Power system node characteristic data deviation correction method

Abstract

The invention discloses an electric power system node characteristic data deviation correction method, which comprises the following steps of: collecting characteristic data related to an electric power system, generally installing related equipment in the electric power system so as to monitor, collect and upload the characteristic data of each node of the electric power system in real time, and simultaneously obtaining historical data; and carrying out trend analysis and comparison, preprocessing the collected feature data, and carrying out data cleaning, data enhancement and data normalization on the collected feature data. According to the method, the node data of the electric power system are collected, and the preprocessed node data of the electric power system are subjected to feature selection, reasonable deviation threshold setting, resampling, suspicious data point identification and sensitivity analysis to identify the deviated node data, so that the correction effect of the deviation of the node data is improved; deviation data in a power system can be identified conveniently, and the situation that deviation node data are missed is prevented.

Description

Technical Field

[0001] The present invention relates to the field of power system node characteristic data deviation correction, and in particular to a power system node characteristic data deviation correction method. Background Art

[0002] Power system node characteristic data usually refers to a series of data sets that describe the characteristics of each node in the power system. Power system node characteristic data is usually of great significance for the operation, planning, optimization and fault analysis of the power system. Power system node characteristic data includes node voltage, node injection power, node load and node unit price.

[0003] During the use of the power system, there will usually be deviations in some node data. Deviations in node data not only affect the stable operation of the power system and the safe use of equipment, but may also lead to decreased production efficiency, reduced equipment performance and increased safety hazards, which requires corrective operations on the node data of the power system.

[0004] The existing method for correcting deviations in characteristic data of power system nodes usually collects data and then corrects the data. However, this method may miss some deviation data, which affects the correction effect. Therefore, it is necessary to identify data deviations in power system node data. Summary of the invention

[0005] The purpose of the present invention is to provide a method for correcting deviations of characteristic data of power system nodes to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: a method for correcting deviation of characteristic data of power system nodes, comprising the following specific steps:

[0007] Step 1: Collect the characteristic data related to the power system. Usually, relevant equipment is installed in the power system to monitor and upload the characteristic data of each node of the power system in real time, obtain historical data, and perform trend analysis and comparison;

[0008] Step 2: Preprocess the collected feature data to perform data cleaning, data enhancement and data normalization;

[0009] Step 3: Perform deviation identification on the preprocessed data, first perform feature selection on the preprocessed node data, then set the deviation threshold based on historical data and system characteristics, then resample the preprocessed data to improve the quality of the node data, then identify suspicious data points on the node data, and then perform sensitivity analysis on the node data to identify deviation data;

[0010] Step 4: Correct the deviation of the identified deviation data, and use the deviation correction algorithm to correct the deviation data that is not easy to directly identify;

[0011] Step 5: Verify and evaluate the corrected data by comparing the corrected node data with the actual measured data.

[0012] Preferably, the power system related sensing equipment in step one is a voltage sensor, a current sensor, a temperature sensor and a humidity sensor, the power system related monitoring equipment in step one is a video surveillance camera, an infrared thermal imager and a data acquisition terminal, and the power system characteristic data in step one includes voltage, current and power.

[0013] Preferably, the data cleaning in step 2 is to remove duplicate data, fill missing values ​​and process outliers in the collected node data, so that the quality of the collected node data is improved, and the data normalization in step 2 is to convert data of different dimensions into the same dimension.

[0014] Preferably, the feature selection operation in step three is to screen features by calculating the correlation between each feature and the label to obtain corresponding data features.

[0015] Preferably, the resampling in step three is performed by nearest neighbor interpolation, bilinear interpolation and cubic convolution interpolation.

[0016] Preferably, the resampling operation steps in step three are to first select and determine the resampling method according to the characteristics and requirements of the data, then use the selected resampling method to process the data, and then improve the node data quality by comparing the quality of the data set before and after resampling.

[0017] Preferably, the suspicious data point identification operation method in step three is to compare real-time data with historical data, and use the statistical method of standard deviation and the machine learning algorithm of BP neural network to perform calculations, and then identify suspicious deviation data points, and then further analyze the identified suspicious data points to confirm that they are true deviation data.

[0018] Preferably, the operating steps of the sensitivity analysis in step three are to first select uncertain factors, and according to the characteristics of the power system node data and the needs of deviation correction, select the uncertain factors that have a greater impact on the data deviation for analysis, and then determine the range of change of the node data, set a reasonable range of change for each uncertain factor, observe the impact of its change on the data deviation, and then implement sensitivity analysis by simulating the changes of uncertain factors and observing the changes in data deviation in real time. Sensitivity analysis tables or sensitivity analysis diagrams can be used to record and display the analysis results to further identify the deviation data.

[0019] Preferably, the deviation identification of the preprocessed data in step 3 can be performed using a node power balance formula. By using the principle of node power balance, it can be determined whether there is a data deviation in the power system. The calculation method of the node power balance formula is as follows: assuming that there are n branches on the bus, the active power and reactive power of the i-th branch are Pi and Qi respectively, then the expression of the node power balance is:

[0020]

[0021] By judging whether the power boxes of all branches on the node are zero, if they are not zero, there may be data deviation.

[0022] Preferably, the deviation correction algorithm in step 4 includes linear deviation, and the linear deviation is corrected in such a way that the actual measured value of a certain characteristic data in the node data of the power system is V meas , and the expected voltage value of the node is V des , and then a correction value DV needs to be calculated so that the adjusted voltage value V adj Close to the expected voltage value V des , the linear deviation correction formula can be:

[0023] DV=κ·(V des -V meas

[0024] Among them, κ is the deviation correction coefficient, which determines the strength and direction of the correction. The value of κ can be adjusted according to the actual situation;

[0025] Adjusted voltage value V adj It can be calculated by the following formula:

[0026] V adj =V meas +ΔV=V meas +κ(V des -V meas ).

[0027] Technical effects and advantages of the present invention:

[0028] The present invention utilizes the collection of power system node data and performs preprocessing operations on the collected power system node data, so that the preprocessed power system node data is successively identified through feature selection, setting a reasonable deviation threshold, resampling, suspicious data point identification and sensitivity analysis on the deviated node data, which is beneficial to improve the correction effect of node data deviation and facilitates the identification of deviation data in the power system to prevent the omission of deviation node data. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a schematic diagram of the process of correcting the deviation of characteristic data of power system nodes according to the present invention.

[0030] Figure 2 The figure is a schematic diagram of the preprocessing process of the collected feature data according to the present invention.

[0031] Figure 3 This is a schematic diagram of the deviation identification process for preprocessed data in the present invention.

[0032] Figure 4 This is a schematic diagram of the sensitivity analysis process of data in the present invention. DETAILED DESCRIPTION

[0033] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] Embodiment 1

[0035] The present invention provides Figure 1-4 A method for correcting deviation of characteristic data of power system nodes shown in the figure comprises the following specific steps:

[0036] Step 1: Collect characteristic data related to the power system. Usually, relevant equipment is installed in the power system to monitor the characteristic data of each node of the power system in real time and collect and upload it. At the same time, historical data is obtained and trend analysis and comparison are performed. The sensing equipment related to the power system is voltage sensor, current sensor, temperature sensor and humidity sensor. The monitoring equipment related to the power system is video surveillance camera, infrared thermal imager and data acquisition terminal. The characteristic data of the power system includes voltage, current and power.

[0037] Step 2: Preprocess the collected feature data to perform data cleaning, data enhancement and data normalization. Data cleaning is to remove duplicate data, fill missing values ​​and process outliers in the collected node data to improve the quality of the collected node data. Data normalization is to convert data of different dimensions into the same dimension. The operation steps of data cleaning are to filter and remove duplicate data by comparing the key fields of the data records. There may be missing values ​​in the power node data, that is, the values ​​of some fields are empty. Missing values ​​will affect the results of data analysis, so they need to be processed. Common processing methods include deleting the records where the missing values ​​are located, Use the mean or median to fill missing values, etc. At the same time, there may be outliers in the power node data, that is, values ​​that are obviously inconsistent with other data or exceed the normal range. Outliers will interfere with data analysis and mining, so statistical methods are needed to identify and process outliers. In order to facilitate subsequent data analysis and mining, the data needs to be converted into the correct data type. Programming languages ​​or data processing tools can be used to perform data conversion operations. After data cleaning is completed, the cleaned data needs to be stored for subsequent data analysis and mining. Databases, data warehouses or other data storage methods can be used, and appropriate storage methods can be selected according to needs.

[0038] Step 3: Perform deviation identification on the preprocessed data. First, perform feature selection on the preprocessed node data. Then, based on historical data and system characteristics, the feature selection operation method is to screen the features by calculating the correlation between each feature and the label to obtain the corresponding data features, set the deviation threshold, and then resample the preprocessed data. The resampling operation is performed through nearest neighbor interpolation, bilinear interpolation and cubic convolution interpolation. The resampling operation steps are to first select and determine the resampling method according to the characteristics and needs of the data, and then use the selected resampling method to process the data. Then, by comparing the quality of the data set before and after resampling, the node data quality is improved. In this way, the node data quality is improved, and then suspicious data points are identified for the node data. The suspicious data point identification operation method is to compare the real-time data with the historical data, and use the statistical method of the standard deviation and the machine learning algorithm of the BP neural network for calculation, and then identify the suspicious deviation data points, and then further analyze the identified suspicious data points to confirm their It is the real deviation data, and then the sensitivity analysis of the node data is carried out. The operation steps of the sensitivity analysis are as follows: first, the uncertain factors are selected. According to the characteristics of the power system node data and the needs of deviation correction, the uncertain factors that have a greater impact on the data deviation are selected for analysis, and then the range of change of the node data is determined. A reasonable range of change is set for each uncertain factor, and the impact of its change on the data deviation is observed. Then, the sensitivity analysis is implemented. By simulating the changes of uncertain factors and observing the changes in data deviation in real time, the sensitivity analysis table or sensitivity analysis chart can be used to record and display the analysis results to further identify the deviation data, and then identify the deviation data. The node power balance formula can be used to identify the deviation of the preprocessed data. Using the principle of node power balance, it can be determined whether there is data deviation in the power system. The calculation method of the node power balance formula is to assume that there are n branches on the bus, and the active power and reactive power of the i-th branch are Pi and Qi respectively. Then the expression of node power balance is:

[0039]

[0040] By judging whether the power boxes of all branches on the node are zero, if they are not zero, there may be data deviation.

[0041] Step 4: Correct the deviation of the identified deviation data. Use the deviation correction algorithm to correct the deviation data that is not easy to directly identify. The deviation correction algorithm includes linear deviation. The correction method of linear deviation is that in the node data of the power system, the actual measurement value of a certain characteristic data is V meas , and the expected voltage value of the node is V des , and then a correction value DV needs to be calculated so that the adjusted voltage value V adjClose to the expected voltage value V des , the linear deviation correction formula can be:

[0042] DV=κ·(V des -V meas )

[0043] Among them, κ is the deviation correction coefficient, which determines the strength and direction of the correction. The value of κ can be adjusted according to the actual situation;

[0044] Adjusted voltage value V adj It can be calculated by the following formula:

[0045] V adj =V meas +DV=V meas +κV des -V meas ).

[0046] Step 5: Verify and evaluate the corrected data by comparing the corrected node data with the actual measured data.

[0047] Embodiment 2

[0048] Step 1: Collect characteristic data related to the power system. Usually, relevant equipment is installed in the power system to monitor the characteristic data of each node of the power system in real time and collect and upload it. At the same time, historical data is obtained and trend analysis and comparison are performed. The sensing equipment related to the power system is voltage sensor, current sensor, temperature sensor and humidity sensor. The monitoring equipment related to the power system is video surveillance camera, infrared thermal imager and data acquisition terminal. The characteristic data of the power system includes voltage, current and power.

[0049] Step 2: Preprocess the collected feature data to perform data cleaning, data enhancement and data normalization. Data cleaning is to remove duplicate data, fill missing values ​​and process outliers in the collected node data to improve the quality of the collected node data. Data normalization is to convert data of different dimensions into the same dimension.

[0050] Step 3: Perform deviation identification on the preprocessed data, so that the preprocessed node data is first subjected to feature selection, and then based on historical data and system characteristics, the feature selection operation is to filter features by calculating the correlation between each feature and the label to obtain the corresponding data features.

[0051] Step 4: Correct the identified deviation data and use the deviation correction algorithm to correct the deviation data that is not easy to identify directly.

[0052] Step 5: Verify and evaluate the corrected data by comparing the corrected node data with the actual measured data.

[0053] Compared with the first embodiment, in the present embodiment, only feature selection is used to identify deviation data in the process of identifying deviation of pre-processed data.

[0054] Implementation Three

[0055] Step 1: Collect characteristic data related to the power system. Usually, relevant equipment is installed in the power system to monitor the characteristic data of each node of the power system in real time and collect and upload it. At the same time, historical data is obtained and trend analysis and comparison are performed. The sensing equipment related to the power system is voltage sensor, current sensor, temperature sensor and humidity sensor. The monitoring equipment related to the power system is video surveillance camera, infrared thermal imager and data acquisition terminal. The characteristic data of the power system includes voltage, current and power.

[0056] Step 2: Preprocess the collected feature data to perform data cleaning, data enhancement and data normalization. Data cleaning is to remove duplicate data, fill missing values ​​and process outliers in the collected node data to improve the quality of the collected node data. Data normalization is to convert data of different dimensions into the same dimension.

[0057] Step 3: Perform deviation identification on the preprocessed data, so that the preprocessed node data is first subjected to feature selection, and then based on historical data and system characteristics, the feature selection operation method is to filter features by calculating the correlation between each feature and the label to obtain the corresponding data features and set the deviation threshold.

[0058] Step 4: Correct the identified deviation data and use the deviation correction algorithm to correct the deviation data that is not easy to identify directly.

[0059] Step 5: Verify and evaluate the corrected data by comparing the corrected node data with the actual measured data.

[0060] Compared with the first embodiment, in the process of identifying the deviation of the pre-processed data, this embodiment only adopts the method of feature selection and setting the deviation threshold to identify the deviation data.

[0061] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for correcting deviation of characteristic data of power system nodes, characterized in that: The specific steps include: Step 1: Collect the characteristic data related to the power system. Usually, relevant equipment is installed in the power system to monitor and upload the characteristic data of each node of the power system in real time, obtain historical data, and perform trend analysis and comparison; Step 2: Preprocess the collected feature data to perform data cleaning, data enhancement and data normalization; Step 3: Perform deviation identification on the preprocessed data, first perform feature selection on the preprocessed node data, then set the deviation threshold based on historical data and system characteristics, then resample the preprocessed data to improve the quality of the node data, then identify suspicious data points on the node data, and then perform sensitivity analysis on the node data to identify deviation data; Step 4: Correct the deviation of the identified deviation data, and use the deviation correction algorithm to correct the deviation data that is not easy to directly identify; Step 5: Verify and evaluate the corrected data by comparing the corrected node data with the actual measured data.

2. A method for correcting deviation of characteristic data of power system nodes according to claim 1, characterized in that: The power system related sensing equipment in step one is a voltage sensor, a current sensor, a temperature sensor and a humidity sensor, the power system related monitoring equipment in step one is a video surveillance camera, an infrared thermal imager and a data acquisition terminal, and the power system characteristic data in step one includes voltage, current and power.

3. A method for correcting deviation of characteristic data of power system nodes according to claim 1, characterized in that: The data cleaning in step 2 is to remove duplicate data, fill missing values ​​and process outliers in the collected node data, so that the quality of the collected node data is improved. The data normalization in step 2 is to convert data of different dimensions into the same dimension.

4. A method for correcting deviation of characteristic data of power system nodes according to claim 1, characterized in that: The feature selection operation in step three is to screen features by calculating the correlation between each feature and the label to obtain corresponding data features.

5. A method for correcting deviation of characteristic data of power system nodes according to claim 1, characterized in that: The resampling in step three is performed by nearest neighbor interpolation, bilinear interpolation and cubic convolution interpolation.

6. A method for correcting deviation of characteristic data of power system nodes according to claim 1, characterized in that: The resampling operation steps in step three are to first select and determine the resampling method according to the characteristics and requirements of the data, then use the selected resampling method to process the data, and then improve the quality of the node data by comparing the quality of the data set before and after resampling.

7. A method for correcting deviation of characteristic data of power system nodes according to claim 1, characterized in that: The suspicious data point identification operation method in step three is to compare the real-time data with the historical data, and use the statistical method of standard deviation and the machine learning algorithm of BP neural network to calculate, and then identify the suspicious deviation data points, and then further analyze the identified suspicious data points to confirm that they are real deviation data.

8. A method for correcting deviation of characteristic data of power system nodes according to claim 1, characterized in that: The operation steps of the sensitivity analysis in step three are as follows: first, select uncertain factors, and according to the characteristics of the power system node data and the need for deviation correction, select the uncertain factors that have a greater impact on the data deviation for analysis, and then determine the range of change of the node data, set a reasonable range of change for each uncertain factor, observe the impact of its change on the data deviation, and then implement sensitivity analysis by simulating the changes of uncertain factors and observing the changes in data deviation in real time. Sensitivity analysis tables or sensitivity analysis diagrams can be used to record and display the analysis results to further identify the deviation data.

9. A method for correcting deviation of characteristic data of power system nodes according to claim 1, characterized in that: The deviation identification of the pre-processed data in step 3 can be performed using the node power balance formula. Using the principle of node power balance, it can be determined whether there is data deviation in the power system. The calculation method of the node power balance formula is as follows: assuming that there are n branches on the bus, the active power and reactive power of the i-th branch are Pi and Qi respectively, then the expression of the node power balance is: By judging whether the power boxes of all branches on the node are zero, if they are not zero, there may be data deviation.

10. A method for correcting deviation of characteristic data of power system nodes according to claim 1, characterized in that: The deviation correction algorithm in step 4 includes linear deviation. The correction method of linear deviation is that in the node data of the power system, the actual measurement value of a certain characteristic data is V meas , and the expected voltage value of the node is V des , and then a correction value DV needs to be calculated so that the adjusted voltage value V adj Close to the expected voltage value V des , the linear deviation correction formula can be: DV=κ×(V des -V meas ) Among them, κ is the deviation correction coefficient, which determines the strength and direction of the correction. The value of κ can be adjusted according to the actual situation; Adjusted voltage value V adj It can be calculated by the following formula: V adj =V meas +DV=V meas +κ(V des -V meas )。

Images