment is .

[0024] In step S05, it is determined whether the difference between the first ratio difference and the second ratio difference is greater than a preset value.

[0025] In step S06, if the difference between the first ratio difference and the second ratio difference is greater than a preset value, it is determined that the all-fiber current transformer has an abnormal situation at the next moment.

[0026] In step S07, if the difference between the first ratio difference and the second ratio difference is less than or equal to a preset value, it is determined that the all-fiber current transformer will operate normally at the next moment.

[0027] In this embodiment, when the sample is currently in the first... At the time step, the key state quantities of the all-fiber current transformer at the current moment are obtained according to step S01. (LED Current) (Light Source Drive) (AverageDeviation) (Second Harmonic) and (Fourth Harmonic), according to step S02, the state variables are... As input, it is fed into the state variable prediction model module. In the middle, the prediction was obtained state quantity at time 1 Then, according to the following steps in S03 State set at time and the preset set of comparisons As a ratio prediction module The training set, and it is used for After several iterations of training, the final prediction model is obtained, and the predicted state variables are... As a ratio prediction model The input is obtained The ratio of time difference That is, the second difference, then Time prediction results That is, the first ratio difference, and The ratio of time difference The data is sent to the early warning module, which compares the first ratio difference with the second ratio difference. If the difference between the first ratio difference and the second ratio difference is greater than a preset value, the all-fiber current transformer will be in an abnormal condition at the next moment, and the system will issue an early warning. If the difference between the first ratio difference and the second ratio difference is less than or equal to the preset value, it indicates that the all-fiber current transformer is operating normally at the next moment.

[0028] In this embodiment, the method for establishing the prediction model based on the set of key state variables and the state variables at the next moment compared with a preset set of ratios can be various methods known to those skilled in the art. In one embodiment of the present invention, such as... Figure 2 As shown, the specific steps are as follows: In step S031, an initial state variable prediction model and an initial ratio difference prediction model are established. In this embodiment, at the time step... At time t, the state variable Comparison and difference set Since the ratio difference cannot be obtained directly, the ratio difference is initially set. Regarding the methods for establishing the initial state variable prediction model and the initial ratio difference prediction model, there are various methods known to those skilled in the art. In one embodiment of the present invention, the initial state variable prediction model is obtained according to formula (1), and the initial ratio difference prediction model is obtained according to formula (2). (1) (2) in, , These are the initialization state variable prediction model and the initialization ratio difference prediction model, respectively. and For the true value, and For predicted values, and For loss function, For the number of samples, , and Both are integer codes.

[0029] In step S032, the state variables at the next time step are predicted based on the state variable prediction model. State quantity at time step Through the state quantity prediction module The next moment is predicted respectively. state quantity at time 1 .

[0030] In step S033, the state quantity prediction model and the ratio difference prediction model are updated based on the predicted state quantity at the next time step. There are various methods known to those skilled in the art for updating the state quantity prediction model based on the predicted state quantity at the next time step. In one embodiment of the present invention, this is achieved by minimizing the predicted state quantity. and the actual state quantity The difference is used to update the state variable model parameters. ,like Figure 3 As shown, the specific steps are as follows: In step S0331, based on the state variables at the current time and the state variable prediction model, the predicted value of the state variables at the next time is predicted.

[0031] In step S0332, the actual state value at the next moment is obtained, and the state value prediction model is updated based on the predicted value and the actual state value. There are various methods known to those skilled in the art for updating the state value prediction model based on the predicted value and the actual state value. In one embodiment of the present invention, the parameters of the state value prediction model are updated to obtain the updated next state prediction model. The updated parameters are obtained according to formula (3). (3) in, For the updated number One parameter, For the first One parameter, This represents the actual state value at the next moment. This is the predicted value for the next moment. For loss function, For the first The gradient of each parameter, For learning efficiency, It is the minimum of the true state value at the next time step and the true state value at the next time step.

[0032] There are various methods known to those skilled in the art for updating the ratio difference prediction model based on the predicted state variables at the next moment. In one embodiment of the present invention, such as... Figure 4 As shown, the specific steps are as follows: In step S0333, the set of state variables and the set of ratio differences at a preset time are obtained.

[0033] In step S0334, a ratio difference prediction model is established based on the set of state variables and the ratio difference set to obtain the ratio difference value.

[0034] In step S0335, it is determined whether the current training count of the difference meets the preset training count.

[0035] In step S0336, if the number of training iterations for the current ratio difference meets the preset number of training iterations, a predicted ratio difference model is established.

[0036] In step SS0337, if the number of training iterations for the current ratio difference does not meet the preset number of training iterations, the training model operation is repeatedly executed until the number of training iterations for the current ratio difference meets the preset number of training iterations.

[0037] In step S034, it is determined whether the number of training iterations of the updated state variable prediction model and the difference prediction model meets the preset number of training iterations.

[0038] In step S035, if the number of training iterations of the updated state variable prediction model and the ratio difference prediction model meets the preset number of training iterations, the final state variable prediction model and the ratio difference prediction model are established. In one embodiment of the present invention, the final state variable prediction model is established according to formula (4), and the final ratio difference prediction model is established according to formula (5). (4) in, To initialize the state variable prediction model, express Each region The maximum number of iterations, The minimum loss function value for each region. For specific input-related values, For input variables, It is a set of regions and iteration counts.

[0039] (5) in, For the final ratio difference prediction model, To initialize the ratio difference prediction model, For the number of iterations, To regress the number of leaf nodes, For the best fit value, For the first The partitioning in the next iteration Each region For input variables.

[0040] In step S036, if it is determined that the number of training times of the updated state quantity prediction model and the ratio difference prediction model does not meet the preset number of training times, the process returns to the step of updating the state quantity prediction model and the ratio difference prediction model according to the predicted state quantity at the next moment, that is, returning to step S033.

[0041] In this embodiment, for the prediction module based on state variables The final state quantity prediction model can be obtained in various ways known to those skilled in the art. In one embodiment of the present invention, the training set... These are the state variables in the corresponding set of state variables, that is, for Predicting different state variables requires... This training process involves running a corresponding prediction module for each state variable, as detailed below: In step S10, training set samples are obtained. That is, the values ​​of a certain state variable at various time intervals. The training set samples are obtained according to formula (1-1).

[0042] (1-1) in, Encoding is for integers. Formula (1-1) can be written as formula (1-2), that is... (1-2) In step S11, the loss function is calculated based on the training set. The loss function is calculated according to formula (1-3). (1-3) in, To output the prediction model, These are the actual training values.

[0043] In step S12, the initial state variable prediction model is obtained according to the loss function. That is, the initial state variable prediction model is obtained according to formula (1). (1) in, To initialize the state variable prediction model, For the true value, For predicted values, The number of samples.

[0044] In step S13, the negative gradient is calculated based on the initial state variable prediction model. In one embodiment of the present invention, the negative gradient is obtained according to formulas (1-4). (1-4) in, For the first The first time step The negative gradient of each sample. For loss function, This is the model from the previous iteration.

[0045] In step S14, the corresponding region value is obtained based on the negative gradient. The corresponding region value is obtained according to formula (1-5). (1-5) In step S15, the region is selected based on the region value. In one embodiment of the invention, an arbitrary dividing point is selected. The region Divided into , Choose the smallest dividing point according to formula (1-6). Then, based on (1-7) selected Divide into areas, (1-6) (1-7) in This is the output after dividing the region.

[0046] The input space is divided into multiple regions, and these regions are combined to obtain a regression tree for the current time step. The regions are collected to obtain A tree of return in time .

[0047] In step S16, the output value is obtained based on the samples in the regression tree for each region. That is, for each region... The minimum value of the loss function is found from the samples, which is the output value that best fits the leaf node. In one embodiment of the present invention, the output value is obtained according to formulas (1-8). , (1-8) In step S17, the decision tree fitting function for this round is calculated based on the output value. In one embodiment of the present invention, the decision tree fitting function for this round is obtained according to formula (1-9). (1-9) In step S18, the initial state variable prediction model is updated according to the decision tree fitting function to obtain the updated model. In one embodiment of the present invention, the updated model is obtained according to formula (1-10). (1-10) In step S19, the final state variable prediction model is obtained according to the updated model. In one embodiment of the present invention, the final state variable prediction model is obtained according to formula (4). (4) in, To initialize the state variable prediction model, express Each region The maximum number of iterations, The minimum loss function value for each region. For specific input-related values, For input variables, It is a set of regions and iteration counts.

[0048] In this embodiment, the method for obtaining the final ratio difference prediction model based on the ratio difference prediction module can be one of various methods known to those skilled in the art. In one embodiment of the present invention, based on the subsequent... state quantity at time 1 and the corresponding set of ratios and differences To train the gradient boosting decision tree, and obtain The final learner at each time step will then predict the... state quantity at time 1 As The learning machine takes input from each time step and predicts the ratio difference for the next time step. and automatically update the state variable set. Sum and Difference Sets The specific steps are as follows: In step S20, the initial ratio difference prediction model is obtained. That is, the weak learner is initialized. In one embodiment of the present invention, the initial ratio difference prediction model is obtained according to formula (2). (2) in, To initialize the ratio difference prediction model, For the true value, For predicted values, The loss function is obtained according to formula (2-1) in this invention. The number of samples.

[0049] (2-1) The loss function is the squared loss function, which is a convex function and can be obtained simply by setting the derivative to zero. value.

[0050] In step S21, the mean of the ratio differences of all current training samples is obtained according to the initialized ratio difference prediction model, thereby obtaining the predicted value. In one embodiment of the present invention, the mean of the ratio differences of all current training samples is obtained according to formulas (2-2) and (2-3). (2-2) (2-3) in, Encoded as an integer. This represents the total number of samples.

[0051] In step S22, a prediction model for the current time difference is established based on the predicted values. (Using...) The final prediction model at time 1 is used as the current In one embodiment of the present invention, a time-time difference prediction model is established according to formula (2-4) to predict the current time-time difference. (2-4) in, This represents the number of iterations.

[0052] In step S23, the negative gradient of each training sample is calculated based on the current ratio difference prediction model. In one embodiment of the invention, the negative gradient of each training sample is calculated according to formula (2-5). (2-5) In step S24, the best-fit value is calculated based on the negative gradient of each training sample. The negative gradient of each training sample is used as the new true label value, and... This data is used as training data for the next round of trees, and a new regression tree is obtained for that round. The leaf node region of this regression tree is, ,in To determine the number of leaf nodes in the regression tree, in one embodiment of the invention, the best-fit value is calculated according to formulas (2-6). (2-6) in, For the first In the next iteration, the region The corresponding best-fit value; For the first The partitioning in the next iteration One region; For the sample of this area ; The loss function measures the true value. and predicted value Differences; For the first The cumulative model after the second iteration applies to the samples The predicted value; This is the step size for gradient descent.

[0053] In step S25, the ratio difference prediction model is updated based on the best-fit value. In one embodiment of the invention, the ratio difference prediction model is updated according to formulas (2-7) and (2-8). (2-7) (2-8) In step S26, the final ratio difference prediction model is established based on the updated ratio difference prediction model. In one embodiment of the present invention, the final ratio difference prediction model is established according to formula (5). (5) in, For the final ratio difference prediction model, To initialize the ratio difference prediction model, For the number of iterations, To regress the number of leaf nodes, For the best fit value, For the first The partitioning in the next iteration Each region For input variables.

[0054] Furthermore, the present invention also includes an online monitoring system for all-fiber current transformers, such as... Figure 5 and Figure 6As shown, the system includes a hardware parameter acquisition module, a status monitoring module, and a host computer. The hardware parameter acquisition module collects various hardware parameters of the all-fiber optic current transformer. The status monitoring module processes and analyzes the hardware parameters collected by the hardware parameter acquisition module. The host computer is connected to the monitoring device and displays both local and remote monitoring interfaces. It also receives status data from the monitoring device and stores, analyzes, and evaluates the data. The hardware parameter acquisition module collects various hardware parameters of the all-fiber optic current transformer, while the status monitoring software runs on the monitoring device, responsible for processing and analyzing the collected hardware parameters. The host computer, connected to the monitoring device, displays both local and remote monitoring interfaces and receives status data from the monitoring device, storing, analyzing, and evaluating the data. This allows for real-time, comprehensive, and reliable monitoring of the all-fiber optic current transformer's operating status, improving the safety and stability of the power system.

[0055] In this implementation, the hardware parameter acquisition module transmits electrical parameters to the monitoring device via high-precision sensors and 485 serial or optical communication. For parameters of components such as light sources, modulators, and detectors, a specially designed IED acquisition circuit ensures the accuracy and stability of the acquired data. The status monitoring module operates within the monitoring device, and its algorithm is based on the working principle and performance requirements of the all-fiber optic current transformer. For example, for monitoring relative and absolute optical power, a specific optical power detection algorithm combined with a preset threshold is used to determine whether the operation is normal. A stable communication protocol, such as the FT3 digital channel, is used for data transmission between the host computer and the monitoring device. The host computer software has a user-friendly interface, allowing staff to easily view real-time data, historical data, and early warning information from both the local and remote monitoring interfaces. Simultaneously, the data analysis module in the host computer periodically mines and analyzes historical data, for example, using machine learning algorithms to predict the operating trend of the all-fiber optic current transformer, enabling proactive maintenance measures.

[0056] On the other hand, the present invention also includes a machine-readable storage medium storing instructions for causing a machine to perform any of the above-described online monitoring methods.

[0057] The above technical solution first obtains the key state variables of the all-fiber optic current transformer at the current moment and a preset set of ratio differences; then, it obtains the state variables for the next moment based on the key state variables; next, it establishes a prediction model based on the set of key state variables, the state variables for the next moment, and the preset set of ratio differences; then, it obtains the first ratio difference value for the current moment and the second ratio difference value for the next moment based on the prediction model; finally, it determines whether the difference between the first and second ratio differences is greater than a preset value. If the difference is greater than the preset value, it is determined that the all-fiber optic current transformer has an abnormal situation at the next moment; if the difference is less than or equal to the preset value, it is determined that the all-fiber optic current transformer is operating normally at the next moment. This method can monitor the operating status of the all-fiber optic current transformer in real time, comprehensively, and reliably, improving the safety and stability of the power system.

[0058] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0059] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0060] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0061] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0062] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0063] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, like read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0064] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0065] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0066] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. An all-fiber current transformer on-line monitoring method, characterized in that, The online monitoring method comprises: obtaining a key state quantity of a full-fiber current transformer at a current moment and a preset ratio difference set; obtaining a state quantity at a next moment according to the key state quantity; establishing an estimation model according to the key state quantity set and the state quantity at the next moment and the preset ratio difference set; obtaining a first ratio difference value at the current moment and a second ratio difference value at the next moment according to the estimation model; judging whether a difference between the first ratio difference value and the second ratio difference value is greater than a preset value; in a case where the difference between the first ratio difference value and the second ratio difference value is greater than the preset value, determining that the full-fiber current transformer at the next moment has an abnormal condition; in a case where the difference between the first ratio difference value and the second ratio difference value is less than or equal to the preset value, determining that the full-fiber current transformer at the next moment is in normal operation.

2. The on-line monitoring method according to claim 1, characterized in that, The method comprises: establishing an initial state quantity prediction model and an initial ratio difference prediction model; predicting the state quantity at the next moment according to the state quantity prediction model; updating the state quantity prediction model and the ratio difference prediction model according to the predicted state quantity at the next moment; judging whether a training number of the updated state quantity prediction model and the ratio difference prediction model meets a preset training number; in a case where the training number of the updated state quantity prediction model and the ratio difference prediction model meets the preset training number, establishing a final state quantity prediction model and a ratio difference prediction model; in a case where the training number of the updated state quantity prediction model and the ratio difference prediction model does not meet the preset training number, returning to the step of updating the state quantity prediction model and the ratio difference prediction model according to the predicted state quantity at the next moment.

3. The on-line monitoring method according to claim 2, characterized in that, The method comprises: obtaining the initial state quantity prediction model according to formula (1) and obtaining the initial ratio difference prediction model according to formula (2), ，（1） ，（2） wherein, , are initialization of state quantity prediction model and initialization of ratio difference prediction model respectively, and are true values, and are predicted values, and are loss functions, is the number of samples, , and are all integer encodings.

4. The on-line monitoring method according to claim 2, characterized in that, The method comprises: predicting a predicted value of the state quantity value at the next moment according to the state quantity at the current moment and the state quantity prediction model; obtaining a real state quantity at the next moment and updating the state quantity prediction model according to the predicted value and the real state value.

5. The on-line monitoring method according to claim 3, characterized in that, The method comprises: updating parameters of the state quantity prediction model to obtain an updated next state prediction model; obtaining the updated parameters according to formula (3), ，（3） in, For the updated number One parameter, For the first One parameter, This represents the actual state value at the next moment. This is the predicted value for the next moment. For loss function, For the first The gradient of each parameter, For learning efficiency, It is the minimum of the true state value at the next time step and the true state value at the next time step.

6. The on-line monitoring method according to claim 4, characterized in that, The method further comprises: obtaining a state quantity set and a ratio difference set at a preset moment; establishing a ratio difference prediction model according to the state quantity set and the ratio difference set to obtain a ratio difference value; judging whether a training number of the current ratio difference value meets a preset training number; in a case where the training number of the current ratio difference value meets the preset training number, establishing an estimation ratio difference model; In a case where the training times of the current ratio difference value do not satisfy the preset training times, the training model operation is repeatedly performed until the training times of the current ratio difference value satisfy the preset training times.

7. The on-line monitoring method according to claim 2, characterized in that, In a case where the training times of the updated state quantity prediction model and the ratio difference prediction model satisfy the preset training times, establishing a final state quantity prediction model and a ratio difference prediction model comprises: The final state quantity prediction model is established according to formula (4), ，（4） wherein, to initialize the state quantity prediction model, denotes regions, is a maximum number of iterations, is a minimum loss function value in each region, is a value related to a specific input, is an input variable, is a set of regions and iteration numbers.

8. The on-line monitoring method according to claim 2, characterized in that, In a case where the training times of the updated state quantity prediction model and the ratio difference prediction model satisfy the preset training times, establishing a final state quantity prediction model and a ratio difference prediction model comprises: The final ratio difference prediction model is established according to formula (5), ，（5） in, For the final ratio difference prediction model, To initialize the ratio difference prediction model, For the number of iterations, To regress the number of leaf nodes, For the best fit value, For the first The partitioning in the next iteration Each region For input variables.

9. An all-fiber current transformer on-line monitoring system characterized by, The system is used for executing the online monitoring method as claimed in any one of claims 1 to 8, and the system comprises: A hardware parameter acquisition module is configured to acquire a plurality of hardware parameters of the full-optical current transformer. A state quantity monitoring module is configured to process and analyze the hardware parameters acquired by the hardware parameter acquisition module. A host computer is connected with the monitoring device, configured to display an on-site monitoring interface and a remote monitoring interface, and configured to receive and store, analyze and evaluate the state quantity data sent by the monitoring device.

10. A machine-readable storage medium, characterized in that, The machine readable storage medium has instructions stored thereon, which are used to cause the machine to execute the online monitoring method as claimed in any one of claims 1 to 8.

Images