CN118709855A - Power system data anomaly detection method and device, electronic equipment and storage medium - Google Patents

Abstract

The invention discloses a method and a device for detecting data anomalies of a power system, electronic equipment and a storage medium. The method comprises the following steps: determining current power system data of the power system in a current period; determining a target prediction model matched with the current power system data from the candidate prediction models; determining next power system data of the power system in a next period after the current period according to the target prediction model; and determining an early warning result according to the next power system data and a preset threshold value to perform early warning. According to the method, the data of the future time are predicted through the target prediction model, early warning can be sent out in advance for the faults of the power system, loss is avoided, the stable operation of the power system is ensured, meanwhile, the training cost is reduced, and the prediction efficiency and accuracy are improved.

Description

Power system data anomaly detection method and device, electronic equipment and storage medium

Technical Field

The present invention relates to the field of data processing technologies, and in particular, to a method and apparatus for detecting data anomalies in a power system, an electronic device, and a storage medium.

Background

With the development of the power system, the power grid scale of the power system is larger and larger, the problem of safety and stability of the power grid is also more and more complex, and particularly the problem of safety and stability of the power system is becoming the focus of attention of the power system. However, the prediction model adopted by the existing methods for detecting the abnormality of the power system often has the problems of higher requirements on data quality, complex calculation, poor generalization capability and the like, and the predicted power system data is inaccurate, so that the accuracy of detecting the abnormality cannot meet the requirements.

Disclosure of Invention

The invention provides a method, a device, electronic equipment and a storage medium for detecting power system data abnormality, which are used for solving the problem that the accuracy of power system abnormality detection cannot meet the requirement due to inaccurate prediction of power system data.

According to an aspect of the present invention, there is provided a method for detecting an abnormality of data of a power system, including:

Determining current power system data of a power system in a current period, wherein the current power system data are obtained by processing original data of the power system in the current period;

Determining a target prediction model matched with the current power system data from candidate prediction models, wherein the candidate prediction models are used for predicting power system data of a power system in a second period according to the power system data in a first period, the second period is an adjacent period after the first period, the candidate prediction models are obtained by training according to reference power system data, and the reference power system data are obtained by processing original power system data of the power system in a historical moment;

Determining next power system data of a power system in a next period after the current period according to the target prediction model;

And determining an early warning result according to the next power system data and a preset threshold value, wherein the preset threshold value is used for judging whether the next power system data has overrun data or not, and the early warning result is used for representing whether the next power system data exceeds the preset threshold value or not.

According to another aspect of the present invention, there is provided an electric power system data abnormality detection apparatus including:

the data determining module is used for determining current power system data of the power system in a current period, wherein the current power system data are obtained by processing original data of the power system in the current period;

The prediction model matching module is used for determining a target prediction model matched with the current power system data from candidate prediction models, the candidate prediction models are used for predicting the power system data of a power system in a second period according to the power system data in a first period, the second period is an adjacent period after the first period, the candidate prediction models are obtained by training according to reference power system data, and the reference power system data are obtained by processing the original power system data of the power system in a historical moment;

the prediction module is used for determining next power system data of a power system in a next period after the current period according to the target prediction model;

The early warning module is used for carrying out early warning according to the data of the next power system and a preset threshold value, the preset threshold value is used for judging whether the data of the next power system have overrun data, and the early warning result is used for representing whether the data of the next power system exceed the preset threshold value.

According to another aspect of the present invention, there is provided an electronic apparatus including:

At least one processor; and

A memory communicatively coupled to the at least one processor; wherein,

The memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the power system data anomaly detection method according to any one of the embodiments of the present invention.

According to another aspect of the present invention, there is provided a computer readable storage medium storing computer instructions for causing a processor to implement the method for detecting an abnormality of power system data according to any one of the embodiments of the present invention when executed.

According to the technical scheme, the current power system data of the power system in the current period is determined, the target prediction model corresponding to the current power system data is matched from the candidate prediction models, the next power system data of the power system in the next period after the current period is determined through the target prediction model, and early warning is carried out according to the next power system data and a preset threshold value. According to the method, the data of the future time are predicted through the target prediction model, early warning can be sent out in advance for the faults of the power system, loss is avoided, the stable operation of the power system is ensured, meanwhile, the training cost is reduced, and the prediction efficiency and accuracy are improved.

It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the invention or to delineate the scope of the invention. Other features of the present invention will become apparent from the description that follows.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flowchart of a method for detecting anomalies in power system data according to an embodiment of the present invention;

fig. 2 is a schematic structural diagram of an apparatus for detecting data anomalies in a power system according to an embodiment of the present invention;

Fig. 3 is a schematic structural diagram of an electronic device for implementing the method for detecting data anomalies in a power system according to an embodiment of the present invention.

Detailed Description

In order that those skilled in the art will better understand the present invention, a technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.

It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

Fig. 1 is a flowchart of a method for detecting an abnormality of power system data according to an embodiment of the present invention, where the method may be performed by a power system data abnormality detection device, and the power system data abnormality detection device may be implemented in hardware and/or software, and the power system data abnormality detection device may be configured in any electronic device having a network communication function. As shown in fig. 1, the method includes:

s110, determining current power system data of the power system in a current period.

The current power system data are obtained by processing the original data of the power system in the current period.

The method comprises the steps of collecting original data of the power system in a current time period through a data collecting device, and preprocessing the collected original data of the power system in the current time period to obtain current power system data of the power system in the current time period.

Further, the preprocessing of the original data of the power system in the current period is because abnormal values or missing values of the data may exist in the process of collecting the data due to equipment failure, shutdown maintenance, measurement errors and the like, so that the quality of the original data is reduced.

Optionally, determining current power system data of the power system in the current period includes steps A1-A2:

and A1, acquiring original data of the power system in the current period.

And collecting the original data of various power systems in the current period through collectors arranged on the edge side, and storing the original data in a time sequence database.

And step A2, preprocessing the original data of the power system in the current period to obtain the current power system data.

Wherein the preprocessing comprises the following steps: missing value completion and outlier processing, wherein outlier is data generated during abnormal operation of the power system, and the abnormal operation at least comprises: and (5) stopping, and performing fault and maintenance.

Wherein the outliers include: unsteady data, null, far exceeding normal thresholds, abnormal fluctuations, etc.

Firstly, carrying out missing value processing on original data of the power system in the current period, and carrying out abnormal value processing after the missing value is completely supplemented.

Further, the missing value processing process includes: firstly, dividing original data of the power system in a current period according to a preset time interval, traversing the divided data, checking whether the divided data have defects, and if so, adopting an interpolation mode to fill.

The interpolation method can adopt mean value, median value, linear interpolation and the like.

Wherein the preset time interval is determined by the accuracy of the prediction. Illustratively, the preset time interval may be 1 hour, 30 minutes, 5 minutes, 60s, 10s.

For example, assuming that the current power system data is gear oil temperature data, among prediction accuracies corresponding to different preset time intervals, the prediction accuracy of 1 hour is the lowest, and the prediction accuracy of 5 minutes is the highest, then the preset time interval of the gear oil temperature data is 5 minutes.

By way of example, assume that the preset time interval of the power system X is 1 hour; the preset time interval of the power system Y is 5 minutes; the preset time interval of the power system Z is 30 minutes, and is stored in the form of a dictionary, i.e., d= { X:1 hour, Y:5 minutes, Z:30 minutes }. And when the original data of the power system in the current period is acquired, searching a corresponding preset time interval according to the dictionary in the category of the current power system.

Further, after the missing value is complemented, traversing the complemented original data, judging whether an abnormal value exists in the original data, and if the abnormal value exists, processing according to a processing method corresponding to the abnormal value type matching.

For example, if the abnormal value is non-steady state data filtering, firstly, acquiring an aggregate object of a state and time combination of the fan, judging whether the running state of the fan is a time period of shutdown, fault and maintenance, and if so, filtering out the data in the time period. If the outlier is Null, it is filtered directly. If the outlier is far above the normal threshold, a threshold (fixed or dynamic) is first set and when the data point exceeds this threshold, the data is automatically filtered out. If the outlier is an outlier fluctuation, a filtering algorithm is used to smooth the outlier fluctuation in the data, wherein the filtering algorithm may be: a moving average filter, a median filter or a kalman filter.

S120, determining a target prediction model matched with the current power system data from the candidate prediction models.

The candidate prediction model is used for predicting the power system data of the power system in a second period according to the power system data in the first period, the second period is an adjacent period after the first period, the candidate prediction model is obtained by training according to reference power system data, and the reference power system data is obtained by processing the original power system data of the power system at the historical moment.

And matching a target prediction model corresponding to the current power system data type from the candidate preset models according to the current power system data.

Optionally, determining a target prediction model matched with the current power system data from the candidate prediction models, including steps B1-B2:

And B1, determining the data type of the current power system data.

The data types of the current power system data comprise: univariate data or multivariate data.

And matching the data type corresponding to the current power system data from the data type database according to the current power system data.

The data type database is built through historical prediction precision.

For example, assuming that the current power system data is gear oil temperature data, and the data type is univariate data, the prediction precision is M; when the data type is multivariable data, the prediction precision is N, M < N, and the prediction precision is high when the data type is univariate data type when the current power system data is gear oil temperature data, and the data type of the gear oil temperature data is univariate data.

By way of example, suppose that the data type of power system X is univariate data; data type multivariable data of the power system Y; the data type univariate data of the power system Z, the data type library is stored in the form of a dictionary, i.e., d= { X: univariate data, Y: multivariate data, Z: univariate data }.

And B2, determining a matched target prediction model from the candidate prediction models according to the data type of the current power system data.

The target prediction model can process current power system data corresponding to the data type of the current power system.

And matching the target prediction model matched with the data type from the candidate prediction models according to the data type of the current power system data.

Optionally, determining a matched target prediction model from the candidate prediction models according to the data type of the current power system data includes steps C1-C3:

and C1, determining the duration of the current power system data.

And determining the duration of the current power system data according to the time tag in the data of the current power system data.

And C2, if the data type of the current power system data is single variable data, matching a prediction model corresponding to the single variable data from the candidate prediction models, and matching a weight value corresponding to the duration of the current power system data from the first preset weight, so as to obtain a target prediction model matched with the single variable data.

The first preset weight is a weight value obtained by training according to univariate reference power system data of different durations in advance.

For example, the first preset weight may be stored in the form of a dictionary, i.e., d= { weight 11:1 hour, weight 21:30 minutes, weight 31:5 minutes, weight 41:60s, weight 51:10s }.

If the data type of the current power system data is single variable data, matching a prediction model corresponding to the single variable data from candidate prediction models, and matching a corresponding weight value from a first preset weight according to the duration of the current power system data, wherein if the duration of the current power system data is 5 minutes, the corresponding weight value is weight 31, and combining the weight value with the prediction model corresponding to the single variable data to obtain a target prediction model matched with the single variable data.

And C3, if the data type of the current power system data is the multivariate data, matching a prediction model corresponding to the multivariate data from the candidate prediction models, and matching a weight value corresponding to the duration of the current power system data from the second preset weight, so as to obtain a target prediction model matched with the multivariate data.

The second preset weight is a weight value obtained by training according to the multi-variable reference power system data of different durations in advance.

For example, the second preset weight may be stored in the form of a dictionary, i.e., d= { weight 12:1 hour, weight 22:30 minutes, weight 32:5 minutes, weight 42:60s, weight 52:10s }.

If the data type of the current power system data is the multi-variable data, matching a prediction model corresponding to the multi-variable data from the candidate prediction models, and matching a corresponding weight value from a second preset weight according to the duration of the current power system data, wherein if the duration of the current power system data is 5 minutes, the corresponding weight value is weight 32, and combining the weight value with the prediction model corresponding to the multi-variable data to obtain a target prediction model matched with the multi-variable data.

Optionally, the construction process of the candidate prediction model includes steps D1-D5:

and D1, acquiring reference power system data of the power system.

And acquiring the original power system data of the power system at the historical moment from the time sequence database, and preprocessing the original power system data at the historical moment to obtain the reference power system data.

And D2, dividing the reference power system data into first time period reference power system data and second time period reference power system data.

Dividing the reference power system data into first-period reference power system data and second-period reference power system data, wherein the first-period reference power system data is used as a training set; the second period references the power system data as a validation set.

And D3, dividing the reference power system data of the first period according to a preset time step to obtain data to be predicted.

The preset time step is a time period preset according to actual requirements. By way of example, the preset time steps may be 1 hour, 30 minutes, 5 minutes, 60s and 10s.

And D4, inputting the data to be predicted into a prediction model to be verified for prediction, and obtaining the reference power system data of the next period.

The prediction model to be verified can adopt Timegpt, lag-Llama, chronos and the like.

And inputting the data to be predicted into a prediction model to be verified, and obtaining the reference power system data of the next period through model calculation.

And D5, adjusting the prediction model to be verified according to the second period reference power system data and the next period reference power system data to obtain candidate prediction models.

And calculating errors between the reference power system data of the second period and the reference power system data of the next period, adjusting the prediction model to be verified according to the errors until the errors meet the requirements, and taking the prediction model to be verified meeting the requirements as a candidate prediction model.

Optionally, adjusting the prediction model to be verified according to the second period reference power system data and the next period reference power system data includes steps E1-E2:

and E1, determining evaluation indexes of the reference power system data of the second period and the reference power system data of the next period.

The evaluation index is used for evaluating the accuracy of prediction of the prediction model to be verified.

Further, the evaluation index adopts accuracy and mean square error.

And calculating the accuracy and the mean square error according to the second time period reference power system data and the next time period reference power system data.

And E2, adjusting parameters of the prediction model to be verified according to the evaluation index, training according to the adjusted parameters until the evaluation index meets the preset requirement, and taking the weight corresponding to the evaluation index meeting the preset requirement as the preset weight.

The preset weights comprise: the first preset weight and the second preset weight.

The preset requirement is a precision value which is required to be reached by a preset evaluation index.

If the calculated evaluation index does not meet the preset requirement, the parameters of the prediction model to be verified are adjusted, the model is trained according to the adjusted parameters until the evaluation index can meet the preset requirement, the model to be verified corresponding to the current evaluation index is used as a candidate prediction model, and the weight corresponding to the current evaluation index is used as a preset weight.

For example, assume that the reference power system data is gearbox oil temperature data; the model to be verified is a Chronos model which is divided into a plurality of models with sizes of Mini (20M), small (46M), base (200M), large (710M) and GPT-2 basic models (90M). According to the resource condition and the business requirement of the power system, a Small (46M) model is selected, the reference power system data is divided into first-period reference power system data and second-period reference power system data, the first-period reference power system data is used as a training set, the second-period reference power system data is used as a verification set, the first-period reference power system data is divided into five training tasks of 1 hour, 30 minutes, 5 minutes, 60s and 10s according to a preset time step, the five training tasks are input into a Chronos model, data of three hours, one day and three days in the future are predicted, and the accuracy and mean square error of the predicted data of the reference power system data of the second period are calculated respectively. The parameters are adjusted according to the accuracy and mean square error, such as the learning rate is linearly set to 0 from the initial value of 0.001.

S130, determining next power system data of a power system in a next period after the current period according to the target prediction model.

And inputting the current power system data into a target prediction model to perform prediction to obtain the next power system data of the power system in the next period after the current period.

Wherein the duration of the next period is not limited.

And S140, determining an early warning result according to the next power system data and a preset threshold value for early warning.

The pre-set threshold is used for judging whether the next power system data has overrun data, and the early warning result is used for representing whether the next power system data exceeds the pre-set threshold.

The early warning result comprises: the overrun needs to be early-warned and operates normally.

And comparing the next power system data with a preset threshold value to obtain an early warning result, and carrying out early warning according to the early warning result.

Optionally, the early warning is performed according to the data of the next power system and a preset threshold value, and the early warning result is determined, including the steps of F1-F2:

and F1, comparing the data of the next power system with a preset threshold value to obtain an early warning result.

And comparing the next power system data with a preset threshold value one by one, and taking the comparison result as an early warning result.

And F2, judging whether to trigger early warning according to the early warning result, and displaying early warning information on a display interface. The early warning information is used for representing whether the abnormality of the power system is processed or not.

Further, the early warning information includes: and (5) an electric power system and an early warning result.

If the early warning result is that the electric power system exceeds the limit, the early warning is triggered, early warning information is generated according to the type of the electric power system and the early warning result, and the early warning information is displayed on a display interface.

And F3, generating an early warning receipt according to the early warning information.

When the inspection of the early warning power system is completed, an early warning receipt is generated according to the processing result and the early warning information, wherein the processing result mode comprises the following steps: early warning confirmation and early warning rejection.

The method comprises the steps of taking an electric power system as a fan, carrying out early warning if the predicted next electric power system data exceeds a preset threshold value, generating early warning information, namely, if the early warning information is fan overrun, checking the fan according to the early warning information, processing if the fan has a fault, generating an early warning receipt after the processing is finished, and carrying out fan overrun-early warning confirmation; if no fault exists, an early warning receipt is directly generated, and the fan is out of limit and early warning is rejected.

According to the technical scheme of the embodiment, through determining the current power system data of the power system in the current period, matching a target prediction model corresponding to the current power system data from candidate prediction models, determining the next power system data of the power system in the next period after the current period through the target prediction model, and determining an early warning result according to the next power system data and a preset threshold value to perform early warning. According to the method, the data of the future time are predicted through the target prediction model, early warning can be sent out in advance for the faults of the power system, loss is avoided, the stable operation of the power system is ensured, meanwhile, the training cost is reduced, and the prediction efficiency and accuracy are improved.

Fig. 2 is a schematic structural diagram of an apparatus for detecting data anomalies in a power system according to an embodiment of the present invention. The embodiment is applicable to the case of abnormality detection of power system data, the power system data abnormality detection device may be implemented in hardware and/or software, and the power system data abnormality detection device may be configured in any electronic device having a network communication function. As shown in fig. 2, the apparatus includes: a data determination module 210, a prediction model matching module 220, a prediction module 230, and an early warning module 240, wherein:

The data determination module 210: the method comprises the steps of determining current power system data of a power system in a current period, wherein the current power system data are obtained by processing original data of the power system in the current period;

The prediction model matching module 220: the method comprises the steps that a target prediction model matched with current power system data is determined from candidate prediction models, the candidate prediction models are used for predicting power system data of a power system in a second period according to the power system data in a first period, the second period is an adjacent period after the first period, the candidate prediction models are obtained by training according to reference power system data, and the reference power system data are obtained by processing original power system data of the power system in a historical moment;

The prediction module 230: determining next power system data of a power system in a next period after the current period according to the target prediction model;

early warning module 240: the early warning method comprises the steps of determining an early warning result according to the next power system data and a preset threshold value, wherein the preset threshold value is used for judging whether the next power system data have overrun data, and the early warning result is used for representing whether the next power system data exceed the preset threshold value.

Optionally, the data determining module 210 includes:

original data acquisition unit: the method comprises the steps of acquiring original data of a power system in a current period;

a data determination unit: the preprocessing method is used for preprocessing the original data of the power system in the current period to obtain the current power system data, and the preprocessing method comprises the following steps: missing value completion and outlier processing, wherein outlier is data generated during abnormal operation of the power system, and the abnormal operation at least comprises: and (5) stopping, and performing fault and maintenance.

Optionally, the prediction model matching module 220 includes:

A data type determining unit: the data type of the current power system data comprises univariate data or multivariate data;

a prediction model matching unit: and the target prediction model is used for determining a matched target prediction model from the candidate prediction models according to the data type of the current power system data, and the target prediction model can process the current power system data corresponding to the data type of the current power system.

Optionally, the prediction model matching unit includes:

A data duration acquisition subunit: a time period for determining current power system data;

A first target prediction model matching subunit: if the data type of the current power system data is single variable data, matching a prediction model corresponding to the single variable data from candidate prediction models, and matching a weight value corresponding to the duration of the current power system data from a first preset weight, so as to obtain a target prediction model matched with the single variable data, wherein the first preset weight is a weight value obtained by training single variable reference power system data according to different durations in advance;

a second target prediction model matching subunit: and if the data type of the current power system data is the multi-variable data, matching a prediction model corresponding to the multi-variable data from the candidate prediction models, and matching a weight value corresponding to the duration of the current power system data from a second preset weight, so as to obtain a target prediction model matched with the multi-variable data, wherein the second preset weight is a weight value obtained by training according to multi-variable reference power system data with different durations in advance.

Optionally, the prediction model matching module 220 includes:

A reference data acquisition unit: the method comprises the steps of acquiring reference power system data of a power system;

A data dividing unit: dividing the reference power system data into first period reference power system data and second period reference power system data;

a data to be predicted determining unit: the method comprises the steps of dividing reference power system data of a first period according to a preset time step to obtain data to be predicted;

Prediction unit: the method comprises the steps of inputting data to be predicted into a prediction model to be verified for prediction to obtain reference power system data of the next period;

Candidate prediction model determination unit: and the method is used for adjusting the prediction model to be verified according to the second period reference power system data and the next period reference power system data to obtain a candidate prediction model.

Optionally, the candidate prediction model determining unit includes:

an evaluation index determination subunit: the evaluation index is used for determining evaluation indexes of the second period reference power system data and the next period reference power system data, and the evaluation indexes are used for evaluating the accuracy rate of prediction of the prediction model to be verified;

Parameter adjustment subunit: the method is used for adjusting parameters of the prediction model to be verified according to the evaluation indexes, training according to the adjusted parameters until the evaluation indexes meet the preset requirements, taking weights corresponding to the evaluation indexes meeting the preset requirements as preset weights, wherein the preset weights comprise: the first preset weight and the second preset weight.

Optionally, the early warning module 240 includes:

early warning result determining unit: the method comprises the steps of comparing the data of the next power system with a preset threshold value to obtain an early warning result;

An early warning unit: the early warning device is used for judging whether to trigger early warning according to the early warning result, displaying early warning information on a display interface, wherein the early warning information is used for representing whether the abnormality of the power system is processed or not;

receipt generation unit: and the early warning receipt is generated according to the early warning information.

The device for detecting the abnormality of the power system data provided by the embodiment of the invention can execute the method for detecting the abnormality of the power system data provided by any embodiment of the invention, has the corresponding functions and beneficial effects of executing the method for detecting the abnormality of the power system data, and the detailed process refers to the related operation of the method for detecting the abnormality of the power system data in the embodiment.

Fig. 3 is a schematic structural diagram of an electronic device for implementing the method for detecting data anomalies in a power system according to an embodiment of the present invention. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. Electronic equipment may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed herein.

As shown in fig. 3, the electronic device 10 includes at least one processor 11, and a memory, such as a Read Only Memory (ROM) 12, a Random Access Memory (RAM) 13, etc., communicatively connected to the at least one processor 11, in which the memory stores a computer program executable by the at least one processor, and the processor 11 may perform various appropriate actions and processes according to the computer program stored in the Read Only Memory (ROM) 12 or the computer program loaded from the storage unit 18 into the Random Access Memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 may also be stored. The processor 11, the ROM 12 and the RAM 13 are connected to each other via a bus 14. An input/output (I/O) interface 15 is also connected to bus 14.

Various components in the electronic device 10 are connected to the I/O interface 15, including: an input unit 16 such as a keyboard, a mouse, etc.; an output unit 17 such as various types of displays, speakers, and the like; a storage unit 18 such as a magnetic disk, an optical disk, or the like; and a communication unit 19 such as a network card, modem, wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.

The processor 11 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various processors running machine learning model algorithms, digital Signal Processors (DSPs), and any suitable processor, controller, microcontroller, etc. The processor 11 performs the respective methods and processes described above, such as the power system data abnormality detection method.

In some embodiments, the power system data anomaly detection method may be implemented as a computer program tangibly embodied on a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 10 via the ROM 12 and/or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the power system data abnormality detection method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform the power system data anomaly detection method by any other suitable means (e.g., by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.

A computer program for carrying out methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions/acts specified in the flowchart and/or block diagram block or blocks to be implemented. The computer program may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of the present invention, a computer-readable storage medium may be a tangible medium that can contain, or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer readable storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, the computer readable storage medium may be a machine readable signal medium. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) through which a user can provide input to the electronic device. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), blockchain networks, and the internet.

The computing system may include clients and servers. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system, so that the defects of high management difficulty and weak service expansibility in the traditional physical hosts and VPS service are overcome.

It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps described in the present invention may be performed in parallel, sequentially, or in a different order, so long as the desired results of the technical solution of the present invention are achieved, and the present invention is not limited herein.

The above embodiments do not limit the scope of the present invention. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of the present invention.

Claims (10)

1. A method for detecting anomalies in power system data, comprising:

Determining current power system data of a power system in a current period, wherein the current power system data are obtained by processing original data of the power system in the current period;

Determining a target prediction model matched with the current power system data from candidate prediction models, wherein the candidate prediction models are used for predicting power system data of a power system in a second period according to the power system data in a first period, the second period is an adjacent period after the first period, the candidate prediction models are obtained by training according to reference power system data, and the reference power system data are obtained by processing original power system data of the power system in a historical moment;

Determining next power system data of a power system in a next period after the current period according to the target prediction model;

And determining an early warning result according to the next power system data and a preset threshold value, wherein the preset threshold value is used for judging whether the next power system data has overrun data or not, and the early warning result is used for representing whether the next power system data exceeds the preset threshold value or not.

2. The method of claim 1, wherein determining current power system data for a current time period for a power system comprises:

Acquiring original data of the power system in a current period;

Preprocessing the original data of the power system in the current period to obtain the current power system data, wherein the preprocessing comprises the following steps: missing value completion and outlier processing, wherein the outlier is data generated during abnormal operation of the power system, and the abnormal operation at least comprises: and (5) stopping, and performing fault and maintenance.

3. The method of claim 1, wherein the determining a target prediction model from candidate prediction models that matches the current power system data comprises:

determining the data type of current power system data, wherein the data type of the current power system data comprises univariate data or multivariate data;

and determining a matched target prediction model from the candidate prediction models according to the data type of the current power system data, wherein the target prediction model can process the current power system data corresponding to the data type of the current power system.

4. A method according to claim 3, wherein said determining a matching target prediction model from candidate prediction models according to the data type of the current power system data comprises:

determining the duration of current power system data;

If the data type of the current power system data is single variable data, matching a prediction model corresponding to the single variable data from a candidate prediction model, and matching a weight value corresponding to the duration of the current power system data from a first preset weight, so as to obtain a target prediction model matched with the single variable data, wherein the first preset weight is a weight value obtained by training single variable reference power system data according to different durations in advance;

And if the data type of the current power system data is the multi-variable data, matching a prediction model corresponding to the multi-variable data from the candidate prediction model, and matching a weight value corresponding to the duration of the current power system data from a second preset weight, so as to obtain a target prediction model matched with the multi-variable data, wherein the second preset weight is a weight value obtained by training according to multi-variable reference power system data with different durations in advance.

5. The method according to claim 1, wherein the process of constructing the candidate predictive model comprises:

acquiring reference power system data of a power system;

Dividing the reference power system data into first period reference power system data and second period reference power system data;

Dividing the first period reference power system data according to a preset time step to obtain data to be predicted;

Inputting the data to be predicted into a prediction model to be verified for prediction to obtain reference power system data of the next period;

And adjusting the prediction model to be verified according to the second period reference power system data and the next period reference power system data to obtain candidate prediction models.

6. The method of claim 5, wherein said adjusting the predictive model to be validated based on the second period reference power system data and the next period reference power system data comprises:

Determining an evaluation index of the second period reference power system data and the next period reference power system data, wherein the evaluation index is used for evaluating the accuracy rate of prediction of a prediction model to be verified;

Adjusting parameters of the prediction model to be verified according to the evaluation index, training according to the adjusted parameters until the evaluation index meets the preset requirement, taking the weight corresponding to the evaluation index meeting the preset requirement as the preset weight, wherein the preset weight comprises: the first preset weight and the second preset weight.

7. The method of claim 1, wherein the determining the pre-warning result according to the next power system data and a preset threshold value comprises:

comparing the data of the next power system with a preset threshold value to obtain an early warning result;

Judging whether to trigger early warning according to the early warning result, and displaying early warning information on a display interface, wherein the early warning information is used for representing whether the abnormality of the power system is processed or not;

And generating an early warning receipt according to the early warning information.

8. An electric power system data abnormality detection device, characterized by comprising:

the data determining module is used for determining current power system data of the power system in a current period, wherein the current power system data are obtained by processing original data of the power system in the current period;

The prediction model matching module is used for determining a target prediction model matched with the current power system data from candidate prediction models, the candidate prediction models are used for predicting the power system data of a power system in a second period according to the power system data in a first period, the second period is an adjacent period after the first period, the candidate prediction models are obtained by training according to reference power system data, and the reference power system data are obtained by processing the original power system data of the power system in a historical moment;

the prediction module is used for determining next power system data of a power system in a next period after the current period according to the target prediction model;

The early warning module is used for carrying out early warning according to the data of the next power system and a preset threshold value, the preset threshold value is used for judging whether the data of the next power system have overrun data, and the early warning result is used for representing whether the data of the next power system exceed the preset threshold value.

9. An electronic device, the electronic device comprising:

At least one processor; and

A memory communicatively coupled to the at least one processor; wherein,

The memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the power system data anomaly detection method of any one of claims 1-7.

10. A computer readable storage medium storing computer instructions for causing a processor to implement the power system data anomaly detection method of any one of claims 1-7 when executed.