CN112269059B - Power grid load test method and device, computer equipment and storage medium - Google Patents
Power grid load test method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN112269059B CN112269059B CN202011029461.3A CN202011029461A CN112269059B CN 112269059 B CN112269059 B CN 112269059B CN 202011029461 A CN202011029461 A CN 202011029461A CN 112269059 B CN112269059 B CN 112269059B
- Authority
- CN
- China
- Prior art keywords
- frequency
- phase voltage
- model
- data
- phase
- 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
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R25/00—Arrangements for measuring phase angle between a voltage and a current or between voltages or currents
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R23/00—Arrangements for measuring frequencies; Arrangements for analysing frequency spectra
- G01R23/02—Arrangements for measuring frequency, e.g. pulse repetition rate; Arrangements for measuring period of current or voltage
- G01R23/12—Arrangements for measuring frequency, e.g. pulse repetition rate; Arrangements for measuring period of current or voltage by converting frequency into phase shift
-
- 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
Abstract
The application relates to a method, a device, computer equipment and a storage medium for testing load of a power grid, wherein the method comprises the following steps: acquiring phase voltage data obtained by continuously sampling each sampling point of a power grid; model training is carried out according to the phase voltage data, and a frequency prediction model is obtained; acquiring a phase voltage phase value and a phase current at an asynchronous phase current sampling moment from a target sampling point; calculating according to the phase voltage phase value and the frequency prediction model to obtain a phase voltage frequency value of a target sampling point at an asynchronous phase current sampling moment; and calculating to obtain the load angle of the target sampling point at the sampling moment of the asynchronous phase current according to the phase voltage frequency value and the phase current at the sampling moment of the asynchronous phase current. The load angle is calculated by using the phase voltage and the phase current measured at different moments through the phase voltage frequency prediction, the testing steps are simplified, the voltage testing line wiring work is not required to be repeated in the on-load testing process, and the working efficiency is greatly improved.
Description
Technical Field
The present disclosure relates to the field of power grid technologies, and in particular, to a method and apparatus for testing load on a power grid, a computer device, and a storage medium.
Background
The electric power system is used as an important infrastructure related to national security and national economy pulse, and is required to meet reliable operation in normal environment at any time. In the power system, the correctness of the wiring of the voltage and current secondary circuits determines whether the protection, measurement and control, self-safety and other devices can operate correctly, and further whether the power grid can operate safely and stably is ensured.
The traditional method for testing the load of the electric power system widely adopts the form of an external current clamp and a voltage connecting wire, and the phase and amplitude are calculated by synchronously sampling the current and the voltage in the same meter.
Disclosure of Invention
In view of the foregoing, it is desirable to provide a method, apparatus, computer device, and storage medium for testing on load of a power grid that can improve the efficiency of testing on load of the power grid.
A method for grid load testing, comprising:
acquiring phase voltage data obtained by continuously sampling each sampling point of a power grid;
model training is carried out according to the phase voltage data, and a frequency prediction model is obtained;
acquiring a phase voltage phase value and a phase current at an asynchronous phase current sampling moment from a target sampling point;
calculating a phase voltage frequency value of the target sampling point at an asynchronous phase current sampling moment according to the phase voltage phase value and the frequency prediction model;
and calculating the load angle of the target sampling point at the sampling moment of the asynchronous phase current according to the phase voltage frequency value and the phase current at the sampling moment of the asynchronous phase current.
In one embodiment, the obtaining phase voltage data obtained by continuously sampling each sampling point of the power grid includes: and continuously sampling each sampling point of the power grid by using the high-stability crystal oscillator to acquire phase voltage within a set time length to obtain phase voltage data.
In one embodiment, the frequency prediction model is an ARMA model; model training is carried out according to the phase voltage data to obtain a frequency prediction model, and the method comprises the following steps:
dividing the phase voltage data into training data and test data according to the sampling time;
model training is carried out according to the training data, and a trained ARMA model is obtained;
analyzing the trained ARMA model according to the test data, and detecting whether the ARMA model meets the measurement requirement; if yes, the ARMA model is completed; and if not, returning to the step of acquiring the phase voltage data obtained by continuously sampling each sampling point of the power grid.
In one embodiment, the training of the model according to the training data, to obtain a trained ARMA model, includes:
calculating the frequency value of each cycle according to the rated frequency to the training data to obtain a frequency sequence;
determining ARMA model parameters according to the frequency sequence;
training according to the ARMA model parameters and the frequency sequence to obtain training result data;
calculating the variance between training result data and the frequency sequence, and judging whether the variance meets a preset condition; if yes, obtaining a trained ARMA model; if not, returning to the step of determining ARMA model parameters according to the frequency sequence.
In one embodiment, the determining the ARMA model parameters from the frequency sequence includes: and carrying out autocorrelation diagram and partial autocorrelation diagram analysis on the frequency sequence to obtain an order serving as an ARMA model parameter.
In one embodiment, after the frequency value of each cycle is calculated according to the rated frequency for the training data to obtain a frequency sequence, before determining the ARMA model parameter according to the frequency sequence, the method further includes:
carrying out unit root test on the frequency sequence, and judging whether the frequency sequence is stable or not; if yes, the step of determining ARMA model parameters according to the frequency sequence is carried out; if not, outputting prompt information of unstable power grid frequency.
In one embodiment, after calculating the load angle of the target sampling point at the asynchronous phase current sampling time according to the phase voltage frequency value and the phase current at the asynchronous phase current sampling time, the method further includes:
after the phase voltage frequency measured value of the target sampling point at the asynchronous phase current sampling moment is obtained, verifying the frequency prediction model according to the phase voltage frequency measured value, and outputting model failure prompt information when verification is not passed.
A power grid on-load testing device, comprising:
the data acquisition module is used for acquiring phase voltage data obtained by continuously sampling each sampling point of the power grid;
the model training module is used for carrying out model training according to the phase voltage data to obtain a frequency prediction model;
the parameter acquisition module is used for acquiring a phase voltage phase value and a phase current at an asynchronous phase current sampling moment from a target sampling point;
the frequency calculation module is used for calculating a phase voltage frequency value of the target sampling point at the asynchronous phase current sampling moment according to the phase voltage phase value and the frequency prediction model;
and the load calculation module is used for calculating the load angle of the target sampling point at the sampling moment of the asynchronous phase current according to the phase voltage frequency value and the phase current at the sampling moment of the asynchronous phase current.
A computer device comprising a memory storing a computer program and a processor implementing the steps of the method described above when the processor executes the computer program.
A computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the method described above.
According to the method, the device, the computer equipment and the storage medium for testing the load of the power grid, the frequency prediction model obtained by training the phase voltage data obtained by continuously sampling all sampling points of the power grid is utilized, the phase voltage phase value obtained by collecting the target sampling points is combined to conduct frequency prediction, the phase voltage frequency value of the target sampling points at the asynchronous phase current sampling moment is obtained, and then the load angle of the target sampling points at the asynchronous phase current sampling moment is calculated according to the predicted phase voltage frequency value and the phase current at the asynchronous phase current sampling moment. The load angle is calculated by using the phase voltage and the phase current measured at different moments through the phase voltage frequency prediction, the testing steps are simplified, the voltage testing line wiring work is not required to be repeated in the on-load testing process, and the working efficiency is greatly improved.
Drawings
FIG. 1 is a flow chart of a method for testing load on a power grid in an embodiment;
FIG. 2 is a flow chart of model training based on phase voltage data to obtain a frequency prediction model in an embodiment;
FIG. 3 is a flow chart of model training based on training data to obtain a trained ARMA model in one embodiment;
FIG. 4 is a flowchart of model training based on training data to obtain a trained ARMA model in another embodiment;
FIG. 5 is a flowchart of a method for grid load testing in another embodiment;
FIG. 6 is a block diagram of a device for testing load on a power grid in an embodiment;
FIG. 7 is an internal block diagram of a computer device in one embodiment;
fig. 8 is a schematic diagram of an application method of machine learning-based frequency prediction in grid load test in an embodiment.
Detailed Description
In order to facilitate an understanding of the present application, a more complete description of the present application will now be provided with reference to the relevant figures. Examples of the present application are given in the accompanying drawings. This application may, however, be embodied in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein in the description of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.
In one embodiment, the application provides a power grid on-load test method which is suitable for on-load test of relay protection equipment in a transformer substation. As shown in fig. 1, the method includes:
step S100: and acquiring phase voltage data obtained by continuously sampling each sampling point of the power grid.
Specifically, sampling points can be set in different positions of the power grid in advance, and phase voltage data in a period of time of continuously collecting the sampling points is used for subsequent model training. In one embodiment, step S100 includes: and continuously sampling each sampling point of the power grid by using the high-stability crystal oscillator to acquire phase voltage within a set time length to obtain phase voltage data. The specific values of the frequency for data acquisition of the sampling points and the set time length are not unique, and can be adjusted according to actual requirements. The continuous sampling can be carried out at 4000Hz, absolute time calibration is carried out on each sampling point, and phase voltage data which are continuously measured for 5 minutes are stored. By adopting the high-stability crystal oscillator, the uniformity of the sampling pulse for on-load test is ensured, thereby meeting the requirement of synchronous sampling.
Step S200: and performing model training according to the phase voltage data to obtain a frequency prediction model.
Correspondingly, the collected phase voltage data can be divided into a training sample and a verification sample, which are respectively used for model training and verification, so as to obtain a frequency prediction model. The specific type of frequency prediction model is not unique, and as the measurement object is a single variable, ARMA (Autoregressive Moving Average, autoregressive moving average model) can be used to implement the prediction of the grid frequency.
Specifically, in one embodiment, the frequency prediction model is an ARMA model; as shown in fig. 2, step S200 includes step S220, step S240, and step S260.
Step S220: the phase voltage data is separated into training data and test data according to the sampling time. Also taking phase voltage data measured continuously for 5 minutes as an example, data sampled for the first 4 minutes can be used as training data for training a machine learning model, and data sampled for the last 1 minute can be used as test data for testing the model prediction result.
Step S240: model training is carried out according to the training data, and a trained ARMA model is obtained. Specifically, an infrastructure of the ARMA model can be built firstly, and model parameters of the ARMA model are calculated through training by utilizing training data until output data of the model meets requirements, so that the trained ARMA model is obtained.
Step S260: and analyzing the trained ARMA model according to the test data, and detecting whether the ARMA model meets the measurement requirements. After the ARMA model is trained, frequency prediction is carried out by using the test data, the prediction result is judged, and when the result meets the measurement requirement, the ARMA model is considered to be trained; if the result does not meet the measurement requirement, returning to the step S100, and collecting the data again to perform model training. The method of detecting whether the ARMA model meets the measurement requirement is not unique, in this embodiment, after the frequency prediction is performed by using the test data to obtain the prediction result of the ARMA model, the prediction result is compared with the actual collected data, and when the difference between the prediction result and the actual collected data is smaller than the set error, the ARMA model can be considered to meet the measurement requirement, and the ARMA model training is completed.
Further, the specific manner in which model training is performed based on training data is also not unique, and in one embodiment, as shown in fig. 3, step S240 includes steps S241 to S246.
Step S241: and calculating the frequency value of each cycle according to the rated frequency to the training data to obtain a frequency sequence. Specifically, the frequency value of each cycle is calculated for the training data at a nominal frequency of 50Hz, thereby obtaining a continuous frequency sequence.
Step S244: ARMA model parameters are determined from the frequency sequence. The values of parameters p and q in the ARMA model are determined by analysis by decomposing the data of the frequency sequence into three parts, namely a period quantity, a trend quantity and a residual error. In one embodiment, step S244 includes: and (3) carrying out autocorrelation diagram and partial autocorrelation diagram analysis on the frequency sequence to obtain the orders p and q as ARMA model parameters.
Step S245: training is carried out according to ARMA model parameters and the frequency sequence, and training result data is obtained. Substituting the values of the parameters p and q into an ARMA model and training by using the values of the frequency sequence to obtain training result data.
Step S246: and calculating the variance of the training result data and the frequency sequence, and judging whether the variance meets the preset condition. If yes, obtaining a trained ARMA model; if not, return to step S244. Specifically, after the variance between the training result data and the original frequency sequence is calculated, if the variance is smaller than a set threshold, the method can be considered to meet the preset condition.
Further, in one embodiment, as shown in fig. 4, after step S241, before step S244, step S240 further includes step S242: and carrying out unit root test on the frequency sequence to judge whether the frequency sequence is stable. If yes, go to step S244; if not, then step S243 is performed: and outputting prompt information of unstable power grid frequency.
The frequency sequence is also checked for ADF (root-by-root) before model training with the frequency sequence to determine if the current frequency sequence is stationary. For an unstable sequence, the user is directly prompted that the current system frequency is unstable, and offline measurement cannot be performed. For the stationary sequence, step S244 is performed to perform model training, so as to ensure accuracy of model training.
Continuous sampling at 4000Hz is continued for 5 minutes of phase voltage data, the first 4 minutes of sampling data being used for training of the machine learning model, and the last 1 minute of sampling data being used for verification of the model prediction results. Specifically, the ARMA model training comprises three parts, wherein the first part calculates the frequency value of each cycle according to the rated frequency of the system of 50Hz for sampling data, so as to obtain a continuous frequency sequence consisting of 12000 points, and the second part calculates the ADF value of the frequency sequence, determines whether the current frequency sequence is stable or not, and directly prompts a user that the current system frequency is unstable for an unstable sequence, so that offline measurement cannot be performed.
The unit root test is to test whether there is a unit root in the sequence, and the existence of the unit root is the non-stationary time sequence. The unit root check is a problem of the random process. Defining a random sequence { x } t T=1, 2, … is a unit root process, if x t ＝ρx t -1+ε, t=1, 2 …, where |ρ|<1, { ε } is a stationary sequence (white noise), and ε]＝0,V(ε)＝σ<∞,Cov(ε,ε)＝μ<And infinity, where t=1, 2 …. In particular, if ρ=1, the above formula becomes a random walk sequence, and thus the random walk sequence is a simplest unit root process. The definition is rewritten as follows: (1- ρL) x t =ε, t=1, 2, …. Where L is a hysteresis operator and 1- ρl is a hysteresis operator polynomial, characterized by the equation 1- ρz=0, rooted z=1/ρ. When ρ=1, there is one unit root in the time series, at which { x } t And is a unity root process. When ρ is<1, { x t And is a plateau sequence. And when ρ > 1, { x t The term "differential" refers to a type of non-stationary process with a so-called explosive root, which is still a non-stationary process after differentiation, and thus is not a single whole process. In general, the single overall process may be referred to as a unit root process.
Unit root test Unit root research of time series is a hotspot problem of time series analysis. The time-varying behavior of the moment characteristics of the time series actually reflects the non-stationary nature of the time series. The non-stationary time series is generally processed by converting it into a stationary sequence, so that a method concerning the stationary time series can be applied for corresponding research. The test of the time sequence unit root is the test of the stability of the time sequence, if the non-stable time sequence has the unit root, the unit root can be eliminated by a differential method to obtain the stable sequence. For time series where a unit root exists, obvious memorability and persistence of fluctuation are generally displayed, so the unit root test is the basis for discussion about the presence test of the synergistic relationship and persistence of the fluctuation of the sequence.
Further, the ARMA model is obtained by combining an AR (Auto Regression) model with a MA (Moving Average) model, and the Auto Regression Moving Average model ARMA (p, q) is:
X t ＝α 1 X t-1 +α 2 X t-2 +...+α p X t-p +ε t +β 1 ε t-1 +...+β q ε t-q
the formula shows that: a random time sequence may be represented by an autoregressive moving average model, i.e. the sequence may be interpreted by its own past or hysteresis values and random disturbance terms. If the sequence is stationary, i.e. its behaviour does not change over time, we can predict the future from the past behaviour of the sequence.
After obtaining the frequency sequence, ADF test is carried out to observe whether the sequence is stable or not; for a stable frequency sequence, the sequence data is decomposed into three parts of a period quantity, a trend quantity and a residual error, the autocorrelation coefficient and the Partial autocorrelation coefficient of the stable time sequence are respectively obtained, an ACF (Auto-Correlation Function, autocorrelation function)/PACF (Partial Auto-Correlation Function, partial autocorrelation function) graph is drawn, and the optimal orders p and q are obtained through analysis of the autocorrelation graph and the Partial autocorrelation graph. And substituting the p and q values into an ARMA model, training by using the sequence values, and solving the variance between the training result and the original sequence. And after the variance value meets the requirement, carrying out frequency prediction on the data of the next 1 minute by adopting a trained model, judging a predicted result, and when the result meets the measurement requirement, considering that ARMA model training is completed, otherwise, repeatedly carrying out data sampling and model training until a usable model is obtained.
Step S300: and acquiring a phase voltage phase value and a phase current at an asynchronous phase current sampling moment from the target sampling point. And taking the power grid node needing to be measured as a target sampling point, and performing asynchronous sampling of the phase voltage and the phase current. After recording the phase voltage phase value, the phase current can be collected separately at other moments and the difference between the phase current and the phase voltage point can be guaranteed to be an integral multiple of 80.
Step S400: and calculating according to the phase voltage phase value and the frequency prediction model to obtain the phase voltage frequency value of the target sampling point at the sampling moment of the asynchronous phase current. After the phase voltage and the phase current are acquired asynchronously, the phase voltage frequency value at the time of the acquisition of the asynchronous phase current is predicted by using a trained model according to the phase voltage phase value.
Step S500: and calculating to obtain the load angle of the target sampling point at the sampling moment of the asynchronous phase current according to the phase voltage frequency value and the phase current at the sampling moment of the asynchronous phase current. And calculating according to the phase current at the sampling time of the asynchronous phase current and the phase voltage frequency value at the predicted sampling time of the asynchronous phase current to obtain the phase of the phase current, and then subtracting the phase current phase from the phase voltage phase to obtain the load angle and recording.
Furthermore, in one embodiment, as shown in fig. 5, after step S500, the method further includes step S600: after the phase voltage frequency measured value of the target sampling point at the asynchronous phase current sampling moment is obtained, verifying the frequency prediction model according to the phase voltage frequency measured value, and outputting model failure prompt information when verification is not passed.
After completion of the power angle measurement record, the phase voltage frequency measurement may be performed again to verify the accuracy of the frequency model. Specifically, the phase voltage frequency measured value at the sampling moment of the actually collected asynchronous phase current can be calculated, the difference value between the actually collected phase voltage frequency measured value and the phase voltage frequency value predicted by the frequency prediction model is judged, whether the difference value is larger than a set error threshold value or not is judged, if yes, the verification is passed, and the model is qualified. And when the phase voltage and the phase current are verified to be qualified, confirming that the results of calculating the load angle by using the phase voltage and the phase current measured at different moments meet the requirements, otherwise prompting a user that the phase voltage frequency is changed greatly to cause the failure of the prediction model, and invalidating the measurement result.
According to the power grid on-load test method, the load angle is calculated by using the phase voltage and the phase current measured at different moments through the phase voltage frequency prediction, the test steps are simplified, the wiring work of the voltage test line is not required to be repeated in the on-load test process, and the work efficiency is greatly improved.
In one embodiment, there is also provided a power grid on-load testing apparatus, as shown in fig. 6, including a data acquisition module 100, a model training module 200, a parameter acquisition module 300, a frequency calculation module 400, and a load calculation module 500.
The data acquisition module 100 is used for acquiring phase voltage data obtained by continuously sampling each sampling point of the power grid; the model training module 200 is used for performing model training according to the phase voltage data to obtain a frequency prediction model; the parameter acquisition module 300 is used for acquiring a phase voltage phase value and a phase current at an asynchronous phase current sampling moment from a target sampling point; the frequency calculation module 400 is configured to calculate a phase voltage frequency value of the target sampling point at the asynchronous phase current sampling time according to the phase voltage phase value and the frequency prediction model; the load calculation module 500 is configured to calculate, according to the phase voltage frequency value and the phase current at the asynchronous phase current sampling time, a load angle of the target sampling point at the asynchronous phase current sampling time.
In one embodiment, the data acquisition module 100 uses the high-stability crystal oscillator to continuously sample each sampling point of the power grid, and acquires the phase voltage within a set time period to obtain the phase voltage data.
In one embodiment, model training module 200 separates phase voltage data into training data and verification data based on sampling time; model training is carried out according to the training data, and a trained ARMA model is obtained; and analyzing the trained ARMA model according to the test data, and detecting whether the ARMA model meets the measurement requirements.
Further, the model training module 200 calculates the frequency value of each cycle according to the rated frequency to the training data to obtain a frequency sequence; determining ARMA model parameters according to the frequency sequence; training according to ARMA model parameters and the frequency sequence to obtain training result data; and calculating the variance of the training result data and the frequency sequence, and judging whether the variance meets the preset condition. If yes, obtaining a trained ARMA model; if not, determining ARMA model parameters again according to the frequency sequence.
In addition, in one embodiment, the model training module 200 calculates the frequency value of each cycle according to the rated frequency for the training data, and after obtaining the frequency sequence, performs a unit root test for the frequency sequence to determine whether the frequency sequence is stable. If yes, determining ARMA model parameters according to the frequency sequence; if not, outputting prompt information of unstable power grid frequency.
In addition, in one embodiment, the load calculation module 500 is further configured to verify the frequency prediction model according to the measured value of the phase voltage frequency after obtaining the measured value of the phase voltage frequency of the target sampling point at the sampling time of the asynchronous phase current, and output the model failure prompt information when the verification is not passed.
According to the power grid on-load testing device, the load angle is calculated by using the phase voltage and the phase current measured at different moments through the phase voltage frequency prediction, the testing steps are simplified, the wiring work of the voltage testing line is not required to be repeated in the on-load testing process, and the working efficiency is greatly improved.
In one embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 7. The computer device includes a processor, a memory, and a network interface connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a method for grid load testing.
It will be appreciated by those skilled in the art that the structure shown in fig. 7 is merely a block diagram of some of the structures associated with the present application and is not limiting of the computer device to which the present application may be applied, and that a particular computer device may include more or fewer components than shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided comprising a memory storing a computer program and a processor that when executing the computer program performs the steps of: acquiring phase voltage data obtained by continuously sampling each sampling point of a power grid; model training is carried out according to the phase voltage data, and a frequency prediction model is obtained; acquiring a phase voltage phase value and a phase current at an asynchronous phase current sampling moment from a target sampling point; calculating according to the phase voltage phase value and the frequency prediction model to obtain a phase voltage frequency value of a target sampling point at an asynchronous phase current sampling moment; and calculating to obtain the load angle of the target sampling point at the sampling moment of the asynchronous phase current according to the phase voltage frequency value and the phase current at the sampling moment of the asynchronous phase current.
In one embodiment, the processor when executing the computer program further performs the steps of: and continuously sampling each sampling point of the power grid by using the high-stability crystal oscillator to acquire phase voltage within a set time length to obtain phase voltage data.
In one embodiment, the processor when executing the computer program further performs the steps of: dividing the phase voltage data into training data and test data according to the sampling time; model training is carried out according to the training data, and a trained ARMA model is obtained; and analyzing the trained ARMA model according to the test data, and detecting whether the ARMA model meets the measurement requirements.
In one embodiment, the processor when executing the computer program further performs the steps of: calculating the frequency value of each cycle according to the rated frequency to the training data to obtain a frequency sequence; determining ARMA model parameters according to the frequency sequence; training according to ARMA model parameters and the frequency sequence to obtain training result data; and calculating the variance of the training result data and the frequency sequence, and judging whether the variance meets the preset condition. If yes, obtaining a trained ARMA model; if not, determining ARMA model parameters again according to the frequency sequence.
In one embodiment, the processor when executing the computer program further performs the steps of: and carrying out unit root test on the frequency sequence to judge whether the frequency sequence is stable. If yes, determining ARMA model parameters according to the frequency sequence; if not, outputting prompt information of unstable power grid frequency.
In one embodiment, the processor when executing the computer program further performs the steps of: after the phase voltage frequency measured value of the target sampling point at the asynchronous phase current sampling moment is obtained, verifying the frequency prediction model according to the phase voltage frequency measured value, and outputting model failure prompt information when verification is not passed.
In one embodiment, there is also provided a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of: acquiring phase voltage data obtained by continuously sampling each sampling point of a power grid; model training is carried out according to the phase voltage data, and a frequency prediction model is obtained; acquiring a phase voltage phase value and a phase current at an asynchronous phase current sampling moment from a target sampling point; calculating according to the phase voltage phase value and the frequency prediction model to obtain a phase voltage frequency value of a target sampling point at an asynchronous phase current sampling moment; and calculating to obtain the load angle of the target sampling point at the sampling moment of the asynchronous phase current according to the phase voltage frequency value and the phase current at the sampling moment of the asynchronous phase current.
In one embodiment, the computer program when executed by the processor further performs the steps of: and continuously sampling each sampling point of the power grid by using the high-stability crystal oscillator to acquire phase voltage within a set time length to obtain phase voltage data.
In one embodiment, the computer program when executed by the processor further performs the steps of: dividing the phase voltage data into training data and test data according to the sampling time; model training is carried out according to the training data, and a trained ARMA model is obtained; and analyzing the trained ARMA model according to the test data, and detecting whether the ARMA model meets the measurement requirements.
In one embodiment, the computer program when executed by the processor further performs the steps of: calculating the frequency value of each cycle according to the rated frequency to the training data to obtain a frequency sequence; determining ARMA model parameters according to the frequency sequence; training according to ARMA model parameters and the frequency sequence to obtain training result data; and calculating the variance of the training result data and the frequency sequence, and judging whether the variance meets the preset condition. If yes, obtaining a trained ARMA model; if not, determining ARMA model parameters again according to the frequency sequence.
In one embodiment, the computer program when executed by the processor further performs the steps of: and carrying out unit root test on the frequency sequence to judge whether the frequency sequence is stable. If yes, determining ARMA model parameters according to the frequency sequence; if not, outputting prompt information of unstable power grid frequency.
In one embodiment, the computer program when executed by the processor further performs the steps of: after the phase voltage frequency measured value of the target sampling point at the asynchronous phase current sampling moment is obtained, verifying the frequency prediction model according to the phase voltage frequency measured value, and outputting model failure prompt information when verification is not passed.
Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the various embodiments provided herein may include non-volatile and/or volatile memory. The nonvolatile memory can include Read Only Memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), memory bus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), among others.
In order to better understand the above power distribution network elasticity evaluation method, device, computer equipment and storage medium, the following detailed explanation is made with reference to specific embodiments.
In the prior art, an external current clamp and a voltage connecting wire are adopted, and a voltage testing wire is required to be connected when the current is measured so as to realize the calculation of the phase difference. The specific operation is as follows: connecting a voltage test line with a secondary voltage loop to be tested and a phase meter, and testing a reference voltage; then connecting the tested CT secondary circuit and the phase meter through the current clamp meter and the test wire, and synchronously testing the corresponding CT secondary circuit; the magnitude and angle of the current need to be recorded after each group of CT secondary loop measurement is completed.
In the on-load test work in the transformer substation, the secondary circuits of different windings and different phases in different intervals are required to be measured, so that the test process is required to repeatedly perform the work of wiring voltage test lines and wiring CT clamp meters, the time and the labor are wasted, and the work efficiency is low. Particularly when the CT secondary circuit and the reference voltage are not in the same screen body or the same position, difficulty is brought to test work, and the trouble is often caused to find a reference voltage, and in this case, the voltage circuit is connected through a long cable, so that certain potential safety hazard exists.
Based on the problems, the application provides a method for applying frequency prediction based on machine learning in power grid on-load test, and by adopting a high-stability crystal oscillator, the time-keeping deviation of a measurement system is ensured to be less than 10us within 2 hours, so that the requirement of synchronous sampling is met, and a foundation is laid for applying the measurement method.
As shown in fig. 8, the specific procedure is as follows:
1) Continuous sampling is carried out at 4000Hz, and absolute time calibration is carried out on each sampling point.
2) The phase voltages are measured continuously for 5 minutes, wherein the first 4 minutes of sampling data are used for training a machine learning model, the last 1 minute of sampling data are used for checking the model prediction result, and ARMA is adopted to realize the prediction of the power grid frequency because the measured object is a single variable.
3) The model training comprises three parts, wherein the first part calculates the frequency value of each cycle according to the rated frequency of the system of 50Hz to obtain a continuous frequency sequence consisting of 12000 points, and the second part calculates the ADF value of the sequence to determine whether the current frequency sequence is stable or not, and for an unstable sequence, the user is directly prompted that the current system frequency is unstable and offline measurement cannot be performed. The p, q values in the ARMA model were determined for the plateau sequences by decomposing the sequence data into three parts of period, trend and residual and plotting ACF/PACF. And substituting the p and q values into an ARMA model and training by using a sequence value, solving the variance between the training result and the original sequence, adopting the trained model to predict the frequency of the data in the next 1 minute after the variance value meets the requirement, judging the predicted result, and considering that the ARMA model is trained when the result meets the measurement requirement, otherwise, repeating the steps 2 and 3 until a usable model is obtained.
4) Recording phase voltage phase, independently collecting phase current at other moments and ensuring that the difference between the phase voltage phase and the phase voltage point is 80 times, predicting the phase voltage frequency value at the phase current collection moment by using a trained model so as to calculate the phase of the phase current, and then obtaining and recording a power angle.
5) After the power angle measurement record is completed, the phase voltage frequency measurement is needed to be performed again so as to verify the accuracy of the frequency model. And when the phase voltage and the phase current are verified to be qualified, confirming that the results of calculating the load angle by using the phase voltage and the phase current measured at different moments meet the requirements, otherwise prompting a user that the phase voltage frequency is changed greatly to cause the failure of the prediction model, and invalidating the measurement result.
The method for applying the frequency prediction based on machine learning in the on-load test of the power grid adopts the high-stability crystal to realize the step-by-step synchronous acquisition of the voltage and the current of the system; adopting an ARMA model to realize phase voltage frequency prediction and using a prediction result for current phase calculation at different moments; and (3) checking by adopting an ARMA model to realize the judgment of the correctness of the load angle result. The uniformity of self sampling pulse of the load test equipment is realized through a high-stability crystal oscillator and time tame algorithm, and the phase angle value of reference voltage at different time is predicted based on the uniformity of self sampling pulse of the load test equipment, so that the phase angle difference between current and reference voltage at any time can be measured, and the trouble that the reference voltage cannot be found under specific conditions is solved. The method simplifies the testing steps, does not need to repeatedly carry out the wiring work of the voltage testing line in the process of the on-load testing, and greatly improves the working efficiency.
The technical features of the above-described embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above-described embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples merely represent a few embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the invention. It should be noted that it would be apparent to those skilled in the art that various modifications and improvements could be made without departing from the spirit of the present application, which would be within the scope of the present application. Accordingly, the scope of protection of the present application is to be determined by the claims appended hereto.
Claims (10)
1. The utility model provides a power grid load test method which is characterized by comprising the following steps:
acquiring phase voltage data obtained by continuously sampling each sampling point of a power grid;
model training is carried out according to the phase voltage data, and a frequency prediction model is obtained;
acquiring a phase voltage phase value and a phase current at an asynchronous phase current sampling moment from a target sampling point;
calculating a phase voltage frequency value of the target sampling point at an asynchronous phase current sampling moment according to the phase voltage phase value and the frequency prediction model;
calculating a load angle of the target sampling point at the sampling moment of the asynchronous phase current according to the phase voltage frequency value and the phase current at the sampling moment of the asynchronous phase current;
wherein the frequency prediction model is an ARMA model; model training is carried out according to the phase voltage data to obtain a frequency prediction model, and the method comprises the following steps: dividing the phase voltage data into training data and test data according to the sampling time; model training is carried out according to the training data, and a trained ARMA model is obtained; analyzing the trained ARMA model according to the test data, and detecting whether the ARMA model meets the measurement requirement; if yes, the ARMA model is completed; if not, returning to the step of acquiring the phase voltage data obtained by continuously sampling each sampling point of the power grid;
the training of the model according to the training data to obtain a trained ARMA model comprises the following steps: calculating the frequency value of each cycle according to the rated frequency to the training data to obtain a frequency sequence; determining ARMA model parameters according to the frequency sequence; training according to the ARMA model parameters and the frequency sequence to obtain training result data; calculating the variance between training result data and the frequency sequence, and judging whether the variance meets a preset condition; if yes, obtaining a trained ARMA model; if not, returning to the step of determining ARMA model parameters according to the frequency sequence.
2. The method for testing the load of the power grid according to claim 1, wherein the step of obtaining phase voltage data obtained by continuously sampling each sampling point of the power grid comprises the steps of: and continuously sampling each sampling point of the power grid by using the high-stability crystal oscillator to acquire phase voltage within a set time length to obtain phase voltage data.
3. The grid on-load testing method according to claim 1, wherein said determining ARMA model parameters from said frequency sequence comprises: and carrying out autocorrelation diagram and partial autocorrelation diagram analysis on the frequency sequence to obtain an order serving as an ARMA model parameter.
4. The method for testing load on a power grid according to claim 1, wherein the calculating the frequency value of each cycle according to the rated frequency for the training data, after obtaining the frequency sequence, before determining the ARMA model parameters according to the frequency sequence, further comprises:
carrying out unit root test on the frequency sequence, and judging whether the frequency sequence is stable or not; if yes, the step of determining ARMA model parameters according to the frequency sequence is carried out; if not, outputting prompt information of unstable power grid frequency.
5. The method for testing load on a power grid according to any one of claims 1 to 4, wherein after calculating the load angle of the target sampling point at the asynchronous phase current sampling time according to the phase voltage frequency value and the phase current at the asynchronous phase current sampling time, the method further comprises:
after the phase voltage frequency measured value of the target sampling point at the asynchronous phase current sampling moment is obtained, verifying the frequency prediction model according to the phase voltage frequency measured value, and outputting model failure prompt information when verification is not passed.
6. A power grid load test device, comprising:
the data acquisition module is used for acquiring phase voltage data obtained by continuously sampling each sampling point of the power grid;
the model training module is used for carrying out model training according to the phase voltage data to obtain a frequency prediction model;
the parameter acquisition module is used for acquiring a phase voltage phase value and a phase current at an asynchronous phase current sampling moment from a target sampling point;
the frequency calculation module is used for calculating a phase voltage frequency value of the target sampling point at the asynchronous phase current sampling moment according to the phase voltage phase value and the frequency prediction model;
the load calculation module is used for calculating the load angle of the target sampling point at the sampling moment of the asynchronous phase current according to the phase voltage frequency value and the phase current at the sampling moment of the asynchronous phase current;
wherein the frequency prediction model is an ARMA model; the model training module is also used for dividing the phase voltage data into training data and checking data according to the sampling time; model training is carried out according to the training data, and a trained ARMA model is obtained; analyzing the trained ARMA model according to the test data, and detecting whether the ARMA model meets the measurement requirement;
the model training module is also used for calculating the frequency value of each cycle according to the rated frequency by training data to obtain a frequency sequence; determining ARMA model parameters according to the frequency sequence; training according to ARMA model parameters and the frequency sequence to obtain training result data; calculating the variance of the training result data and the frequency sequence, judging whether the variance meets preset conditions, and if so, obtaining a trained ARMA model; if not, determining ARMA model parameters again according to the frequency sequence.
7. The device of claim 6, wherein the data acquisition module is configured to acquire phase voltage data by continuously sampling each sampling point of the power grid with the high-stability crystal oscillator and acquiring phase voltage within a set period of time.
8. The apparatus of claim 6, wherein the model training module is further configured to calculate a frequency value of each cycle according to a rated frequency for training data, and further perform a unit root test for the frequency sequence after obtaining the frequency sequence, to determine whether the frequency sequence is stable, and if so, determine an ARMA model parameter according to the frequency sequence; if not, outputting prompt information of unstable power grid frequency.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any one of claims 1 to 5 when the computer program is executed.
10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011029461.3A CN112269059B (en) | 2020-09-27 | 2020-09-27 | Power grid load test method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011029461.3A CN112269059B (en) | 2020-09-27 | 2020-09-27 | Power grid load test method and device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112269059A CN112269059A (en) | 2021-01-26 |
| CN112269059B true CN112269059B (en) | 2023-06-06 |
Family
ID=74348855
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011029461.3A Active CN112269059B (en) | 2020-09-27 | 2020-09-27 | Power grid load test method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112269059B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113608041B (en) * | 2021-06-25 | 2023-12-29 | 深圳供电局有限公司 | On-load test method, device, equipment and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10257674A (en) * | 1997-03-07 | 1998-09-25 | Mitsubishi Electric Corp | Modeling method, prediction method, and system stabilization control method |
| CN102237691A (en) * | 2011-07-06 | 2011-11-09 | 东北大学 | Wind energy and solar energy grid-connected generation system and control method thereof |
| CN103604992A (en) * | 2013-11-28 | 2014-02-26 | 国家电网公司 | Wireless nuclear phase on-load correction protection method and system for secondary circuit of substation |
| CN110264001A (en) * | 2019-06-20 | 2019-09-20 | 国网上海市电力公司 | Electro-load forecast method based on multiple timings |
| CN110829487A (en) * | 2019-10-22 | 2020-02-21 | 电子科技大学 | Dynamic frequency prediction method for power system |
| CN111030137A (en) * | 2019-12-06 | 2020-04-17 | 中国电力科学研究院有限公司 | Power grid frequency prediction method and device with load participating in primary frequency modulation |
| CN111474477A (en) * | 2020-04-29 | 2020-07-31 | 西安工业大学 | Method for mapping process of asynchronous motor parameters and fault information |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6850426B2 (en) * | 2002-04-30 | 2005-02-01 | Honeywell International Inc. | Synchronous and bi-directional variable frequency power conversion systems |
-
2020
- 2020-09-27 CN CN202011029461.3A patent/CN112269059B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10257674A (en) * | 1997-03-07 | 1998-09-25 | Mitsubishi Electric Corp | Modeling method, prediction method, and system stabilization control method |
| CN102237691A (en) * | 2011-07-06 | 2011-11-09 | 东北大学 | Wind energy and solar energy grid-connected generation system and control method thereof |
| CN103604992A (en) * | 2013-11-28 | 2014-02-26 | 国家电网公司 | Wireless nuclear phase on-load correction protection method and system for secondary circuit of substation |
| CN110264001A (en) * | 2019-06-20 | 2019-09-20 | 国网上海市电力公司 | Electro-load forecast method based on multiple timings |
| CN110829487A (en) * | 2019-10-22 | 2020-02-21 | 电子科技大学 | Dynamic frequency prediction method for power system |
| CN111030137A (en) * | 2019-12-06 | 2020-04-17 | 中国电力科学研究院有限公司 | Power grid frequency prediction method and device with load participating in primary frequency modulation |
| CN111474477A (en) * | 2020-04-29 | 2020-07-31 | 西安工业大学 | Method for mapping process of asynchronous motor parameters and fault information |
Non-Patent Citations (1)
| Title |
|---|
| 基于ARMA模型的晶振时钟频率预测方法;马少林 等;《电气技术》;20131231(第3期);第5-8页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112269059A (en) | 2021-01-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3477404B1 (en) | Power system model parameter conditioning tool | |
| CN101907868B (en) | Intelligent trouble diagnosis method for tractive power supply system and system thereof | |
| CN111313405B (en) | Medium-voltage distribution network topology identification method based on multiple measurement sections | |
| AT513189A2 (en) | Method for determining a control-technical observer for the SoC | |
| CN108830456B (en) | Sensitive equipment and power grid voltage sag compatibility analysis method and device | |
| CN112269059B (en) | Power grid load test method and device, computer equipment and storage medium | |
| US11029344B2 (en) | Sensitivity based Thevenin index for voltage stability assessment considering N-1 contingency | |
| CN111458661A (en) | Power distribution network line variation relation diagnosis method, device and system | |
| CN110554347B (en) | Current transformer transformation ratio detection method and power detection method | |
| CN111913039B (en) | Background harmonic voltage calculation method and device of alternating current filter | |
| CN109375145B (en) | Voltage loss fault monitoring method and device based on three-phase three-wire meter equipment | |
| EP3531526A1 (en) | Estimation of electrical grid parameters | |
| Dambrowski | Validation of impedance-data and of impedance-based modeling approach of electrochemical cells by means of mathematical system theory | |
| CN112946421B (en) | Three-phase power grid fault diagnosis method and device, computer equipment and storage medium | |
| CN109375144B (en) | Current loss fault monitoring method and device based on three-phase four-wire meter equipment | |
| CN112952838B (en) | Intelligent tide analysis method and device based on eye movement equipment | |
| CN109193563B (en) | Current loss fault monitoring method and device based on three-phase three-wire meter equipment | |
| CN110275076B (en) | Method and device for testing immunity of equipment voltage sag, computer equipment and medium | |
| CN109212451B (en) | Voltage loss fault monitoring method and device based on three-phase four-wire meter measuring equipment | |
| Hilbrich et al. | Optimized grid reduction for state estimation algorithms in under-determined distribution grids | |
| KR102226677B1 (en) | Apparatus for measuring harmonic impedance of electric power system and method for the same | |
| Lyamets et al. | Description of 6–35 kV lines in recognition of earth fault in a distribution network | |
| Eisenbarth et al. | A data-driven method for estimating parameter uncertainty in PMU-based Power Plant Model Validation | |
| CN111274687A (en) | Component failure rate prediction method and device, computer equipment and storage medium | |
| Clarkson et al. | The calibration of IEC standard flickermeters using complex modulated signals |
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 |