Method, device and equipment for predicting residual life of low-voltage current transformer and medium

Abstract

The invention provides a low-voltage current transformer residual life prediction method and device, equipment and a medium, and relates to the technical field of current transformers, and the method comprises the steps: obtaining a multi-stress operation information parameter and a performance index parameter of a to-be-predicted low-voltage current transformer; calculating performance degradation index parameters according to the performance index parameters; training to obtain a preset residual life prediction model based on the multi-stress operation information parameter historical sample value, the performance index parameter historical sample value and the residual life historical sample value, and predicting the residual life according to the multi-stress operation information parameter, the performance degradation index parameter and the preset residual life prediction model. The residual life prediction value of the to-be-predicted low-voltage current transformer is obtained, multiple stress factors are comprehensively considered, the health condition of the transformer is comprehensively evaluated, the accuracy and reliability of residual life prediction of the low-voltage current transformer are improved, production shutdown and safety risks caused by equipment faults are effectively reduced, and the production efficiency is improved. And continuity and stability of production are guaranteed.

Description

Technical Field

[0001] The present application relates to the technical field of current transformers, and in particular to a method, device, equipment and medium for predicting the remaining life of a low-voltage current transformer. Background Art

[0002] In the power system, current transformers (CT) are an indispensable component, especially low-voltage current transformers, which are widely used in substations, distribution networks and various industrial facilities. The main function of these low-voltage current transformers is to convert high currents into lower-level currents to facilitate the measurement, monitoring and control of equipment safe operation, while providing accurate data for power metering.

[0003] Low voltage current transformers are usually used to work within a certain rated current range and can accurately convert current values. However, over time and with continued use, low voltage current transformers may suffer from various types of degradation, such as core aging, insulation aging, mechanical damage and environmental factors. Usually, a current transformer with an accuracy level of 0.1 will degrade to level 0.2 after 3 years of use and to level 0.5 after 10 years.

[0004] Since the performance degradation of low-voltage current transformers will directly affect the reliability of the protection system and the accuracy of power metering, and according to the use regulations, low-voltage current transformers need to be replaced after degradation to a certain extent. Therefore, it is particularly important to conduct real-time monitoring and predictive evaluation of their health status. The traditional current transformer remaining life prediction method in the prior art generally only considers a single stress factor, which greatly affects the accuracy and reliability of the remaining life prediction, and thus causes production downtime and safety risks due to equipment failure. Summary of the invention

[0005] The purpose of this application is to address the above problems and provide a method, device, equipment and medium for predicting the remaining life of a low-voltage current transformer, so as to improve the accuracy and reliability of the prediction of the remaining life of the low-voltage current transformer, effectively reduce production downtime and safety risks caused by equipment failure, and ensure the continuity and stability of production.

[0006] In a first aspect, the present application provides a method for predicting the remaining life of a low-voltage current transformer, comprising:

[0007] Obtaining multi-stress operation information parameters and performance index parameters of the low-voltage current transformer to be predicted;

[0008] Calculating a performance degradation index parameter according to the performance index parameter;

[0009] According to the multi-stress operation information parameters, the performance degradation index parameters and the preset remaining life prediction model, the remaining life prediction value of the low-voltage current transformer to be predicted is obtained; wherein, the preset remaining life prediction model is obtained by training based on historical sample values ​​of multi-stress operation information parameters, historical sample values ​​of performance index parameters and historical sample values ​​of remaining life.

[0010] According to the technical solution provided by this application, the method also includes:

[0011] Obtain multiple groups of historical sample values ​​of multi-stress operation information parameters and historical sample values ​​of performance index parameters and corresponding historical sample values ​​of remaining life;

[0012] Calculating a performance degradation indicator parameter historical sample value according to the performance indicator parameter historical sample value;

[0013] The historical sample values ​​of stress operation information parameters and the historical sample values ​​of performance degradation index parameters are used as input, and the historical sample values ​​of remaining life are used as output to perform neural network model training to obtain the preset remaining life prediction model.

[0014] According to the technical solution provided by the present application, the historical sample values ​​of stress operation information parameters and the historical sample values ​​of performance degradation index parameters are used as model inputs, the historical sample values ​​of remaining life are used as model outputs, and a neural network model is trained to obtain the preset remaining life prediction model, including:

[0015] Dividing a plurality of groups of historical sample values ​​of the stress operation information parameters, historical sample values ​​of the performance degradation index parameters and historical sample values ​​of the remaining life into a training set, a validation set and a test set;

[0016] Performing model training on the neural network model based on the training set to obtain an initial neural network model;

[0017] Performing performance verification on the initial neural network model based on the verification set;

[0018] Calculating a prediction error index parameter of the initial neural network model based on the test set;

[0019] Taking the mean square error as the loss function, the parameters of the initial neural network model are optimized based on the gradient descent method, and the hyperparameters of the initial neural network model are optimized based on the Bayesian optimization algorithm until the prediction error index parameter is less than the preset threshold, so as to obtain the preset remaining life prediction model.

[0020] According to the technical solution provided by the present application, the performance indicator parameter includes a current measured value, and the current measured value includes a primary winding current measured value and a secondary winding current measured value; accordingly, the performance degradation indicator parameter is calculated according to the performance indicator parameter, including:

[0021] Calculating a basic error based on the actual current value and the current standard value;

[0022] Calculating a measured transformation ratio according to the measured value of the primary winding current and the measured value of the secondary winding current, and optimizing the measured transformation ratio to obtain a corrected transformation ratio;

[0023] The basic error and the modified transformation ratio are used as the performance degradation index parameters.

[0024] According to the technical solution provided by the present application, the multi-stress operation information parameters include current load, ambient temperature and ambient humidity; accordingly, the measured transformation ratio is calculated according to the measured value of the primary winding current and the measured value of the secondary winding current, and the measured transformation ratio is optimized to obtain the corrected transformation ratio, including:

[0025] The measured transformation ratio is calculated according to k=I1 / I2, wherein I1 is the measured value of the primary winding current, I2 is the measured value of the secondary winding current, and k is the measured transformation ratio;

[0026] The measured transformation ratio is optimized based on a multi-stress operation influence weight function to obtain the modified transformation ratio.

[0027] According to the technical solution provided by the present application, the method of optimizing the measured transformation ratio based on the multi-stress operation influence weight function to obtain the modified transformation ratio includes:

[0028] according to The measured transformation ratio is optimized to obtain a modified transformation ratio; wherein k' is the modified transformation ratio, k is the measured transformation ratio, P(t) is the performance of the low-voltage current transformer to be predicted at time t, and W load (t) is the current load influence weight at time t, W temp (t) is the ambient temperature influence weight at time t, W humid (t) is the influence weight of the ambient humidity at time t, H(t) is the historical effect function that characterizes the historical influence relationship between the multi-stress operation information parameters and the performance of the low-voltage current transformer to be predicted, T(t) is the time factor that characterizes the degree of degradation of the performance of the low-voltage current transformer to be predicted over time, and I(t) is the dynamic response function that characterizes the response of the performance of the low-voltage current transformer to be predicted to the multi-stress operation information parameters.

[0029] According to the technical solution provided by this application, the method also includes:

[0030] According to the predicted value of the remaining life of the low-voltage current transformer to be predicted, obtaining an equipment maintenance strategy for the low-voltage current transformer to be predicted;

[0031] Generate equipment maintenance prompt information according to the equipment maintenance strategy for display.

[0032] In a second aspect, the present application provides a low-voltage current transformer remaining life prediction device, comprising:

[0033] An acquisition module, used to acquire multi-stress operation information parameters and performance index parameters of the low-voltage current transformer to be predicted;

[0034] A calculation module, used to calculate a performance degradation index parameter according to the performance index parameter;

[0035] A prediction module is used to obtain the remaining life prediction value of the low-voltage current transformer to be predicted based on the multi-stress operation information parameters, the performance degradation index parameters and a preset remaining life prediction model; wherein the preset remaining life prediction model is obtained by training based on historical sample values ​​of multi-stress operation information parameters, historical sample values ​​of performance index parameters and historical sample values ​​of remaining life.

[0036] In a third aspect, the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method described above when executing the computer program.

[0037] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, and the computer program implements the method described above when executed by a processor.

[0038] Compared with the prior art, the beneficial effects of the present application are as follows: obtaining multi-stress operation information parameters and performance index parameters of the low-voltage current transformer to be predicted; calculating the performance degradation index parameters according to the performance index parameters; obtaining a preset remaining life prediction model based on the historical sample values ​​of the multi-stress operation information parameters, the historical sample values ​​of the performance index parameters and the historical sample values ​​of the remaining life, and obtaining the remaining life prediction value of the low-voltage current transformer to be predicted according to the multi-stress operation information parameters, the performance degradation index parameters and the preset remaining life prediction model, comprehensively considering multiple stress factors and avoiding the limitations of single factor prediction; using the gradient descent method as the optimization algorithm, combined with the mean square error The difference is used as the loss function, which can effectively optimize the parameters of the neural network model so that it can better fit the model training data; Bayesian optimization is used to dynamically explore the hyperparameter space and adjust it according to previous observations, which avoids overfitting and helps to find better hyperparameter combinations, thereby improving the generalization ability of the prediction model and improving training efficiency; the health status of the transformer is comprehensively evaluated, which improves the accuracy and reliability of the remaining life prediction of the low-voltage current transformer, effectively reduces the production downtime and safety risks caused by equipment failure, ensures the continuity and stability of production, and is conducive to timely acquisition of maintenance strategies, extending the service life of equipment and reducing maintenance costs.

[0039] It should be understood that the description of technical features, technical solutions, beneficial effects or similar language in this application does not imply that all features and advantages can be realized in any single embodiment. On the contrary, it is understood that the description of features or beneficial effects means that specific technical features, technical solutions or beneficial effects are included in at least one embodiment. Therefore, the description of technical features, technical solutions or beneficial effects in this specification does not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions and beneficial effects described in the present embodiment can also be combined in any appropriate manner. Those skilled in the art will understand that the embodiment can be realized without one or more specific technical features, technical solutions or beneficial effects of a specific embodiment. In other embodiments, additional technical features and beneficial effects can also be identified in a specific embodiment that does not embody all embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solution in this embodiment, the drawings required for use in the description of the embodiment will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 This is a schematic diagram of the working principle of a low voltage current transformer;

[0042] Figure 2 A flowchart of a method for predicting the remaining life of a low-voltage current transformer provided in an embodiment of the present application;

[0043] Figure 3 A schematic diagram of a low-voltage current transformer remaining life prediction device provided in an embodiment of the present application;

[0044] Figure 4 A schematic diagram of a computer system of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0045] In order to enable those skilled in the art to better understand the technical solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings. The description in this section is only exemplary and explanatory and should not have any limiting effect on the protection scope of the present application. Specifically, the described embodiments are only embodiments of a part of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work should fall within the scope of protection of the present application.

[0046] It should be noted that similar reference numerals and letters represent similar items in the following figures, so once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0047] In order to facilitate the understanding of this application, a brief introduction to the related technology is first given:

[0048] A low voltage current transformer is a measuring device used in power systems that converts high current into a smaller, easier-to-measure current. This type of transformer is typically used in low voltage power distribution systems, such as 380V, for current measurement, energy metering, and protection. The operation of a low voltage current transformer is based on the principle of electromagnetic induction. Figure 1 As shown, the primary winding is connected in series in the circuit under test. When current passes through the primary winding, a magnetic field is generated, which in turn induces a voltage in the secondary winding, thereby obtaining a secondary current proportional to the primary current.

[0049] Low-voltage current transformers can be used to monitor current in distribution cabinets to ensure the normal operation of the power system. They can also be used for billing by power supply bureaus to ensure accurate measurement of energy consumption. They can also trigger protection devices by monitoring abnormal currents to prevent faults such as overload and short circuit. It is very necessary to ensure the normal operation of low-voltage current transformers. However, in actual use, low-voltage current transformers will degrade over time and with continuous use. The reasons for the degradation of low-voltage current transformers include: 1. Core aging: The core material may gradually lose its original magnetic properties due to long-term exposure to the magnetic field, resulting in a decrease in magnetic permeability; 2. Insulation aging: The insulation material of the transformer may age under long-term working voltage, reducing the insulation performance and increasing the risk of short circuit or breakdown; 3. Mechanical damage: Vibration, shock or temperature changes may cause tiny cracks or deformation in the physical structure of the transformer, affecting its accuracy; 4. Environmental factors: Environmental conditions such as humidity, temperature, corrosive gases or liquids may corrode the external materials or internal components of the transformer; 5. Electrical overload: Current beyond the design range may cause overheating, damage the windings or core, and thus reduce the life of the transformer. The performance degradation of low-voltage current transformers will directly affect the reliability of the protection system and the accuracy of electricity metering. Therefore, it is necessary to conduct a timely and comprehensive assessment of the remaining life of the low-voltage current transformers, so as to detect problems in time and take preventive measures or corresponding maintenance decisions to extend the service life of the equipment and improve the stability and reliability of equipment operation.

[0050] In view of this, an embodiment of the present application provides a method for predicting the remaining life of a low-voltage current transformer, which obtains multi-stress operating information parameters and performance index parameters of the low-voltage current transformer to be predicted; calculates performance degradation index parameters based on the performance index parameters; obtains a preset remaining life prediction model based on historical sample values ​​of multi-stress operating information parameters, historical sample values ​​of performance index parameters, and historical sample values ​​of remaining life, and obtains the remaining life prediction value of the low-voltage current transformer to be predicted based on the multi-stress operating information parameters, the performance degradation index parameters, and the preset remaining life prediction model, comprehensively considering multiple stress factors to avoid the limitations of single factor prediction; uses gradient descent method as optimization The algorithm, combined with the mean square error as the loss function, can effectively optimize the parameters of the neural network model so that it can better fit the model training data; it uses Bayesian optimization to dynamically explore the hyperparameter space and adjust it according to previous observations, avoiding overfitting and helping to find better hyperparameter combinations, thereby improving the generalization ability of the prediction model and improving training efficiency; it comprehensively evaluates the health status of the transformer, improves the accuracy and reliability of the remaining life prediction of the low-voltage current transformer, effectively reduces production downtime and safety risks caused by equipment failure, ensures the continuity and stability of production, and is conducive to timely acquisition of maintenance strategies, extending the service life of equipment and reducing maintenance costs.

[0051] In order to make the technical solution of the present application clearer and easier to understand, the remaining life prediction method of a low-voltage current transformer provided in an embodiment of the present application is introduced below.

[0052] like Figure 2 As shown in the figure, this figure is a flow chart of a method for predicting the remaining life of a low-voltage current transformer provided in this embodiment, and the method includes the following steps:

[0053] S101, obtaining multi-stress operation information parameters and performance index parameters of the low-voltage current transformer to be predicted;

[0054] Specifically, the multi-stress operation information parameters of the low-voltage current transformer to be predicted are collected, and the multi-stress operation information parameters are characteristic parameters that have an impact on the remaining life prediction of the low-voltage current transformer to be predicted, including current load, ambient temperature, ambient humidity, and other parameters such as vibration. They can be set and adjusted according to actual conditions, and are not specifically limited here; at the same time, the performance index parameters of the low-voltage current transformer to be predicted are collected, and the performance index parameters include the actual measured value of the current. It can be understood that in the actual application process, a variety of multi-stress operation information parameters including but not limited to current load, ambient temperature, ambient humidity, etc. can be collected, and the collected data can be pre-processed by cleaning, denoising, outlier processing and other pre-processing operations, and the pre-processed data can be subjected to feature selection, transformation, combination and other operations to extract the characteristic parameters that have an impact on the remaining life prediction as the multi-stress operation information parameters used in the present application for the remaining life prediction of the low-voltage current transformer to be predicted.

[0055] S102, calculating a performance degradation index parameter according to the performance index parameter;

[0056] Specifically, the performance index parameter includes the measured current value, and the performance degradation index parameter is used to characterize the impact of environmental factors including but not limited to current load, temperature change, humidity, etc. on the performance degradation of the low-voltage current transformer. The performance degradation index parameter includes basic error and corrected transformation ratio. Specifically, the basic error can be calculated based on the measured current value and the current standard value; the measured transformation ratio is calculated based on the measured current value of the primary winding and the measured current value of the secondary winding, and the measured transformation ratio is optimized to obtain the corrected transformation ratio. The performance degradation index parameter may also include other parameters, which can be set and adjusted according to actual conditions, and are not specifically limited here.

[0057] S103. Obtain the remaining life prediction value of the low-voltage current transformer to be predicted according to the multi-stress operation information parameters, the performance degradation index parameters and a preset remaining life prediction model; wherein the preset remaining life prediction model is obtained by training based on historical sample values ​​of multi-stress operation information parameters, historical sample values ​​of performance index parameters and historical sample values ​​of remaining life.

[0058] Specifically, first obtain multiple groups of historical sample values ​​of multi-stress operation information parameters and historical sample values ​​of performance index parameters and corresponding historical sample values ​​of remaining life, then calculate historical sample values ​​of performance degradation index parameters according to the historical sample values ​​of performance index parameters, and then use the historical sample values ​​of stress operation information parameters and historical sample values ​​of performance degradation index parameters as inputs and the historical sample values ​​of remaining life as outputs to train a neural network model and obtain the preset remaining life prediction model. Input the multi-stress operation information parameters and the performance degradation index parameters into the preset remaining life prediction model to obtain the remaining life prediction value of the low-voltage current transformer to be predicted.

[0059] The present application provides a method for predicting the remaining life of a low-voltage current transformer, which obtains multi-stress operation information parameters and performance index parameters of the low-voltage current transformer to be predicted; calculates performance degradation index parameters according to the performance index parameters; obtains a preset remaining life prediction model based on historical sample values ​​of the multi-stress operation information parameters, historical sample values ​​of the performance index parameters, and historical sample values ​​of the remaining life, and obtains the remaining life prediction value of the low-voltage current transformer to be predicted according to the multi-stress operation information parameters, the performance degradation index parameters, and the preset remaining life prediction model, comprehensively considers various stress factors, comprehensively evaluates the health status of the transformer, improves the accuracy and reliability of the prediction of the remaining life of the low-voltage current transformer, effectively reduces production downtime and safety risks caused by equipment failure, and ensures the continuity and stability of production.

[0060] On the basis of the above embodiment, the method further includes:

[0061] Obtain multiple groups of historical sample values ​​of multi-stress operation information parameters and historical sample values ​​of performance index parameters and corresponding historical sample values ​​of remaining life;

[0062] Specifically, multiple groups of multi-stress operation information parameter historical sample values ​​and performance indicator parameter historical sample values ​​and corresponding remaining life historical sample values ​​are obtained, and each group of the multi-stress operation information parameter historical sample values, performance indicator parameter historical sample values ​​and remaining life historical sample values ​​correspond to each other. The multi-stress operation information parameter historical sample values ​​include parameters such as current load historical sample values, ambient temperature historical sample values, and ambient humidity historical sample values, and the performance indicator parameter historical sample values ​​include current measured historical sample values. The collected multi-stress operation information parameter historical sample values, performance indicator parameter historical sample values ​​and remaining life historical sample values ​​are cleaned, including processing missing values, abnormal values ​​and duplicate values, and the cleaned data are standardized and normalized.

[0063] Calculating a performance degradation indicator parameter historical sample value according to the performance indicator parameter historical sample value;

[0064] Specifically, the performance degradation index parameter historical sample value is calculated according to the current measured historical sample value, and the performance degradation index parameter historical sample value includes the basic error historical sample value and the transformation ratio historical sample value, which is used to evaluate the influence of multiple stress operation information parameters on the measured current value of the low voltage current transformer.

[0065] The historical sample values ​​of stress operation information parameters and the historical sample values ​​of performance degradation index parameters are used as model inputs, and the historical sample values ​​of remaining life are used as model outputs, and neural network model training is performed to obtain the preset remaining life prediction model.

[0066] Specifically, historical sample values ​​of stress operation information parameters and historical sample values ​​of performance degradation index parameters are used as model inputs, and historical sample values ​​of remaining life are used as model outputs. Multiple groups of historical sample values ​​of stress operation information parameters, historical sample values ​​of performance degradation index parameters and historical sample values ​​of remaining life are divided into training set, verification set and test set. Model training is performed on the neural network model based on the training set to obtain an initial neural network model. Performance verification is performed on the initial neural network model based on the verification set. Prediction error index parameters of the initial neural network model are calculated based on the test set to evaluate the credibility of the model prediction results. Mean square error is used as the loss function. Parameters of the initial neural network model are optimized based on the gradient descent method. Hyperparameter optimization is performed on the initial neural network model based on the Bayesian optimization algorithm until the prediction error index parameter is less than a preset threshold, so as to obtain the preset remaining life prediction model.

[0067] On the basis of the above embodiment, further, the method of taking the historical sample values ​​of the stress operation information parameters and the historical sample values ​​of the performance degradation index parameters as inputs and taking the historical sample values ​​of the remaining life as outputs to perform neural network model training to obtain the preset remaining life prediction model includes:

[0068] Dividing a plurality of groups of historical sample values ​​of the stress operation information parameters, historical sample values ​​of the performance degradation index parameters and historical sample values ​​of the remaining life into a training set, a validation set and a test set;

[0069] Specifically, the data set consisting of the collected multiple groups of current load historical sample values, ambient temperature historical sample values, ambient humidity historical sample values, and the calculated basic error historical sample values, transformation ratio historical sample values, and remaining life historical sample values ​​is divided into a training set, a verification set, and a test set. The ratio usually used is 70% of the historical data for training, 20% of the historical data for verification, and 10% of the historical data for testing. Of course, it can also be set and adjusted according to the actual situation, and is not specifically limited here.

[0070] Performing model training on the neural network model based on the training set to obtain an initial neural network model;

[0071] Specifically, a recurrent neural network (RNN) is selected, and the current load historical sample values, ambient temperature historical sample values, ambient humidity historical sample values, basic error historical sample values, and transformation ratio historical sample values ​​in the training set are used as model inputs, and the corresponding remaining life historical sample values ​​in the training set are used as outputs. The recurrent neural network is trained to obtain an initial neural network model.

[0072] Performing performance verification on the initial neural network model based on the verification set;

[0073] Specifically, the trained model is verified using the current load historical sample values, ambient temperature historical sample values, ambient humidity historical sample values, basic error historical sample values, transformation ratio historical sample values, and corresponding remaining life historical sample values ​​in the verification set to evaluate the generalization ability and performance of the initial neural network model.

[0074] Calculating a prediction error index parameter of the initial neural network model based on the test set;

[0075] Specifically, the current load historical sample values, ambient temperature historical sample values, ambient humidity historical sample values, basic error historical sample values, and transformation ratio historical sample values ​​in the test set are input into the initial neural network model, and the corresponding predicted life sample values ​​are output. The predicted error index parameter is calculated based on the predicted life sample values ​​and the remaining life historical sample values. The error index parameter value is used to characterize the error between the predicted value and the actual value. The error percentage between the predicted value and the actual value corresponding to each group of data in the test set can be calculated respectively, and then the error percentage between the predicted value and the actual value corresponding to multiple groups of data in the test set is averaged as the predicted error index parameter. Of course, other methods can also be used to calculate the predicted error index parameter, which can be set and adjusted according to actual conditions, and is not specifically limited here. It can be understood that the predicted error index parameter is used to characterize the credibility of the initial neural network model and affects the accuracy of the prediction results.

[0076] Taking the mean square error as the loss function, the parameters of the initial neural network model are optimized based on the gradient descent method, and the hyperparameters of the initial neural network model are optimized based on the Bayesian optimization algorithm until the prediction error index parameter is less than the preset threshold, so as to obtain the preset remaining life prediction model.

[0077] Specifically, mean-square error (MSE) is used as the loss function, and the specific calculation formula is: Among them, Y i is the i-th historical sample value; is the i-th prediction value; n is the number of samples. A historical sample value is a set of current load historical sample values, ambient temperature historical sample values, ambient humidity historical sample values, basic error historical sample values ​​and transformation ratio historical sample values, and a prediction value is the predicted remaining life output by the neural network model based on the historical sample values.

[0078] The parameters of the initial neural network model are optimized based on the gradient descent method. The calculation formula of the gradient descent method is specifically as follows: in, is the value of the jth parameter after the t+1th iteration; is the value of the jth parameter at the tth iteration; a is the learning rate, which is used to determine the step size of parameter update; n is the number of samples; J(θ (t) ) is the value of the loss function with respect to the parameter θ; is the partial derivative of the loss function with respect to the jth parameter, i.e., the gradient; in each iteration, the parameter θ j It will be updated according to the negative gradient direction of the loss function in order to find the parameter value that minimizes the loss function.

[0079] The initial neural network model is optimized for hyperparameters based on the Bayesian optimization algorithm. First, the hyperparameter space of the neural network model is determined, including the number of hidden layers and the number of neurons in each hidden layer, the learning rate, the regularization parameter, etc.; then, initial sampling is performed in the hyperparameter space to obtain a set of initial sample points for training the proxy model; according to the proxy model, the next sampling point is selected in each iteration, and the next hyperparameter combination is selected by maximizing or minimizing the objective function; the proxy model is updated with the new sampling point to obtain a more accurate approximation of the objective function; the optimization process is stopped when the maximum number of iterations is reached or the objective function change is less than the set threshold, which in this application means that the prediction error index parameter is less than the preset threshold; after the optimization is completed, the hyperparameter combination with the best prediction result of the proxy model is selected as the final model hyperparameter to obtain the preset remaining life prediction model.

[0080] This application uses the gradient descent method as an optimization algorithm, combined with the mean square error as a loss function, which can effectively optimize the parameters of the neural network model so that it can better fit the training data and improve the prediction accuracy. By minimizing the loss function, the gradient descent method can find the parameter combination that makes the model perform best on the training data; using Bayesian optimization to adjust the hyperparameters of the neural network model, such as the learning rate, the number of hidden layer nodes, etc., can help the model better generalize to unseen data. Bayesian optimization avoids overfitting by dynamically exploring the hyperparameter space and adjusting it according to previous observations, and helps to find better hyperparameter combinations, thereby improving the generalization ability of the model; and by adjusting the hyperparameters through Bayesian optimization, this efficient search method can find a better hyperparameter combination faster under limited computing resources, and can more effectively search the hyperparameter space, avoiding the blindness of traditional grid search or random search, thereby improving training efficiency. Combining these three components can build a more robust model. By using the gradient descent method to optimize the parameters and Bayesian optimization to adjust the hyperparameters, the model can better adapt to changes in data and perform more reliably in the face of different stress factors.

[0081] On the basis of the above embodiment, further, the performance indicator parameter includes a current measured value, and the current measured value includes a primary winding current measured value and a secondary winding current measured value; accordingly, the performance degradation indicator parameter is calculated according to the performance indicator parameter, including:

[0082] Calculating a basic error based on the actual current value and the current standard value;

[0083] The measured transformation ratio is calculated according to the measured value of the primary winding current and the measured value of the secondary winding current, and the measured transformation ratio is optimized to obtain a corrected transformation ratio, and the basic error and the corrected transformation ratio are used as the performance degradation indicator parameters.

[0084] Specifically, the difference between the actual current value and the standard current value is taken as the basic error; the actual current value includes the actual current value of the primary winding and the actual current value of the secondary winding, and the actual transformation ratio is calculated according to the formula: k=I1 / I2, wherein I1 is the actual current value of the primary winding, I2 is the actual current value of the secondary winding, and k is the actual transformation ratio. Since the transformation ratio is affected by stress operating parameters such as current load, ambient temperature, and ambient humidity, the actual transformation ratio is optimized based on the influence weight function corresponding to the stress operating parameters such as current load, ambient temperature, and ambient humidity to obtain a corrected transformation ratio, and the basic error and the corrected transformation ratio are used as the performance degradation indicator parameters.

[0085] On the basis of the above embodiment, further, the multi-stress operation information parameters include current load, ambient temperature and ambient humidity; accordingly, the measured transformation ratio is calculated according to the measured value of the primary winding current and the measured value of the secondary winding current, and the measured transformation ratio is optimized to obtain the corrected transformation ratio, including:

[0086] The measured transformation ratio is calculated according to k=I1 / I2, wherein I1 is the measured value of the primary winding current, I2 is the measured value of the secondary winding current, and k is the measured transformation ratio;

[0087] The measured transformation ratio is optimized based on a multi-stress operation influence weight function to obtain the modified transformation ratio.

[0088] Specifically, the measured transformation ratio is calculated according to k=I1 / I2, wherein I1 is the measured value of the primary winding current, I2 is the measured value of the secondary winding current, and k is the measured transformation ratio. The multi-stress operation information parameters include current load, ambient temperature, and ambient humidity. It can be understood that when the current load on the primary side exceeds the standard value of the low-voltage current transformer, it may cause the core to saturate, and the measured value of the primary winding current I1 will increase, causing the measured transformation ratio k to increase, and the measured value of the secondary winding current will decrease accordingly, thereby affecting the stability of the transformation ratio. Similarly, an increase in ambient temperature will affect the magnetic properties of the low-voltage current transformer, especially the magnetic permeability, which will cause the saturation characteristics of the core to change, thereby affecting the transformation ratio. This relationship can be indirectly expressed by the following formula: Among them, U2 is the voltage on the secondary side, and Z2 is the impedance on the secondary side. The impedance on the secondary side changes due to the influence of the ambient temperature. The increase in ambient temperature may reduce the impedance Z2 on the secondary side, resulting in a change in the transformation ratio. The increase in ambient humidity may reduce the insulation performance of the low-voltage current transformer, especially in a high humidity environment, which may cause leakage or short circuit between windings and affect the stability of the transformation ratio. The influence of ambient humidity on insulation resistance can be expressed as: Among them, R load is the load resistance, U 2eff is the effective voltage on the secondary side, Vp is the peak voltage; increased ambient humidity may cause R load The ratio is reduced, affecting the transformation ratio. Therefore, the transformation ratio is used as one of the performance degradation index parameters of the low-voltage current transformer affected by stress operating parameters such as current load, ambient temperature, and ambient humidity. In the model training process, there are three influencing factors: the weight of the current load influence W load , the weight of ambient temperature influence W temp , the weight of environmental humidity influence W humid , the prediction model defines an overall performance factor P, which is a function of these weights P = f(W load ,W temp ,W humid ), the overall performance factor is fitted through historical data, and then the measured transformation ratio is optimized by the overall performance factor P to obtain the modified transformation ratio.

[0089] On the basis of the above embodiment, further, the step of optimizing the measured transformation ratio based on the multi-stress operation influence weight function to obtain the modified transformation ratio includes:

[0090] according to The measured transformation ratio is optimized to obtain a modified transformation ratio; wherein k' is the modified transformation ratio, k is the measured transformation ratio, P(t) is the performance of the low-voltage current transformer to be predicted at time t, and W load (t) is the current load influence weight at time t, W temp (t) is the ambient temperature influence weight at time t, W humid (t) is the influence weight of the ambient humidity at time t, H(t) is the historical effect function that characterizes the historical influence relationship between the multi-stress operation information parameters and the performance of the low-voltage current transformer to be predicted, T(t) is the time factor that characterizes the degree of degradation of the performance of the low-voltage current transformer to be predicted over time, and I(t) is the dynamic response function that characterizes the response of the performance of the low-voltage current transformer to be predicted to the multi-stress operation information parameters.

[0091] Specifically, based on the above embodiment, the weight W of the current load influence is load , the weight of ambient temperature influence W temp , the weight of environmental humidity influence W humid It changes with time, and P(t) is used to represent the performance of the low-voltage current transformer to be predicted at time t, W load (t) represents the current load influence weight at time t, W temp (t) represents the influence weight of the ambient temperature at time t, W humid(t) represents the weight of the environmental humidity at time t. At the same time, H(t) is used to represent the historical effect function of the historical influence relationship between the multi-stress operation information parameters and the performance of the low-voltage current transformer to be predicted. T(t) is used to represent the time factor of the degree of degradation of the performance of the low-voltage current transformer to be predicted over time. I(t) is used to represent the dynamic response function of the performance of the low-voltage current transformer to be predicted to the multi-stress operation information parameters. The overall performance factor function P(t)=f(W load (t),W temp (t),W humid (t),H(t),T(t),I(t)), and through The measured transformation ratio is optimized to obtain the corrected transformation ratio.

[0092] On the basis of the above embodiment, the method further includes:

[0093] According to the predicted value of the remaining life of the low-voltage current transformer to be predicted, obtaining an equipment maintenance strategy for the low-voltage current transformer to be predicted;

[0094] Generate equipment maintenance prompt information according to the equipment maintenance strategy for display.

[0095] Specifically, according to the remaining life prediction value of the low-voltage current transformer to be predicted output by the preset remaining life prediction model, the equipment maintenance strategy for the low-voltage current transformer to be predicted is obtained; specifically, according to the use time of the low-voltage current transformer to be predicted and the pre-stored equipment maintenance strategy correspondence table, the equipment maintenance strategy of the low-voltage current transformer to be predicted can be obtained, and the equipment maintenance strategy correspondence table includes the use time of the low-voltage current transformer, the standard remaining life, and multiple non-standard remaining life intervals, each of the non-standard remaining life intervals corresponds to a type of equipment maintenance strategy, such as checking and adjusting the current load, checking and adjusting the ambient temperature, checking and adjusting the ambient humidity, replacing the low-voltage current transformer, etc., which can be set, adjusted and updated according to the actual situation, and no specific limitation is made here. Then, equipment maintenance prompt information can be generated according to the equipment maintenance strategy, and displayed through a display or a remote client (computer, mobile phone, etc.). By predicting the remaining life of low-voltage current transformers in advance, we can effectively reduce production downtime and safety risks caused by equipment failure, ensure production continuity and stability, detect problems in a timely manner and take preventive and maintenance measures, extend the service life of equipment, reduce maintenance costs, and improve equipment reliability and operating efficiency.

[0096] Combination of the above Figure 2 The low-voltage current transformer remaining life prediction method provided in the embodiment of the present application is introduced in detail. The low-voltage current transformer remaining life prediction device, electronic device and computer-readable storage medium provided in the embodiment of the present application will be introduced in conjunction with the accompanying drawings.

[0097] like Figure 3 As shown in the figure, this figure is a schematic diagram of a low-voltage current transformer remaining life prediction device provided by the present application, and the device includes:

[0098] An acquisition module 301 is used to acquire multi-stress operation information parameters and performance index parameters of the low-voltage current transformer to be predicted;

[0099] A calculation module 302, configured to calculate a performance degradation indicator parameter according to the performance indicator parameter;

[0100] The prediction module 303 is used to obtain the remaining life prediction value of the low-voltage current transformer to be predicted based on the multi-stress operation information parameters, the performance degradation index parameters and the preset remaining life prediction model; wherein the preset remaining life prediction model is obtained by training based on the historical sample values ​​of the multi-stress operation information parameters, the historical sample values ​​of the performance index parameters and the historical sample values ​​of the remaining life.

[0101] The low-voltage current transformer remaining life prediction device provided in the embodiment of the present application can correspond to the execution of the low-voltage current transformer remaining life prediction method described in the embodiment of the present application, and the above functions of each module of the device correspond to the implementation of Figure 2 For the sake of brevity, the corresponding process of the method shown will not be repeated here.

[0102] An embodiment of the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for predicting the remaining life of a low-voltage current transformer as described in the above embodiment is implemented.

[0103] like Figure 4 As shown, the computer system 400 of the electronic device includes a CPU 401, which can perform various appropriate actions and processes according to the program stored in the ROM 402 or the program loaded from the storage part 408 to the RAM 403. In the RAM 403, various programs and data required for system operation are also stored. The CPU 401, the ROM 402 and the RAM 403 are connected to each other through a bus 404. The I / O interface 405 is also connected to the bus 404. Among them, CPU 401 represents a central processing unit, ROM 402 represents a read-only memory, RAM 403 represents a random access memory, and I / O represents input / output.

[0104] The following components are connected to the I / O interface 405: an input section 406 including a keyboard, a mouse, etc.; an output section 407 including a cathode ray tube, a liquid crystal display, etc., and a speaker, etc.; a storage section 408 including a hard disk, etc.; and a communication section 409 including a network interface card such as a LAN card, a modem, etc. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the I / O interface 405 as needed. A removable medium 411, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 410 as needed, so that a computer program read therefrom is installed into the storage section 408 as needed.

[0105] In particular, the process of the low-voltage current transformer remaining life prediction method described in the above embodiment can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer-readable storage medium, and the computer program contains a program code for executing the low-voltage current transformer remaining life prediction method described in the above embodiment. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 409, and / or installed from the removable medium 411. When the computer program is executed by the CPU 401, the above functions defined in the present computer system 400 are executed.

[0106] An embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method for predicting the remaining life of a low-voltage current transformer as described in the above embodiment is implemented.

[0107] Specifically, the computer-readable storage medium may be included in the electronic device described in the above embodiment; or it may exist independently without being assembled into the electronic device. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed by an electronic device, the electronic device implements the method for predicting the remaining life of a low-voltage current transformer as described in the above embodiment.

[0108] It should be noted that the computer-readable storage medium shown in the present application may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical fiber, a portable compact disk read-only memory, an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above.

[0109] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and its core ideas of this application. The above is only the preferred implementation method of this application. It should be pointed out that due to the limitations of textual expression and the objective existence of infinite specific structures, ordinary technicians in this technical field can make several improvements, modifications or changes without departing from the principles of the present invention, and can also combine the above technical features in an appropriate manner; these improvements, modifications, changes or combinations, or the direct application of the inventive concept and technical solution to other occasions without improvement, should be regarded as the scope of protection of this application.

Claims

1. A method for predicting the remaining life of a low voltage current transformer, characterized in that: include: Obtaining multi-stress operation information parameters and performance index parameters of the low-voltage current transformer to be predicted; Calculating a performance degradation index parameter according to the performance index parameter; According to the multi-stress operation information parameters, the performance degradation index parameters and the preset remaining life prediction model, the remaining life prediction value of the low-voltage current transformer to be predicted is obtained; wherein, the preset remaining life prediction model is obtained by training based on historical sample values ​​of multi-stress operation information parameters, historical sample values ​​of performance index parameters and historical sample values ​​of remaining life.

2. The method according to claim 1, characterized in that The method further comprises: Obtain multiple groups of historical sample values ​​of multi-stress operation information parameters and historical sample values ​​of performance index parameters and corresponding historical sample values ​​of remaining life; Calculating a performance degradation indicator parameter historical sample value according to the performance indicator parameter historical sample value; The historical sample values ​​of stress operation information parameters and the historical sample values ​​of performance degradation index parameters are used as input, and the historical sample values ​​of remaining life are used as output to perform neural network model training to obtain the preset remaining life prediction model.

3. The method according to claim 2, characterized in that The method uses the historical sample values ​​of stress operation information parameters and the historical sample values ​​of performance degradation index parameters as model inputs, uses the historical sample values ​​of remaining life as model outputs, performs neural network model training, and obtains the preset remaining life prediction model, including: Dividing a plurality of groups of historical sample values ​​of the stress operation information parameters, historical sample values ​​of the performance degradation index parameters and historical sample values ​​of the remaining life into a training set, a validation set and a test set; Performing model training on the neural network model based on the training set to obtain an initial neural network model; Performing performance verification on the initial neural network model based on the verification set; Calculating a prediction error index parameter of the initial neural network model based on the test set; Taking the mean square error as the loss function, the parameters of the initial neural network model are optimized based on the gradient descent method, and the hyperparameters of the initial neural network model are optimized based on the Bayesian optimization algorithm until the prediction error index parameter is less than the preset threshold, so as to obtain the preset remaining life prediction model.

4. The method according to claim 1, characterized in that The performance indicator parameter includes a current measured value, and the current measured value includes a primary winding current measured value and a secondary winding current measured value; accordingly, the performance degradation indicator parameter is calculated according to the performance indicator parameter, including: Calculating a basic error based on the actual current value and the current standard value; The measured transformation ratio is calculated according to the measured value of the primary winding current and the measured value of the secondary winding current, and the measured transformation ratio is optimized to obtain a corrected transformation ratio, and the basic error and the corrected transformation ratio are used as the performance degradation indicator parameters.

5. The method according to claim 4, characterized in that The multi-stress operation information parameters include current load, ambient temperature and ambient humidity; accordingly, the measured transformation ratio is calculated according to the measured value of the primary winding current and the measured value of the secondary winding current, and the measured transformation ratio is optimized to obtain the corrected transformation ratio, including: The measured transformation ratio is calculated according to k=I1 / I2, wherein I1 is the measured value of the primary winding current, I2 is the measured value of the secondary winding current, and k is the measured transformation ratio; The measured transformation ratio is optimized based on a multi-stress operation influence weight function to obtain the modified transformation ratio.

6. The method according to claim 5, characterized in that The step of optimizing the measured transformation ratio based on the multi-stress operation influence weight function to obtain the modified transformation ratio includes: according to The measured transformation ratio is optimized to obtain a modified transformation ratio; wherein k' is the modified transformation ratio, k is the measured transformation ratio, P(t) is the performance of the low-voltage current transformer to be predicted at time t, and W load (t) is the current load influence weight at time t, W temp (t) is the ambient temperature influence weight at time t, W humid (t) is the influence weight of the ambient humidity at time t, H(t) is the historical effect function that characterizes the historical influence relationship between the multi-stress operation information parameters and the performance of the low-voltage current transformer to be predicted, T(t) is the time factor that characterizes the degree of degradation of the performance of the low-voltage current transformer to be predicted over time, and I(t) is the dynamic response function that characterizes the response of the performance of the low-voltage current transformer to be predicted to the multi-stress operation information parameters.

7. The method according to any one of claims 1 to 6, characterized in that: The method further comprises: According to the predicted value of the remaining life of the low-voltage current transformer to be predicted, obtaining an equipment maintenance strategy for the low-voltage current transformer to be predicted; Generate equipment maintenance prompt information according to the equipment maintenance strategy for display.

8. A low voltage current transformer remaining life prediction device, characterized in that: include: An acquisition module, used to acquire multi-stress operation information parameters and performance index parameters of the low-voltage current transformer to be predicted; A calculation module, used to calculate a performance degradation index parameter according to the performance index parameter; A prediction module is used to obtain the remaining life prediction value of the low-voltage current transformer to be predicted based on the multi-stress operation information parameters, the performance degradation index parameters and a preset remaining life prediction model; wherein the preset remaining life prediction model is obtained by training based on historical sample values ​​of multi-stress operation information parameters, historical sample values ​​of performance index parameters and historical sample values ​​of remaining life.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images