A method and device for suppressing fire during testing of polypropylene film capacitors

Abstract

The present invention provides a method and device for suppressing fire during the testing process of polypropylene film capacitors. The method comprises: obtaining the internal temperature of the capacitor through a temperature sensor, and obtaining the first sample surface temperature imaging data through an infrared imager; establishing a relationship model between the internal temperature and the surface temperature of the capacitor to calibrate the first sample surface temperature imaging data; using the calibrated first sample surface temperature imaging data to train a recurrent neural network model to obtain a recurrent neural network prediction model; using an infrared imager to scan all capacitors to obtain the second sample surface temperature imaging data, and calibrating the second sample surface temperature imaging data; using the recurrent neural network prediction model to predict the calibrated second sample surface temperature imaging data, and judging whether to disconnect the circuit according to the prediction result. During the test process, the capacitor can be temperature monitored to suppress fire, and the number of integrated temperature sensors can be reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of capacitors, and in particular to a method and a device capable of suppressing fire during a testing process of a polypropylene film capacitor. Background Art

[0002] At present, metallized polypropylene film is the most widely used capacitor film at home and abroad. In recent years, polypropylene film capacitor manufacturers have also invested a certain amount of manpower and material resources in the research of high-temperature polypropylene film. The purpose of studying high-temperature polypropylene film is to make up for the defect that the maximum operating temperature of metallized capacitors is about +85℃. Now a high-temperature resistant polypropylene film has been developed, which can increase the upper temperature limit of polypropylene film during operation to +105℃, which is a major breakthrough. The operating temperature of metallized capacitors made of high-temperature resistant polypropylene film is −55~125℃, and the long-term operating temperature is +105℃. However, metallized capacitors also have certain limitations. When they are subjected to frequent overvoltage shocks, especially when the capacitors are tested for a long time, the long-term effect of frequent partial discharge will affect the life of the capacitors and even cause spontaneous combustion, which will cause serious troubles to the capacitor testing process.

[0003] Temperature has an important impact on capacitor production and safety testing. Measures must be taken to deal with temperature changes to prevent damage to the capacitor and unnecessary losses. In order to accurately monitor the temperature changes of the capacitor during operation, a temperature sensor is usually integrated inside or on the surface of the capacitor. However, in the design and manufacturing process of the capacitor, the integrated temperature sensor requires additional design and process steps, which will also increase the production cost. The sensor and its connecting wires may become a new source of electromagnetic interference, affecting the normal operation of the capacitor and other electrical components. Therefore, there is an urgent need for a method that can monitor the temperature of the capacitor to prevent the capacitor from catching fire and reduce the number of integrated temperature sensors to save costs. Summary of the invention

[0004] The present invention describes a method and a device for suppressing fire during the testing process of polypropylene film capacitors, which can solve the above technical problems.

[0005] According to a first aspect, a method for suppressing fire during a polypropylene film capacitor testing process is provided, the method comprising:

[0006] The internal temperature of the capacitor is monitored by a temperature sensor inside the target capacitor, and the temperature data thereof is recorded, and the surface of the target capacitor is scanned by an infrared imager to obtain first sample surface temperature imaging data; wherein the target capacitor refers to a portion of capacitors selected from all capacitors for measuring the internal temperature;

[0007] A temperature relationship model is established using the capacitor internal temperature data and the first sample surface temperature imaging data to calibrate the first sample surface temperature imaging data; wherein the first sample surface temperature imaging data is calibrated to make the first sample surface temperature imaging data closer to the real capacitor internal temperature data, and the temperature relationship model is:

[0008]

[0009] in, represents the first sample surface temperature, represents the internal temperature of the capacitor, , , and represents the model parameters, represents the model error, represents the thermal resistance of the capacitor, represents the thermal capacity of the capacitor;

[0010] Using the calibrated first sample surface temperature imaging data to train a recurrent neural network model to obtain a recurrent neural network prediction model;

[0011] In the process of testing all capacitors, an infrared imager is used to scan all capacitors to obtain surface temperature imaging data of the second sample, and the surface temperature imaging data of the second sample is calibrated using the established temperature relationship model;

[0012] The recurrent neural network prediction model is used to predict the calibrated second sample surface temperature imaging data, and whether to disconnect the circuit is determined based on the prediction result.

[0013] In some embodiments, the method for establishing a temperature relationship model specifically includes:

[0014] Establish the thermal impedance model of the capacitor, the thermal impedance model is:

[0015]

[0016] in, represents the thermal resistance of the capacitor, represents the thermal capacity of the capacitor, Represents angular frequency , represents the imaginary unit;

[0017] A temperature relationship model is established based on the internal temperature data of the capacitor and the surface temperature imaging data of the first sample, and the parameters in the thermal impedance model are and As an additional input feature of the temperature relationship model, the temperature relationship model is:

[0018]

[0019] in, represents the first sample surface temperature, represents the internal temperature of the capacitor, , , and represents the model parameters, represents the model error;

[0020] The least squares method is used to define an objective function to minimize the square error between the predicted value and the actual value. The objective function is:

[0021]

[0022] in, represents the surface temperature of the capacitor at the i-th observation point, represents the observed value of the internal temperature of the capacitor at the i-th observation point;

[0023] Use an optimization algorithm to minimize the objective function and find The optimal parameters to achieve the minimum , , and , and the obtained parameters 、 、 and Substitute the value of into the linear regression model to obtain the established temperature relationship model.

[0024] In some embodiments, before establishing the temperature relationship model, the internal temperature data of the capacitor and the first sample surface temperature imaging data need to be cleaned to remove abnormal values ​​and missing values, and normalized.

[0025] In some embodiments, the method for normalizing the internal temperature data of the cleaned capacitor and the first sample surface temperature imaging data specifically includes:

[0026] Converting the first sample surface temperature imaging data into a numerical matrix form;

[0027] Using the standard normal distribution scaling method, the internal temperature data of the capacitor and the surface temperature imaging data of the first sample are converted into a standard normal distribution form with a mean of 0 and a standard deviation of 1. The formula is:

[0028]

[0029] in, represents the sample mean, represents the sample standard deviation, represents the original data, Represents the transformed data.

[0030] In some embodiments, the infrared imager also has a light imaging function for collecting light imaging of the capacitor surface to detect whether the capacitor surface is damaged.

[0031] In some embodiments, the test environment of the capacitor is a closed space, and whether nitrogen is introduced into the test environment can be determined based on the prediction results of the model to suppress the capacitor from catching fire.

[0032] According to a second aspect, a device capable of simulating a fire during a test process of a polypropylene film capacitor is provided, the device comprising:

[0033] A temperature data acquisition module, used to acquire the internal temperature data of the capacitor, the surface temperature imaging data of the first sample and the surface temperature imaging data of the second sample;

[0034] A temperature relationship model building module is used to build a temperature relationship model between the inside and the surface of the capacitor and calibrate the surface temperature of the capacitor;

[0035] A recurrent neural network training module is used to train a recurrent neural network model using the calibrated first sample surface temperature imaging data to obtain a recurrent neural network prediction model;

[0036] A recurrent neural network prediction module, used for predicting the calibrated second sample surface temperature imaging data;

[0037] The circuit control module is used to receive the prediction results of the model and determine whether to disconnect the circuit according to the prediction results of the model.

[0038] The present invention provides a method and device for simulating a fire in a polypropylene film capacitor test process. First, the temperature data of the interior and surface of the target capacitor is obtained, and then the temperature relationship model is established using the temperature data of the interior and surface of the capacitor. Next, the surface temperature of the capacitor is calibrated to the internal temperature, and then the surface temperature data of the target capacitor is used to train a recurrent neural network model. The trained model is used to predict the temperature change results of the capacitor over a period of time. Finally, the prediction results of the model are used to determine whether the circuit needs to be disconnected to prevent the capacitor from catching fire. In summary, according to the above method, the temperature of the capacitor can be conveniently monitored during the test process to prevent the capacitor from catching fire, and the number of integrated temperature sensors can be reduced, thereby saving costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0040] Figure 1 A schematic flow chart of a method for simulating a capacitor test process fire provided by an embodiment of the present specification is shown;

[0041] Figure 2 A schematic diagram of a fire ignition device capable of simulating a capacitor testing process provided by an embodiment of the present specification is shown. DETAILED DESCRIPTION

[0042] The solution provided in this specification is described below in conjunction with the accompanying drawings.

[0043] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0044] In the description of the embodiments of the present application, words such as "exemplary", "for example" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary", "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary", "for example" or "for example" is intended to present related concepts in a concrete way.

[0045] In the description of the embodiments of the present application, the term "and / or" is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, B exists alone, and A and B exist at the same time. In addition, unless otherwise specified, the term "plurality" means two or more.

[0046] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. The terms "include", "comprises", "has" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0047] In order to make it easier to understand this solution, some of the principles involved in this solution will be explained below.

[0048] Figure 1A schematic diagram of a method for simulating a fire in a polypropylene film capacitor test process provided by an embodiment of this specification is shown, such as Figure 1 As shown, the method comprises the following steps:

[0049] Step S101, monitor the internal temperature of the capacitor through the temperature sensor inside the target capacitor, and record its temperature data, use an infrared imager to scan the surface of the target capacitor to obtain first sample surface temperature imaging data; wherein, the target capacitor refers to a part of the capacitors selected from all capacitors for measuring the internal temperature.

[0050] A portion of samples are selected from all capacitors as target capacitors, and a temperature sensor is used to monitor the temperature inside the target capacitor, and its temperature data is recorded in detail, wherein the temperature data includes a time series. At the same time, an infrared imager with an infrared temperature imaging function is used to scan the surface of the target capacitor to obtain its surface temperature data, which is used as the first sample surface temperature imaging data.

[0051] It is worth noting that the purpose of selecting some samples as target capacitors is to save the cost of manufacturing temperature sensors inside the capacitors. If the cost is not taken into account, temperature sensors can also be placed inside all capacitors, and this solution does not impose any restrictions on this.

[0052] In one embodiment, the infrared imager may also have a light imaging function for collecting light imaging of the capacitor surface to detect whether the capacitor surface is damaged.

[0053] On the one hand, damage on the surface of the capacitor will increase the leakage current, which may cause the capacitor to heat up more during operation, thus affecting its life and stability. Severe damage may cause the internal metallization layer to be exposed or directly contacted, thus forming a short circuit path. Under high load or long-term use, the capacitor may be damaged or even burn due to overheating. On the other hand, the use of optical imaging can reduce manual inspection and save labor costs.

[0054] Step S102, using the capacitor internal temperature data and the first sample surface temperature imaging data to establish a temperature relationship model to calibrate the first sample surface temperature imaging data. The purpose of calibrating the first sample surface temperature imaging data is to make the first sample surface temperature imaging data closer to the real capacitor internal temperature data, and the temperature relationship model is:

[0055]

[0056] in, represents the first sample surface temperature, represents the internal temperature of the capacitor, , , and represents the model parameters, represents the model error, represents the thermal resistance of the capacitor, represents the thermal capacity of the capacitor;

[0057] In one embodiment, before establishing the temperature relationship model, the internal temperature data of the capacitor and the first sample surface temperature imaging data need to be cleaned to remove abnormal values ​​and missing values, and normalized.

[0058] To obtain the internal temperature data of the capacitor and the surface temperature imaging data of the first sample, it is necessary to first check the data for outliers and missing values. After the inspection, statistical methods or visualization methods can be used to clean the data. After cleaning, the internal temperature data of the capacitor and the surface temperature imaging data of the first sample are further normalized.

[0059] In one embodiment, the method for normalizing the internal temperature data of the cleaned capacitor and the first sample surface temperature imaging data specifically includes:

[0060] First, the first sample surface temperature imaging data is converted into a numerical matrix form. Because the first sample surface temperature imaging data is image data, which is inconsistent with the internal temperature data of the capacitor, it needs to be converted into a numerical matrix form.

[0061] Then, the internal temperature data of the capacitor and the surface temperature imaging data of the first sample are converted into a standard normal distribution form with a mean of 0 and a standard deviation of 1 using the standard normal distribution scaling method, and the formula is:

[0062]

[0063] in, represents the sample mean, represents the sample standard deviation, represents the original data, Represents the transformed data.

[0064] In one embodiment, the method for establishing a temperature relationship model specifically includes:

[0065] Step S1021, establishing a thermal impedance model of the capacitor, the thermal impedance model is:

[0066]

[0067] in, represents the thermal resistance of the capacitor, represents the thermal capacity of the capacitor, Represents angular frequency , Represents the imaginary unit.

[0068] parameter and It can be measured experimentally, for example, by changing the frequency of the input signal and measuring the temperature response of the capacitor, using the thermal impedance model To fit the experimental data, we can infer and , or by applying a step current or voltage to the capacitor and measuring the change in capacitor temperature over time, the and .

[0069] The difference between the surface temperature and the internal temperature of the capacitor has a certain relationship with the thermal resistance and thermal capacitance of the capacitor. When calibrating the temperature of the capacitor surface, it is necessary to consider these two characteristics. The purpose of establishing the thermal impedance model of the capacitor in this step is to eliminate the influence of the thermal resistance and thermal capacitance of the capacitor and further increase the accuracy of the capacitor surface temperature calibration.

[0070] Step S1022: establish a temperature relationship model based on the internal temperature data of the capacitor and the first sample surface temperature imaging data, and convert the parameters in the thermal impedance model into and As an additional input feature of the temperature relationship model, the temperature relationship model is:

[0071]

[0072] in, represents the first sample surface temperature, represents the internal temperature of the capacitor, , , and represents the model parameters, Represents the model error.

[0073] Step S1023, using the least squares method to define an objective function for minimizing the square error between the predicted value and the actual value, the objective function is:

[0074]

[0075] in, represents the surface temperature of the capacitor at the i-th observation point, Represents the observed value of the internal temperature of the capacitor at the i-th observation point.

[0076] Step S1024, using an optimization algorithm to minimize the objective function, find The optimal parameters to achieve the minimum , , and , and the obtained parameters 、 、 and Substitute the values ​​into the linear regression model to obtain the established temperature relationship model.

[0077] The optimization algorithm can use the gradient descent method or the optimizer in the Scipy library to minimize the objective function and calculate the optimal parameters. , , and , and then these four optimal parameters and the ones derived from step S1021 are and Substituting into the temperature relationship model, that is, the formula in step S1022, the established temperature relationship model is obtained. It is easy to understand that through the established temperature relationship model, the internal temperature of the capacitor can be calculated based on the measured data of the capacitor surface temperature.

[0078] In this embodiment, a temperature relationship model is established by combining the internal temperature data of the capacitor with the surface temperature imaging data of the first sample, and the relationship between the internal temperature of the capacitor and the temperature on the surface of the capacitor is analyzed and obtained. Then, based on the established temperature relationship model, the temperature data inside the capacitor can be calculated using the surface temperature data of the capacitor actually measured, thereby reducing the difficulty of measuring the internal temperature of the capacitor and saving the cost of manufacturing the temperature sensor inside the capacitor.

[0079] Step S103, using the calibrated first sample surface temperature imaging data to train a recurrent neural network model to obtain a recurrent neural network prediction model.

[0080] First, divide the data, one part of the data is used for training, one part of the data is used for verification, and the model performance is evaluated during the training process. The other part of the data is used for testing to evaluate the generalization ability of the model. Then, select a suitable recurrent neural network architecture, including but not limited to: RNN (Recurrent Neural Network), LSTM (Long Short-Term Memory) or GRU (Gated Recurrent Unit), which can be selected according to the training data. Before training, continuous time point data needs to be divided into time windows of fixed length. Each window contains data from several time points in the past, and the temperature of the next time point as a label. Then, use the training data set to train the recurrent neural network model, and then verify, test, and adjust the model until the model training is completed.

[0081] Recurrent neural networks are used to predict capacitor temperature changes in the future. Recurrent neural networks are suitable for processing time series data, which can effectively capture the time dependency in capacitor temperature data.

[0082] Step S104, in the process of testing all capacitors, use an infrared imager to scan all capacitors to obtain second sample surface temperature imaging data, and use the established temperature relationship model to calibrate the second sample surface temperature imaging data.

[0083] In the process of formally testing the capacitors, an infrared imager is used to scan all capacitors to obtain temperature imaging data on the surfaces of all capacitors, and the data is recorded as the second sample surface temperature imaging data. It is worth noting that the second sample surface temperature imaging data is image data. Before calibrating the data, it is necessary to convert it into the form of a numerical matrix. The conversion method is consistent with the second sample surface temperature imaging data conversion method in step S102, and will not be repeated here. Finally, the converted second sample surface temperature imaging data is calibrated using the temperature relationship model established in step S102. It is easy to understand that the purpose of the calibration is to calibrate the surface temperature of the capacitor to the internal temperature of the capacitor.

[0084] Step S105, using a recurrent neural network prediction model to predict the calibrated second sample surface temperature imaging data, and judging whether to disconnect the circuit according to the prediction result.

[0085] The calibrated second sample surface temperature imaging data obtained in step S104 is predicted according to the recurrent neural network prediction model trained in step S103, and the prediction result is output. According to the prediction result of the model, it is determined whether to disconnect the test circuit of the capacitor.

[0086] Exemplarily, a temperature threshold is set to disconnect the capacitor test circuit, for example 85°C. The model's prediction results show that a certain area in the data may exceed 85°C in the future. The result is transmitted as a signal to the circuit control module, which immediately responds by disconnecting the circuit to prevent fire.

[0087] In another embodiment, the test environment of the capacitor is a closed space, and whether nitrogen is introduced into the test environment can be determined based on the prediction result of the model to suppress the burning of the capacitor.

[0088] Exemplarily, a temperature threshold is set to disconnect the capacitor test circuit, for example, 85°C. The model's prediction results show that a certain area in the data may exceed 85°C in the future. The result is transmitted as a signal to the nitrogen control module, and the nitrogen control module immediately responds by releasing nitrogen into the test environment to prevent fire. The actual verification shows that the nitrogen concentration that can suppress capacitor fire is 82%. When the released nitrogen concentration is above 82%, the nitrogen control module can stop releasing nitrogen. It is worth noting that while releasing nitrogen, the circuit control module can also respond and cut off the circuit at the same time, and this solution does not limit this.

[0089] Corresponding to the above method provided by the present invention, the present invention also provides a device. Figure 2 FIG. 2 is a schematic diagram showing a structure of a device for simulating a fire in a capacitor test process provided by an embodiment of the present specification. Figure 2 As shown, the device comprises:

[0090] The temperature data acquisition module 201 is used to acquire the internal temperature data of the capacitor, the first sample surface temperature imaging data and the second sample surface temperature imaging data.

[0091] The temperature relationship model building module 202 is used to build a temperature relationship model between the inside and the surface of the capacitor and calibrate the surface temperature of the capacitor.

[0092] The recurrent neural network training module 203 is used to train the recurrent neural network model using the calibrated first sample surface temperature imaging data to obtain a recurrent neural network prediction model.

[0093] The recurrent neural network prediction module 204 is used to predict the calibrated second sample surface temperature imaging data and determine whether to disconnect the circuit according to the prediction result.

[0094] The circuit control module 205 is used to receive the prediction result of the model and determine whether to disconnect the circuit according to the prediction result of the model.

[0095] It is worth mentioning that Figure 2 The device shown is Figure 1 The method embodiments shown correspond to and can be applied to Figure 1 The acquisition of the internal temperature and surface temperature imaging data of the capacitor in the illustrated method embodiment, the establishment of the temperature relationship model between the inside and the surface of the capacitor, the training of the recurrent neural network model, the prediction of the recurrent neural network model, and the judgment of whether to disconnect the circuit based on the prediction results of the model are not repeated here. Those skilled in the art should easily realize that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0096] According to another embodiment of the present invention, a computing device is provided, including a memory and a processor, wherein the memory stores executable code, and when the processor executes the executable code, the Figure 1 The method described.

[0097] According to another embodiment, there is also provided a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to execute a combination of Figure 1 The method described.

[0098] Those skilled in the art should be aware that in one or more of the above examples, the functions described in the present invention can be implemented by hardware, software, firmware or any combination thereof. When implemented by software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium.

[0099] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for suppressing fire during the testing process of polypropylene film capacitors, characterized in that: The method comprises: The internal temperature of the capacitor is monitored by a temperature sensor inside the target capacitor, and the temperature data thereof is recorded, and the surface of the target capacitor is scanned by an infrared imager to obtain first sample surface temperature imaging data; wherein the target capacitor refers to a portion of capacitors selected from all capacitors for measuring the internal temperature; A temperature relationship model is established using the capacitor internal temperature data and the first sample surface temperature imaging data to calibrate the first sample surface temperature imaging data; wherein the first sample surface temperature imaging data is calibrated to make the first sample surface temperature imaging data closer to the real capacitor internal temperature data; the method for establishing the temperature relationship model specifically includes: Establish the thermal impedance model of the capacitor, the thermal impedance model is: in, represents the thermal resistance of the capacitor, represents the thermal capacity of the capacitor, Represents angular frequency , represents the imaginary unit; A temperature relationship model is established based on the internal temperature data of the capacitor and the surface temperature imaging data of the first sample, and the parameters in the thermal impedance model are and As an additional input feature of the temperature relationship model, the temperature relationship model is: in, represents the capacitor surface temperature, represents the internal temperature of the capacitor, , , and represents the model parameters, represents the model error; The least squares method is used to define an objective function to minimize the square error between the predicted value and the actual value. The objective function is: in, represents the surface temperature of the capacitor at the i-th observation point, represents the observed value of the internal temperature of the capacitor at the i-th observation point, and n represents the number of observation points, that is, for each observation point i, from 1 to n, there will be a corresponding capacitor surface temperature value and the observed value of the capacitor internal temperature ; Use an optimization algorithm to minimize the objective function and find The optimal parameters to achieve the minimum , , and , and the obtained parameters 、 、 and Substitute the value of into the linear regression model to obtain the established temperature relationship model; Using the calibrated first sample surface temperature imaging data to train a recurrent neural network model to obtain a recurrent neural network prediction model; In the process of testing all capacitors, an infrared imager is used to scan all capacitors to obtain surface temperature imaging data of the second sample, and the surface temperature imaging data of the second sample is calibrated using the established temperature relationship model; The recurrent neural network prediction model is used to predict the calibrated second sample surface temperature imaging data, and whether to disconnect the circuit is determined based on the prediction result.

2. The method according to claim 1, characterized in that Before establishing the temperature relationship model, the internal temperature data of the capacitor and the surface temperature imaging data of the first sample need to be cleaned, outliers and missing values ​​removed, and normalized.

3. The method according to claim 2, characterized in that The method for normalizing the internal temperature data of the cleaned capacitor and the surface temperature imaging data of the first sample specifically includes: Converting the first sample surface temperature imaging data into a numerical matrix form; Using the standard normal distribution scaling method, the internal temperature data of the capacitor and the surface temperature imaging data of the first sample are converted into a standard normal distribution form with a mean of 0 and a standard deviation of 1. The formula is: in, represents the sample mean, represents the sample standard deviation, Represents the original data, Represents the transformed data.

4. The method according to claim 1, characterized in that: The infrared imager also has a light imaging function, which is used to collect light imaging of the capacitor surface to detect whether the capacitor surface is damaged.

5. The method according to claim 1, characterized in that The capacitor test environment is a closed space. The model's prediction results are used to determine whether nitrogen should be introduced into the test environment to suppress capacitor fire.

6. A device capable of simulating a fire during the testing process of a polypropylene film capacitor, characterized in that: The method according to any one of claims 1 to 5 is used to carry out a simulated polypropylene film capacitor test process, wherein the device comprises: A temperature data acquisition module, used to acquire the internal temperature data of the capacitor, the surface temperature imaging data of the first sample and the surface temperature imaging data of the second sample; A temperature relationship model building module is used to build a temperature relationship model between the inside and the surface of the capacitor and calibrate the surface temperature of the capacitor; A recurrent neural network training module is used to train a recurrent neural network model using the calibrated first sample surface temperature imaging data to obtain a recurrent neural network prediction model; A recurrent neural network prediction module, used for predicting the calibrated second sample surface temperature imaging data; The circuit control module is used to receive the prediction results of the model and determine whether to disconnect the circuit according to the prediction results of the model.

Images