CN114970373B - Method and device for predicting voltage distortion rate, electronic equipment and readable storage medium - Google Patents
Method and device for predicting voltage distortion rate, electronic equipment and readable storage medium Download PDFInfo
- Publication number
- CN114970373B CN114970373B CN202210861157.8A CN202210861157A CN114970373B CN 114970373 B CN114970373 B CN 114970373B CN 202210861157 A CN202210861157 A CN 202210861157A CN 114970373 B CN114970373 B CN 114970373B
- Authority
- CN
- China
- Prior art keywords
- distortion rate
- voltage distortion
- data
- rate data
- mean difference
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R19/00—Arrangements for measuring currents or voltages or for indicating presence or sign thereof
- G01R19/0084—Arrangements for measuring currents or voltages or for indicating presence or sign thereof measuring voltage only
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/40—Testing power supplies
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The present disclosure relates to the field of computer processing technologies, and in particular, to a method and an apparatus for predicting a voltage distortion rate, an electronic device, and a readable storage medium, where the method for predicting a voltage distortion rate includes: acquiring voltage distortion rate data of a target power supply; obtaining predicted input data according to the voltage distortion rate data; and inputting the predicted input data into a pre-trained voltage distortion rate prediction model to obtain a predicted value of the voltage distortion rate data of the target power supply. The embodiment provided by the disclosure is used for accurately predicting the power quality of the target power supply.
Description
Technical Field
The disclosure relates to the technical field of computer processing, and in particular to a method and an apparatus for predicting a voltage distortion rate, an electronic device, and a readable storage medium.
Background
With the development of power electronic technology, nonlinear power electronic devices and devices are widely applied in modern industry, direct current Transmission and FACTS (Flexible AC Transmission Systems) technology is continuously put into practical engineering application, speed regulating motors and reactive power compensation capacitors are also largely put into operation, and the operation of the devices causes waveform distortion of voltage and current in a power grid to be more and more serious, and harmonic level to be continuously increased. Further, a load having shock and/or fluctuation (for example, an arc furnace, a large-sized rolling mill, an electric locomotive, etc.) generates not only a large amount of harmonics but also power quality problems such as voltage fluctuation, flicker, three-phase imbalance, etc. in operation.
The quality of electric energy refers to the quality of electric energy in an electric power system, and ideal electric energy should be a sine wave with perfect symmetry. The above-mentioned interference in many aspects may degrade the quality of the power supplied by the power grid, and affect the normal operation of the load devices in the power grid. Therefore, the power quality of the power supply grid needs to be predicted, and the power quality can be adjusted in advance according to the prediction result. At present, the power quality of a power supply grid can be detected by using power monitoring equipment, but the power quality cannot be predicted more accurately.
Therefore, how to accurately predict the power quality of the power supply is a technical problem to be solved urgently.
Disclosure of Invention
In order to solve the problems in the related art, embodiments of the present disclosure provide a method and apparatus for predicting a voltage distortion rate, an electronic device, and a readable storage medium.
In a first aspect, an embodiment of the present disclosure provides a method for predicting a voltage distortion rate, including: acquiring voltage distortion rate data of a target power supply; obtaining predicted input data according to the voltage distortion rate data; inputting the predicted input data into a pre-trained voltage distortion rate prediction model to obtain a predicted value of the voltage distortion rate data of the target power supply; the voltage distortion rate prediction model is a machine learning model obtained by training an initial prediction model by using a training data set, and is used for executing a prediction task of the voltage distortion rate data of the target power supply, and the training data set is generated according to a plurality of historical voltage distortion rate data of the target power supply and data obtained by performing mean difference processing on the plurality of historical voltage distortion rate data.
In some embodiments, said deriving predicted input data from said voltage distortion rate data comprises: carrying out mean difference processing on the voltage distortion rate data to obtain mean difference data corresponding to any voltage distortion rate data in the voltage distortion rate data; and taking the arbitrary voltage distortion rate data and mean difference data corresponding to the arbitrary voltage distortion rate data as the prediction input data.
In some embodiments, the performing mean difference processing on the voltage distortion rate data to obtain mean difference data corresponding to any voltage distortion rate data in the voltage distortion rate data includes: calculating the average value of the voltage distortion rate data to obtain a mean difference reference value; and obtaining mean difference data corresponding to the arbitrary voltage distortion rate data according to the arbitrary voltage distortion rate data and the mean difference reference value.
In some embodiments, the method further comprises: normalizing the voltage distortion rate data.
In some embodiments, the normalizing the voltage distortion rate data comprises: determining the voltage distortion rate data with the largest value in the voltage distortion rate data to obtain a normalized reference value; and carrying out normalization processing on any voltage distortion rate data in the voltage distortion rate data according to the normalization reference value.
In some embodiments, the step of inputting the prediction input data into a pre-trained voltage distortion rate prediction model to obtain a prediction value of the voltage distortion rate data of the target power supply includes: and inputting the predicted input data into the long-short term memory neural network model, and obtaining a predicted value of the voltage distortion rate data according to the output of the last hidden layer of the long-short term memory neural network model.
In some embodiments, the voltage distortion rate prediction model is trained by: acquiring a plurality of historical voltage distortion rate data of a target power supply; performing mean difference processing on the plurality of historical voltage distortion rate data; generating a training data set according to the plurality of historical voltage distortion rate data and the data after mean difference processing; and training the initial prediction model by using the training data set to obtain a voltage distortion rate prediction model.
In some embodiments, the method further comprises: dividing the plurality of historical voltage distortion rate data into at least one voltage distortion rate data set; and taking a next element of the plurality of historical voltage distortion rate data, which is adjacent to a last element of any voltage distortion rate data set in the at least one voltage distortion rate data set, as annotation data corresponding to the any voltage distortion rate data set.
In some embodiments, the mean difference processing of the plurality of historical voltage distortion rate data comprises: calculating the average value of all data of any voltage distortion rate data set in the at least one voltage distortion rate data set to obtain a mean difference reference value corresponding to the any voltage distortion rate data set; and obtaining a mean difference data group corresponding to the arbitrary voltage distortion rate data group according to the arbitrary voltage distortion rate data group and the mean difference reference value.
In some embodiments, the obtaining a mean difference data set corresponding to the arbitrary voltage distortion rate data set according to the arbitrary voltage distortion rate data set and the mean difference reference value includes: comparing any voltage distortion rate data in the any voltage distortion rate data group with the mean differential reference value; and determining mean difference data in the mean difference data group corresponding to the arbitrary voltage distortion rate data according to the comparison result.
In some embodiments, the determining, according to the comparison result, mean difference data in the mean difference data group corresponding to the arbitrary voltage distortion rate data includes: if the value of the arbitrary voltage distortion rate data is greater than or equal to the mean difference reference value, determining that the value of the mean difference data is 1, otherwise, determining that the value of the mean difference data is 0.
In some embodiments, the method further comprises: and normalizing any voltage distortion rate data set in the at least one voltage distortion rate data set and the annotation data corresponding to the any voltage distortion rate data set.
In some embodiments, the normalizing any voltage distortion rate data set of the at least one voltage distortion rate data set and the annotation data corresponding to the any voltage distortion rate data set includes: determining voltage distortion rate data with the largest value in the plurality of historical voltage distortion rate data as a normalization reference value; and according to the normalization reference value, normalizing the arbitrary voltage distortion rate data group and the labeled data corresponding to the arbitrary voltage distortion rate data group.
In some embodiments, the normalizing the arbitrary voltage distortion rate data set and the labeled data corresponding to the arbitrary voltage distortion rate data set according to the normalized reference value includes: calculating the ratio of any voltage distortion rate data in the any voltage distortion rate data group to the normalization reference value to obtain normalized voltage distortion rate data corresponding to the any voltage distortion rate data; and calculating the ratio of the annotation data to the normalized reference value to obtain normalized annotation data corresponding to the annotation data.
In some embodiments, the training data set includes sample data and a label of the sample data, and the generating the training data set from the plurality of historical voltage distortion rate data and the mean difference processed data includes: obtaining a group of sample data according to any voltage distortion rate data group in the at least one voltage distortion rate data group and a mean difference data group corresponding to the any voltage distortion rate data group; and taking the labeled data corresponding to the arbitrary voltage distortion rate data group as a label of the sample data, wherein the label is used for representing the predicted value of the voltage distortion rate data expected by using the initial prediction model to process the arbitrary voltage distortion rate data group.
In some embodiments, the obtaining a set of sample data according to any voltage distortion rate data set of the at least one voltage distortion rate data set and a mean difference data set corresponding to the any voltage distortion rate data set includes: and taking the ith voltage distortion rate data in the arbitrary voltage distortion rate data group and the ith mean difference data in the mean difference data group as the ith group element of the sample data, wherein i is an integer greater than 0.
In some embodiments, the training the initial prediction model using the training data set to obtain the voltage distortion rate prediction model includes: inputting any sample data in the training data set into the initial prediction model to obtain a voltage distortion rate data initial identification result; calculating a loss value of the training by using a loss function according to the initial identification result of the voltage distortion rate data and the label of the any sample data; judging whether the loss function is converged or not according to the loss value; and if so, taking the initial prediction model as the voltage distortion rate prediction model, otherwise, adjusting the initial prediction model, carrying out next training until the loss function is converged, and taking the adjusted initial prediction model as the voltage distortion rate prediction model.
In some embodiments, the initial prediction model is a long-short term memory neural network model, and the inputting any sample data in the training data set into the initial prediction model to obtain the initial identification result of the voltage distortion rate data includes: and inputting the any sample data into the long-short term memory neural network model, and obtaining the initial identification result of the voltage distortion rate data according to the output of the last hidden layer of the long-short term memory neural network model.
In some embodiments, said determining whether said loss function converges based on said loss value comprises: judging whether the loss value is smaller than a preset loss threshold value or not; or judging whether the difference value between the loss value and the loss value obtained by the last training is smaller than a preset loss difference threshold value.
In a second aspect, an embodiment of the present disclosure provides an apparatus for predicting a voltage distortion rate, including: the voltage distortion rate data acquisition module is used for acquiring voltage distortion rate data of the target power supply; the predicted input data acquisition module is used for obtaining predicted input data according to the voltage distortion rate data; the prediction module is used for inputting the prediction input data into a pre-trained voltage distortion rate prediction model to obtain a prediction value of the voltage distortion rate data of the target power supply; the voltage distortion rate prediction model is a machine learning model obtained by training an initial prediction model by using a training data set, and is used for executing a prediction task of the voltage distortion rate data of the target power supply, and the training data set is generated according to a plurality of historical voltage distortion rate data of the target power supply and data obtained by performing mean difference processing on the plurality of historical voltage distortion rate data.
In some embodiments, the apparatus further comprises: the training data set acquisition module is used for acquiring a plurality of historical voltage distortion rate data of the target power supply; performing mean difference processing on the plurality of historical voltage distortion rate data; generating a training data set according to the plurality of historical voltage distortion rate data and the data after mean difference processing; and the training module is used for training the initial prediction model by using the training data set to obtain a voltage distortion rate prediction model.
In a third aspect, the embodiments of the present disclosure provide a chip including the apparatus according to any one of the embodiments of the second aspect.
In a fourth aspect, the disclosed embodiments provide an electronic device comprising a memory and a processor, wherein the memory is configured to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the method as described above.
In a fifth aspect, embodiments of the present disclosure provide a computer-readable storage medium having stored thereon computer instructions, which when executed by a processor, implement the method as described above.
According to the technical scheme provided by the embodiment of the disclosure, the voltage distortion rate data of the target power supply is obtained, the prediction input data is obtained according to the voltage distortion rate data, and then the prediction input data is input into the pre-trained voltage distortion rate prediction model, so that the prediction value of the voltage distortion rate data of the target power supply is obtained. The voltage distortion rate data can accurately reflect the electric energy quality of the target power supply, and the trained voltage distortion rate prediction model can be used for obtaining the voltage distortion rate data of the target power supply at a future moment from the obtained voltage distortion rate data, so that the electric energy quality of the target power supply can be accurately predicted by the method provided by the embodiment of the disclosure.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.
Drawings
Other features, objects, and advantages of the present disclosure will become more apparent from the following detailed description of non-limiting embodiments when taken in conjunction with the accompanying drawings. In the drawings.
Fig. 1 illustrates an application scenario of a method for predicting a voltage distortion rate according to an embodiment of the present disclosure.
Fig. 2 illustrates an exemplary schematic diagram of a method of predicting a voltage distortion rate according to an embodiment of the present disclosure.
Fig. 3 illustrates an exemplary schematic diagram of a training method of a voltage distortion rate prediction model according to an embodiment of the disclosure.
Fig. 4 illustrates a block diagram of a structure of a prediction apparatus of a voltage distortion rate according to an embodiment of the present disclosure.
Fig. 5 shows a block diagram of an electronic device according to an embodiment of the present disclosure.
FIG. 6 shows a schematic block diagram of a computer system suitable for use in implementing a method according to an embodiment of the present disclosure.
Detailed Description
Hereinafter, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily implement them. Furthermore, parts that are not relevant to the description of the exemplary embodiments have been omitted from the drawings for the sake of clarity.
In the present disclosure, it is to be understood that terms such as "including" or "having," etc., are intended to indicate the presence of the disclosed features, numbers, steps, behaviors, components, parts, or combinations thereof, and are not intended to preclude the possibility that one or more other features, numbers, steps, behaviors, components, parts, or combinations thereof may be present or added.
It should also be noted that the embodiments and features of the embodiments in the present disclosure may be combined with each other without conflict. The present disclosure will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.
In the present disclosure, if an operation of acquiring user information or user data or an operation of presenting user information or user data to others is involved, the operations are all operations authorized, confirmed by a user, or actively selected by the user.
According to the technical scheme provided by the embodiment of the disclosure, the voltage distortion rate data of the target power supply is obtained, the prediction input data is obtained according to the voltage distortion rate data, and then the prediction input data is input into a pre-trained voltage distortion rate prediction model, so that the prediction value of the voltage distortion rate data of the target power supply is obtained. The voltage distortion rate data can accurately reflect the electric energy quality of the target power supply, and the trained voltage distortion rate prediction model can be used for obtaining the voltage distortion rate data of the target power supply at a future moment from the obtained voltage distortion rate data, so that the electric energy quality of the target power supply can be accurately predicted by the method provided by the embodiment of the disclosure.
Fig. 1 illustrates an application scenario of a method for predicting a voltage distortion rate according to an embodiment of the present disclosure.
The method for predicting the voltage distortion rate can be applied to the task of predicting the power quality of various power supplies, for example, the power quality of various power supplies such as an alternating current stabilized power supply, a direct current stabilized power supply, an inverter stabilized power supply, a switch stabilized power supply and the like can be predicted. By way of example only, fig. 1 illustrates a low-voltage transformer area of a distribution transformer in a substation as a target power supply. The low-voltage transformer area is an area of low-voltage power supply in the transformer, and the low-voltage transformer area is divided for the requirement of power utilization management, so that the management in the aspects of personnel division, equipment maintenance, electric quantity calculation, line loss statistics and the like is more standard and scientific.
As shown in fig. 1, a service end 110, a terminal 120 and a network 130 may be included in an application scenario.
In some embodiments, data or information may be exchanged between the server 110 and the terminal 120 through the network 130. For example, the server 110 may obtain information and/or data in the terminal 120 through the network 130, or may transmit information and/or data to the terminal 120 through the network 130.
The terminal 120 may be an electronic device that receives harmonic data of a target power supply from a harmonic tester. The terminal 120 may be one or any combination of a mobile device, a tablet computer, and the like having input and/or output capabilities. In fig. 1, a terminal 120 is described as an example of a personal computer.
In some embodiments, the personnel may continuously monitor the target power supply using the harmonic tester for a predetermined period of time (e.g., 72 hours) while transmitting the measured harmonic data (e.g., via data line transmission or network transmission, etc.) to the terminal 120. The terminal 120 uploads the harmonic data to the server 110 for processing.
In some embodiments, the terminal 120 may also locally calculate voltage distortion rate data according to the acquired harmonic data, and process the voltage distortion rate data by using the voltage distortion rate prediction model without sending the harmonic data to the server 110, which is not limited in this embodiment. For convenience of description, the following description will be given taking an example in which a server executes an embodiment of the present disclosure.
The server 110 may be a single server or a group of servers. The set of servers may be centralized or distributed (e.g., the server 110 may be a distributed system), may be dedicated, or may be serviced by other devices or systems at the same time. In some embodiments, the server 110 may be regional or remote. In some embodiments, the server 110 may be implemented on a cloud platform or provided in a virtual manner. By way of example only, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multi-tiered cloud, and the like, or any combination thereof.
In some embodiments, the service end 110 may obtain harmonic data of one or more target power supplies (e.g., low-voltage transformer areas of multiple distribution transformers) from the terminal 120 through the network 130, calculate voltage distortion rate data according to the obtained harmonic data, obtain predicted input data according to the voltage distortion rate data, process the predicted input data by using a voltage distortion rate prediction model, and feed the prediction result back to the terminal 120 or send the prediction result to relevant personnel for further processing.
In some embodiments, the server 110 may obtain, from the terminal 120 through the network 130, multiple sets of harmonic data of the target power supply within a predetermined time period, calculate to obtain voltage distortion rate data according to the obtained harmonic data, and train the initial prediction model based on the calculated voltage distortion rate data to obtain a trained voltage distortion rate prediction model. In some embodiments, the voltage distortion rate prediction model may be deployed at the terminal 120. In some embodiments, the voltage distortion rate prediction model may also be deployed in the server 110, and is not limited by the description herein. For example, in the case that the hardware configuration of the terminal 120 is not high enough, the voltage distortion rate prediction model may be deployed and used at the server 110.
In some embodiments, the network 130 may be any one or more of a wired network or a wireless network. For example, the network 130 may include a Local Area Network (LAN), a Wide Area Network (WAN), a Wireless Local Area Network (WLAN), a Metropolitan Area Network (MAN), etc., or any combination thereof.
For the convenience of understanding, the technical solutions of the present disclosure are described below with reference to the accompanying drawings and embodiments.
Fig. 2 illustrates an exemplary schematic diagram of a method of predicting a voltage distortion rate according to an embodiment of the present disclosure. As shown in fig. 2, the method for predicting the voltage distortion rate includes the following steps.
In step S201, voltage distortion rate data of the target power supply source is acquired.
The target power supply is a power supply for which power quality prediction is required. For example, the target power supply may be a low-voltage bay of a distribution transformer B in substation a.
The voltage distortion rate data is data for characterizing a voltage distortion condition of the power supply. The quality of electric energy is the quality of electric energy in an electric power system, and electric energy refers to the ability to do work in various forms using electricity. The power is in direct proportion to the voltage at two ends of the circuit, the current in the circuit and the electrifying time, so that the power quality of the target power supply can be reflected by using the voltage distortion rate data.
The ideal power corresponds to a perfectly symmetrical sine wave current or voltage, for example, the standard for domestic electricity is a sine wave voltage with a frequency of 50hz and an amplitude of 220V. The phenomenon that the actual voltage waveform deviates from the standard sine wave due to the large number of harmonic sources present in the grid is called voltage sine wave distortion. The degree of voltage waveform distortion can be measured using the rate of voltage distortion. The voltage distortion rate is expressed as a percentage of a ratio of a root mean square value of each harmonic (for example, a second harmonic, a third harmonic, etc.) existing in a voltage sine wave output from the power supply source to an effective value of the fundamental voltage. The waveforms of different frequencies can be obtained by decomposing the voltage waveform by using fourier analysis, wherein the waveform corresponding to the lowest frequency is called a fundamental wave, the voltage (peak-to-peak value or effective value) of the fundamental wave is the fundamental wave voltage, and the waveforms corresponding to other frequencies are collectively called harmonics.
In a specific implementation process, a harmonic tester can be used for monitoring the target power supply to obtain harmonic data of the target power supply. For example, an original waveform of the voltage of the target power supply may be acquired, and fourier transform may be performed on the acquired original waveform to obtain effective values of each harmonic existing in the original waveform. After obtaining the effective value of each harmonic in the original waveform of the voltage of the target power supply, the root mean square value of each harmonic can be obtained by the following calculation: calculating a square value of the effective value of each harmonic; carrying out averaging calculation on the result of the square value calculation; and performing evolution operation on the result of the averaging calculation. And then calculating the ratio of the root mean square value of each obtained subharmonic to the effective value of the fundamental wave voltage of the target power supply to obtain voltage distortion rate data.
In a specific implementation process, the voltage distortion rate data of the target power supply within a predetermined time period may be acquired for processing. For example, voltage distortion rate data for the target power supply for the current day may be obtained. As another example, voltage distortion rate data for a target power supply source for three days may be acquired.
In step S202, from the voltage distortion rate data, prediction input data is obtained.
In particular implementations, the predictive input data may be derived from the voltage distortion rate data in a variety of ways.
In some embodiments, the voltage distortion rate data may be used as prediction input data.
In some embodiments, the voltage distortion rate data may be subjected to a mean difference process to obtain mean difference data corresponding to any voltage distortion rate data in the voltage distortion rate data, and then the any voltage distortion rate data and the mean difference data corresponding to any voltage distortion rate data are used as prediction input data. For example, for m voltage distortion rate data: x is a radical of a fluorine atom 1 、x 2 、x 3 …x m Corresponding m mean difference data can be calculated: c. C 1 、c 2 、c 3 …c m Then the prediction input data may be: [ x ] of 1 ，c 1 ]、[x 2 ，c 2 ]、[x 3 ，c 3 ]…[x m ，c m ]。
In a specific implementation, the mean difference processing may be performed on the voltage distortion rate data in the following manner: calculating the average value of the voltage distortion rate data to obtain an average difference reference value; and obtaining mean difference data corresponding to the arbitrary voltage distortion rate data according to the arbitrary voltage distortion rate data and the mean difference reference value.
In a specific implementation process, the arbitrary voltage distortion rate data may be compared with the mean difference reference value, and the mean difference data corresponding to the arbitrary voltage distortion rate data may be determined according to the comparison result. For example, if the value of the arbitrary voltage distortion-rate data is greater than or equal to the mean difference reference value, the value of the mean difference data corresponding to the arbitrary voltage distortion-rate data is determined to be 1, otherwise, the value of the mean difference data corresponding to the arbitrary voltage distortion-rate data is determined to be 0.
In some embodiments, before the voltage distortion rate data is subjected to the mean difference processing, the voltage distortion rate data may be subjected to normalization processing, and the normalized voltage distortion rate data and the mean difference data are taken as prediction input data. For example, for m voltage distortion rate data: x is the number of 1 、x 2 、x 3 …x m And calculating to obtain corresponding m normalized voltage distortion rate data: b 1 、b 2 、b 3 …b m And m mean difference data: c. C 1 、c 2 、c 3 …c m Then the prediction input data may be: [ b ] A 1 ，c 1 ]、[b 2 ，c 2 ]、[b 3 ，c 3 ]…[b m ，c m ]。
In a specific implementation, the voltage distortion rate data may be normalized by: determining the voltage distortion rate data with the largest value in the voltage distortion rate data as a normalization reference value; and calculating the ratio of any voltage distortion rate data in the voltage distortion rate data to the normalization reference value to obtain normalized voltage distortion rate data corresponding to the any voltage distortion rate data.
In some embodiments of the present disclosure, the voltage distortion rate data of the target power supply is processed by using the voltage distortion rate prediction model, so as to obtain a predicted value of the voltage distortion rate data of the target power supply at a future time, so that the power quality of the target power supply can be predicted, and further measures can be taken according to the predicted result to prevent the power quality from continuously deteriorating.
In some embodiments of the present disclosure, the voltage distortion rate data is subjected to a mean difference processing to obtain mean difference data corresponding to any voltage distortion rate data in the voltage distortion rate data, and then the any voltage distortion rate data and the mean difference data corresponding to the any voltage distortion rate data are used as prediction input data. For the sudden change data which may appear in the voltage distortion rate data (for example, a certain voltage distortion rate data is higher than other voltage distortion rate data by a larger value), the sudden change data can be effectively smoothed by adding the average difference data into the prediction input data, so that the sudden change data is prevented from influencing the prediction result of the model.
In step S203, the prediction input data is input to a pre-trained voltage distortion rate prediction model, and a prediction value of the voltage distortion rate data of the target power supply is obtained.
The voltage distortion rate prediction model is a machine learning model obtained by training an initial prediction model by using a training data set and is used for executing a prediction task of voltage distortion rate data of a target power supply. The training data set is generated from a plurality of historical voltage distortion rate data of the target power supply and data obtained by performing mean difference processing on the plurality of historical voltage distortion rate data. The details of the training process of the voltage distortion rate prediction model are described in fig. 3, and are not repeated herein.
In a specific implementation process, the predicted input data obtained in step S202 may be input into the voltage distortion rate prediction model, and voltage distortion rate data of the target power supply in a future time period may be obtained according to an output of the voltage distortion rate prediction model.
The voltage distortion rate prediction model may be a timing-based machine learning model, including but not limited to: a Recurrent Neural Network (RNN) model, a Long Short-Term Memory Neural Network (LSTM) model, a gated cyclic Unit (GRU) model, and the like.
In some embodiments, the voltage distortion rate prediction model is an LSTM model, and the prediction input data may be input into the LSTM model, and the prediction value of the voltage distortion rate data may be obtained according to an output of a last hidden layer of the LSTM model. For example, the output of the last hidden layer of the LSTM model may be used as a predicted value of the voltage distortion rate data. For another example, the output data of the last hidden layer of the LSTM model may be multiplied by a coefficient to obtain a predicted value of the voltage distortion rate data.
The LSTM model is a variant of the RNN model, which again calculates the hidden layer output at time t based on the input at time t and the hidden layer output at time t-1. Compared with an RNN model, the LSTM model designs the internal structure more elaborately, and adds an input gate it, a forgetting gate ft and an output gate ot, and an internal memory unit ct. The input gate controls how much the new state currently calculated is updated into the memory unit; the forgetting door controls how much the information in the memory unit of the previous step is forgotten; the output gate controls how much the current output depends on the current memory cell. The LSTM has a powerful gating system, so that the method has a better prediction effect on long sequence data.
Fig. 3 illustrates an exemplary schematic diagram of a training method of a voltage distortion rate prediction model according to an embodiment of the disclosure. As shown in fig. 3, the training method of the voltage distortion rate prediction model includes the following steps.
In step S301, a plurality of pieces of historical voltage distortion rate data of the target power supply source are acquired.
The historical voltage distortion rate data may be voltage distortion rate data of a plurality of consecutive target power supply sources acquired over a certain historical period of time. For example, the historical voltage distortion rate data may be voltage distortion rate data of the target power supply acquired within the past 72 hours. For another example, the historical voltage distortion rate data may be voltage distortion rate data of the target power supply acquired within the past 48 hours. For a detailed description of obtaining the voltage distortion rate data of the target power supply, reference is made to relevant contents in fig. 2, and details are not repeated here.
In some embodiments, the plurality of historical voltage distortion rate data may be divided into at least one voltage distortion rate data group, and a next element of the plurality of historical voltage distortion rate data, which is adjacent to a last element of any voltage distortion rate data group of the at least one voltage distortion rate data group, is taken as anyAnd the voltage distortion rate data set corresponds to the labeled data. For example only, if n pieces of historical voltage distortion rate data X = { X ] are obtained in total 1 ，x 2 ，x 3 ，…，x n } the n pieces of voltage distortion rate data may be divided into n-m (n)>m, and n and m are positive integers), each group including m voltage distortion rate data, to obtain at least one voltage distortion rate data group: x train [1]={x 1 ，x 2 ，x 3 ，…，x m }，X train [2]={x 2 ，x 3 ，x 4 ，…，x m+1 }，…，X train [n-m]={x n-m ，x n-m+1 ，x n-m+2 ，…，x n-1 }，X train [1]The corresponding label data is x m+1 ， X train [2]The corresponding label data is x m+2 ，X train [1]The corresponding label data is x n . It is understood that, when n-m is 1, the number of the at least one voltage distortion rate data set is 1.
In step S302, the mean difference processing is performed on a plurality of pieces of history voltage distortion rate data.
Abrupt change data (for example, a certain voltage distortion rate data is higher than other voltage distortion rate data in a larger value) may occur in the historical voltage distortion rate data for some reason (for example, errors of instruments or transient large-amplitude interference generated by interference sources). In order to smooth the mutation data and avoid the mutation data from affecting the training effect of the model, in the embodiment provided by the present disclosure, the voltage distortion rate data is subjected to mean difference processing, and then the obtained mean difference data and the voltage distortion rate data are used together as training data.
In a specific implementation, the mean difference processing may be performed on the historical voltage distortion rate data in the following manner.
In some embodiments, an average value of all data in any voltage distortion rate data set in the at least one voltage distortion rate data set can be calculated to obtain a mean difference reference value corresponding to any voltage distortion rate data set.
By way of example only, with respect to the aboveAt least one voltage distortion rate data set of the example: x train [1]={x 1 ，x 2 ，x 3 ，…，x m }，X train [2] ={x 2 ，x 3 ，x 4 ，…，x m+1 }，…，X train [n-m]={x n-m ，x n-m+1 ，x n-m+2 ，…，x n-1 The average value of all the data in any voltage distortion rate data set can be calculated respectively, and then:
X 1 average =（x1+x2+…+x m ）/m，X 2 average =（x2+x3+…+x m+1 ）/m，…，X n-m average =（x n-m +x n-m+1 +…+x n-1 ) M, is a reaction of X 1 average As a mean difference reference value of the group 1 voltage distortion rate data group, X 2 average As the mean difference reference value for the group 2 voltage distortion rate data set, and so on.
In some embodiments, the mean difference data set corresponding to the arbitrary voltage distortion rate data set can be obtained according to the arbitrary voltage distortion rate data set and the corresponding mean difference reference value.
In a specific implementation process, arbitrary voltage distortion rate data in an arbitrary voltage distortion rate data group may be compared with a mean difference reference value, and mean difference data in a mean difference data group corresponding to the arbitrary voltage distortion rate data may be determined according to a comparison result.
By way of example only, X may be used c train [i]Data set X representing the distortion rate of the ith voltage train [i]And (3) corresponding mean difference data groups, wherein i is a positive integer greater than 0. For the 1 st group voltage distortion rate data group X in the above example train [1]Any voltage distortion rate data x in the set of data can be used 1 ，x 2 ，x 3 ，…，x m Respectively differencing the reference value X from the mean value 1 average Comparing, and determining arbitrary voltage distortion data (e.g. x) based on the comparison result 3 ) Corresponding mean differencePartial data set X c train [1]Mean difference data (e.g., X) of (1) c train [1][3]）。
In some embodiments, if the value of the arbitrary voltage distortion rate data is greater than or equal to the mean difference reference value, the value of the mean difference data corresponding to the arbitrary voltage distortion rate data is determined to be 1, otherwise, the value of the mean difference data corresponding to the arbitrary voltage distortion rate data is determined to be 0.
By way of example only, for the 1 st set of voltage distortion rate data sets X in the above example train [1]： x 1 ，x 2 ，x 3 ，…，x m If x is 1 >X 1 average ，x 2 <X 1 average ，x 3 ==X 1 average ，…，x m >X 1 average Then, a mean difference data set corresponding to the set of voltage distortion rate data sets can be obtained according to the above calculation: x c train [1]={1，0，1，…，1}。
In some embodiments, in order to obtain better model prediction effect, normalization processing may be performed on any voltage distortion rate data set in the at least one voltage distortion rate data set and the annotation data corresponding to the any voltage distortion rate data set. In particular implementations, the voltage distortion rate data may be normalized using the following method.
In some embodiments, the voltage distortion rate data having the largest value among the plurality of historical voltage distortion rate data may be determined as the normalized reference value. For example, X = { X for n pieces of history voltage distortion rate data in the above example 1 ，x 2 ，x 3 ，…，x n Is given as x if the voltage distortion rate data with the largest value among them is x 3 Then it is used as the normalized reference value
。
In some embodiments, normalization processing may be performed on the arbitrary voltage distortion rate data group and the annotation data corresponding to the arbitrary voltage distortion rate data group according to the normalization reference value. For example, the ratio of any voltage distortion rate data in any voltage distortion rate data group to the normalization reference value can be calculated to obtain normalized voltage distortion rate data corresponding to any voltage distortion rate data. For another example, the ratio of the labeled data corresponding to any voltage distortion rate data set to the normalized reference value may be calculated to obtain the normalized labeled data corresponding to the labeled data.
By way of example only, for any of the voltage distortion rate data sets X in the above examples train [i]And the corresponding label data Y label 
The normalized calculation formula of (a) is as follows:
X b train [i][j]=X train [i][j]/x max （1）
Y b label [i]=Y label [i]/x max （2）
in formula (1), i is used as an index of any voltage distortion rate data set in at least one voltage distortion rate data set, and j is used as an index of data in the voltage distortion rate data set. X train [i][j]Representing the jth data, X, in the ith set of voltage distortion rate data b train [i][j]Normalized voltage distortion rate data Y corresponding to j-th data in the i-th group of voltage distortion rate data sets b label [i]Is X b train [i][j]And corresponding normalized marking data.
In the embodiments provided by the present disclosure, the training data set may be generated according to the at least one voltage distortion rate data group subjected to the normalization processing, and the training data set may also be generated according to the at least one voltage distortion rate data group not subjected to the normalization processing, which is not limited by the description of the present specification. For convenience of description, the following steps are described by taking as an example the case where the training data set is generated from the at least one voltage distortion rate data group subjected to the normalization processing.
In step S303, a training data set is generated from the plurality of historical voltage distortion rate data and the data after the mean difference processing.
In some embodiments, a set of sample data may be obtained according to any voltage distortion rate data set of the at least one voltage distortion rate data set and a mean difference data set corresponding to the any voltage distortion rate data set. In a specific implementation process, the ith voltage distortion rate data in any voltage distortion rate data group and the ith mean difference data in the mean difference data group can be used as the ith group element of the sample data, and i is an integer greater than 0.
By way of example only, for any of the voltage distortion rate data sets X in the above examples train [i]The corresponding normalized voltage distortion rate data set is X b train [i]Corresponding to a mean difference data set of X c train [i]A set of sample data X can be obtained by combining the following ways r train [i]：X r train [i][j]= [X b train [i][j]，X c train [i][j]]Where i is the index of the sample data set, X r train [i]Representing the ith sample data set, j being the index of the data in the sample data set, X r train [i][j]Represents the jth element in the ith sample data group, and each element contains two data: x b train [i][j]And X c train [i][j]。
In some embodiments, the label data corresponding to the arbitrary voltage distortion rate data set can be used as a label of the sample data, and the label is used for representing the predicted value of the voltage distortion rate data expected by processing the arbitrary voltage distortion rate data set by using the initial prediction model. For example, sample data X in the above example r train [i]Is labeled as a voltage distortion rate data set X train [i]Corresponding annotation data Y label [i]。
In step S304, the initial prediction model is trained using the training data set to obtain a voltage distortion rate prediction model, and the voltage distortion rate prediction model is used to perform a task of predicting voltage distortion rate data of the target power supply.
In some embodiments, the initial predictive model may be an untrained LSTM model. The initial prediction model may include at least one neuron, and parameters (e.g., W) of individual neurons in the initial prediction model may be updated based on the training data set obtained as described above f 、W i 、W c 、W o ). The purpose of the training is to determine parameters corresponding to each neuron in the initial prediction model, so that the voltage distortion rate prediction model is obtained according to the parameters determined in the training.
In some embodiments, any sample data in the training data set may be input to the initial prediction model, resulting in the initial identification result of the voltage distortion rate data. In the specific implementation process, any sample data can be input into the long-short term memory neural network model, and the initial identification result of the voltage distortion rate data is obtained according to the output of the last hidden layer of the long-short term memory neural network model.
Then, a loss value of the training can be calculated by using a loss function (for example, a mean square error loss function) according to the initial identification result of the voltage distortion rate data and the label of the arbitrary sample data. After the loss value of the training is obtained, whether the loss function is converged can be judged according to the loss value; if so, the initial prediction model is used as a voltage distortion rate prediction model, otherwise, parameters of the initial prediction model are adjusted (for example, a back propagation algorithm is used), next training is carried out until the loss function is converged, and the adjusted initial prediction model is used as the voltage distortion rate prediction model. In a specific implementation process, whether the loss function converges or not can be judged according to the loss value in the following way: judging whether the loss value is smaller than a preset loss threshold value or not; or judging whether the difference value between the loss value and the loss value obtained by the last training is smaller than a preset loss difference threshold value.
Fig. 4 illustrates a block diagram of a structure of a prediction apparatus of a voltage distortion rate according to an embodiment of the present disclosure. The apparatus may be implemented as part or all of an electronic device through software, hardware, or a combination of both.
As shown in fig. 4, the apparatus 400 for predicting a voltage distortion rate includes a voltage distortion rate data acquisition module 410, a prediction input data acquisition module 410, and a prediction module 430.
A voltage distortion rate data obtaining module 410, configured to obtain voltage distortion rate data of the target power supply.
And a predicted input data obtaining module 420, configured to obtain predicted input data according to the voltage distortion rate data.
The prediction module 430 is configured to input the predicted input data into a pre-trained voltage distortion rate prediction model to obtain a predicted value of voltage distortion rate data of the target power supply; the voltage distortion rate prediction model is a machine learning model obtained by training an initial prediction model by using a training data set, and is used for executing a prediction task of the voltage distortion rate data of the target power supply, and the training data set is generated according to a plurality of historical voltage distortion rate data of the target power supply and data obtained by performing mean difference processing on the plurality of historical voltage distortion rate data.
In some embodiments, the voltage distortion rate predicting device further comprises:
the training data set acquisition module is used for acquiring a plurality of historical voltage distortion rate data of the target power supply; performing mean difference processing on the plurality of historical voltage distortion rate data; and generating a training data set according to the plurality of historical voltage distortion rate data and the data after mean difference processing.
And the training module is used for training the initial prediction model by using the training data set to obtain a voltage distortion rate prediction model.
In the embodiment of the apparatus for predicting voltage distortion rate, the detailed processing of each module and the technical effects thereof can refer to the related descriptions in the corresponding method embodiments, and are not repeated herein.
The embodiment of the present disclosure also provides a chip, which includes the above device for predicting voltage distortion rate, and the device can be implemented as part or all of the chip through software, hardware or a combination of the two.
The present disclosure also discloses an electronic device, and fig. 5 shows a block diagram of the electronic device according to an embodiment of the present disclosure.
As shown in fig. 5, the electronic device includes a memory and a processor, where the memory is configured to store one or more computer instructions, where the one or more computer instructions are executed by the processor to implement a method according to an embodiment of the disclosure.
FIG. 6 shows a schematic block diagram of a computer system suitable for use in implementing a method according to an embodiment of the present disclosure.
As shown in fig. 6, the computer system includes a processing unit that can execute the various methods in the above-described embodiments according to a program stored in a Read Only Memory (ROM) or a program loaded from a storage section into a Random Access Memory (RAM). In the RAM, various programs and data necessary for the operation of the computer system are also stored. The processing unit, the ROM, and the RAM are connected to each other by a bus. An input/output (I/O) interface is also connected to the bus.
The following components are connected to the I/O interface: an input section including a keyboard, a mouse, and the like; an output section including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage section including a hard disk and the like; and a communication section including a network interface card such as a LAN card, a modem, or the like. The communication section performs a communication process via a network such as the internet. The drive is also connected to the I/O interface as needed. A removable medium such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive as needed, so that the computer program read out therefrom is mounted into the storage section as needed. The processing unit can be realized as a CPU, a GPU, a TPU, an FPGA, an NPU and other processing units.
In particular, the above described methods may be implemented as computer software programs according to embodiments of the present disclosure. For example, embodiments of the present disclosure include a computer program product comprising a computer program tangibly embodied on a machine-readable medium, the computer program comprising program code for performing the above-described method. In such an embodiment, the computer program may be downloaded and installed from a network via the communication section, and/or installed from a removable medium.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems that perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The units or modules described in the embodiments of the present disclosure may be implemented by software or by programmable hardware. The units or modules described may also be provided in a processor, and the names of the units or modules do not in some cases constitute a limitation of the units or modules themselves.
As another aspect, the present disclosure also provides a computer-readable storage medium, which may be a computer-readable storage medium included in the electronic device or the computer system in the above embodiments; or it may be a separate computer readable storage medium not incorporated into the device. The computer readable storage medium stores one or more programs for use by one or more processors in performing the methods described in the present disclosure.
The foregoing description is only exemplary of the preferred embodiments of the disclosure and is illustrative of the principles of the technology employed. It will be appreciated by those skilled in the art that the scope of the invention in the present disclosure is not limited to the specific combination of the above-mentioned features, but also encompasses other embodiments in which any combination of the above-mentioned features or their equivalents is made without departing from the inventive concept. For example, the above features and the technical features disclosed in the present disclosure (but not limited to) having similar functions are replaced with each other to form the technical solution.
Claims (23)
1. A method for predicting a voltage distortion rate, comprising:
acquiring voltage distortion rate data of a target power supply;
obtaining predicted input data from the voltage distortion rate data, comprising: carrying out mean difference processing on the voltage distortion rate data to obtain mean difference data corresponding to any voltage distortion rate data in the voltage distortion rate data; taking the arbitrary voltage distortion rate data and mean difference data corresponding to the arbitrary voltage distortion rate data as the prediction input data;
inputting the predicted input data into a pre-trained voltage distortion rate prediction model to obtain a predicted value of the voltage distortion rate data of the target power supply;
the voltage distortion rate prediction model is a machine learning model obtained by training an initial prediction model by using a training data set, and is used for executing a prediction task of the voltage distortion rate data of the target power supply, and the training data set is generated according to a plurality of historical voltage distortion rate data of the target power supply and data obtained by performing mean difference processing on the plurality of historical voltage distortion rate data.
2. The method according to claim 1, wherein the performing mean difference processing on the voltage distortion rate data to obtain mean difference data corresponding to any voltage distortion rate data in the voltage distortion rate data comprises:
calculating the average value of the voltage distortion rate data to obtain an average difference reference value;
and obtaining mean difference data corresponding to the arbitrary voltage distortion rate data according to the arbitrary voltage distortion rate data and the mean difference reference value.
3. The method of claim 2, further comprising:
normalizing the voltage distortion rate data.
4. The method according to claim 3, wherein the normalizing the voltage distortion rate data comprises:
determining the voltage distortion rate data with the largest value in the voltage distortion rate data to obtain a normalized reference value;
and carrying out normalization processing on any voltage distortion rate data in the voltage distortion rate data according to the normalization reference value.
5. The method according to claim 1, wherein the voltage distortion rate prediction model is a long-short term memory neural network model, and the inputting the prediction input data into a pre-trained voltage distortion rate prediction model to obtain a prediction value of the voltage distortion rate data of the target power supply comprises:
and inputting the prediction input data into the long-short term memory neural network model, and obtaining a prediction value of the voltage distortion rate data according to the output of the last hidden layer of the long-short term memory neural network model.
6. The method according to claim 1, wherein the voltage distortion rate prediction model is trained by:
acquiring a plurality of historical voltage distortion rate data of a target power supply;
performing mean difference processing on the plurality of historical voltage distortion rate data;
generating a training data set according to the plurality of historical voltage distortion rate data and the data after mean difference processing;
and training the initial prediction model by using the training data set to obtain a voltage distortion rate prediction model.
7. The method of claim 6, further comprising:
dividing the plurality of historical voltage distortion rate data into at least one voltage distortion rate data set;
and taking a next element of the plurality of historical voltage distortion rate data, which is adjacent to a last element of any voltage distortion rate data set in the at least one voltage distortion rate data set, as annotation data corresponding to the any voltage distortion rate data set.
8. The method according to claim 7, wherein the averaging difference processing of the plurality of historical voltage distortion rate data comprises:
calculating the average value of all data of any voltage distortion rate data set in the at least one voltage distortion rate data set to obtain a mean difference reference value corresponding to the any voltage distortion rate data set;
and obtaining a mean difference data group corresponding to the arbitrary voltage distortion rate data group according to the arbitrary voltage distortion rate data group and the mean difference reference value.
9. The method according to claim 8, wherein the obtaining a mean difference data set corresponding to the arbitrary voltage distortion rate data set according to the arbitrary voltage distortion rate data set and the mean difference reference value comprises:
comparing any voltage distortion rate data in the any voltage distortion rate data set with the mean difference reference value;
and determining mean difference data in the mean difference data group corresponding to the arbitrary voltage distortion rate data according to the comparison result.
10. The method according to claim 9, wherein the determining mean difference data in the mean difference data group corresponding to the arbitrary voltage distortion rate data according to the comparison result comprises:
if the value of the arbitrary voltage distortion rate data is greater than or equal to the mean difference reference value, determining that the value of the mean difference data is 1, otherwise, determining that the value of the mean difference data is 0.
11. The method of claim 7, further comprising:
and normalizing any voltage distortion rate data set in the at least one voltage distortion rate data set and the annotation data corresponding to the any voltage distortion rate data set.
12. The method according to claim 11, wherein the normalizing any voltage distortion rate data set of the at least one voltage distortion rate data set and the annotation data corresponding to the any voltage distortion rate data set comprises:
determining voltage distortion rate data with the largest value in the plurality of historical voltage distortion rate data as a normalization reference value;
and according to the normalization reference value, normalizing the arbitrary voltage distortion rate data group and the labeled data corresponding to the arbitrary voltage distortion rate data group.
13. The method according to claim 12, wherein the normalizing the arbitrary voltage distortion rate data set and the labeled data corresponding to the arbitrary voltage distortion rate data set according to the normalized reference value comprises:
calculating the ratio of any voltage distortion rate data in the any voltage distortion rate data group to the normalization reference value to obtain normalized voltage distortion rate data corresponding to the any voltage distortion rate data;
and calculating the ratio of the annotation data to the normalized reference value to obtain normalized annotation data corresponding to the annotation data.
14. The method according to claim 7 or 11, wherein the training data set comprises sample data and a label of the sample data, and the generating the training data set according to the plurality of historical voltage distortion rate data and the data after mean difference processing comprises:
obtaining a group of sample data according to any voltage distortion rate data group in the at least one voltage distortion rate data group and a mean difference data group corresponding to the any voltage distortion rate data group;
and taking the labeled data corresponding to the arbitrary voltage distortion rate data group as a label of the sample data, wherein the label is used for representing the predicted value of the voltage distortion rate data expected by using the initial prediction model to process the arbitrary voltage distortion rate data group.
15. The method according to claim 14, wherein obtaining a set of sample data from any voltage distortion rate data set of the at least one voltage distortion rate data set and a mean difference data set corresponding to the any voltage distortion rate data set comprises:
and taking the ith voltage distortion rate data in the arbitrary voltage distortion rate data group and the ith mean difference data in the mean difference data group as the ith group element of the sample data, wherein i is an integer greater than 0.
16. The method of claim 14, wherein the training an initial prediction model using the training data set to obtain a voltage distortion rate prediction model comprises:
inputting any sample data in the training data set into the initial prediction model to obtain a voltage distortion rate data initial identification result;
calculating a loss value of the training by using a loss function according to the initial identification result of the voltage distortion rate data and the label of the any sample data;
judging whether the loss function is converged or not according to the loss value;
and if so, taking the initial prediction model as the voltage distortion rate prediction model, otherwise, adjusting the initial prediction model, carrying out next training until the loss function is converged, and taking the adjusted initial prediction model as the voltage distortion rate prediction model.
17. The method of claim 16, wherein the initial prediction model is a long-short term memory neural network model, and the inputting any sample data in the training data set into the initial prediction model to obtain the initial identification result of the voltage distortion rate data comprises:
and inputting the any sample data into the long-short term memory neural network model, and obtaining the initial identification result of the voltage distortion rate data according to the output of the last hidden layer of the long-short term memory neural network model.
18. The method of claim 16, wherein said determining whether the loss function converges based on the loss value comprises:
judging whether the loss value is smaller than a preset loss threshold value or not;
or judging whether the difference value between the loss value and the loss value obtained by the last training is smaller than a preset loss difference threshold value.
19. An apparatus for predicting a voltage distortion rate, comprising:
the voltage distortion rate data acquisition module is used for acquiring voltage distortion rate data of the target power supply;
the predicted input data acquisition module is used for obtaining predicted input data according to the voltage distortion rate data;
the prediction module is used for inputting the prediction input data into a pre-trained voltage distortion rate prediction model to obtain a prediction value of the voltage distortion rate data of the target power supply;
the voltage distortion rate prediction model is a machine learning model obtained by training an initial prediction model by using a training data set, and is used for executing a prediction task of the voltage distortion rate data of the target power supply, and the training data set is generated according to a plurality of historical voltage distortion rate data of the target power supply and data obtained by performing mean difference processing on the plurality of historical voltage distortion rate data;
the predicted input data acquisition module is specifically used for carrying out mean difference processing on the voltage distortion rate data to obtain mean difference data corresponding to any voltage distortion rate data in the voltage distortion rate data; and taking the arbitrary voltage distortion rate data and mean difference data corresponding to the arbitrary voltage distortion rate data as the prediction input data.
20. The apparatus of claim 19, further comprising:
the training data set acquisition module is used for acquiring a plurality of historical voltage distortion rate data of the target power supply; performing mean difference processing on the plurality of historical voltage distortion rate data; generating a training data set according to the plurality of historical voltage distortion rate data and the data after mean difference processing;
and the training module is used for training the initial prediction model by using the training data set to obtain a voltage distortion rate prediction model.
21. A chip, which is characterized in that,
the chip comprising the device of any one of claims 19-20.
22. An electronic device comprising a memory and a processor; wherein the memory is configured to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the method steps of any of claims 1 to 18.
23. A computer-readable storage medium, on which computer instructions are stored, characterized in that the computer instructions, when executed by a processor, implement the method steps of any of claims 1 to 18.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210861157.8A CN114970373B (en) | 2022-07-22 | 2022-07-22 | Method and device for predicting voltage distortion rate, electronic equipment and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210861157.8A CN114970373B (en) | 2022-07-22 | 2022-07-22 | Method and device for predicting voltage distortion rate, electronic equipment and readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114970373A CN114970373A (en) | 2022-08-30 |
| CN114970373B true CN114970373B (en) | 2022-10-25 |
Family
ID=82970148
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210861157.8A Active CN114970373B (en) | 2022-07-22 | 2022-07-22 | Method and device for predicting voltage distortion rate, electronic equipment and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114970373B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107491832A (en) * | 2017-07-12 | 2017-12-19 | 国网上海市电力公司 | Energy quality steady-state index prediction method based on chaology |
| CN108414848A (en) * | 2018-01-08 | 2018-08-17 | 浙江工业大学 | Electric energy quality multi-period comprehensive early warning method for power distribution network with distributed power supply |
| CN109871976A (en) * | 2018-12-20 | 2019-06-11 | 浙江工业大学 | A kind of prediction technique of power quality containing distributed power distribution network based on cluster and neural network |
| WO2020136576A1 (en) * | 2018-12-28 | 2020-07-02 | Abb Schweiz Ag | Power quality monitoring in a distribution grid |
| CN112734106A (en) * | 2021-01-08 | 2021-04-30 | 深圳市国电科技通信有限公司 | Method and device for predicting energy load |
| CN113107626A (en) * | 2021-05-25 | 2021-07-13 | 浙大城市学院 | Combined cycle generator set load prediction method based on multivariable LSTM |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106885572B (en) * | 2015-12-15 | 2019-06-21 | 中国电信股份有限公司 | Utilize the assisted location method and system of time series forecasting |
-
2022
- 2022-07-22 CN CN202210861157.8A patent/CN114970373B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107491832A (en) * | 2017-07-12 | 2017-12-19 | 国网上海市电力公司 | Energy quality steady-state index prediction method based on chaology |
| CN108414848A (en) * | 2018-01-08 | 2018-08-17 | 浙江工业大学 | Electric energy quality multi-period comprehensive early warning method for power distribution network with distributed power supply |
| CN109871976A (en) * | 2018-12-20 | 2019-06-11 | 浙江工业大学 | A kind of prediction technique of power quality containing distributed power distribution network based on cluster and neural network |
| WO2020136576A1 (en) * | 2018-12-28 | 2020-07-02 | Abb Schweiz Ag | Power quality monitoring in a distribution grid |
| CN112734106A (en) * | 2021-01-08 | 2021-04-30 | 深圳市国电科技通信有限公司 | Method and device for predicting energy load |
| CN113107626A (en) * | 2021-05-25 | 2021-07-13 | 浙大城市学院 | Combined cycle generator set load prediction method based on multivariable LSTM |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114970373A (en) | 2022-08-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Heo et al. | Implementation of reliability-centered maintenance for transmission components using particle swarm optimization | |
| Malekshah et al. | Reliability-driven distribution power network dynamic reconfiguration in presence of distributed generation by the deep reinforcement learning method | |
| CN103745276B (en) | A kind of distribution running state analysis method of electrical network | |
| CN109409658A (en) | A kind of power distribution network operation risk cognitive method and device | |
| CN110826228B (en) | Regional power grid operation quality limit evaluation method | |
| JP2009124913A (en) | Voltage imbalance elimination supporting device, supporting method, and program thereof | |
| CN111967738B (en) | Risk early warning method, system and medium of power grid information energy fusion system | |
| Chang et al. | An advanced EAF model for voltage fluctuation propagation study | |
| US20180166880A1 (en) | Data-Driven Demand Charge Management Solution | |
| Majzoobi et al. | Machine learning applications in estimating transformer loss of life | |
| CN110084502A (en) | A kind of power quality controlling equipment running status appraisal procedure and device | |
| CN114970373B (en) | Method and device for predicting voltage distortion rate, electronic equipment and readable storage medium | |
| Pessanha et al. | Benchmarking the operational expenditures of Brazilian transmission utilities by using DEA models | |
| Hinojosa et al. | A computational comparison of 2 mathematical formulations to handle transmission network constraints in the unit commitment problem | |
| Yin et al. | TOPSIS power quality comprehensive assessment based on a combination weighting method | |
| Wang | Risk-based security-constrained optimal power flow: Mathematical fundamentals, computational strategies, validation, and use within electricity markets | |
| Hua et al. | Short-term reliability assessment of UHVdc systems based on state aggregation with SMP | |
| CN106960262A (en) | UHVDC Transmission Lines forecast failure active reserve capacity appraisal procedure and device | |
| CN112016838B (en) | Contribution rate calculation method, system and terminal equipment of power distribution network energy efficiency index system | |
| Wahab | Artificial neural network-based prediction technique for transformer oil breakdown voltage | |
| CN112215394B (en) | Converter transformer vibration signal prediction method, device, equipment and storage medium | |
| Kumar et al. | Fuzzy logic based stability index power system voltage stability enhancement | |
| CN114118587A (en) | Electric energy quality evaluation method, system, equipment and storage medium for distributed photovoltaic | |
| CN114066269A (en) | Running risk assessment technology for old assets | |
| Pinchukov et al. | Estimation of digital substation reliability indices |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |