CN117996755A

CN117996755A - Transformer operation control method and device, computer equipment and storage medium

Info

Publication number: CN117996755A
Application number: CN202410385329.8A
Authority: CN
Inventors: 夏湘滨; 夏付却; 阳典意; 吴利仁; 陈森; 王雅程; 李世军
Original assignee: Hunan Huaxia Tebian Co ltd
Current assignee: Hunan Huaxia Tebian Co ltd
Priority date: 2024-04-01
Filing date: 2024-04-01
Publication date: 2024-05-07

Abstract

The invention belongs to the field of transformers, and relates to a transformer operation control method, a device, computer equipment and a storage medium, wherein the method comprises the following steps: acquiring historical grid information of a grid associated with the transformer, wherein the historical grid information comprises load information, weather information, time information, space information and event information; determining a characteristic processing mode of the sub-information according to the information type of each sub-information, and carrying out characteristic engineering processing on the sub-information according to the determined characteristic processing mode to obtain initial characteristic information; performing characteristic cross processing on the obtained initial characteristic information to obtain characteristic information; training according to the characteristic information to obtain a load change prediction model; acquiring current power grid information of a power grid, inputting a load change prediction model, and obtaining load change prediction information of a transformer; and generating an operation control strategy of the transformer based on the load change prediction information, and performing operation control on the transformer according to the operation control strategy. The invention improves the accuracy of the operation control of the transformer.

Description

Transformer operation control method and device, computer equipment and storage medium

Technical Field

The present invention relates to the field of transformers, and in particular, to a method and apparatus for controlling transformer operation, a computer device, and a storage medium.

Background

Electric power is a necessity for production and living, and is an important support and guarantee for normal operation of society. Alternating current is most commonly used in production and life, and a transformer is an electrical device for changing alternating voltage. The transformer mainly comprises an iron core (or a magnetic core) and a coil, wherein the coil is provided with two or more windings, the winding connected with a power supply is called a primary coil, and the other windings are called secondary coils. When an input voltage is applied to the primary coil, electrical energy is transferred from the primary coil to the secondary coil by magnetic field induction of the core, thereby producing the required voltage and current at the output.

In the control and maintenance of transformers, improving the efficiency of the transformers is a very important issue. The operating efficiency of a transformer refers to the efficiency of the transformer in converting input electrical energy into output electrical energy in actual operation, which is typically measured as the ratio of output power to input power. The output power of a transformer is usually determined by a load, and the load of the transformer refers to the electric energy consumed by an electric device or a load actually connected to the circuit to which the transformer is connected. The load of the transformer may change during different time periods. Current transformer operation control techniques simply step up or step down the output voltage of the transformer at different time periods to accommodate load changes. However, the adjustment mode is rough, and the accuracy of the operation control of the transformer is low.

Disclosure of Invention

The technical scheme of the invention aims to provide a transformer operation control method, a device, computer equipment and a storage medium, so as to solve the problem of lower accuracy of transformer operation control.

In order to solve the technical problems, the technical scheme of the invention provides a transformer operation control method, which adopts the following technical scheme:

Acquiring historical grid information of a grid associated with the transformer, wherein the historical grid information comprises load information, weather information, time information, space information and event information;

For each piece of sub-information in the historical power grid information, determining a characteristic processing mode of the sub-information according to the information type of the sub-information, and performing characteristic engineering processing on the sub-information according to the determined characteristic processing mode to obtain initial characteristic information;

performing characteristic cross processing on the obtained initial characteristic information to obtain characteristic information;

Training an initial load change prediction model according to the characteristic information to obtain a load change prediction model;

acquiring current power grid information of the power grid, and inputting the current power grid information into the load change prediction model to obtain load change prediction information of the transformer;

And generating an operation control strategy of the transformer based on the load change prediction information, and performing operation control on the transformer according to the operation control strategy.

Further, the step of performing feature engineering processing on the sub-information according to the determined feature processing mode to obtain initial feature information includes:

When the sub information is load information, carrying out time sequence feature extraction on the load information to obtain at least one time sequence feature, carrying out frequency domain feature extraction on the load information to obtain at least one frequency domain feature, and carrying out statistic feature extraction on the load information to obtain at least one statistic feature;

When the sub information is weather information and time information, extracting features of the weather information and the time information according to a preset feature processing strategy to obtain weather time features; the weather time features comprise weather features, time features and composite features;

When the sub information is space information and event information, extracting space characteristics of the space information and the event information to obtain at least one space event characteristic;

And determining the obtained time sequence characteristics, frequency domain characteristics, statistical characteristics, weather time characteristics and spatial event characteristics as initial characteristic information.

Further, the step of extracting the time sequence feature of the load information to obtain at least one time sequence feature includes:

Carrying out time sequence analysis on the load information, and extracting to obtain at least one first time sequence feature;

and identifying long-term dependence and emergency events in the load information, and obtaining a second time sequence characteristic.

Further, the step of extracting the frequency domain features of the load information to obtain at least one frequency domain feature includes:

converting the load information into a frequency domain to obtain frequency domain load information, and extracting frequency domain features of the frequency domain load information to obtain at least one first frequency domain feature;

performing wavelet transformation on the load information to obtain a plurality of groups of sub-signals;

and calculating the frequency distribution map of each group of sub-signals, and respectively extracting the second frequency domain characteristics of each frequency distribution map.

Further, the step of performing feature cross processing on the obtained initial feature information to obtain feature information includes:

Acquiring time stamps of various initial characteristic information;

The initial feature information with the same time stamp is subjected to cross combination to obtain crossed features;

and determining the obtained crossed features and the rest of initial feature information as feature information.

Further, the step of performing operation control on the transformer according to the operation control strategy includes:

When the operation control strategy is a voltage control strategy, acquiring a voltage adjustment mode in the voltage control strategy;

and adjusting the output voltage of the transformer according to the voltage adjustment mode.

Further, the step of performing operation control on the transformer according to the operation control strategy further includes:

And when the operation control strategy is a transfer control strategy, controlling the transformer and the candidate transformer to carry out load transfer according to the transfer control strategy.

In order to solve the technical problems, the technical scheme of the invention also provides a transformer operation control device, which adopts the following technical scheme:

the system comprises a history acquisition module, a control module and a control module, wherein the history acquisition module is used for acquiring history power grid information of a power grid associated with a transformer, and the history power grid information comprises load information, weather information, time information, space information and event information;

The characteristic engineering module is used for determining a characteristic processing mode of the sub-information according to the information type of the sub-information for each sub-information in the historical power grid information, and carrying out characteristic engineering processing on the sub-information according to the determined characteristic processing mode to obtain initial characteristic information;

The feature intersection module is used for performing feature intersection processing on the obtained initial feature information to obtain feature information;

The model training module is used for training an initial load change prediction model according to the characteristic information to obtain a load change prediction model;

The load prediction module is used for acquiring current power grid information of the power grid, inputting the current power grid information into the load change prediction model and obtaining load change prediction information of the transformer;

And the operation control module is used for generating an operation control strategy of the transformer based on the load change prediction information and performing operation control on the transformer according to the operation control strategy.

In order to solve the above technical problem, the technical solution of the present invention further provides a computer device, where the computer device includes a memory and a processor, where the memory stores computer readable instructions, and the processor implements the steps of the transformer operation control method as described above when executing the computer readable instructions.

In order to solve the above technical problem, the present invention further provides a computer readable storage medium, where computer readable instructions are stored on the computer readable storage medium, and the computer readable instructions implement the steps of the transformer operation control method as described above when being executed by a processor.

Compared with the prior art, the technical scheme of the invention has the following main beneficial effects: acquiring historical grid information of a grid associated with the transformer, wherein the historical grid information comprises load information, weather information, time information, space information and event information; for each piece of sub-information in the historical power grid information, determining a characteristic processing mode of the sub-information according to the information type of the sub-information; carrying out feature engineering processing on the sub-information according to the determined feature processing mode, ensuring the accuracy of the feature engineering processing, and obtaining feature initial feature information which can be understood and utilized by the model; performing feature cross processing on the obtained initial feature information to obtain feature information, which is beneficial to capturing the relevance and interaction relation between features; training an initial load change prediction model according to the characteristic information to obtain a load change prediction model; the current power grid information of the power grid is obtained and is input into a load change prediction model, so that the load change prediction information of the transformer can be accurately obtained; the load change prediction information predicts the condition that the load of the transformer is increased or decreased; and generating an operation control strategy of the transformer based on the load change prediction information, and performing operation control on the transformer according to the operation control strategy, so that the transformer operates in a load matching state as much as possible, meanwhile, the accurate control on the transformer is realized through a model, and the operation safety and efficiency of the transformer are improved.

Drawings

In order to more clearly illustrate the solution of the present invention, a brief description will be given below of the drawings required for the description of the technical solution of the present invention, it being obvious that the drawings in the following description are some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is an exemplary system architecture diagram in which the present invention may be applied;

FIG. 2 is a flow chart of one embodiment of a method of controlling operation of a transformer in accordance with the present invention;

FIG. 3 is a schematic diagram of one embodiment of a transformer operation control device according to the present invention;

FIG. 4 is a schematic structural diagram of one embodiment of a computer device in accordance with the present invention.

Reference numerals: 100. a transformer operation control system; 101. a first terminal device; 102. a second terminal device; 103. a third terminal device; 104. a network; 105. a server; 300. a transformer operation control device; 301. a history acquisition module; 302. a feature engineering module; 303. a feature crossing module; 304. a model training module; 305. a load prediction module; 306. a running control module; 4. a computer device; 41. a memory; 42. a processor; 43. a network interface.

Detailed Description

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs; the terminology used in the description is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention; the terms "comprising" and "having" and any variations thereof in the description of the invention and the claims and the description of the drawings above are intended to cover a non-exclusive inclusion. The terms first, second and the like in the description and in the claims or in the above-described figures, are used for distinguishing between different objects and not necessarily for describing a sequential or chronological order.

Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the invention. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly appreciate that the embodiments described herein may be combined with other embodiments.

In order to make the person skilled in the art better understand the technical solutions of the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings.

As shown in fig. 1, the transformer operation control system 100 may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 is a medium used to provide a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, among others.

The first terminal device 101, the second terminal device 102, and the third terminal device 103 may collect various kinds of grid information of the grid, and interact with the server 105 through the network 104 to receive or send a message, etc. The first terminal device 101, the second terminal device 102, the third terminal device 103 may be, but not limited to, various industrial computers, personal computers, notebook computers, sensors, and the like.

It should be noted that, the method for controlling operation of the transformer provided by the technical scheme of the present invention is generally executed by a server, and accordingly, the device for controlling operation of the transformer is generally disposed in the server.

It should be understood that the number of terminal devices, networks and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.

With continued reference to fig. 2, a flow chart of one embodiment of a method of transformer operation control according to the present invention is shown. The transformer operation control method comprises the following steps:

step S201, obtaining historical grid information of a grid associated with the transformer, where the historical grid information includes load information, weather information, time information, space information, and event information.

In this embodiment, the electronic device (for example, the server shown in fig. 1) on which the transformer operation control method operates may communicate with the terminal device through a wired connection manner or a wireless connection manner. It should be noted that the wireless connection may include, but is not limited to, a 3G/4G/5G connection, a WiFi connection, a bluetooth connection, a WiMAX connection, a Zigbee connection, and a UWB (ultra wideband) connection.

In particular, the transformer is associated with a power grid, and the server obtains historical power grid information of the power grid associated with the transformer, which may be a collection of power grid related information including load information, weather information, time information, space information, and event information.

The load information may be detailed information of electric energy consumed by an electric device or a load actually connected in the power grid connected with the transformer. The weather information refers to weather condition information of the region where the power grid is located. The time information indicates date and time. The spatial information may be spatial data such as geographical information, for example, a traffic map of an area where the power grid is located. Event information is used to record some important events in the region where the grid is located, such as user distribution, load distribution, population flow, emergency events, natural disasters, etc. It should be noted that the load information, weather information, space information and event information may also be provided with corresponding time stamps.

Step S202, for each sub-information in the historical power grid information, determining a characteristic processing mode of the sub-information according to the information type of the sub-information, and carrying out characteristic engineering processing on the sub-information according to the determined characteristic processing mode to obtain initial characteristic information.

In particular, the historical grid information contains a plurality of sub-information, and the above-mentioned types of information can be regarded as sub-information, and they can contain more specific information, and these more specific information are also sub-information of the historical grid information.

And for each piece of sub-information in the historical power grid information, determining a characteristic processing mode of the sub-information according to the information type of the sub-information (the information type comprises five major categories of load information, weather information, time information, space information and event information), carrying out characteristic engineering processing on the sub-information according to the determined characteristic processing mode, and selecting characteristics closely related to load change to obtain initial characteristic information.

Further, the step of performing feature engineering processing on the sub-information according to the determined feature processing mode to obtain initial feature information may include: when the sub information is load information, carrying out time sequence feature extraction on the load information to obtain at least one time sequence feature, carrying out frequency domain feature extraction on the load information to obtain at least one frequency domain feature, and carrying out statistic feature extraction on the load information to obtain at least one statistic feature; when the sub information is weather information and time information, extracting features of the weather information and the time information according to a preset feature processing strategy to obtain weather time features; the weather time features include weather features, time features, and composite features; when the sub information is space information and event information, extracting space characteristics of the space information and the event information to obtain at least one space event characteristic; and determining the obtained time sequence characteristics, frequency domain characteristics, statistical characteristics, weather time characteristics and spatial event characteristics as initial characteristic information.

Specifically, when the information type of the sub information is load information, time sequence analysis is performed on the load information, and time sequence feature extraction is performed to obtain at least one time sequence feature. The load information can also be processed in the frequency domain, and frequency domain feature extraction is performed to obtain at least one frequency domain feature.

And carrying out statistical feature extraction on the load information to obtain at least one statistical feature such as mean, variance, kurtosis, skewness and the like. In addition to the above statistical features, more higher-order statistical features, such as gaussian mixture model parameters, multivariate gaussian distribution features, etc., can be introduced to increase the diversity and expression capability of the extracted statistical features.

And when the sub information is weather information and time information, extracting the characteristics of the weather information and the time information according to a preset characteristic processing strategy to obtain weather time characteristics. The weather time features include weather features, time features, and composite features; the time feature may be a feature extracted for time information alone, for example, information such as year, month, day, hour, etc. of extracting a time stamp as a feature; calculating a time interval or time difference between the time stamp and a specific date (such as holidays, weekends) as a feature; judging whether the time stamp is a specific holiday or working day, and generating binary characteristics; sliding window processing is performed on the time series data, and statistical features (such as mean, variance, maximum value, minimum value and the like) in the sliding window are calculated. Weather features are weather-related features, such as associating weather information (e.g., temperature, humidity, precipitation, etc.) with a timestamp to generate new features; discretizing continuous weather data, converting the continuous weather data into category characteristics, such as dividing the temperature into a plurality of different temperature intervals; carrying out time sequence analysis on weather information by using sliding window processing, and extracting statistical features in the sliding window; seasonal features are generated according to seasonal changes in weather, for example, weather data is divided into seasons such as spring, summer, autumn, winter, and the like according to seasons. The weather information may also be interleaved with the time information to generate new combination features, such as combining temperature and season.

And when the sub information is the space information and the event information, extracting the space characteristics of the space information and the event information to obtain at least one space event characteristic. The server can extract spatial characteristics such as regional load density, user distribution condition and the like by combining spatial information such as geographic information and the like. The server may also calculate more complex spatial correlation features, such as traffic conditions between regions, population flow conditions, etc., based on the event information to improve the predictive power of the model for regional load changes.

The server determines the obtained time sequence characteristics, frequency domain characteristics, statistical characteristics, weather time characteristics and spatial event characteristics as initial characteristic information.

In the embodiment, for the sub-information among different information types, feature engineering processing is performed according to different feature processing modes, so that features which can be understood and utilized by the model are obtained, and a data basis is laid for subsequent model training and prediction of load change by the model.

Further, the step of extracting the time sequence feature of the load information to obtain at least one time sequence feature may include: carrying out time sequence analysis on the load information, and extracting to obtain at least one first time sequence feature; and identifying long-term dependence and emergency events in the load information, and obtaining a second time sequence characteristic.

Specifically, the load information is analyzed in time series, and at least one first timing characteristic, such as trend, periodicity, seasonal, etc., is extracted. The trend refers to a continuous increasing or decreasing trend of data with time, and linear fitting, moving average and other methods can be used for extracting trend features. For example, a linear regression model may be used to fit a trend line of historical load information and extract the slope of the trend line as a trend feature.

Periodicity is the repetitive fluctuation exhibited by the data, i.e., a change with a fixed period. The periodic features can be extracted by using fourier transform, autocorrelation function and the like. For example, the historical load information may be converted to the frequency domain using a fourier transform, from which the amplitude and phase of the periodic component are extracted as periodic features.

Seasonal refers to the periodic variation that data exhibits over a fixed season, typically due to seasonal factors (e.g., seasonal activity, climate change, etc.). The seasonal feature may be extracted using a differencing method. For example, the seasonal component in the historical load information may be extracted using a difference method, and then the average or the variation amplitude of the seasonal component may be calculated as the seasonal feature.

The server may also identify long-term dependencies and incidents in the load information, resulting in a second timing characteristic. Long-term dependence is the existence of long-term dependencies or dependencies in time-series data, i.e., data at past moments has a greater impact on future moments. When dealing with long-term dependencies, sequence reconstruction may be employed: the time series data is converted into a series of samples so that the model can capture the dependency between different moments. For example, a sequence sample is formed by using data of the past times as input features and data of the next time as output tags; the recurrent neural network or the long-term memory network and the like are used for processing the sequence data, and the models have memory capacity and can effectively capture long-term dependency relationship in the time sequence data.

Emergency events refer to abnormal or atypical situations occurring in time-series data, which may be caused by external factors such as natural disasters, equipment malfunctions, etc. Anomaly detection algorithms (e.g., statistical-based methods, machine-learning-based methods, etc.) can be used to identify outliers in the time-series data. In addition, if the time of occurrence of the incident or related information is known, such information may be input as additional features into the model to assist the model in identifying and coping with the incident.

Both the first timing feature and the second timing feature are timing features extracted from the load information.

In this embodiment, time sequence analysis is performed on the load information, and at least one first time sequence feature is extracted and obtained; and identifying long-term dependence and emergency in the load information, obtaining a second time sequence feature, and ensuring the accuracy and comprehensiveness of time sequence feature extraction.

Further, the step of extracting the frequency domain feature of the load information to obtain at least one frequency domain feature may include: converting the load information into a frequency domain to obtain frequency domain load information, and extracting frequency domain features of the frequency domain load information to obtain at least one first frequency domain feature; carrying out wavelet transformation on the load information to obtain a plurality of groups of sub-signals; and calculating the frequency distribution map of each group of sub-signals, and respectively extracting the second frequency domain characteristics of each frequency distribution map.

Specifically, the loading information is converted into a frequency domain to obtain frequency domain loading information, and frequency domain feature extraction is performed on the frequency domain loading information, for example, a frequency domain analysis technology such as fourier transform can be adopted to obtain at least one first frequency domain feature.

In addition, frequency domain analysis techniques of wavelet transformation may also be employed. The load information is subjected to wavelet transformation, and the original signal is decomposed into sub-signals with different scales and frequencies. The sub-signals contain components of different frequency ranges in the original signal, so that details of load change can be better captured. For each group of sub-signals, its frequency profile may be calculated to determine the dominant frequency component and frequency range present in the load variation. For each frequency profile, a series of second frequency domain features, such as primary frequency, frequency range, frequency distribution, etc., can be extracted that more fully describe the frequency characteristics of the load variation, providing more information for prediction.

The first frequency domain feature and the second frequency domain feature are both frequency domain features extracted from the payload information.

In this embodiment, load information is converted to a frequency domain to obtain frequency domain load information, and frequency domain feature extraction is performed on the frequency domain load information to obtain at least one first frequency domain feature; carrying out wavelet transformation on the load information to obtain a plurality of groups of sub-signals; and calculating the frequency distribution diagram of each group of sub-signals, and respectively extracting the second frequency domain characteristic of each frequency distribution diagram to obtain the frequency characteristic capable of comprehensively describing the load change, thereby ensuring the accuracy of the follow-up load change prediction.

Step S203, performing feature cross processing on the obtained initial feature information to obtain feature information.

Specifically, different types of initial feature information may be feature-interleaved to generate new feature information to capture correlations and interactions between the initial feature information. For example, the time-domain features and the frequency-domain features may be interleaved to generate time-domain interleaved features that are used to more fully describe the feature information of the load change.

Further, the step S203 may include: acquiring time stamps of various initial characteristic information; the initial feature information with the same time stamp is subjected to cross combination to obtain crossed features; and determining the obtained crossed features and the rest of initial feature information as feature information.

In particular, as mentioned above, the load information, weather information, space information and event information may also be provided with corresponding time stamps, thereby indicating when the corresponding sub-information is generated. According to the time stamps of the load information, the weather information, the space information and the event information, the time stamps of various initial characteristic information are obtained, and different kinds of initial characteristic information with the same time stamp can be combined in a crossing way to obtain crossed characteristics, so that the characteristics generated in the same time period can be associated; and determining the obtained crossed features and the rest initial feature information which is not subjected to feature crossing processing as feature information.

In one embodiment, the initial feature information may be added to the feature information even after feature crossing processing.

In this embodiment, timestamps of various initial feature information are obtained, and the initial feature information with the same timestamp is cross-combined to obtain the cross feature, so that features generated in the same time period are associated, interaction between features is facilitated to be captured, and accuracy of feature information is improved.

And S204, training an initial load change prediction model according to the characteristic information to obtain a load change prediction model.

Specifically, the characteristic information is input into an initial load change prediction model, the model is trained to obtain a load change prediction model, and the load change prediction model obtained through training can predict the load change condition of the power grid in the future.

In one embodiment, the model is trained by inputting the characteristic information before time t as the model and the load information after time t as the model tag according to the time stamp.

Step S205, current power grid information of the power grid is obtained, and the current power grid information is input into a load change prediction model to obtain load change prediction information of the transformer.

Specifically, when the model is applied, current power grid information of the power grid associated with the transformer is obtained, and the sub-information contained in the current power grid information can be the same as or different from the historical power grid information.

And inputting the current power grid information into a load change prediction model, predicting the next load change condition by the load change prediction model, and outputting load change prediction information of the transformer.

Step S206, generating an operation control strategy of the transformer based on the load change prediction information, and performing operation control on the transformer according to the operation control strategy.

In particular, the load change prediction information may show the proportion/value of the load increase or decrease of the transformer. Variations in the load of the transformer can affect the efficiency and operating conditions of the transformer, and therefore, an operational control strategy of the transformer can be determined based on the load variation prediction information. The operation control strategy is used for controlling future operation of the transformer so as to improve the efficiency of the transformer or enable the transformer to reach a load matching state when the power of the transformer is overloaded. Load matching means that the load of the transformer is matched to the designed load range, so that overload or underload states are avoided, the rated loss of the transformer can be reduced, and the efficiency is improved.

Further, the step of performing operation control on the transformer according to the operation control policy may include: when the operation control strategy is a voltage control strategy, a voltage adjustment mode in the voltage control strategy is obtained; and adjusting the output voltage of the transformer according to the voltage adjustment mode.

Specifically, the operation control strategy includes a voltage control strategy and a transfer control strategy, wherein the voltage control strategy refers to a control strategy for adjusting the output voltage of the transformer.

When the operation control strategy is a voltage control strategy, a voltage adjustment mode in the voltage control strategy is obtained, wherein the voltage adjustment mode records whether the transformer should adjust the output voltage to be high or low, and the proportion/value of the high or low. In general, there is a close relationship between the load and the voltage, and the power consumption of the load is directly affected by the voltage variation. When the voltage increases, the power consumption of the load increases; and when the voltage decreases, the power consumption of the load decreases. By adjusting the output voltage of the transformer, control of the load power consumption can be achieved.

When the load increases to cause the transformer to be in an overload state, the power consumed by the load can be reduced by properly adjusting the output voltage of the transformer, thereby reducing the load of the transformer, preventing the overload of the transformer, and reducing the power consumed by the load to operate within the rated range of the transformer.

When the load is reduced and the transformer is in an underload state, the output voltage of the transformer is properly regulated, so that the power consumed by the load can be increased, the transformer can be better utilized, and the underload state is avoided. By increasing the output voltage, the power consumed by the load can be increased, making full use of the transformer.

In one embodiment, the relation between the load change and the output voltage adjustment is preset, and the voltage control strategy can be generated according to the load change prediction information in a table look-up mode. Or training in advance to obtain a transformer control model, and generating a voltage control strategy by the transformer control model according to the load change prediction information. The transformer control model and the load change prediction model may be the same model or different models. If the transformer control model and the load change prediction model are the same model, training content generated by a voltage control strategy needs to be added in the training process of the initial load change model.

In this embodiment, when the operation control policy is a voltage control policy, a voltage adjustment manner in the voltage control policy is obtained; the output voltage of the transformer is regulated up or reduced according to the voltage regulation mode, so that the power consumption of a load is effectively controlled, and overload or lower efficiency of the transformer is avoided.

Further, the step of performing operation control on the transformer according to the operation control policy may further include: when the operation control strategy is a transfer control strategy, the load transfer between the transformer and the candidate transformer is controlled according to the transfer control strategy.

Specifically, the operation control policy further includes a transfer control policy. There may be multiple transformers in the grid system, different transformers being associated with different grids. The load/load conditions of different transformers are different, and the transfer control strategy needs to acquire load change prediction information and current load conditions of a plurality of transformers, and perform load transfer between the transformers and candidate transformers (i.e. other transformers), so that each transformer operates in a load matching state as much as possible.

In one embodiment, the transfer control strategy may be generated by a transformer control model, which may construct constraint conditions and objective functions according to load change prediction information and current load conditions of each transformer, and solve the constraint conditions and objective functions to obtain the transfer control strategy.

In this embodiment, when the operation control policy is a transfer control policy, the load transfer between the transformer and the candidate transformer is controlled according to the transfer control policy, so that each transformer operates in a state where the load matches as much as possible.

In this embodiment, historical grid information of a grid associated with a transformer is obtained, including load information, weather information, time information, space information and event information; for each piece of sub-information in the historical power grid information, determining a characteristic processing mode of the sub-information according to the information type of the sub-information; carrying out feature engineering processing on the sub-information according to the determined feature processing mode, ensuring the accuracy of the feature engineering processing, and obtaining feature initial feature information which can be understood and utilized by the model; performing feature cross processing on the obtained initial feature information to obtain feature information, which is beneficial to capturing the relevance and interaction relation between features; training an initial load change prediction model according to the characteristic information to obtain a load change prediction model; the current power grid information of the power grid is obtained and is input into a load change prediction model, so that the load change prediction information of the transformer can be accurately obtained; the load change prediction information predicts the condition that the load of the transformer is increased or decreased; and generating an operation control strategy of the transformer based on the load change prediction information, and performing operation control on the transformer according to the operation control strategy, so that the transformer operates in a load matching state as much as possible, meanwhile, the accurate control on the transformer is realized through a model, and the operation safety and efficiency of the transformer are improved. The load change prediction model adopts a gradient lifting tree model or a linear regression model.

Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by computer readable instructions stored in a computer readable storage medium that, when executed, may comprise the steps of the embodiments of the methods described above. The storage medium may be a nonvolatile storage medium such as a magnetic disk, an optical disk, a Read-Only Memory (ROM), or a random access Memory (Random Access Memory, RAM).

It should be understood that, although the steps in the flowcharts of the figures are shown in order as indicated by the arrows, these steps are not necessarily performed in order as indicated by the arrows. The steps are not strictly limited in order and may be performed in other orders, unless explicitly stated herein. Moreover, at least some of the steps in the flowcharts of the figures may include a plurality of sub-steps or stages that are not necessarily performed at the same time, but may be performed at different times, the order of their execution not necessarily being sequential, but may be performed in turn or alternately with other steps or at least a portion of the other steps or stages.

With further reference to fig. 3, as an implementation of the method shown in fig. 2, the present invention provides an embodiment of a transformer operation control device, which corresponds to the embodiment of the method shown in fig. 2, and the device is particularly applicable to various electronic devices.

As shown in fig. 3, the transformer operation control device 300 according to the present embodiment includes: a history acquisition module 301, a feature engineering module 302, a feature intersection module 303, a model training module 304, a load prediction module 305, and a run control module 306, wherein:

The history obtaining module 301 is configured to obtain history grid information of a grid associated with the transformer, where the history grid information includes load information, weather information, time information, space information, and event information.

The feature engineering module 302 is configured to determine, for each sub-information in the historical power grid information, a feature processing manner of the sub-information according to an information type of the sub-information, and perform feature engineering processing on the sub-information according to the determined feature processing manner, so as to obtain initial feature information.

And the feature intersection module 303 is configured to perform feature intersection processing on the obtained initial feature information to obtain feature information.

The model training module 304 is configured to train the initial load change prediction model according to the feature information, and obtain a load change prediction model.

The load prediction module 305 is configured to obtain current grid information of the power grid, and input the current grid information into the load change prediction model to obtain load change prediction information of the transformer.

The operation control module 306 is configured to generate an operation control policy of the transformer based on the load change prediction information, and perform operation control on the transformer according to the operation control policy.

In this embodiment, historical grid information of a grid associated with a transformer is obtained, including load information, weather information, time information, space information and event information; for each piece of sub-information in the historical power grid information, determining a characteristic processing mode of the sub-information according to the information type of the sub-information; carrying out feature engineering processing on the sub-information according to the determined feature processing mode, ensuring the accuracy of the feature engineering processing, and obtaining feature initial feature information which can be understood and utilized by the model; performing feature cross processing on the obtained initial feature information to obtain feature information, which is beneficial to capturing the relevance and interaction relation between features; training an initial load change prediction model according to the characteristic information to obtain a load change prediction model; the current power grid information of the power grid is obtained and is input into a load change prediction model, so that the load change prediction information of the transformer can be accurately obtained; the load change prediction information predicts the condition that the load of the transformer is increased or decreased; and generating an operation control strategy of the transformer based on the load change prediction information, and performing operation control on the transformer according to the operation control strategy, so that the transformer operates in a load matching state as much as possible, meanwhile, the accurate control on the transformer is realized through a model, and the operation safety and efficiency of the transformer are improved.

In some alternative implementations of the present embodiment, the feature engineering module 302 may include: the device comprises a first extraction sub-module, a second extraction sub-module, a third extraction sub-module and an initial determination sub-module, wherein:

the first extraction sub-module is used for extracting time sequence characteristics of the load information to obtain at least one time sequence characteristic when the sub-information is the load information, extracting frequency domain characteristics of the load information to obtain at least one frequency domain characteristic, and extracting statistical characteristics of the load information to obtain at least one statistical characteristic.

The second extraction sub-module is used for extracting the characteristics of the weather information and the time information according to a preset characteristic processing strategy when the sub-information is the weather information and the time information, so as to obtain weather time characteristics; weather time features include weather features, time features, and composite features.

And the third extraction sub-module is used for extracting the spatial characteristics of the spatial information and the event information when the sub-information is the spatial information and the event information, so as to obtain at least one spatial event characteristic.

And the initial determining submodule is used for determining the obtained time sequence characteristics, frequency domain characteristics, statistical characteristics, weather time characteristics and space event characteristics as initial characteristic information.

In some optional implementations of this embodiment, the first extraction sub-module may include: time analysis unit and recognition element, wherein:

and the time analysis unit is used for carrying out time sequence analysis on the load information and extracting at least one first time sequence feature.

And the identification unit is used for identifying long-term dependence and emergency in the load information and obtaining a second time sequence characteristic.

In some optional implementations of this embodiment, the first extraction sub-module may further include: frequency domain extraction unit, wavelet transformation unit and graph extraction unit, wherein:

The frequency domain extraction unit is used for converting the load information into a frequency domain to obtain frequency domain load information, and extracting frequency domain features of the frequency domain load information to obtain at least one first frequency domain feature.

And the wavelet transformation unit is used for carrying out wavelet transformation on the load information to obtain a plurality of groups of sub-signals.

And the diagram extracting unit is used for calculating the frequency distribution diagram of each group of sub-signals and respectively extracting the second frequency domain characteristics of each frequency distribution diagram.

In some alternative implementations of the present embodiment, the feature crossing module 303 may include: the device comprises a timestamp acquisition sub-module, a cross combination sub-module and a feature determination sub-module, wherein:

and the time stamp obtaining sub-module is used for obtaining time stamps of various initial characteristic information.

And the cross combination sub-module is used for cross-combining the initial feature information with the same time stamp to obtain the cross feature.

And the feature determination submodule is used for determining the obtained crossed features and the rest initial feature information as feature information.

In some alternative implementations of the present embodiment, the operation control module 306 may include: the method comprises the steps of obtaining a submodule and a voltage adjustment submodule, wherein:

The mode acquisition sub-module is used for acquiring a voltage adjustment mode in the voltage control strategy when the operation control strategy is the voltage control strategy.

The voltage adjusting sub-module is used for adjusting the output voltage of the transformer according to the voltage adjusting mode.

In some alternative implementations of the present embodiment, the operation control module 306 may include: a transfer control sub-module, wherein:

and the transfer control sub-module is used for controlling the load transfer between the transformer and the candidate transformer according to the transfer control strategy when the operation control strategy is the transfer control strategy.

In order to solve the technical problems, the technical scheme of the invention also provides computer equipment. Referring specifically to fig. 4, fig. 4 is a basic structural block diagram of a computer device according to the present embodiment.

The computer device 4 comprises a memory 41, a processor 42, a network interface 43 communicatively connected to each other via a system bus. It should be noted that only computer device 4 having components 41-43 is shown in the figures, but it should be understood that not all of the illustrated components are required to be implemented and that more or fewer components may be implemented instead. It will be appreciated by those skilled in the art that the computer device herein is a device capable of automatically performing numerical calculation and/or information processing according to a preset or stored instruction, and its hardware includes, but is not limited to, a microprocessor, an Application SPECIFIC INTEGRATED Circuit (ASIC), a Programmable gate array (Field-Programmable GATE ARRAY, FPGA), a digital Processor (DIGITAL SIGNAL Processor, DSP), an embedded device, and the like.

The computer equipment can be a desktop computer, a notebook computer, a palm computer, a cloud server and other computing equipment. The computer equipment can perform man-machine interaction with a user through a keyboard, a mouse, a remote controller, a touch pad or voice control equipment and the like.

The memory 41 includes at least one type of readable storage medium including flash memory, hard disk, multimedia card, card memory (e.g., SD or DX memory, etc.), random Access Memory (RAM), static Random Access Memory (SRAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), programmable Read Only Memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the storage 41 may be an internal storage unit of the computer device 4, such as a hard disk or a memory of the computer device 4. In other embodiments, the memory 41 may also be an external storage device of the computer device 4, such as a plug-in hard disk, a smart memory card (SMART MEDIA CARD, SMC), a Secure Digital (SD) card, a flash memory card (FLASH CARD) or the like, which are provided on the computer device 4. Of course, the memory 41 may also comprise both an internal memory unit of the computer device 4 and an external memory device. In this embodiment, the memory 41 is typically used to store an operating system and various application software installed on the computer device 4, such as computer readable instructions of a transformer operation control method, and the like. Further, the memory 41 may be used to temporarily store various types of data that have been output or are to be output.

The processor 42 may be a central processing unit (Central Processing Unit, CPU), controller, microcontroller, microprocessor, or other data processing chip in some embodiments. The processor 42 is typically used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is configured to execute computer readable instructions stored in the memory 41 or process data, such as computer readable instructions for executing the transformer operation control method.

The network interface 43 may comprise a wireless network interface or a wired network interface, which network interface 43 is typically used for establishing a communication connection between the computer device 4 and other electronic devices.

The computer device provided in the present embodiment may execute the above-described transformer operation control method. The transformer operation control method here may be the transformer operation control method of each of the above embodiments.

The present invention also provides another embodiment, namely, a computer-readable storage medium storing computer-readable instructions executable by at least one processor to cause the at least one processor to perform the steps of the transformer operation control method as described above.

From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) comprising instructions for causing a terminal device (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the method according to the embodiments of the present invention.

It is apparent that the above-described embodiments are only some embodiments of the present invention, but not all embodiments, and the preferred embodiments of the present invention are shown in the drawings, which do not limit the scope of the patent claims. This invention may be embodied in many different forms, but rather, embodiments are provided in order to provide a thorough and complete understanding of the present disclosure. Although the invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that modifications may be made to the embodiments described in the foregoing description, or equivalents may be substituted for elements thereof. All equivalent structures made by the content of the specification and the drawings of the invention are directly or indirectly applied to other related technical fields, and are also within the scope of the invention.

Claims

1. A method for controlling operation of a transformer, comprising the steps of:

Acquiring current power grid information of the power grid, and inputting the current power grid information into the load change prediction model to obtain load change prediction information of the transformer; the load change prediction model adopts a gradient lifting tree model or a linear regression model;

generating an operation control strategy of the transformer based on the load change prediction information, and performing operation control on the transformer according to the operation control strategy;

The step of carrying out feature engineering processing on the sub-information according to the determined feature processing mode to obtain initial feature information comprises the following steps:

When the sub information is load information, carrying out time sequence feature extraction on the load information to obtain at least one time sequence feature, carrying out frequency domain feature extraction on the load information to obtain at least one frequency domain feature, and carrying out statistic feature extraction on the load information to obtain at least one statistic feature; the timing characteristics include trend characteristics, periodic characteristics, and seasonal characteristics; fitting a trend line of the historical load information by using a linear regression model, extracting the slope of the trend line as a trend feature, and extracting periodic features through Fourier change or autocorrelation function; the step of extracting the time sequence characteristics of the load information to obtain at least one time sequence characteristic comprises the following steps:

Identifying long-term dependence and emergency in the load information, obtaining a second time sequence characteristic, taking data at the past moments as input characteristics, and taking data at the next moment as output labels to form a sequence sample; processing the sequence data using a recurrent neural network or a long-short-term memory network;

When the sub information is weather information and time information, extracting features of the weather information and the time information according to a preset feature processing strategy to obtain weather time features; the weather time features comprise weather features, time features and composite features; using fourier transforms or autocorrelation functions;

When the sub information is space information and event information, extracting space characteristics of the space information and the event information to obtain at least one space event characteristic; calculating more complex spatial correlation features including traffic conditions and population flow conditions between regions according to the event information;

Determining the obtained time sequence characteristics, frequency domain characteristics, statistical characteristics, weather time characteristics and space event characteristics as initial characteristic information;

the step of performing feature cross processing on the obtained initial feature information to obtain feature information comprises the following steps:

Acquiring time stamps of various initial characteristic information;

The initial feature information with the same time stamp is subjected to cross combination to obtain crossed features; crossing the time sequence features and the frequency domain features to generate time-frequency domain crossing features;

2. The transformer operation control method according to claim 1, wherein the step of extracting the frequency domain features of the load information to obtain at least one frequency domain feature comprises:

3. The transformer operation control method according to claim 1, wherein the step of performing operation control of the transformer according to the operation control strategy comprises:

4. The transformer operation control method according to claim 3, wherein the step of performing operation control on the transformer according to the operation control strategy further comprises:

5. A transformer operation control device, comprising:

the model training module is used for training an initial load change prediction model according to the characteristic information to obtain a load change prediction model; the load change prediction model adopts a gradient lifting tree model or a linear regression model;

The operation control module is used for generating an operation control strategy of the transformer based on the load change prediction information and performing operation control on the transformer according to the operation control strategy;

The characteristic engineering module is further used for extracting time sequence characteristics of the load information to obtain at least one time sequence characteristic when the sub information is the load information, extracting frequency domain characteristics of the load information to obtain at least one frequency domain characteristic, and extracting statistical characteristics of the load information to obtain at least one statistical characteristic; the timing characteristics include trend characteristics, periodic characteristics, and seasonal characteristics; fitting a trend line of the historical load information by using a linear regression model, extracting the slope of the trend line as a trend feature, and extracting periodic features through Fourier change or autocorrelation function; the step of extracting the time sequence characteristics of the load information to obtain at least one time sequence characteristic comprises the following steps:

Identifying long-term dependence and emergency in the load information, obtaining a second time sequence characteristic, taking data at the past moments as input characteristics, and taking data at the next moment as output labels to form a sequence sample; processing the sequence data using a recurrent neural network or a long-short-term memory network; when the sub information is weather information and time information, extracting features of the weather information and the time information according to a preset feature processing strategy to obtain weather time features; the weather time features comprise weather features, time features and composite features; using fourier transforms or autocorrelation functions; when the sub information is space information and event information, extracting space characteristics of the space information and the event information to obtain at least one space event characteristic; calculating more complex spatial correlation features including traffic conditions and population flow conditions between regions according to the event information; determining the obtained time sequence characteristics, frequency domain characteristics, statistical characteristics, weather time characteristics and space event characteristics as initial characteristic information;

The feature crossing module is also used for acquiring time stamps of various initial feature information; the initial feature information with the same time stamp is subjected to cross combination to obtain crossed features; crossing the time sequence features and the frequency domain features to generate time-frequency domain crossing features; and determining the obtained crossed features and the rest of initial feature information as feature information.

6. A computer device comprising a memory having stored therein computer readable instructions, and a processor which when executed implements the steps of the transformer operation control method of any of claims 1 to 4.

7. A computer readable storage medium having stored thereon computer readable instructions which when executed by a processor implement the steps of the transformer operation control method according to any of claims 1 to 4.