A data acquisition method and device for power transformation and distribution system equipment

Abstract

The present invention relates to the field of data processing technology, and more specifically to a method and device for collecting data from equipment in a power distribution system, comprising: screening out similar historical parent time periods of a parent time period of a time period to be predicted, as well as similar electricity consumption units and corresponding similarity levels, obtaining predicted power consumption for the time period to be predicted, updating the predicted power consumption based on actual power consumption, adjusting the sampling frequency based on the power consumption similarity of similar historical parent time periods, and allocating power resources based on the adjusted sampling frequency of each electricity consumption unit and the updated predicted power consumption for the time period to be predicted. The present invention obtains an initial prediction curve by analyzing historical power consumption habits, compares it with actual power consumption, and updates the prediction results, making the prediction more accurate and performing more rigorous data collection intervals for subsequent data collection.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method and device for collecting data of power transformation and distribution system equipment. Background Art

[0002] Electricity is widely used as a clean, efficient, and easily transportable energy source. Power systems need to distribute electricity based on user consumption. This resource allocation avoids waste while improving user experience. This can be achieved by collecting daily electricity data, performing predictive analysis on the data, and scheduling electricity based on predicted usage. However, electricity consumption forecasting involves the issue of accuracy; inaccurate forecasts can also lead to erroneous impacts on power distribution.

[0003] Currently, electricity data is often collected at a single frequency and then directly forecasted using historical data. Due to the complexity of electricity usage, the varying usage data of electricity users over time, and the large fluctuations in power load (due to the uncertainty of commercial activities and the rapid changes in power load), using historical data at a single frequency as forecast data is less accurate and cannot accurately estimate future electricity usage. Summary of the Invention

[0004] The present invention provides a method and device for collecting data of power transformation and distribution system equipment to solve the existing problems.

[0005] The present invention provides a method and device for collecting data from power distribution system equipment using the following technical solutions:

[0006] An embodiment of the present invention provides a method for collecting data from equipment in a power transformation and distribution system, the method comprising the following steps:

[0007] Obtaining electricity consumption data of several electricity users in any power transformation and distribution system, wherein the electricity consumption data includes several sampling times and corresponding electricity consumption;

[0008] Divide the electricity consumption data of any electricity user into several time periods, wherein the time periods include several historical time periods and one time period to be predicted, and any time period and several time periods before the time period constitute the parent time period of the time period;

[0009] Based on the electricity consumption difference between the time period to be predicted and the parent time period of any historical time period, similar historical parent time periods of the parent time period of the time period to be predicted are selected;

[0010] Based on the similarity between any electricity consumer and other electricity consumers with respect to similar historical parent time periods, similar electricity consumers of any electricity consumer with respect to any similar historical parent time period are obtained, as well as the degree of similarity between the electricity consumer and any similar electricity consumer with respect to any corresponding similar historical parent time period; based on the degree of similarity between similar electricity consumers with respect to any corresponding similar historical parent time period, combined with the electricity consumption of similar historical parent time periods, a prediction is made to obtain the predicted electricity consumption for the time period to be predicted;

[0011] Based on the difference between the predicted power consumption of the time period to be predicted and the actual power consumption at the time of sampling, the predicted power consumption of the time period to be predicted is updated; the sampling frequency is adjusted based on the similar power consumption of similar historical parent time periods to obtain the adjusted sampling frequency;

[0012] Electricity resources are allocated according to the adjusted sampling frequency of each electricity user and the updated predicted power consumption for the time period to be predicted.

[0013] Furthermore, the method of selecting similar historical parent time periods of the parent time period of the time period to be predicted based on the electricity consumption difference between the time period to be predicted and the parent time period of any historical time period includes the following specific methods:

[0014] Obtain electricity consumption fluctuation curves for any historical time period;

[0015] Performing first-order difference processing on the electricity consumption fluctuation curve of any historical time period to obtain the first-order difference curve of the historical time period;

[0016] Obtain the similarity of electricity consumption between any two historical time periods;

[0017] Calculate the average of the power consumption similarity of each corresponding time period between the parent time period of the time period to be predicted and the parent time period of all historical time periods, and record the parent time period of the historical time period whose average of the power consumption similarity of each corresponding time period is greater than the preset first parameter as the similar historical parent time period of the parent time period of the time period to be predicted.

[0018] Furthermore, the electricity consumption fluctuation curve of any historical time period includes the following specific methods:

[0019] For any historical period, As the origin, the sampling time is used as the horizontal axis, and the power consumption is used as the vertical axis to establish a coordinate system; obtain a number of data points in the historical time period, wherein the data points include the sampling time and the corresponding power consumption, and fit the data points into a curve through spline interpolation, which is recorded as the power consumption fluctuation curve of the historical time period.

[0020] Furthermore, the specific method for obtaining the similarity of electricity consumption between any two historical time periods is as follows:

[0021] For the first-order difference curves of any two historical time periods, use the dynamic time warping algorithm with Euclidean distance as the distance metric to obtain the corresponding matching points of the two historical time periods, which are recorded as the DTW matching pairs of the first-order difference curves of the two historical time periods;

[0022]

[0023] in, Indicates the similarity of electricity consumption between any two historical time periods; The number of DTW matching pairs of the first-order difference curves of any two historical time periods; The first order difference curve of any two historical time periods The distance between DTW matching pairs; Represents an exponential function with a natural constant as its base.

[0024] Furthermore, the method of obtaining similar electricity users of any electricity user for any similar historical parent time period and the degree of similarity between the electricity user and any similar electricity user for any corresponding similar historical parent time period based on similar situations between any electricity user and other electricity users for similar historical parent time periods includes the following specific methods:

[0025] For any similar historical parent time period of the parent time period of the time period to be predicted of any electricity consumption unit, obtain the electricity consumption units with the same similar historical parent time period among all electricity consumption units, and record them as similar electricity consumption units of the electricity consumption unit with respect to the similar historical parent time period; calculate the average value of the electricity consumption similarity between any similar electricity consumption unit and each corresponding time period of the similar historical parent time period of the electricity consumption unit, and record it as the similarity between the electricity consumption unit and any similar electricity consumption unit with respect to any corresponding similar historical parent time period.

[0026] Furthermore, the method of performing prediction based on the similarity between similar electricity users with respect to any corresponding similar historical parent time period and combining the electricity consumption of similar historical parent time periods to obtain the predicted electricity consumption for the time period to be predicted includes the following specific methods:

[0027] Obtain the reference weight of the electricity consumption forecast value of any similar historical parent time period for the forecast time period;

[0028] The first time period to be predicted The method for obtaining the predicted power consumption at a sampling time is:

[0029]

[0030] in, Indicates the number of similar historical parent time periods of the parent time period of the time period to be predicted. Indicates the The reference weight of the electricity consumption forecast value of the forecast period for the similar historical parent time period is Indicates the The first time period in the similar historical parent time periods that corresponds to the time period to be predicted The power consumption during each sampling period;

[0031] The predicted power consumption at each sampling time in the time period to be predicted is obtained to form the predicted power consumption data for the time period to be predicted.

[0032] Furthermore, the specific method for obtaining the reference weight of the electricity consumption forecast value of any similar historical parent time period to the forecast time period is as follows:

[0033]

[0034] in, Indicates the The reference weight of the electricity consumption forecast value of the forecast time period for the similar historical parent time period; Indicates the parent time period of the time period to be predicted and the The average value of the electricity consumption similarity of each corresponding time period in similar historical parent time periods; Indicates the electricity consumption unit and the The first of similar electricity users The similarity between similar historical time periods, Indicates the number of similar electricity consumers to the electricity consumer in the forecast period.

[0035] Furthermore, the specific method for updating the predicted power consumption of the time period to be predicted based on the difference between the predicted power consumption of the time period to be predicted and the actual power consumption at the time of collection is:

[0036] The difference between the actual power consumption at the time of collection and the predicted power consumption at the corresponding time is calculated, and the absolute value is taken, and then normalization is performed to obtain the difference between the actual power consumption at the time of collection and the predicted power consumption at the corresponding time. When the difference is greater than the preset second parameter, the actual power consumption at the time of collection is incorporated into the historical time period according to the method of obtaining the predicted power consumption for the time period to be predicted, and the predicted power consumption for the time period to be predicted is obtained again.

[0037] Furthermore, the sampling frequency is adjusted based on the electricity consumption of similar historical parent time periods to obtain the adjusted sampling frequency, including the following specific methods:

[0038] Select the most similar historical parent time period of the time period to be predicted Similar historical time periods;

[0039] described is the default value;

[0040]

[0041] in, Indicates the degree of adjustment of the sampling frequency, Indicates the difference between the actual power consumption at the latest sampling time and the predicted power consumption at the corresponding sampling time; Indicates the number of similar historical parent time periods with the greatest similarity. The one with the greatest similarity The average value of the electricity consumption similarity of each corresponding time period of the parent time period of the similar historical time period and the parent time period of the time period to be predicted; represents an exponential function with a natural constant as its base;

[0042] Multiply the sampling frequency by , and get the adjusted sampling frequency.

[0043] A data acquisition device for power transformation and distribution system equipment includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the processor implements any one of the steps of a data acquisition method for power transformation and distribution system equipment.

[0044] The beneficial effects of the technical solution of the present invention are: by analyzing historical electricity usage habits, the current stage of electricity usage can be predicted to a certain extent, and the daily historical electricity consumption fluctuation curves obtained can be calculated and weighted averaged to obtain the initial prediction curve. When using the prediction data, the weighted similar historical data collection can be recalculated based on the difference between the currently collected data and the predicted data, and the data of the time period already obtained on the current day, so that the obtained prediction results are more accurate when the data increases with time. At the same time, the special situation of electricity consumption obtained in the current stage can also be analyzed through the prediction deviation, and a more rigorous data collection interval is required for subsequent data collection. The technical solution provided by the present invention requires less data to be collected, and the equipment pressure brought by data transmission, storage, and calculation is relatively small. There is no need to configure a large number of smart meters and high-end servers. It is particularly suitable for upgrading and transformation of old power distribution systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0046] Figure 1 The present invention is a flowchart of the steps of a method for collecting data of power distribution system equipment. DETAILED DESCRIPTION

[0047] To further illustrate the technical means and effectiveness of the present invention in achieving its intended objectives, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail a method and apparatus for collecting data from power distribution system equipment according to the present invention, including its specific implementation, structure, features, and effectiveness. In the following description, references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0048] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0049] The following describes in detail a method and device for collecting data on power distribution system equipment provided by the present invention with reference to the accompanying drawings.

[0050] See also Figure 1 , which shows a flowchart of a method for collecting data of power distribution system equipment provided by one embodiment of the present invention, the method comprising the following steps:

[0051] Step S001: Obtain power consumption data of several power consumption units in any power transformation and distribution system.

[0052] It should be noted that this embodiment takes any commercial center as an example, and the commercial center includes several electricity-consuming units of different types or sizes, such as shopping malls and office buildings.

[0053] Specifically, in order to implement the data collection method for power distribution system equipment proposed in this embodiment, it is first necessary to collect power consumption data of the power distribution system. The specific process is as follows:

[0054] In this embodiment, the power system distributes power through low-voltage substation distribution equipment, and power sensors can be used to obtain power consumption data.

[0055] The power sensor is connected to the bus of the low-voltage substation and the sub-lines of several power-consuming units, and the power consumption data of the power sensor is sampled at a preset sampling frequency to obtain the power consumption data of each power-consuming unit. The power consumption data includes several sampling times and corresponding power consumption.

[0056] The electricity consumption data of any electricity consuming unit is divided into several time periods, wherein the time periods include several historical time periods and one time period to be predicted. Any time period and several time periods before the time period constitute the parent time period of the time period.

[0057] It should be noted that although the higher the sampling frequency, the higher the accuracy, the data calculation amount will increase exponentially. Taking into account the sampling accuracy and data calculation amount, the optimal preset sampling frequency is 10 minutes / time, but it can also be adjusted according to actual conditions. This embodiment does not make specific limitations.

[0058] It should be noted that because electricity consumption by different electricity users varies periodically, the optimal solution for residential power distribution is to set the time period to one day, i.e., 24 hours, and the parent time period to one week, i.e., 7 days. However, this can be adjusted based on actual conditions and is not specifically limited in this embodiment. For example, for an industrial park, the time period can be set to one week, i.e., 7 days, and the parent time period can be set to one year, i.e., 365 days.

[0059] So far, the power consumption data of several power consuming units in any power transformation and distribution system have been obtained through the above method.

[0060] Step S002: based on the power consumption difference between the time period to be predicted and the parent time period of any historical time period, similar historical parent time periods of the parent time period of the time period to be predicted are selected.

[0061] It should be noted that when dispatching power resources, it is necessary to analyze the power demand of each power user and adjust resource scheduling based on the power user's historical power usage habits. The power demand of each power user is obtained based on the historical power usage data of the power user, and the historical power usage data of the stage that has not yet occurred is predicted by historical data. The predicted value can be used as the power demand of each power user. At the same time, based on the prediction results, it can also be judged that if there is a large difference between the currently collected power data and the predicted value, it indicates that there may be a significant change in power demand, which can be used to adjust the data collection and closely monitor it.

[0062] It should be noted that when calculating the predicted data through the acquired historical data, it is possible to analyze whether there is similar data in the history through the data close to the current stage, and perform a joint weighted average of the acquired similar historical data and the data acquired within a period of time to determine the initial predicted power consumption. The model can be adjusted based on the judgment of the actual collected data and the predicted data, and the similar historical data can be recalculated based on the data already collected in the current stage.

[0063] First, obtain the electricity consumption fluctuation curve of any historical time period.

[0064] It's important to note that electricity consumption in commercial centers exhibits certain cyclical distribution characteristics. Analyzing historical electricity usage patterns can, to a certain extent, predict current electricity usage. Commercial centers have similar electricity demands during normal business hours. For example, a shopping mall's peak electricity demand occurs between 10:00 AM and 11:00 PM. Furthermore, electricity usage exhibits cyclical characteristics due to the presence of weekdays and holidays.

[0065] For any historical period, As the origin, the sampling time is used as the horizontal axis, and the power consumption is used as the vertical axis to establish a coordinate system; obtain a number of data points in the historical time period, wherein the data points include the sampling time and the corresponding power consumption, and fit the data points into a curve through spline interpolation, which is recorded as the power consumption fluctuation curve of the historical time period.

[0066] It should be noted that spline interpolation is an existing mathematical method for generating a smooth curve through a series of data points, and therefore will not be described in detail in this embodiment.

[0067] Secondly, a first-order difference process is performed on the electricity consumption fluctuation curve of any historical time period to obtain a first-order difference curve of the historical time period.

[0068] It should be noted that the electricity consumption data of shopping malls has a strong seasonal trend, and differential processing can reduce the impact of seasonality.

[0069] It should be noted that the first-order difference is an existing calculation method, and therefore will not be described in detail in this embodiment.

[0070] Then, the similarity of electricity consumption between any two historical time periods is obtained.

[0071] It should be noted that when obtaining forecast data for the time period to be predicted, since no electricity usage data has yet been generated for the time period to be predicted, it is impossible to compare the known data for the time period to the historical data for similarity, and it is impossible to determine similar historical data to obtain forecast data. However, due to the existence of periodicity, it is possible to compare other time periods within the parent time period of the time period to historical time periods to obtain similarities with other time periods within the parent time period of the time period to be predicted and any historical time periods, thereby obtaining similar historical time periods for other time periods within the parent time period of the time period to be predicted.

[0072] In one embodiment of the present application, a method for obtaining the similarity of electricity usage between any two historical time periods is as follows:

[0073] For the first-order difference curves of any two historical time periods, use the dynamic time warping algorithm with Euclidean distance as the distance metric to obtain the corresponding matching points of the two historical time periods, which are recorded as the DTW matching pairs of the first-order difference curves of the two historical time periods;

[0074]

[0075] in, Indicates the similarity of electricity consumption between any two historical time periods; The number of DTW matching pairs of the first-order difference curves of any two historical time periods; The first order difference curve of any two historical time periods The distance between DTW matching pairs; Represents an exponential function with a natural constant as its base.

[0076] It should be noted that the Euclidean distance is an existing way of expressing the distance between two points, and the dynamic time warping algorithm is an existing method for calculating the distance between two time series, so it is not described in detail in this embodiment.

[0077] It should be noted that The average distance between each DTW matching pair of the first-order difference curves of any two historical time periods shows the overall difference in electricity consumption data between the two historical time periods and is inversely proportional to the similarity of electricity consumption between the two historical time periods.

[0078] In another embodiment of the present application, the sampling frequency is 10 minutes per time, the time period is 24 hours, the main time period is one week, and 1007 data points need to be collected for each electricity consumer in one main time period; the calculation of one main time period including 1000 electricity consumers only needs to process 1.007 million data, and the data calculation work can be completed using an ordinary PC.

[0079] Finally, the mean of the similarity of electricity consumption between the parent time period of the time period to be predicted and the parent time period of all historical time periods is calculated, and the parent time period of the historical time period whose mean of the similarity of electricity consumption in each corresponding time period is greater than the preset first parameter is recorded as the similar historical parent time period of the parent time period of the time period to be predicted.

[0080] It should be noted that the preset first parameter is set to 0.7 based on experience, and can be adjusted according to actual conditions. This embodiment does not impose any specific limitation.

[0081] Step S003: Based on the similarity between any electricity consuming unit and other electricity consuming units regarding similar historical mother time periods, obtain similar electricity consuming units of any electricity consuming unit regarding any similar historical mother time period, and the degree of similarity between the electricity consuming unit and any similar electricity consuming unit regarding any corresponding similar historical mother time period; based on the degree of similarity between similar electricity consuming units regarding any corresponding similar historical mother time period, and combined with the electricity consumption of similar historical mother time periods, make a prediction to obtain the predicted electricity consumption of the time period to be predicted.

[0082] It should be noted that, in step S002, similar historical parent time periods of the parent time period of the time period to be predicted of any electricity consumption unit are obtained, but the similarity of each similar historical parent time period is not the same, and the availability of the predicted power consumption for the time period to be predicted is also different. Therefore, it is necessary to obtain the availability of the predicted power consumption of each similar historical parent time period for the predicted time period. The availability of each similar historical parent time period can be analyzed by other electricity consumption units. When there are other electricity consumption units that have the same similar historical parent time period as the electricity consumption unit, it can be considered that the electricity consumption unit has similar electricity consumption units and similar power consumption conditions. Then, the availability of each similar historical parent time period can be obtained through the similarity of similar electricity consumption units with respect to each similar historical parent time period.

[0083] Step 3.1, based on the similarity between any electricity consuming unit and other electricity consuming units regarding similar historical mother time periods, obtain similar electricity consuming units of any electricity consuming unit regarding any similar historical mother time period, as well as the degree of similarity between the electricity consuming unit and any similar electricity consuming unit regarding any corresponding similar historical mother time period.

[0084] It should be noted that if there are the same similar historical parent time periods between the time periods to be predicted of different electricity consuming units, then the time periods to be predicted should also be similar. At the same time, the more electricity consuming units that have the same similar historical parent time periods, the higher the degree of similarity of the similar historical parent time periods.

[0085] Specifically, for any similar historical parent time period of the parent time period of the time period to be predicted of any electricity consumption unit, obtain the electricity consumption units with the same similar historical parent time period among all electricity consumption units, and record them as similar electricity consumption units of the electricity consumption unit with respect to the similar historical parent time period; calculate the average value of the electricity consumption similarity of any similar electricity consumption unit and each corresponding time period of the similar historical parent time period of the electricity consumption unit, and record it as the similarity degree between the electricity consumption unit and any similar electricity consumption unit with respect to any corresponding similar historical parent time period.

[0086] Step 3.2: Based on the similarity of similar electricity users with respect to any corresponding similar historical parent time period, and in combination with the electricity consumption of similar historical parent time periods, a prediction is made to obtain the predicted electricity consumption of the time period to be predicted.

[0087] First, obtain the reference weight of the electricity consumption forecast value of the forecast time period for any similar historical parent time period.

[0088] It should be noted that the reference weight of any similar historical parent time period of the time period to be predicted for any electricity consuming unit is related to the degree of similarity in electricity consumption between the parent time period of the time period to be predicted and each corresponding time period of the similar historical parent time period. If the degree of similarity in electricity consumption of the corresponding time periods is large, then the degree of similarity between the parent time period of the time period to be predicted and the similar historical parent time period will be higher, and the reference weight of the similar historical parent time period will be greater. At the same time, the more similar electricity consuming units there are and the higher the degree of similarity, the greater the reference weight of the corresponding similar historical parent time period.

[0089] As an embodiment, a method for obtaining a reference weight of the electricity consumption forecast value of the forecast time period for any similar historical parent time period is as follows:

[0090]

[0091] in, Indicates the The reference weight of the electricity consumption forecast value of the forecast time period for the similar historical parent time period; Indicates the parent time period of the time period to be predicted and the The average value of the electricity consumption similarity of each corresponding time period in similar historical parent time periods; Indicates the electricity consumption unit and the The first of similar electricity users The similarity between similar historical time periods, The number of similar electricity users to the electricity user in the time period to be predicted; Represents the normalization function.

[0092] It should be noted that It represents the sum of the similarity of all similar units of the electricity consumption unit belonging to the time period to be predicted with respect to the same similar historical parent time period, and represents the occurrence of the similar historical parent time period.

[0093] Secondly, the reference weight of each similar historical parent time period of the time period to be predicted is combined with the power consumption data of the time period corresponding to the time period to be predicted in the similar historical parent time period to obtain the predicted power consumption data of the time period to be predicted.

[0094] As an embodiment, a method for obtaining the predicted power consumption at the fth sampling time in the time period to be predicted is as follows:

[0095]

[0096] in, represents the predicted power consumption at the fth sampling time in the time period to be predicted, Indicates the number of similar historical parent time periods of the parent time period of the time period to be predicted. Indicates the The reference weight of the electricity consumption forecast value of the forecast period for the similar historical parent time period is Indicates the The power consumption at the fth sampling time of the time period corresponding to the time period to be predicted in the similar historical parent time periods.

[0097] It should be noted that It means that the reference weight of the electricity consumption forecast value of the time period to be predicted in similar historical parent time periods is used as the weight, and the weighted average of the electricity consumption at the corresponding sampling time of the time period corresponding to the time period to be predicted in all similar historical parent time periods is taken as the predicted electricity consumption at any sampling time of the time period to be predicted.

[0098] The predicted power consumption at each sampling time in the time period to be predicted is obtained to form the predicted power consumption data for the time period to be predicted.

[0099] Step S004: Update the predicted power consumption of the time period to be predicted based on the difference between the predicted power consumption of the time period to be predicted and the actual power consumption at the time of collection; adjust the sampling frequency based on the similar power consumption of similar historical parent time periods to obtain the adjusted sampling frequency.

[0100] It should be noted that due to the large fluctuations in power load in commercial centers, the predicted power consumption data obtained for the forecasted time period may differ significantly from the actual power consumption data. Therefore, during power distribution, the actual power consumption data already obtained can be compared with the predicted power consumption data, and similar historical parent time periods can be recalculated. The power consumption data that has not yet been obtained can be re-forecasted. This makes the prediction results more accurate as more actual power consumption data is obtained. At the same time, the deviation of the predicted power consumption data can also be used to analyze the special power consumption conditions of the current stage, and the subsequent data collection frequency can be adjusted to further increase the accuracy of the prediction.

[0101] Specifically, in step 4.1, the predicted power consumption for the time period to be predicted is updated according to the difference between the predicted power consumption for the time period to be predicted and the actual power consumption at the time of collection.

[0102] It should be noted that when the predicted electricity consumption is significantly different from the actual electricity consumption, there will be deviations in the allocation of electricity consumption and the adjustment of the collection frequency. Therefore, it is necessary to update the subsequent predicted electricity consumption in a timely manner and incorporate the actual electricity consumption obtained into the historical time period to make the prediction results more accurate.

[0103] The difference between the actual power consumption at the time of collection and the predicted power consumption at the corresponding time is calculated, and the absolute value is taken, and then normalization is performed to obtain the difference between the actual power consumption at the time of collection and the predicted power consumption at the corresponding time. When the difference is greater than the preset second parameter, the actual power consumption at the time of collection is incorporated into the historical time period according to the method of obtaining the predicted power consumption for the time period to be predicted, and the predicted power consumption for the time period to be predicted is obtained again.

[0104] It should be noted that the preset second parameter is set to 0.5 based on experience, and can be adjusted according to actual conditions. This embodiment does not specifically limit it.

[0105] In step 4.2, the sampling frequency is adjusted based on the electricity consumption in similar historical time periods to obtain an adjusted sampling frequency.

[0106] First, select the most similar historical parent time period of the time period to be predicted. A similar historical period.

[0107] It should be noted that N is a preset value. Not all similar historical mother time periods are available. Similar historical mother time periods with low similarity have little reference value for adjusting the sampling frequency. Based on experience, N is set to 5 and can be adjusted according to actual conditions. This embodiment does not make any specific limitations.

[0108] Secondly, the degree of adjustment of the sampling frequency of any electricity consumption unit is obtained based on the difference between the actual electricity consumption and the predicted electricity consumption, and the similarity of electricity consumption between all similar historical parent time periods and the parent time period of the time period to be predicted.

[0109] It should be noted that when the difference between actual electricity consumption and predicted electricity consumption is large, it means that the current sampling frequency is not appropriate and the degree of adjustment should be large; at the same time, the greater the similarity between the electricity consumption of all similar historical parent time periods and the parent time period of the time period to be predicted, the smaller the degree of adjustment of the sampling frequency should be.

[0110] As an embodiment, a method for obtaining the adjustment degree of the sampling frequency of any electricity consumption unit is as follows:

[0111]

[0112] in, Indicates the degree of adjustment of the sampling frequency, Indicates the difference between the actual power consumption at the latest sampling time and the predicted power consumption at the corresponding sampling time; Indicates the number of similar historical parent time periods with the greatest similarity. The one with the greatest similarity The average value of the electricity consumption similarity of each corresponding time period of the parent time period of the similar historical time period and the parent time period of the time period to be predicted; Represents an exponential function with a natural constant as its base.

[0113] It should be noted that Indicates the greatest similarity The average values ​​of the power consumption similarity between the similar historical parent time periods and the parent time period of the time period to be predicted are summed to represent the overall similarity between the similar historical parent time periods and the parent time period of the time period to be predicted.

[0114] Finally, multiply the sampling frequency by , obtaining an adjusted sampling frequency, and using the adjusted sampling frequency of any electricity consumption unit to subsequently collect the electricity consumption data of the electricity consumption unit.

[0115] Through the above steps, the data collection of the power transformation and distribution system equipment is completed.

[0116] An embodiment of the present invention provides a data acquisition device for power transformation and distribution system equipment, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any one of the steps of the data acquisition method for power transformation and distribution system equipment.

[0117] This embodiment can predict the current electricity usage to a certain extent by analyzing historical electricity usage habits. The initial prediction curve is obtained by calculating the weighted average of the daily historical electricity consumption fluctuation curve. When using the predicted data, the difference between the currently collected data and the predicted data and the data of the time period already acquired on the current day can be used to recalculate the weighted collection of similar historical data. This makes the prediction results more accurate as time goes by and the data increases. At the same time, the prediction deviation can also be used to analyze the special circumstances of the electricity usage acquired in the current stage, requiring more stringent data collection intervals for subsequent data collection.

[0118] It should be noted that the The model is only used to represent negative correlation and constrain the output of the model to be in In the specific implementation, it can be replaced by other models with the same purpose. This embodiment is only based on The model is described as an example without any specific limitation. is the input to the model.

[0119] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A data collection method for power transformation and distribution system equipment, characterized in that: The method comprises the following steps: Obtaining electricity consumption data of several electricity users in any power transformation and distribution system, wherein the electricity consumption data includes several sampling times and corresponding electricity consumption; Divide the electricity consumption data of any electricity user into several time periods, wherein the several time periods include several historical time periods and one time period to be predicted, and any time period and several time periods before the any time period constitute the parent time period of the any time period; Based on the electricity consumption difference between the time period to be predicted and the parent time period of any historical time period, similar historical parent time periods of the parent time period of the time period to be predicted are selected; Based on the similarity between any electricity consumer and other electricity consumers with respect to similar historical parent time periods, similar electricity consumers of any electricity consumer with respect to any similar historical parent time period are obtained, as well as the degree of similarity between the electricity consumer and any similar electricity consumer with respect to any corresponding similar historical parent time period; based on the degree of similarity between similar electricity consumers with respect to any corresponding similar historical parent time period, combined with the electricity consumption of similar historical parent time periods, a prediction is made to obtain the predicted electricity consumption for the time period to be predicted; Based on the difference between the predicted power consumption of the time period to be predicted and the actual power consumption at the time of sampling, the predicted power consumption of the time period to be predicted is updated; the sampling frequency is adjusted based on the similar power consumption of similar historical parent time periods to obtain the adjusted sampling frequency; Allocate power resources based on the adjusted sampling frequency of each power user and the updated predicted power consumption for the time period to be predicted; The method of performing a prediction based on the similarity between similar electricity users with respect to any corresponding similar historical parent time period and combining the electricity consumption of similar historical parent time periods to obtain the predicted electricity consumption for the time period to be predicted includes the following specific methods: Obtain the reference weight of the electricity consumption forecast value of any similar historical parent time period for the forecast time period; The first time period to be predicted The method for obtaining the predicted power consumption at a sampling time is: in, Indicates the number of similar historical parent time periods of the parent time period of the time period to be predicted. Indicates the The reference weight of the electricity consumption forecast value of the forecast period for the similar historical parent time period is Indicates the The first time period in the similar historical parent time periods that corresponds to the time period to be predicted The power consumption during each sampling period; Obtain the predicted power consumption at each sampling time in the time period to be predicted, and form the predicted power consumption data for the time period to be predicted; The sampling frequency is adjusted based on the electricity consumption of similar historical time periods to obtain the adjusted sampling frequency, and the specific method includes: Select the most similar historical parent time period of the time period to be predicted Similar historical time periods; described is the default value; in, Indicates the degree of adjustment of the sampling frequency, Indicates the difference between the actual power consumption at the latest sampling time and the predicted power consumption at the corresponding sampling time; Indicates the number of similar historical parent time periods with the greatest similarity. The one with the greatest similarity The average value of the electricity consumption similarity of each corresponding time period of the parent time period of the similar historical time period and the parent time period of the time period to be predicted; represents an exponential function with a natural constant as its base; Multiply the sampling frequency by , and get the adjusted sampling frequency.

2. A method for collecting data from power distribution system equipment according to claim 1, characterized in that: The method of selecting similar historical parent time periods of the parent time period of the time period to be predicted based on the electricity consumption difference between the time period to be predicted and the parent time period of any historical time period includes the following specific methods: Obtain electricity consumption fluctuation curves for any historical time period; Performing first-order difference processing on the electricity consumption fluctuation curve of any historical time period to obtain the first-order difference curve of the historical time period; Obtain the similarity of electricity consumption between any two historical time periods; Calculate the average of the power consumption similarity of each corresponding time period between the parent time period of the time period to be predicted and the parent time period of all historical time periods, and record the parent time period of the historical time period whose average of the power consumption similarity of each corresponding time period is greater than the preset first parameter as the similar historical parent time period of the parent time period of the time period to be predicted.

3. A method for collecting data from power distribution system equipment according to claim 2, characterized in that: The electricity consumption fluctuation curve of any historical time period includes the following specific methods: For any historical period, As the origin, the sampling time is used as the horizontal axis, and the power consumption is used as the vertical axis to establish a coordinate system; obtain a number of data points in the historical time period, wherein the data points include the sampling time and the corresponding power consumption, and fit the data points into a curve through spline interpolation, which is recorded as the power consumption fluctuation curve of the historical time period.

4. A method for collecting data from power distribution system equipment according to claim 2, characterized in that: The specific method for obtaining the similarity of electricity consumption between any two historical time periods is as follows: For the first-order difference curves of any two historical time periods, use the dynamic time warping algorithm with Euclidean distance as the distance metric to obtain the corresponding matching points of the two historical time periods, which are recorded as the DTW matching pairs of the first-order difference curves of the two historical time periods; in, Indicates the similarity of electricity consumption between any two historical time periods; The number of DTW matching pairs of the first-order difference curves of any two historical time periods; The first order difference curve of any two historical time periods The distance between DTW matching pairs; Represents an exponential function with a natural constant as its base.

5. The data collection method for power transformation and distribution system equipment according to claim 1, characterized in that: The method of obtaining similar electricity users of any electricity user for any similar historical parent time period and the degree of similarity between the electricity user and any similar electricity user for any corresponding similar historical parent time period based on similar situations of any electricity user and other electricity users for similar historical parent time periods includes the following specific methods: For any similar historical parent time period of the parent time period of the time period to be predicted of any electricity consumer, obtain the electricity consumer units with the same similar historical parent time period among all electricity consumers, and record them as similar electricity consumer units of the electricity consumer unit with respect to the similar historical parent time period; Calculate the average of the electricity consumption similarity between any similar electricity consuming unit and each corresponding time period of the similar historical parent time period of the electricity consuming unit, and record it as the similarity between the electricity consuming unit and any similar electricity consuming unit with respect to any corresponding similar historical parent time period.

6. The data collection method for power transformation and distribution system equipment according to claim 1, characterized in that: The specific method for obtaining the reference weight of the electricity consumption forecast value of the forecast time period in any similar historical parent time period is as follows: in, Indicates the The reference weight of the electricity consumption forecast value of the forecast time period for the similar historical parent time period; Indicates the parent time period of the time period to be predicted and the The average value of the electricity consumption similarity of each corresponding time period in similar historical parent time periods; Indicates the electricity consumption unit and the The first of similar electricity users The similarity between similar historical time periods, Indicates the number of similar electricity consumers to the electricity consumer in the forecast period.

7. The data collection method for power transformation and distribution system equipment according to claim 1, characterized in that: The specific method for updating the predicted power consumption of the time period to be predicted based on the difference between the predicted power consumption of the time period to be predicted and the actual power consumption at the time of collection is: The difference between the actual power consumption at the time of collection and the predicted power consumption at the corresponding time is calculated, and the absolute value is taken, and then normalization is performed to obtain the difference between the actual power consumption at the time of collection and the predicted power consumption at the corresponding time. When the difference is greater than the preset second parameter, the actual power consumption at the time of collection is incorporated into the historical time period according to the method of obtaining the predicted power consumption for the time period to be predicted, and the predicted power consumption for the time period to be predicted is obtained again.

8. A data acquisition device for power transformation and distribution system equipment, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

Images