CN112580187A

CN112580187A - Dry-type transformer overheating early warning method and device, computer equipment and storage medium

Info

Publication number: CN112580187A
Application number: CN202011247292.0A
Authority: CN
Inventors: 孔令明; 莫文雄; 栾乐; 许中; 马智远; 周凯; 肖天为; 彭和平; 范伟男; 刘田; 郭倩雯
Original assignee: Guangzhou Power Supply Bureau of Guangdong Power Grid Co Ltd
Current assignee: Guangzhou Power Supply Bureau of Guangdong Power Grid Co Ltd
Priority date: 2020-11-10
Filing date: 2020-11-10
Publication date: 2021-03-30
Anticipated expiration: 2040-11-10
Also published as: CN112580187B

Abstract

The method comprises the steps of obtaining a plurality of items of first historical load data and second historical load data with the generation time after the first historical load data; performing correlation analysis on the first historical load data, and screening out third historical load data which are correlated with the second historical load data from the first historical load data; inputting the third history load data into a target load prediction model to obtain target prediction load data in a future preset time period; and acquiring environment predicted temperature data in a future preset time period, and inputting the target predicted load data and the environment predicted temperature into the hot spot temperature calculation model to obtain hot spot temperature distribution of the dry-type transformer in the future preset time period. According to the method and the device, overheating early warning can be carried out on the dry-type transformer in operation, and corresponding protective measures can be taken for the transformer in time according to overheating prediction results.

Description

Dry-type transformer overheating early warning method and device, computer equipment and storage medium

Technical Field

The application relates to the technical field of electrical detection, in particular to an overheating early warning method and device for a dry-type transformer, computer equipment and a storage medium.

Background

The dry-type transformer has good fireproof performance, so the dry-type transformer is widely applied to places with high fireproof requirements, such as local illumination, residential areas, high-rise buildings or public places. However, for most dry-type transformers, due to the non-uniform temperature distribution during operation, the highest temperature point, i.e., the highest temperature within the equipment, is reached at some point during operation (the highest point being the thermal limit used to measure the maximum load capacity of the dry-type distribution transformer). When the hot spot temperature of the dry-type transformer reaches the hottest spot, the dry-type transformer is easy to cause overheating fault.

In the calculation of the hot spot temperature of the traditional dry-type transformer, a temperature model calculation mode is often adopted, and the dry-type transformer is modeled according to a thermoelectric analogy principle and a heat and mass transfer principle so as to obtain the internal hot spot temperature. Although the on-line monitoring of the hot spot temperature of the dry-type transformer can be realized by using a temperature model calculation mode, the method has the problem of inaccurate prediction, so that the thermal fault of the dry-type transformer cannot be prevented in advance.

Disclosure of Invention

In view of the above, it is necessary to provide an overheat warning method and apparatus for a dry-type transformer, a computer device and a storage medium, which can warn a dry-type transformer of future overheat.

An overheating early warning method for a dry-type transformer, the method comprising:

acquiring historical load data related to a dry-type transformer, wherein the historical load data comprises a plurality of items of first historical load data and second historical load data with generation time after the first historical load data;

performing correlation analysis on the plurality of items of first historical load data, and screening out a plurality of items of third historical load data which have correlation with the second historical load data from the plurality of items of first historical load data;

inputting the plurality of items of third history load data into a target load prediction model, and processing the plurality of items of third history load data through the target load prediction model to obtain target prediction load data in a future preset time period;

acquiring environment predicted temperature data in the future preset time period, and inputting target predicted load data and environment predicted temperature corresponding to the future preset time period into a hot spot temperature calculation model to obtain hot spot temperature distribution of the dry-type transformer in the future preset time period;

and when the hot spot temperature distribution meets the temperature early warning condition, triggering the early warning action on the dry type transformer.

In one embodiment, the obtaining historical load data related to the dry-type transformer comprises:

acquiring original load data related to the dry-type transformer in a historical preset time period;

deleting repeated data in the original load data to obtain intermediate processing data;

determining a target data segment with a load value missing condition in the intermediate processing data;

when the time period length corresponding to the target data segment is less than or equal to a preset time length, performing interpolation processing on the target data segment to fill up the missing load value in the target data segment;

when the time period length corresponding to the target data period is greater than the preset time period length, filling the load value according to the load average value in the same time period in the historical time period;

and taking the intermediate processing data after the completion of the load value filling as historical load data related to the dry-type transformer.

In one embodiment, the performing a correlation analysis on the plurality of items of first historical load data and screening out a plurality of items of third historical load data having a correlation with the second historical load data from the plurality of items of first historical load data includes:

for each item of first historical load data, calculating covariance between the first historical load data and the second historical load data according to expected values corresponding to the first historical load data and the second historical load data respectively;

calculating a correlation coefficient between each first historical load data and each second historical load data according to the variance and the covariance corresponding to the first historical load data and the second historical load data respectively;

and taking the first historical load data corresponding to the relation number which is more than or equal to a preset coefficient threshold value as third historical load data which has correlation with the second historical load data.

In one embodiment, the inputting the target predicted load data and the environmental predicted temperature corresponding to the future preset time period into a hot spot temperature calculation model to obtain a hot spot temperature distribution of the dry-type transformer in the future preset time period includes:

acquiring equipment parameters of the dry-type transformer;

calculating the hot point temperature rise of the hot point temperature under different load conditions relative to the environment predicted temperature at each time point according to the target predicted load data and the equipment parameters in the future preset time period; the different load conditions include at least one of a rated load condition, a given load rate condition, a continuous load condition, and a transient overload condition;

determining the hot spot temperature corresponding to each time point in the future preset time period according to the environment predicted temperature and the hot spot temperature rise at each time point in the future preset time period;

and counting the hot spot temperatures at all time points to obtain the hot spot temperature distribution of the dry-type transformer in the future preset time period.

In one embodiment, the calculating the hot spot temperature rise of the hot spot temperature under different load conditions relative to the environmental predicted temperature at each time point according to the target predicted load data in the future preset time period and the device parameters includes:

when the load condition is a rated load condition, determining a hot spot coefficient according to the equipment parameters, and calculating the hot spot temperature rise of the hot spot temperature relative to the environment predicted temperature at each time point according to the hot spot coefficient and the average temperature rise under the rated load;

when the load condition is a given load rate condition, calculating the hot spot temperature rise of the hot spot temperature relative to the environment predicted temperature at each time point according to the hot spot coefficient determined by the equipment parameters, the thermoelectric coefficient, the average temperature rise under the rated load and the given load rate determined according to target predicted load data;

when the load condition is a continuous load condition, determining the temperature rise under the rated load according to the equipment parameters, and calculating the hot spot temperature rise of the hot spot temperature relative to the environment predicted temperature at each time point according to the temperature rise under the rated load and the corresponding load rate under the continuous load determined according to the target predicted load data;

and when the load condition is a transient overload condition, determining a hot spot temperature rise change value according to the initial hot spot temperature rise at the beginning of any load rate and the termination hot spot temperature rise when the load rate is not changed, and calculating the hot spot temperature rise of the hot spot temperature at each time point relative to the environment predicted temperature according to the hot spot temperature rise change value and the initial hot spot temperature rise.

In one embodiment, the triggering an early warning action on the dry-type transformer when the hot spot temperature distribution meets a temperature early warning condition includes:

when the hot spot temperature corresponding to any time period in the hot spot temperature distribution is greater than a preset first temperature early warning threshold value and the corresponding time period is greater than or equal to the first time period threshold value, triggering a first-level early warning prompt action;

when the hot spot temperature corresponding to any time period in the hot spot temperature distribution is greater than a preset second temperature early warning threshold value and the corresponding time period is greater than or equal to the second time period threshold value, triggering early warning prompting actions for early warning a second level; the second temperature early warning threshold is larger than the first temperature early warning threshold, and the second time period threshold is smaller than the first time period threshold.

In one embodiment, the method further comprises:

obtaining sample historical load data and sample reference load data generated after the sample historical load data;

inputting the sample historical load data into a target load prediction model to be trained, processing the sample historical load data through the target load prediction model to be trained, and outputting a prediction result;

and adjusting the model parameters of the target load prediction model to be trained according to the difference between the prediction result and the sample parameter load data until the training stop condition is reached, and stopping training to obtain the trained target load prediction model.

An overheating early warning device for a dry type transformer, the device comprising:

the system comprises a historical load data acquisition module, a load management module and a load management module, wherein the historical load data acquisition module is used for acquiring historical load data related to the dry-type transformer, and the historical load data comprises a plurality of items of first historical load data and second historical load data with the generation time after the first historical load data;

the correlation analysis module is used for carrying out correlation analysis on the plurality of items of first historical load data and screening out a plurality of items of third historical load data which have correlation with the second historical load data from the plurality of items of first historical load data;

the load prediction module is used for inputting the third history load data into a target load prediction model, and processing the third history load data through the target load prediction model to obtain target prediction load data in a future preset time period;

the hot spot temperature prediction module is used for acquiring environment predicted temperature data in the future preset time period, inputting target predicted load data and environment predicted temperature corresponding to the future preset time period to a hot spot temperature calculation model, and obtaining hot spot temperature distribution of the dry-type transformer in the future preset time period;

and the temperature early warning module is used for triggering early warning action on the dry-type transformer when the hot spot temperature distribution meets the temperature early warning condition.

A computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:

A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:

According to the dry-type transformer overheating early warning method, the dry-type transformer overheating early warning device, the computer equipment and the storage medium, through correlation analysis of the multiple items of first historical load data, multiple items of third historical load data which are correlated with the second historical load data are screened out from the multiple items of first historical load data. And processing the obtained multiple items of third history load data through a pre-constructed load prediction model so as to obtain prediction data of the target load data. Therefore, correlation analysis is carried out based on the periodic change rule of the first historical load data in time, and some third historical load data which can reflect the internal incidence relation of the load can be screened out from the plurality of items of first historical load data, so that the prediction accuracy can be effectively improved under the condition of reducing the data input quantity. In addition, the load prediction result and the hot spot temperature calculation model are combined, the hot spot temperature prediction in the future preset time period is realized, the overheating early warning can be performed on the running dry-type transformer based on the preset temperature early warning condition, in addition, corresponding preventive measures can be taken for the dry-type transformer in time according to the overheating early warning result, and the reliability of the power distribution network is effectively improved.

Drawings

Fig. 1 is an application environment diagram of an overheating warning method for a dry-type transformer in one embodiment;

fig. 2 is a schematic flow chart illustrating an overheat warning method for a dry-type transformer according to an embodiment;

FIG. 3 illustrates load rates predicted by the dry-type transformer overheating early warning method within a certain three days in the future;

FIG. 4 is a predicted hotspot temperature within a certain three days in the future based on the dry-type transformer overheating warning method;

fig. 5 is a block diagram illustrating a dry-type transformer overheat warning device according to an embodiment;

FIG. 6 is a diagram illustrating an internal structure of a computer device according to an embodiment.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.

The overheating early warning method for the dry-type transformer can be applied to the application environment shown in fig. 1. Wherein dry-type transformer 102 communicates with computer device 104 over a network. The computer device mentioned in the present application may be a terminal or a server, and the dry-type transformer overheating warning method mentioned in the embodiments may be implemented by being executed by the computer device. Taking an example that the server cooperatively executes to implement the dry-type transformer overheating early-warning method in the present application, the server acquires a plurality of items of first historical load data related to the dry-type transformer, second historical load data with generation time after the first historical load data, and environmental prediction temperature data within a future preset time period. Performing correlation analysis on each item of first historical load data in the server, and further screening a plurality of items of third historical load data which are correlated with the second historical load data from the plurality of items of first historical load data; the server can also input each item of the obtained third history load data into the target load prediction model to predict target predicted load data in a future preset time period. The server can also input target predicted load data and environment predicted temperature into the hot spot temperature calculation model when hot spot temperature prediction is carried out, hot spot temperature distribution of the dry-type transformer in a future preset time period is calculated, and when the hot spot temperature distribution meets temperature early warning conditions, early warning action on the dry-type transformer is triggered. The terminal can be, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers and portable wearable devices, and the server can be implemented by an independent server or a server cluster formed by a plurality of servers.

In an embodiment, as shown in fig. 2, there is provided a dry-type transformer overheating warning method, which is exemplified by applying the method to the computer device in fig. 1, and includes the following steps:

step S202, historical load data related to the dry-type transformer is obtained, and the historical load data comprises a plurality of items of first historical load data and second historical load data with the generation time after the first historical load data.

The load data refers to an electric load, which is also called an "electrical load," and may be, in this application, specifically, a sum of electric powers that the dry-type transformer takes from the electric power system at a certain time. Because the load changes randomly, the corresponding load changes every time the electric equipment is started or stopped, so that certain regularity can be found by analyzing load data in a certain time period to some extent.

The historical load data is the sum of the electrical power taken by the dry-type transformer to the electrical power system over the historical period of time. In one embodiment, the computer device may select the uniform time interval from the historical load data at the uniform time interval, for example, in units of minutes, hours, or days, to obtain the first historical load data corresponding to different time points respectively.

Specifically, the computer device may collect in advance first historical load data and generate second historical load data at a time subsequent to the first historical load data. When the computer device is a terminal, the terminal may provide an information input page, and collect first historical load data and second historical load data whose generation time is after the first historical load data through the information input page. The input page may be a page provided by an application program, or a web page provided by a programming web page, and the like, which is not limited in the embodiment of the present application. When the computer device is a server, the server may acquire first historical load data acquired by the terminal through the information input page and second historical load data whose generation time is after the first historical load data.

In one embodiment, the terminal may provide an information input page and collect first historical load data through the information input page, and generate second historical load data at a time subsequent to the first historical load data.

In a specific embodiment, the first historical load data may be historical load data such as yesterday historical load data (1 day), previous three-day historical load data (3 days), previous seven-day load data (7 days), previous week and week historical load data (1 day), previous week and week type historical load data (4 days), previous thirty days historical load data (30 days), and the like. The data type corresponding to yesterday historical load data (1 day) can be hourly load data, daily load mean, daily load minimum and daily load maximum; the data type corresponding to the first three calendar history load data (3 days) can be the average value of the loads of the first three days per hour; the data type corresponding to the historical load data (7 days) of the previous seven days can be the average value of the loads per hour of the previous seven days; the data type corresponding to the load data (1 day) in the same week of the previous week can be the average value of the load data per hour and the load data per day; the data type corresponding to the historical load data (4 days) of the same week type in the previous four weeks can be the average value of the load data every time; the data type corresponding to the previous thirty days of historical load data (30 days) may be a load average value per time, which is not limited in the embodiment of the present application.

In one embodiment, the historical load data may be pre-stored in a database or stored on the terminal device. When the historical load data needs to be acquired, the computer equipment only needs to acquire and call from a database or terminal equipment.

Step S204, carrying out correlation analysis on the plurality of items of first historical load data, and screening out a plurality of items of third historical load data which have correlation with the second historical load data from the plurality of items of first historical load data.

Specifically, the computer device may perform correlation analysis on each item of first historical load data based on a periodic variation rule of the load data, so as to calculate a correlation coefficient between each item of first historical load data and each item of second historical load data. And the computer equipment can screen out first historical load data corresponding to the correlation coefficient which is greater than or equal to a preset coefficient threshold from the first historical load data according to the obtained correlation coefficients, and the first historical load data is used as third historical load data which is relevant to the second historical load data.

It can be understood from the prior experience that the periodic variation law of the load data is mainly reflected in the following aspects: (1) the overall regularity of the daily load curves of different days is similar, and the correlation between the daily load of the previous day and the daily load of the previous seven days is strongest; (2) the load laws of the same week type day are similar; (3) the load laws of the working day and the rest day are respectively similar; (4) the law of statutory festival and holiday is similar in different years.

In one embodiment, the present application provides further guidance for obtaining the third history load data based on the above-mentioned periodic variation rule of the load data. Wherein, the correlation analysis is carried out on each item of first historical load data, and a plurality of items of third historical load data which have correlation with the second historical load data are screened out from the plurality of items of first historical load data, and the method comprises the following steps: for each item of first historical load data, calculating covariance between the first historical load data and the second historical load data according to expected values corresponding to the first historical load data and the second historical load data respectively; calculating a correlation coefficient between each first historical load data and each second historical load data according to the corresponding variance and covariance of the first historical load data and the second historical load data; and taking the first historical load data corresponding to the relation number more than or equal to the preset coefficient threshold value as third historical load data having correlation with the second historical load data.

Specifically, the computer device may determine a covariance between the first historical load data and the second historical load data, and determine a variance corresponding to each of the first historical load data and the second historical load data. And calculating a correlation coefficient between each first historical load data and each second historical load data based on each covariance and each variance. Furthermore, the computer device may set the first historical load data corresponding to the relationship number equal to or greater than the preset coefficient threshold as third historical load data having a correlation with the second historical load data.

In a particular embodiment, the computer device may determine the covariance between the first historical load data and the second historical load data by equation (1):

cov(X,Y)＝E[(X-μ_X)(Y-μ_Y)]； (1)

in the formula (1), X is a sequence of first historical load data, and Y is a sequence of second historical load data; mu.s_XFor the expected value, mu, corresponding to the first historical load data_YAnd the expected value is corresponding to the second historical load data.

In a specific embodiment, the computer device may determine the correlation coefficient ρ between each of the first historical load data and the second historical load data using a person correlation coefficient, i.e., by equation (2):

in the formula (2), σ_XIs the corresponding variance, sigma, of the first historical load data_YThe variance corresponding to the second historical load data.

In a specific embodiment, the computer device may select 15 items of the first historical load data, wherein the selected first historical load data types include: average of yesterday load rate

(average of daily load rate), and minimum daily load rate V_t-1(min)(minimum value of yesterday load rate) and maximum value of day load rate V_t-1(max)(highest value of yesterday load factor), load factor V at time t every day of the first 7 days_t(d-1)、V_t(d-2)、V_t(d-3)、V_t(d-4)、V_t(d-5)、V_t(d-6)And V_t(d-7)(from the first 1 to the first 7 days), and a daily load factor V at time t of the same week type in the first 4 weeks_t(w-1)、V_t(w-2)、V_t(w-3)、V_t(w-4)And average load rate of load data of previous 30 days at time t

If the preset coefficient threshold is set to 0.4 in the current embodiment, the first historical load data with the relationship number ρ greater than or equal to 0.4 is used as the third historical load data with which the second historical load data is most related.

In the above embodiment, the computer device screens out a plurality of items of third historical load data having a correlation with the second historical load data from the plurality of items of first historical load data, so that the computer device can screen out some load data more reflecting the internal association relationship of the load, and thus the prediction accuracy can be effectively improved under the condition of reducing the data input amount.

And step S206, inputting the multiple items of third history load data into the target load prediction model, and processing the multiple items of third history load data through the target load prediction model to obtain target prediction load data in a future preset time period.

Specifically, the target load prediction model may be constructed based on a GRU (Gated Recurrent Unit) neural network, or may be constructed by using other algorithms, such as a time series algorithm and an artificial neural network, and the embodiment of the present application is not limited thereto.

In one embodiment, a practical application environment of the GRU neural network is built by using computer equipment, wherein the equipment carrying the load prediction algorithm can be a Windows10 (a computer operating system) 64-bit operating system, and the computer equipment can also build a GRU neural network model based on a Keras (an open source artificial neural network library) deep learning framework of a Python (a cross-platform computer programming language) development environment. The bottom layer of the GRU neural network model may use tensrflow (symbolic mathematical system based on data flow programming), which is not limited in this application. It should be noted that the tensrflow system is an end-to-end open source machine learning platform, and it has a comprehensive and flexible ecosystem, which contains various tools, libraries and community resources, so that developers can easily build and deploy applications supported by machine learning.

In a specific embodiment, the target load prediction model is built based on a GRU neural network, and the target load prediction model can be obtained by training through the following steps: acquiring sample historical load data and sample reference load data generated after the sample historical load data; inputting the sample historical load data into a target load prediction model to be trained, processing the sample historical load data through the target load prediction model to be trained, and outputting a prediction result; and adjusting the model parameters of the target load prediction model to be trained according to the difference between the prediction result and the sample parameter load data until the training stop condition is reached, and stopping training to obtain the trained target load prediction model. Please refer to fig. 3, which is a predicted load rate in a third day in the future based on the dry-type transformer overheating warning method, wherein a straight line denoted by reference numeral 310 in fig. 3 represents the predicted load rate in the third day in the future in advance, and a straight line denoted by reference numeral 312 represents the actually generated load rate in the third day. As can be seen from fig. 3, when the load rate predicted in advance by the present application in a certain three days in the future is compared with the load rate actually generated in the three days, it can be seen that the difference between the two load rates is not more than 0.3 at most, which further illustrates the effectiveness and accuracy of predicting the load data in the preset time period in the future by using the scheme disclosed in this embodiment.

The training stopping condition is a condition for stopping training, and specifically may be that a preset iteration number is reached, a preset iteration time is reached, or the prediction performance of the model obtained by training reaches a preset index, and the like. Specifically, the building process of the target load prediction model can be implemented by using computer equipment. When the target load prediction model is built by using computer equipment, the structure of the model may be composed of 1 input layer, 2 GRU layers and 1 output layer, the output layer is a fully connected layer, the output dimension is 1, and a sigmoid function (a S-type function common in biology, also referred to as an S-type growth curve) is used as an activation function, and of course, the target load prediction model may have other model structures, and the like, which is not limited in the embodiment of the present application.

In one embodiment, when the target load prediction model is compiled and trained by using a computer device, an Adam optimization algorithm (a first-order optimization algorithm that can replace a conventional random gradient descent process) may also be used to iteratively update the weight values of the neural network based on the sample historical load data and the sample reference load data, where in the above iterative process, a mean _ squared _ error function may be used as a loss function, and epochs (the number of single training iterations of all batches in the forward and backward propagation processes) is set to 700 times, and batch _ size (the number of samples selected in one training) is set to 50, which is not limited in the embodiments of the present application. It should be noted that, in order to obtain more realistic prediction data, the computer device may train a more complex GRU neural network model by increasing the number of epochs.

In the embodiment, the computer equipment is used for building the target load prediction model by integrating the GRU neural network, and based on the characteristic that the GRU neural network can efficiently process the problem highly related to the time sequence, various change rules in the historical load data can be effectively learned, so that the prediction accuracy can be effectively improved while the target prediction load data is predicted by using the historical load data.

Step S208, obtaining environment prediction temperature data in a future preset time period, inputting target prediction load data and environment prediction temperature corresponding to the future preset time period to the hot spot temperature calculation model, and obtaining hot spot temperature distribution of the dry-type transformer in the future preset time period.

The environment predicted temperature data refers to predicted data acquired by the computer device from a meteorological resource database.

Specifically, the environmental prediction temperature data stored in the weather resource database may be data acquired from weather forecast websites such as chinese weather bureau websites; the environment temperature data may also be obtained by performing prediction processing based on historical environment temperature data, which is not limited in the embodiment of the present application.

In one embodiment, the computer device may obtain the district future weather forecast data, i.e., the environmental forecast temperature data, from the chinese weather station website. In another embodiment, to simplify future ambient predicted temperature data, the ambient predicted temperature data may be uniformly set to 30 degrees celsius.

Specifically, the computer device may establish the hot spot temperature calculation model according to the dry type transformer load guidance provided in the power transformer design specification and the consequences caused by the overload operation of the dry type transformer provided in the design specification.

In one embodiment, the computer device may build a hot spot temperature calculation model for calculating the hot spot temperature of the dry transformer over time under different load conditions based on the dry transformer load guidelines provided in IEC 60076-12:2008 and GB 1094.11-2008 design specifications, and the consequences that may result from the dry transformer overload operation given in the design specifications. The target predicted load data and the environment predicted temperature corresponding to the future preset time period can be processed through the currently established hot spot temperature calculation model, and then hot spot temperature distribution of the dry-type transformer corresponding to the future preset time period is calculated.

In a specific embodiment, the hot spot temperature calculation model can be constructed by formula (3):

T＝273+θ_a+Δθ_HSn (3)

in the formula (3), θ_aThe ambient temperature is given in units of "° c"; delta theta_HSnThe unit is K for the temperature rise corresponding to the hot spot temperature relative to the environment temperature under different load conditions determined based on the load data; t is the hotspot temperature.

Specifically, with reference to the above specific embodiment, the computer device calculates, based on the obtained target predicted load data, a hot-spot temperature rise at each time point of the hot-spot temperature relative to the environment predicted temperature under different load conditions, and further obtains, according to the obtained hot-spot temperature rise and the environment predicted temperature, a hot-spot temperature distribution of the dry-type transformer in a future preset time period.

In one embodiment, inputting, by the computer device, the target predicted load data and the environment predicted temperature corresponding to the future preset time period to the hot spot temperature calculation model to obtain the hot spot temperature distribution of the dry-type transformer in the future preset time period includes: acquiring equipment parameters of the dry-type transformer; according to target predicted load data and equipment parameters in a future preset time period, calculating the hot point temperature rise of the hot point temperature relative to the environment predicted temperature at each time point under different load conditions; the different load conditions include at least one of a rated load condition, a given load rate condition, a continuous load condition, and a transient overload condition; determining the hot spot temperature corresponding to each time point in the future preset time period according to the environment predicted temperature and the hot spot temperature rise at each time point in the future preset time period; and counting the hot spot temperatures at all time points to obtain the hot spot temperature distribution of the dry-type transformer in a future preset time period. The obtained hot spot temperature distribution result can refer to fig. 4, in which a bar graph indicates the predicted load rate in a certain three days in the future, a straight line indicates the predicted hot spot temperature in the certain three days in the future, a broken line numbered 410 indicates the preset temperature warning value in the certain three days in the future, a broken line numbered 412 indicates the preset temperature risk value in the certain three days in the future, and a broken line numbered 414 indicates the predicted ambient temperature in the certain three days in the future (in the current embodiment, the ambient predicted temperature data is uniformly set to 30 degrees celsius). As can be seen from fig. 4, in the present embodiment, the temperature warning value is set to 105 degrees celsius and 130 degrees celsius. The change trend of the hot spot predicted temperature can be intuitively reflected through the hot spot temperature distribution diagram, so that a user can conveniently master the prediction progress in real time and take corresponding protection measures for the dry-type transformer in time.

Specifically, the computer device calculates and obtains the hot point temperature rise of the hot point temperature relative to the environment predicted temperature at each time point under different load conditions based on the obtained target predicted load data.

In one embodiment, the calculating, by the computer device, the hot spot temperature rise of the hot spot temperature at each time point relative to the environmental predicted temperature under different load conditions according to the target predicted load data and the device parameters within the preset time period in the future includes:

and when the load condition is a rated load condition, the computer equipment determines a hot spot coefficient according to the equipment parameters, and calculates the hot spot temperature rise of the hot spot temperature at each time point relative to the environment predicted temperature according to the hot spot coefficient and the average temperature rise under the rated load. The computer device can calculate the hot spot temperature rise of the hot spot temperature under the current load condition relative to the environment predicted temperature at each time point according to the formula (4):

Δθ_HS,r＝Z×Δθ_wr； (4)

in the formula (4), Delta Theta_HS,rFor the hot spot temperature rise,. DELTA.theta_wrTaking the unit of K as the average temperature rise under the rated load; z is a hot spot coefficient, and the value of the parameter Z can be determined to be 1.25 through design parameters.

And when the load condition is a given load rate condition, calculating the hot spot temperature rise of the hot spot temperature relative to the environment predicted temperature at each time point according to the hot spot coefficient determined by the equipment parameters, the thermoelectric coefficient, the average temperature rise under the rated load and the given load rate determined according to the target predicted load data. The computer device can calculate the hot spot temperature rise of the hot spot temperature under the current load condition relative to the environment predicted temperature at each time point according to the formula (5):

in the formula (5), the reaction mixture is,

for a given load rate determined from the target predicted load data, the coefficient q takes a value of 1.6 in the case of natural cooling (AN) and 2 in the case of forced air cooling (AF).

And when the load condition is a continuous load condition, determining the temperature rise under the rated load according to the equipment parameters, and calculating the hot point temperature rise of the hot point temperature relative to the environment predicted temperature at each time point according to the temperature rise under the rated load and the corresponding load rate under the continuous load determined by the target predicted load data. The computer device can calculate the hot spot temperature rise of the hot spot temperature under the current load condition relative to the environment predicted temperature at each time point according to the formula (6):

Δθ_HS＝θ_HS,r×I^2m； (6)

in the formula (6), θ_HS,rIn order to increase the temperature under the rated load, I is the load factor under the continuous load, and m is the operation index of the dry-type transformer for natural cooling under the continuous load, in an embodiment, the value is 0.8, which is not limited in this embodiment.

When the load condition is a transient overload condition, determining a hot spot temperature rise change value according to the initial hot spot temperature rise when any load rate starts and the termination hot spot temperature rise when the load rate does not change, and calculating the hot spot temperature rise of the hot spot temperature at each time point relative to the environment predicted temperature according to the hot spot temperature rise change value and the initial hot spot temperature rise. The computer device can calculate the hot spot temperature rise of the hot spot temperature under the current load condition relative to the environment predicted temperature at each time point according to the formula (7):

Δθ_t＝(Δθ_U-Δθ_i)[1-e^-t/τ]+Δθ_i； (7)

in the formula (7), Δ θ_iThe temperature rise (K) of the hot spot at the beginning of a certain load factor; delta theta_tThe temperature rise of the hot spot after the load changes for t time; delta theta_UThe final hot point temperature rise under the condition that the load rate is not changed; t is the load change time; τ is the time constant at a given load. Wherein the time constant τ is a time required for the temperature rise to reach 63.2% over a stable value of the ambient temperature when the load of the dry type transformer is changed. It is usually 5^τAt this point in time the dry-type transformer reaches a steady state.

Specifically, the time constant τ can be determined by the computer device according to the parameter m, and it can be understood that τ will change when the hot spot temperature rise is changed.

In one embodiment, the computer device may perform the calculation of the time constant τ by equation (8):

in the formula (8), τ_RA time constant corresponding to a dry-type transformer under a rated load, C an effective heat capacity, theta_eIs the influence of the iron core on the temperature rise during no load, P_rThe total loss at rated load and rated temperature rise. When m ≠ 1, the time constant τ is calculated according to equation (9):

specifically, the computer equipment can acquire the equipment parameters based on the nameplate information and the design parameters of the transformer. In one embodiment, the device parameters that need to be obtained are shown in table 1.

Table 1: equipment parameter data required by hot spot temperature calculation model construction

In one embodiment, the hot spot temperature corresponding to the dry-type transformer can be used for evaluating the life loss of the dry-type transformer in a specific time period, so that the effective evaluation and prediction of the hot spot temperature can enable a user to comprehensively master the loss condition of the dry-type transformer.

In one embodiment, equipment parameters required for constructing a hot spot temperature calculation model are determined by acquiring nameplate information and design parameters of a transformer, and under different load conditions, the hot spot temperature prediction method combining predicted load data and the hot spot temperature calculation model is provided, so that the defects of high manufacturing cost, poor feasibility, incapability of realizing hot spot temperature online monitoring and real-time prediction caused by the traditional temperature monitoring device are overcome.

Step S210, when the hot spot temperature distribution meets the temperature early warning condition, triggering the early warning action on the dry type transformer.

Specifically, the computer device sets a temperature early warning condition according to the temperature, namely the heat resistance level, of the insulation system of the dry-type transformer, wherein the temperature early warning condition comprises a temperature early warning value and a temperature danger value. The heat resistance levels can refer to table 2, and table 2 gives the temperature warning values and the value criteria of the temperature risk values corresponding to different heat resistance levels, for example, according to the heat resistance level "105 (a)", the set temperature warning value in the shift period can be determined to be 105, the temperature risk value is 130, and the like.

TABLE 2 temperature limits of dry transformers for different heat resistance ratings

In an embodiment, when the temperature pre-warning condition is set according to the heat-resistant grade of the dry-type transformer, the following two ways may be considered:

the first method is as follows: the computer device sets a temperature early warning condition based on a set temperature early warning value, that is, when a hot spot predicted temperature of a dry-type transformer exceeds the temperature early warning value at a certain time period on a certain day and the exceeding time is maintained above a first preset time threshold (in a specific embodiment, a value of the first preset time threshold may be 3 hours, which is not limited in this embodiment), the computer device outputs prompt type information through a terminal or other transmission devices to prompt a user. If the terminal is provided with a display screen, the prompt information can be directly displayed on the screen. In a specific embodiment, the prompt type information may be terms such as "notice" and the like, which are not limited in the embodiment of the present application.

The second method comprises the following steps: when a temperature early warning condition is set based on the temperature danger value, that is, when the hot spot of a dry-type transformer in a certain time period on a certain day predicts that the temperature exceeds the temperature danger value and the time of exceeding exceeds the time is maintained above a second preset time threshold (in a specific embodiment, the value of the second preset time threshold may be 1 hour, which is not limited in this embodiment of the present application), the temperature early warning condition is output through a terminal or other transmission equipment to prompt a user. If the terminal is provided with a display screen, the alarm information can be directly displayed on the screen. In a specific embodiment, the alarm information may be a term such as "danger", and the embodiment of the present application does not limit this.

In the dry-type transformer overheating early warning method, correlation analysis is performed on the plurality of items of first historical load data, and a plurality of items of third historical load data with correlation of the second historical load data are screened out from the plurality of items of first historical load data. And processing the obtained multiple items of third history load data through a pre-constructed load prediction model so as to obtain target prediction load data in a future preset time period. Therefore, correlation analysis is carried out based on the periodic change rule of the first historical load data in time, some load data which can reflect the internal incidence relation of the load can be screened from the plurality of items of first historical load data, and therefore the prediction precision can be effectively improved under the condition that the data input quantity is reduced. In addition, the load prediction result and the hot spot temperature calculation model are combined, the hot spot temperature prediction in the future preset time period is realized, the overheating early warning can be performed on the running dry-type transformer based on the preset temperature early warning condition, in addition, corresponding preventive measures can be taken for the dry-type transformer in time according to the overheating early warning result, and the reliability of the power distribution network is effectively improved.

In one embodiment, historical load data associated with a dry-type transformer is obtained, the steps comprising: acquiring original load data related to the dry-type transformer in a historical preset time period; deleting repeated data in the original load data to obtain intermediate processing data; determining a target data segment with a load value missing condition in the intermediate processing data; when the time period length corresponding to the target data segment is less than or equal to the preset time length, performing interpolation processing on the target data segment to fill up the missing load value in the target data segment; when the time period length corresponding to the target data segment is greater than the preset time period length, filling the load value according to the load average value in the same time period in the historical time period; and taking the intermediate processing data after the completion of the load value filling as historical load data related to the dry-type transformer.

Specifically, the computer device may obtain the intermediate processing data, and perform interpolation processing on the target data segment based on the target data segment with the load value missing condition in the intermediate processing data, so as to fill up the missing load value in the target data segment.

In one embodiment, the computer device can find missing data segments in the historical load data from the intermediate processing data and specify the time length corresponding to each missing data segment. The data interval is compared with a preset time length by analyzing the time lengths corresponding to the missing data segments one by one. And according to the comparison result, filling the missing load value by adopting an interpolation method for the first historical load data with the data interval smaller than the preset time length. And for the first historical load data with the data interval being greater than or equal to the preset time length, filling the load value by adopting the average value of the load data in the historical time period in the same time period or other data indexes capable of reflecting the average value.

In one embodiment, the historical time period and the preset time length may be set according to actual conditions, and the embodiment of the present application is not limited by this.

In the above embodiment, the interpolation method is an important method for discrete function approximation, and by using the method, the approximate values of the discrete function at other points can be estimated according to the value conditions of the discrete function at a limited number of points, so that missing data is filled based on the estimated approximate values.

In the present embodiment, an implementation flow of preprocessing historical load data is specifically described, it should be noted that a method of preprocessing historical load data is not unique, and in other embodiments, a general method may be summarized according to a problem in load data generated in an actual operation condition of a dry-type transformer, which is not limited in the present embodiment. In the embodiment, the problems that redundancy and deficiency often occur to the generated historical load data in the actual operation process of the dry-type transformer are considered, the historical load data are preprocessed in advance before load prediction is performed, and the preprocessed historical load data can reflect the load change rule more intuitively, so that the accuracy of load prediction can be effectively improved based on the embodiment.

It should be understood that, although the steps in the flowchart of fig. 2 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least a portion of the steps in fig. 2 may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, which are not necessarily performed in sequence, but may be performed in turn or alternately with other steps or at least a portion of the other steps or stages.

In one embodiment, as shown in fig. 5, there is provided an overheating early warning device 500 for a dry type transformer, including: the system comprises a historical load data acquisition module 501, a correlation analysis module 502, a load prediction module 503, a hot spot temperature prediction module 504 and a temperature early warning module 505, wherein:

the historical load data acquiring module 501 is configured to acquire historical load data related to the dry-type transformer, where the historical load data includes a plurality of items of first historical load data and second historical load data generated at a time subsequent to the first historical load data.

The correlation analysis module 502 is configured to perform correlation analysis on the first historical load data, and screen out a plurality of third historical load data having correlations with the second historical load data from the plurality of first historical load data.

And the load prediction module 503 is configured to input the multiple items of third history load data to the target load prediction model, and process the multiple items of third history load data through the target load prediction model to obtain target predicted load data in a future preset time period.

The hot spot temperature prediction module 504 is configured to obtain environment predicted temperature data in a future preset time period, and input target predicted load data and an environment predicted temperature corresponding to the future preset time period to the hot spot temperature calculation model to obtain hot spot temperature distribution of the dry-type transformer in the future preset time period;

and the temperature early warning module 505 is configured to trigger an early warning action on the dry-type transformer when the hot spot temperature distribution meets a temperature early warning condition.

In one embodiment, the historical load data obtaining module 501 includes an original load data obtaining module, a redundancy processing module, a target data segment searching module, a first target data segment padding module, a second target data segment padding module, and an output module, where:

the system comprises an original load data acquisition module, a data processing module and a data processing module, wherein the original load data acquisition module is used for acquiring original load data related to the dry-type transformer in a historical preset time period;

the redundancy processing module is used for deleting repeated data in the original load data to obtain intermediate processing data;

the target data segment searching module is used for determining a target data segment with a load value missing condition in the intermediate processing data;

the first filling module is used for performing interpolation processing on the target data segment to fill up the missing load value in the target data segment when the time segment length corresponding to the target data segment is less than or equal to the preset time length;

the second filling module is used for filling the load value according to the load average value in the historical time period under the same time period when the time period length corresponding to the target data segment is greater than the preset time period;

and the output module is used for taking the intermediate processing data after the completion of the load value filling as historical load data related to the dry-type transformer.

In one embodiment, the correlation analysis module 502 includes a covariance calculation module, a correlation coefficient calculation module, and a filtering module, wherein:

the covariance calculation module is used for calculating covariance between the first historical load data and the second historical load data according to expected values corresponding to the first historical load data and the second historical load data respectively for each item of first historical load data;

the correlation coefficient calculation module is used for calculating the correlation coefficient between each first historical load data and each second historical load data according to the corresponding variance and covariance of the first historical load data and the second historical load data;

and the screening module is used for taking the first historical load data corresponding to the relation number greater than or equal to the preset coefficient threshold value as third historical load data which has correlation with the second historical load data.

In one embodiment, the hot spot temperature prediction module 504 includes a device parameter obtaining module, a hot spot temperature rise calculating module, a hot spot temperature calculating module, and a hot spot temperature distribution statistical module, where:

the equipment parameter acquisition module is used for acquiring the equipment parameters of the dry-type transformer;

the hot spot temperature rise calculation module is used for calculating the hot spot temperature rise of the hot spot temperature relative to the environment predicted temperature at each time point under different load conditions according to target predicted load data and equipment parameters in a future preset time period; the different load conditions include at least one of a rated load condition, a given load rate condition, a continuous load condition, and a transient overload condition;

the hot spot temperature calculation module is used for determining the hot spot temperature corresponding to each time point in the future preset time period according to the environment predicted temperature and the hot spot temperature rise at each time point in the future preset time period;

and the hot spot temperature distribution counting module is used for counting the hot spot temperatures at all time points to obtain the hot spot temperature distribution of the dry-type transformer in a future preset time period. The method for calculating the hot spot temperature rise of the hot spot temperature relative to the environmental predicted temperature under different load conditions at each time point comprises the following steps:

when the load condition is a rated load condition, determining a hot spot coefficient according to equipment parameters, and calculating the hot spot temperature rise of the hot spot temperature relative to the environment predicted temperature at each time point according to the hot spot coefficient and the average temperature rise under the rated load;

when the load condition is a given load rate condition, calculating the hot point temperature rise of the hot point temperature relative to the environment predicted temperature at each time point according to the hot point coefficient determined by the equipment parameters, the thermoelectric coefficient, the average temperature rise under the rated load and the given load rate determined according to the target predicted load data;

when the load condition is a continuous load condition, determining the temperature rise under the rated load according to the equipment parameters, and calculating the hot point temperature rise of the hot point temperature relative to the environment predicted temperature at each time point according to the temperature rise under the rated load and the corresponding load rate under the continuous load determined according to the target predicted load data;

when the load condition is a transient overload condition, determining a hot spot temperature rise change value according to the initial hot spot temperature rise when any load rate starts and the termination hot spot temperature rise when the load rate does not change, and calculating the hot spot temperature rise of the hot spot temperature at each time point relative to the environment predicted temperature according to the hot spot temperature rise change value and the initial hot spot temperature rise.

In one embodiment, the dry-type transformer overheating early warning device 500 disclosed in the present application further includes a model training module. The model training module comprises a sample obtaining module, a sample processing module and a model parameter adjusting module, wherein:

the sample acquisition module is used for acquiring sample historical load data and sample reference load data generated after the sample historical load data;

the sample processing module is used for inputting the sample historical load data into a target load prediction model to be trained, processing the sample historical load data through the target load prediction model to be trained and outputting a prediction result;

and the model training module is used for adjusting the model parameters of the target load prediction model to be trained according to the difference between the prediction result and the sample parameter load data, and stopping training until the training stopping condition is reached to obtain the trained target load prediction model.

In one embodiment, the temperature pre-warning module 505 includes a first level pre-warning trigger module and a second level pre-warning trigger module, wherein:

the first-level early warning triggering module is used for triggering a first-level early warning prompting action when the hot spot temperature corresponding to any time period in the hot spot temperature distribution is greater than a preset first temperature early warning threshold value and the corresponding time period is greater than or equal to the first time period threshold value;

the second-level early warning triggering module is used for triggering early warning prompting actions of early warning the second level when the hot spot temperature corresponding to any time period in the hot spot temperature distribution is greater than a preset second temperature early warning threshold value and the corresponding time period is greater than or equal to the second time period threshold value; the second temperature early warning threshold value is larger than the first temperature early warning threshold value, and the second time period threshold value is smaller than the first time period threshold value.

The specific definition of the device for the dry-type transformer overheating warning can be referred to the definition of the method for the dry-type transformer overheating warning, and is not described herein again. All or part of each module in the device for early warning of overheating of the dry-type transformer can be realized by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.

In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as shown in fig. 6. The computer device includes a processor, a memory, and a network interface connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer equipment is used for storing processing data for realizing the overheating early warning of the dry-type transformer. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a dry-type transformer overheating warning method.

Those skilled in the art will appreciate that the architecture shown in fig. 6 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.

In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program: acquiring historical load data related to the dry-type transformer, wherein the historical load data comprises a plurality of items of first historical load data and second historical load data with the generation time after the first historical load data; performing correlation analysis on the first historical load data, and screening multiple items of third historical load data which have correlation with the second historical load data from the multiple items of first historical load data; inputting the multiple items of third history load data into a target load prediction model, and processing the multiple items of third history load data through the target load prediction model to obtain target prediction load data in a future preset time period; acquiring environment predicted temperature data in a future preset time period, and inputting target predicted load data and environment predicted temperature corresponding to the future preset time period into a hot spot temperature calculation model to obtain hot spot temperature distribution of the dry-type transformer in the future preset time period; and when the hot spot temperature distribution meets the temperature early warning condition, triggering the early warning action on the dry type transformer.

In one embodiment, the processor, when executing the computer program, further performs the steps of:

acquiring original load data related to the dry-type transformer in a historical preset time period; deleting repeated data in the original load data to obtain intermediate processing data; determining a target data segment with a load value missing condition in the intermediate processing data; when the time period length corresponding to the target data segment is less than or equal to the preset time length, performing interpolation processing on the target data segment to fill up the missing load value in the target data segment; when the time period length corresponding to the target data segment is greater than the preset time period length, filling the load value according to the load average value in the same time period in the historical time period; and taking the intermediate processing data after the completion of the load value filling as historical load data related to the dry-type transformer.

for each item of first historical load data, calculating covariance between the first historical load data and the second historical load data according to expected values corresponding to the first historical load data and the second historical load data respectively; calculating a correlation coefficient between each first historical load data and each second historical load data according to the corresponding variance and covariance of the first historical load data and the second historical load data; and taking the first historical load data corresponding to the relation number more than or equal to the preset coefficient threshold value as third historical load data having correlation with the second historical load data.

acquiring equipment parameters of the dry-type transformer; according to target predicted load data and equipment parameters in a future preset time period, calculating the hot point temperature rise of the hot point temperature relative to the environment predicted temperature at each time point under different load conditions; the different load conditions include at least one of a rated load condition, a given load rate condition, a continuous load condition, and a transient overload condition; determining the hot spot temperature corresponding to each time point in the future preset time period according to the environment predicted temperature and the hot spot temperature rise at each time point in the future preset time period; and counting the hot spot temperatures at all time points to obtain the hot spot temperature distribution of the dry-type transformer in a future preset time period.

when the load condition is a rated load condition, determining a hot spot coefficient according to equipment parameters, and calculating the hot spot temperature rise of the hot spot temperature relative to the environment predicted temperature at each time point according to the hot spot coefficient and the average temperature rise under the rated load; when the load condition is a given load rate condition, calculating the hot point temperature rise of the hot point temperature relative to the environment predicted temperature at each time point according to the hot point coefficient determined by the equipment parameters, the thermoelectric coefficient, the average temperature rise under the rated load and the given load rate determined according to the target predicted load data; when the load condition is a continuous load condition, determining the temperature rise under the rated load according to the equipment parameters, and calculating the hot point temperature rise of the hot point temperature relative to the environment predicted temperature at each time point according to the temperature rise under the rated load and the corresponding load rate under the continuous load determined according to the target predicted load data; when the load condition is a transient overload condition, determining a hot spot temperature rise change value according to the initial hot spot temperature rise when any load rate starts and the termination hot spot temperature rise when the load rate does not change, and calculating the hot spot temperature rise of the hot spot temperature at each time point relative to the environment predicted temperature according to the hot spot temperature rise change value and the initial hot spot temperature rise.

when the hot spot temperature corresponding to any time period in the hot spot temperature distribution is greater than a preset first temperature early warning threshold value and the corresponding time period is greater than or equal to the first time period threshold value, triggering a first-level early warning prompt action; when the hot spot temperature corresponding to any time period in the hot spot temperature distribution is greater than a preset second temperature early warning threshold value and the corresponding time period is greater than or equal to the second time period threshold value, triggering early warning prompting actions for early warning a second level; the second temperature early warning threshold value is larger than the first temperature early warning threshold value, and the second time period threshold value is smaller than the first time period threshold value.

acquiring sample historical load data and sample reference load data generated after the sample historical load data; inputting the sample historical load data into a target load prediction model to be trained, processing the sample historical load data through the target load prediction model to be trained, and outputting a prediction result; and adjusting the model parameters of the target load prediction model to be trained according to the difference between the prediction result and the sample parameter load data until the training stop condition is reached, and stopping training to obtain the trained target load prediction model.

In an embodiment, a computer-readable storage medium is provided, on which a computer program is stored which, when being executed by a processor, carries out the steps of the above-mentioned method embodiments.

It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database or other medium used in the embodiments provided herein can include at least one of non-volatile and volatile memory. Non-volatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical storage, or the like. Volatile Memory can include Random Access Memory (RAM) or external cache Memory. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others.

The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.

The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.

Claims

1. An overheating early warning method for a dry-type transformer, the method comprising:

2. The method of claim 1, wherein the obtaining historical load data related to a dry-type transformer comprises:

3. The method according to claim 1, wherein the performing a correlation analysis on the plurality of items of first historical load data and screening out a plurality of items of third historical load data having a correlation with the second historical load data from the plurality of items of first historical load data comprises:

4. The method according to claim 1, wherein the step of inputting the target predicted load data and the environmental predicted temperature corresponding to the future preset time period into a hot spot temperature calculation model to obtain a hot spot temperature distribution of the dry-type transformer in the future preset time period comprises:

acquiring equipment parameters of the dry-type transformer;

5. The method of claim 4, wherein calculating the hot-spot temperature rise of the hot-spot temperature at different load conditions relative to the environmental predicted temperature at each time point according to the target predicted load data and the equipment parameters in the future preset time period comprises:

6. The method according to claim 1, wherein the triggering an early warning action on the dry-type transformer when the hot spot temperature distribution meets a temperature early warning condition comprises:

7. The method according to any one of claims 1 to 6, further comprising:

8. An overheating early warning device for a dry type transformer, the device comprising:

9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method of any of claims 1 to 7.

10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.