Low-voltage reactive compensation method and system for power distribution device

Abstract

The invention discloses a low-voltage reactive power compensation method and system for a power distribution device. The method comprises the steps that S10, reactive load data of all phases of the low-voltage side of the power distribution device are collected; s20, constructing a time sequence model of the reactive load of each phase of the low-voltage side of the power distribution device based on the reactive load data of each phase of a plurality of time nodes; s30, on the basis of the time sequence model, reactive load data of each phase of the next time node is predicted through exponential smoothing; and S40, based on the reactive load data of each phase of the next time node, determining the grouping switching condition of the reactive compensation capacitor of each phase of the power distribution device. According to the method, the reactive load data in the future is predicted based on the existing data through the exponential smoothing method, so that the adaptability to load fluctuation is improved, the response to emergency situations is enhanced, and the timeliness is higher.

Description

Technical Field

[0001] The present application relates to the field of electric power, and in particular to a low-voltage reactive power compensation method and system for a power distribution device. Background Art

[0002] In the power system, distribution devices are usually used to distribute electric energy. However, when the distribution device is working, some of the electric energy cannot be used to do work, that is, reactive power, such as the electric energy occupied by electromagnetic components to establish a magnetic field, the electric energy occupied by capacitors to establish an electric field, etc. In order to ensure the safe and efficient transportation of electricity in the power grid and distribution area, reactive power compensation for the distribution device is one of the main measures.

[0003] The most common method for reactive power compensation of power distribution equipment is to install capacitors on the low-voltage side. However, the reactive power output by the capacitors cannot be adjusted, so the capacitors need to be switched in groups according to the load conditions. However, the existing automatic switching methods for capacitors are not perfect, and most of them rely on the probability statistics of historical data for adjustment. However, in actual applications, the load of the power grid is time-sensitive and cannot respond quickly to sudden load fluctuations. Therefore, it is necessary to optimize the capacitor switching method for power distribution equipment. Summary of the invention

[0004] The present application mainly provides a low-voltage reactive power compensation method and system for a power distribution device to solve the defects of existing low-voltage reactive power compensation methods, such as lack of timeliness, predictability, and slow response to emergency situations.

[0005] In order to solve the above technical problems, a technical solution adopted in the present application is: to provide a low-voltage reactive power compensation method for a power distribution device, comprising: S10: Collect reactive load data of each phase on the low voltage side of the power distribution device; S20: constructing a time series model of reactive load of each phase on the low-voltage side of the power distribution device based on the reactive load data of each phase at a number of time nodes; S30: Based on the time series model, predict the reactive load data of each phase at the next time node by exponential smoothing; S40: Based on the reactive load data of each phase at the next time node, determine the group switching status of the capacitors for reactive compensation of each phase of the power distribution device.

[0006] In a possible implementation manner, the step of predicting reactive load data of each phase at the next time by exponential smoothing based on the time series model includes: S31: Obtain reactive load data of each phase at the next time node through an exponential smoothing prediction, as shown in the following formula:

[0007] Where St+1 is the predicted value of the next time node, S t is the reactive load data of each phase at the tth time node, P t is the smoothing value at the tth time node, and a is the weighting coefficient.

[0008] In a possible implementation manner, after the step of obtaining the reactive load data of each phase at the next time node through a first exponential smoothing prediction, the method includes: S32: When the deviation of the prediction value of the next time node exceeds a first threshold, the prediction value of the next time node is corrected by quadratic exponential smoothing prediction, a corrected prediction value of the next time node is obtained, and reactive load data of each phase of the next time node is corrected.

[0009] In a possible implementation, after the step of correcting the next time node prediction value by using a quadratic exponential smoothing prediction when the next time node prediction value deviation exceeds a first threshold, the method further includes: S33: When the deviation of the first-corrected prediction value of the next time node exceeds the second threshold, the first-corrected prediction value of the next time node is corrected by triple exponential smoothing prediction, the second-corrected prediction value of the next time node is obtained, and the reactive load data of each phase of the next time node is corrected.

[0010] In a possible implementation manner, the step of collecting reactive load data of each phase on the low-voltage side of the power distribution device includes: S11: Use edge computing technology to collect reactive load data of each phase on the low-voltage side of the power distribution device.

[0011] In a possible implementation, the total length of time of the multiple time nodes includes days, weeks, months or quarters.

[0012] In a possible implementation manner, the step of determining the grouping of capacitors for reactive power compensation of each phase of the power distribution device based on the reactive load data of each phase at the next time node includes: S41: Collect and count the fluctuation values ​​of reactive load data of each phase on the low-voltage side of the power distribution device and form a probability distribution function, and correct the reactive load data of each phase at the next time node based on the probability distribution function.

[0013] In a possible implementation manner, after the step of determining the grouping of capacitors for reactive power compensation of each phase of the power distribution device based on the reactive load data of each phase at the next time node, the step further includes: S42: Perform abnormality detection on reactive load data of each phase on the low-voltage side of the power distribution device, and remove abnormal data segments.

[0014] In a possible implementation manner, the step of performing abnormality detection on reactive load data of each phase on the low-voltage side of the power distribution device and removing abnormal data segments includes: S43: After removing the abnormal data, the corresponding data is completed by interpolating adjacent data.

[0015] In order to solve the above technical problems, another technical solution adopted by the present application is: to provide a low-voltage reactive power compensation system for a power distribution device, which is applicable to a low-voltage reactive power compensation method for a power distribution device as described above, comprising: A collection module, used to collect reactive load data of each phase on the low-voltage side of the power distribution device; An analysis module, used for processing the reactive load data of each phase and predicting the reactive load data of each phase at the next time node; The grouping module determines the group switching conditions of the reactive compensation capacitors of each phase of the power distribution device based on the reactive load data of each phase at the next time node.

[0016] The beneficial effects of the present application are as follows: Different from the prior art, the present application discloses a low-voltage reactive power compensation method and system for a power distribution device, which predicts future reactive load data based on existing data through an exponential smoothing method, thereby improving adaptability to load fluctuations, strengthening response to emergencies, and being more timely. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work, among which: Figure 1 It is a schematic diagram of a method flow of an embodiment of the present application. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0019] The terms "first", "second", "third" in the embodiments of the present application are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Thus, the features defined as "first", "second", "third" can expressly or implicitly include at least one of the features. In the description of the present application, the meaning of "multiple" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes other steps or units inherent to these processes, methods, products or devices.

[0020] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0021] See also Figure 1 , the embodiment of the present application includes a low-voltage reactive power compensation method for a power distribution device, comprising: S10: Collect reactive load data of each phase on the low voltage side of the power distribution device; This embodiment adopts edge computing technology to quickly collect and process information by locally deploying a lightweight stream data processing platform (such as Apache Kafka, Flink, etc.), so as to process data and update models more quickly and efficiently, thereby improving response speed and timeliness.

[0022] S20: constructing a time series model of reactive load of each phase on the low voltage side of the power distribution device based on reactive load data of each phase at a number of time nodes; A time series model is constructed for the collected reactive load data of each phase according to the time of collection, so as to predict the data change trend in the short term in the future based on historical data and real-time data, and improve the timeliness of the reactive compensation method.

[0023] S30: Based on the time series model, the reactive load data of each phase at the next time node is predicted by exponential smoothing; The exponential smoothing method is used to predict the data for the next time node. The exponential smoothing method weights the known data and gives higher weight to recent data to capture changing trends, enhance timeliness, and better respond to sudden changes.

[0024] S40: Based on the reactive load data of each phase at the next time node, determine the group switching status of the capacitors for reactive compensation of each phase of the power distribution device.

[0025] The prediction data can be used to adjust the group switching of reactive compensation capacitors in a targeted manner to better adapt to sudden situations in the power grid.

[0026] In a possible implementation, based on the time series model, the step of predicting the reactive load data of each phase at the next time by exponential smoothing includes: S31: Obtain the reactive load data of each phase at the next time node through an exponential smoothing prediction, as shown in the following formula:

[0027] Where S t+1 is the predicted value of the next time node, S t is the reactive load data of each phase at the tth time node, P t is the smoothing value at the tth time node, and a is the weighting coefficient.

[0028] Usually we select a piece of data as the initial value P0 of the prediction value. The initial value can be the average of the first few data values, and then the prediction value corresponding to the next node is calculated through the data of each time node in turn.

[0029] Then judge based on the distribution trend in the series scatter plot. The greater the change in the trend, the more exponential smoothing is applied. For example, when the trend is a straight line, the quadratic exponential smoothing method is used, and when the trend is a parabola, the cubic exponential smoothing method is used. The trend can be judged by the change in curvature.

[0030] The smaller the weighting coefficient a is, the stronger the smoothing effect is, which is reflected in the embodiment as a slower response to changes in the reactive load parameters of each phase.

[0031] In a possible implementation manner, after the step of obtaining reactive load data of each phase at the next time node through an exponential smoothing prediction, the method includes: S32: When the deviation of the prediction value of the next time node exceeds the first threshold, the prediction value of the next time node is corrected by quadratic exponential smoothing prediction, a corrected prediction value of the next time node (that is, the quadratic smoothing value of the next time node) is obtained, and the reactive load data of each phase at the next time node is corrected.

[0032] Even if the trend of the smoothed value is relatively gentle after the first exponential smoothing method, its predicted value still has obvious lag and slower response. Therefore, an interval is set for the predicted value of the next node. When the first threshold is exceeded and the predicted value exceeds the range of the interval, the second exponential smoothing is used for correction. The calculation formula of the second exponential smoothing is:

[0033] in is the quadratic exponential smoothing value at the next time node, is the quadratic exponential smoothing value at the tth time node, and the initial value of the quadratic exponential smoothing value can be the average of the initial several data. The quadratic exponential smoothing must be smoothed on the basis of the primary exponential smoothing and cannot be used alone.

[0034] In a possible implementation, when the deviation of the prediction value of the next time node exceeds the first threshold, after the step of correcting the prediction value of the next time node by using the secondary exponential smoothing prediction, the method includes: S33: When the deviation of the first-corrected prediction value of the next time node exceeds the second threshold, the first-corrected prediction value of the next time node is corrected by triple exponential smoothing prediction, the second-corrected prediction value of the next time node is obtained, and the reactive load data of each phase at the next time node is corrected.

[0035] Similarly, the second threshold is set. If the predicted value still exceeds the second threshold after the second exponential smoothing, a third correction is performed to improve the prediction accuracy and response speed. The calculation formula for the third exponential smoothing is:

[0036] in is the triple exponential smoothing value of the next time node, is the triple exponential smoothing value at the tth time node, and the initial value of the triple exponential smoothing value can be the average of the initial several data. The triple exponential smoothing must be smoothed on the basis of the double exponential smoothing and cannot be used alone.

[0037] In a possible implementation manner, the step of collecting reactive load data of each phase on the low voltage side of the power distribution device includes: S11: Use edge computing technology to collect reactive load data of each phase on the low-voltage side of the distribution device.

[0038] Edge computing technology is a distributed computing framework that sets up devices or servers near the source of data. It improves the speed and efficiency of data processing by reducing the physical transmission distance to achieve faster response time and better bandwidth availability.

[0039] In a possible implementation, the total length of time of several time nodes includes days, weeks, months or quarters. The historical data can be processed on a daily, weekly, monthly or quarterly basis to analyze different time lengths to derive corresponding time-varying rules, thereby better targeting the changing trend due to the periodicity of the time period.

[0040] In a possible implementation, the step of determining the grouping of capacitors for reactive power compensation of each phase of the power distribution device based on reactive load data of each phase at the next time node includes: S41: Collect and count the fluctuation values ​​of reactive load data of each phase on the low-voltage side of the power distribution device and form a probability distribution function, and correct the reactive load data of each phase at the next time node based on the probability distribution function.

[0041] The difference between the predicted value and the actual detection value is statistically analyzed, and the difference data is used to correct subsequent predictions through the probability distribution function to further improve the prediction accuracy for the next time node.

[0042] In a possible implementation manner, after the step of determining the grouping of capacitors for reactive power compensation of each phase of the power distribution device based on the reactive load data of each phase at the next time node, the step further includes: S42: Perform abnormality detection on reactive load data of each phase on the low-voltage side of the power distribution device, and remove abnormal data segments.

[0043] By setting an abnormal detection algorithm to detect data collection, when an abnormal data segment is found, it is removed to avoid the abnormal data from interfering with the prediction. The judgment of abnormal data includes but is not limited to data interval judgment, abnormal detection algorithm and setting detection device.

[0044] In a possible implementation manner, the step of performing abnormality detection on reactive load data of each phase on the low-voltage side of the power distribution device and removing abnormal data segments includes: S43: After removing abnormal data, the corresponding data is completed by interpolating adjacent data.

[0045] After removing the abnormal data, the data can be supplemented by interpolating adjacent data to maintain the continued prediction of the next time node, thereby improving the applicability and stability of the method.

[0046] In order to solve the above technical problems, another technical solution adopted by the present application is: to provide a low-voltage reactive power compensation system for a power distribution device, which is applicable to the low-voltage reactive power compensation method for a power distribution device as mentioned above, comprising: A collection module is used to collect reactive load data of each phase on the low-voltage side of the power distribution device; An analysis module is used to process the reactive load data of each phase and predict the reactive load data of each phase at the next time node; The grouping module determines the group switching conditions of the reactive compensation capacitors of each phase of the distribution device based on the reactive load data of each phase at the next time node.

[0047] The embodiment of the present application uses the exponential smoothing method to predict future reactive load data based on existing detection data, thereby improving adaptability and predictability to grid load fluctuations, strengthening rapid response and reaction to emergency situations, and being more timely and accurate.

[0048] The above descriptions are merely embodiments of the present application and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A low voltage reactive power compensation method for a power distribution device, characterized in that: include: S10: Collect reactive load data of each phase on the low voltage side of the power distribution device; S20: constructing a time series model of reactive load of each phase on the low-voltage side of the power distribution device based on the reactive load data of each phase at a number of time nodes; S30: Based on the time series model, predict the reactive load data of each phase at the next time node by exponential smoothing; S40: Based on the reactive load data of each phase at the next time node, determine the group switching status of the capacitors for reactive compensation of each phase of the power distribution device.

2. A low voltage reactive power compensation method for a power distribution device as claimed in claim 1, characterized in that: The step of predicting reactive load data of each phase at the next time by exponential smoothing based on the time series model comprises: S31: Obtain reactive load data of each phase at the next time node through an exponential smoothing prediction, as shown in the following formula: Where S t+1 is the predicted value of the next time node, S t is the reactive load data of each phase at the tth time node, P t is the smoothing value at the tth time node, and a is the weighting coefficient.

3. A low voltage reactive power compensation method for a power distribution device as claimed in claim 2, characterized in that: After the step of obtaining the reactive load data of each phase at the next time node through a first exponential smoothing prediction, the method further comprises: S32: When the deviation of the prediction value of the next time node exceeds a first threshold, the prediction value of the next time node is corrected by quadratic exponential smoothing prediction, a corrected prediction value of the next time node is obtained, and reactive load data of each phase of the next time node is corrected.

4. A low voltage reactive power compensation method for a power distribution device as claimed in claim 3, characterized in that: After the step of correcting the prediction value of the next time node by using a secondary exponential smoothing prediction when the deviation of the prediction value of the next time node exceeds the first threshold, the method includes: S33: When the deviation of the first-corrected prediction value of the next time node exceeds the second threshold, the first-corrected prediction value of the next time node is corrected by triple exponential smoothing prediction, the second-corrected prediction value of the next time node is obtained, and the reactive load data of each phase of the next time node is corrected.

5. A low voltage reactive power compensation method for a power distribution device as claimed in claim 2, characterized in that: After the step of determining the grouping of capacitors for reactive power compensation of each phase of the power distribution device based on the reactive load data of each phase at the next time node, the method further comprises: S41: Collect the actual value of reactive load data of each phase at the next time node and compare it with the predicted value, and correct the weighting coefficient based on the error.

6. A low voltage reactive power compensation method for a power distribution device as claimed in claim 1, characterized in that: The step of collecting reactive load data of each phase on the low voltage side of the power distribution device comprises: S11: Use edge computing technology to collect reactive load data of each phase on the low-voltage side of the power distribution device.

7. A low voltage reactive power compensation method for a power distribution device as claimed in claim 1, characterized in that: The total length of time of the several time nodes includes days, weeks, months or quarters.

8. A low voltage reactive power compensation method for a power distribution device as claimed in claim 1, characterized in that: After the step of determining the grouping of capacitors for reactive power compensation of each phase of the power distribution device based on the reactive load data of each phase at the next time node, the method further includes: S42: Perform abnormality detection on reactive load data of each phase on the low-voltage side of the power distribution device, and remove abnormal data segments.

9. A low voltage reactive power compensation method for a power distribution device as claimed in claim 8, characterized in that: The step of detecting abnormalities in the reactive load data of each phase on the low-voltage side of the power distribution device and removing abnormal data segments comprises: S43: After removing the abnormal data, the corresponding data is completed by interpolating adjacent data.

10. A low-voltage reactive power compensation system for a power distribution device, applicable to a low-voltage reactive power compensation method for a power distribution device as claimed in any one of claims 1 to 9, characterized in that: include: A collection module, used to collect reactive load data of each phase on the low-voltage side of the power distribution device; An analysis module, used for processing the reactive load data of each phase and predicting the reactive load data of each phase at the next time node; The grouping module determines the group switching conditions of the reactive compensation capacitors of each phase of the power distribution device based on the reactive load data of each phase at the next time node.

Images