Load flexible adjustment method, device and system for smart energy unit

By constructing a short-term asynchronous feature matrix and exponential smoothing algorithm, the user's time period characteristics and load adjustment coefficient are determined, and the problem of deviation in the power consumption demand side analysis in traditional methods is solved, and the accuracy and effect of load flexibility adjustment of smart energy units is improved.

CN118630767BActive Publication Date: 2024-12-06JIANGSU ZHIGE HI TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411080938.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-08
Publication Date
2024-12-06
Estimated Expiration
2044-08-08

AI Technical Summary

Technical Problem

When existing smart energy units deal with complex electric use scenarios, traditional data monitoring and prediction methods lead to deviations in the power demand side analysis, reducing the accuracy and effect of flexible load regulation.

Method used

By obtaining the real-time power consumption data of the user, a short-term asynchronous feature matrix is ​​constructed, the user's period characteristic coefficient and load period adjustment coefficient are determined, and load prediction is carried out in combination with the exponential smoothing algorithm to achieve flexible adjustment of smart energy units.

Benefits of technology

It improves the accuracy of load analysis, reduces data analysis errors, enhances the ability to cope with complex interference factors, and improves the accuracy of flexible load regulation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118630767B_ABST
    Figure CN118630767B_ABST
Patent Text Reader

Abstract

The present application relates to the field of intelligent electricity consumption technology, and specifically to a method, device and system for flexible load regulation of a smart energy unit, the method comprising: obtaining real-time electricity consumption data at the user end; determining the user's periodic characteristic coefficient based on the time-varying differences in the electricity consumption data at the user end; determining the load periodic adjustment coefficient based on the user's periodic characteristic coefficient and the change in the electricity load at the user end; predicting the actual electricity load at the user end according to the load periodic adjustment coefficient, and flexibly regulating the smart energy unit at the user end in combination with the electricity load prediction result at the user end. The present application takes into account the influence of interference factors of different users on the electricity consumption data at the user end, reduces the error of the smart energy unit in analyzing the user end data, and improves the accuracy of load regulation of the smart energy unit.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of intelligent electricity utilization technology, and specifically to a load flexible regulation method, device and system for a smart energy unit. Background Art

[0002] The flexible load regulation of smart energy units refers to the process of real-time monitoring, analysis and optimization of the load in the power system using a smart energy management system. The goal of this regulation is to achieve efficient energy utilization and demand-side management while ensuring the stable operation of the power system. Nowadays, the load regulation of smart units mainly includes real-time monitoring, load forecasting, demand response, optimized scheduling, remote control, user interaction and energy efficiency management. Among them, real-time monitoring and load forecasting are used to provide basic data for load regulation to smart energy units, thereby completing the flexible load regulation of smart energy units and improving energy utilization efficiency and the stability of power system operation.

[0003] Nowadays, with the expansion of the energy market, peak loads are rising, and the risks of periodic and local supply and demand gaps are prominent. Therefore, in response to complex electricity consumption scenarios, the accuracy of analysis of electricity consumption data collected on the electricity demand side and real-time monitoring and prediction is greatly affected. The traditional data monitoring and prediction methods are used to analyze data of electricity consumption change scenarios with characteristics such as periodicity. For example, the traditional time series prediction method may lead to large deviations in the electricity consumption status analysis and load demand analysis on the electricity demand side when dealing with changeable electricity consumption scenarios, thereby reducing the accuracy of the load analysis of the smart energy unit, and thus reducing the load flexibility regulation effect of the smart energy unit. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of this application is to provide a load flexible adjustment method, device and system for a smart energy unit. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present application provides a load flexible adjustment method for a smart energy unit, the method comprising the following steps:

[0006] Obtain real-time power consumption data from the user end;

[0007] Determine the user's time characteristic coefficient based on the time-varying differences of each user's electricity consumption data;

[0008] Determine the load period adjustment coefficient based on the user's period characteristic coefficient and the change of the user's end power load;

[0009] The power load at the user end is predicted according to the load period adjustment coefficient, and the smart energy unit at the user end is flexibly adjusted based on the power load prediction results at the user end.

[0010] Furthermore, the process of determining the user's time period characteristic coefficient is as follows:

[0011] Constructing various short-term asynchronous feature matrices based on various power consumption data collected at different times from the user end;

[0012] Analyze the relationship between each group of modal components of any two rows of elements in the short-term asynchronous characteristic matrix, and determine the first asynchronous characteristic coefficients of each group of modal components of the any two rows of elements;

[0013] The trend change difference between any two rows of elements in the short-term asynchronous characteristic matrix is ​​used as the second asynchronous characteristic coefficient of the any two rows of elements;

[0014] The user time period characteristic coefficient between any two rows of elements is determined based on the first asynchronous characteristic coefficient and the second asynchronous characteristic coefficient, wherein the user time period characteristic coefficient between any two rows of elements is negatively correlated with the mean of the first asynchronous characteristic coefficients of all group modal components of the any two rows of elements, and is positively correlated with the second asynchronous characteristic coefficient.

[0015] Furthermore, the construction process of the short-term asynchronous feature matrix is ​​as follows:

[0016] Arrange all collected data of each type of electricity consumption data in ascending order according to the collection time to form each electricity consumption data sequence, form all electricity consumption data sequences into a matrix and perform standardization processing to obtain an electricity consumption data analysis matrix;

[0017] The mean value of each column of data in the power consumption data analysis matrix is ​​used as the superposition feature data at the corresponding collection time, and all the superposition feature data corresponding to the power consumption data analysis matrix are combined into a superposition feature data sequence in ascending time order;

[0018] The sequence segmentation algorithm is used to divide the superimposed feature data sequence into subsequences, and the column vectors corresponding to the power consumption data analysis matrix of all superimposed feature data in each subsequence form a short-term asynchronous feature matrix.

[0019] Furthermore, the process of determining the first asynchronous characteristic coefficient is as follows:

[0020] Multiple modal decompositions are performed on each row element in the short-term asynchronous characteristic matrix. Each row element corresponds to multiple modal components. The modal components of any two row elements at the corresponding decomposition times form a group of modal components.

[0021] The correlation coefficients of each group of modal components of any two rows of elements in the short-term asynchronous characteristic matrix are analyzed and recorded as the first asynchronous characteristic coefficients of each group of modal components of any two rows of elements.

[0022] Furthermore, the process of determining the load period adjustment coefficient is as follows:

[0023] The user period feature coefficients between all two row elements in each short-term asynchronous feature matrix are used to form a short-term asynchronous period feature sequence of each short-term asynchronous feature matrix;

[0024] The mean of the periodic characteristic coefficients of all users in the short-term asynchronous feature sequence is taken as the first periodic contrast coefficient of each short-term asynchronous feature matrix;

[0025] The mean of the difference between the short-term asynchronous feature sequence and other short-term asynchronous feature sequences is used as the second period contrast coefficient of each short-term asynchronous feature matrix;

[0026] The load periodic adjustment coefficient of each short-term asynchronous characteristic matrix is ​​determined based on the first periodic comparison coefficient and the second periodic comparison coefficient.

[0027] Furthermore, the load periodic adjustment coefficient of each short-term asynchronous characteristic matrix is ​​a forward fusion of a first periodic comparison coefficient and a second periodic comparison coefficient of each short-term asynchronous characteristic matrix.

[0028] Furthermore, the process of predicting the power load at the user end is as follows:

[0029] Normalize the load period adjustment coefficients of all short-term asynchronous characteristic matrices, and obtain the average of the normalized results of all load period adjustment coefficients, which is recorded as the smoothing reference coefficient;

[0030] The weight parameter of the exponential smoothing algorithm is determined based on the smoothing reference coefficient, and the load forecasting result is obtained by combining the exponential smoothing algorithm with the power data collected by the user end.

[0031] Furthermore, the calculation formulas of the weight parameters of the exponential smoothing algorithm are respectively: , , , where , and are the three weight parameters of the exponential smoothing algorithm, is the smoothing reference coefficient.

[0032] In a second aspect, an embodiment of the present application provides a load flexibility adjustment device for a smart energy unit, wherein the load flexibility adjustment device comprises:

[0033] The load flexible regulation device includes: an energy storage data acquisition module, a data feature analysis module, and a load regulation distribution module.

[0034] Energy storage data acquisition module, used to obtain real-time power consumption data at the user end;

[0035] The data characteristic analysis module is used to determine the user's periodic characteristic coefficient based on the periodic variation difference of each power consumption data of the user end; determine the load periodic adjustment coefficient based on the user's periodic characteristic coefficient and the change of the power load at the user end;

[0036] The load regulation and allocation module is used to predict the actual power load at the user end according to the load period adjustment coefficient, and flexibly adjust the smart energy unit at the user end based on the power load prediction results at the user end.

[0037] In a third aspect, an embodiment of the present application further provides a load flexible regulation system for a smart energy unit, the system comprising a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the steps of any one of the above methods when executing the computer program.

[0038] It can be seen from the above embodiments that the load flexible adjustment method, device and system of the smart energy unit provided in the embodiments of the present application have at least the following beneficial effects:

[0039] Because when the smart energy unit is applied to the client, the collected electricity consumption data may be affected by the user's electricity consumption behavior habits and differences in electricity facilities. Therefore, the present application conducts long-term and short-term asynchronous deviation feature analysis on the user's electricity consumption data, optimizes and adjusts the user's load forecast based on the analysis results, improves the accuracy of the electricity consumption data feature analysis when the smart energy unit is used on the user side to deal with complex interference, and thus improves the accuracy of the smart energy unit's flexible adjustment of the user-side load. Compared with the traditional smart energy unit's flexible adjustment process, its beneficial effect is that it takes into account the influence of interference factors of different users on the user's electricity consumption data, reduces the error of the smart energy unit's analysis of the user-side data, and thus completes the precise adjustment of the load by the smart energy unit. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or 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 ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0041] Figure 1 A flowchart of the steps of a load flexible adjustment method of a smart energy unit provided in one embodiment of the present application;

[0042] Figure 2 Construct a flow chart for the user's time period characteristic coefficient;

[0043] Figure 3 Construct a flow diagram for load period adjustment coefficient;

[0044] Figure 4 It is a schematic diagram of load change after load adjustment;

[0045] Figure 5 A schematic diagram of the structure of a load flexible adjustment device for a smart energy unit provided in one embodiment of the present application. DETAILED DESCRIPTION

[0046] In order to further explain the technical means and effects adopted by this application to achieve the predetermined invention purpose, the following is a detailed description of the load flexible adjustment method, device and system of the smart energy unit proposed in this application, its specific implementation method, structure, characteristics and effects in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0047] Unless otherwise specified and limited, terms such as "comprises", "comprising" or any other variants thereof are intended to cover non-exclusive inclusion, so that a circuit structure, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such articles or devices. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the article or device including the element. In addition, the term "and\or" used herein includes any and all combinations of one or more related listed items. All technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of this application.

[0048] The specific scheme of the load flexible adjustment method, device and system of the smart energy unit provided in this application is described in detail below with reference to the accompanying drawings.

[0049] See also Figure 1 , which shows a flow chart of the steps of a load flexible adjustment method of a smart energy unit provided by an embodiment of the present application, the method comprising the following steps:

[0050] Step 1: Obtain real-time power consumption data from the user end.

[0051] The smart energy unit integrates advanced information and communication technologies, analyzes power consumption status and load changes by collecting power consumption data from the user in real time, and implements flexible adjustment based on the analysis results. The smart energy unit collects the user's power consumption data in real time through intelligent sensors and metering devices installed at the user's end. The power consumption data includes electricity data, power data and temperature data. The implementers of the three types of power consumption data collection can select them according to the actual usage scenario requirements. In one embodiment of the present application, the devices for collecting electricity data, power data and temperature data may be electricity collectors, power sensors and temperature sensors.

[0052] Furthermore, the smart energy unit is provided with an energy storage data acquisition module, a data feature analysis module and a load regulation and distribution module, wherein the collected user's electricity consumption data is transmitted to the data feature analysis module for analysis. Specifically, the acquired electricity data, power data and temperature data are sequenced in ascending order of collection time as electricity data sequence, power data sequence and temperature data sequence, respectively.

[0053] At this point, the real-time electricity consumption data of the user end has been obtained, including the electricity data series, power data series and temperature data series.

[0054] Step 2: Determine the user's time-period characteristic coefficient based on the time-period variation difference of each user's electricity consumption data.

[0055] Using smart energy units to flexibly adjust the load from the user side requires dealing with the complex characteristics of electricity consumption data changes caused by differences in electricity consumption behavior habits and electricity consumption scenario facilities at different users. Therefore, the traditional data analysis method has a large error in feature analysis of user-side electricity consumption data. It is necessary to further adjust the user's electricity consumption status and load change analysis based on the actual change characteristics of the user's electricity consumption data to improve the effect of flexible adjustment at the user side through smart energy units.

[0056] Specifically, since the electricity consumption data collected at the user end is affected by the use of electricity facilities and electricity consumption habits of different users, the user's electricity consumption data may show greatly different short-term or long-term trends and seasonal changes, and may even show irregular trend change characteristics. Therefore, when the smart energy unit is applied to the user end, the analysis of the characteristics of the user's electricity consumption data is optimized and adjusted by analyzing the deviations that may be caused by the above-mentioned characteristic changes in the user's electricity consumption data; under normal circumstances, the electricity data, power data and temperature data at the user end show synchronous change characteristics in short-term or long-term changes, and with the interference of influencing factors such as changes in the user's electricity consumption behavior, the user's electricity consumption data may show short-term or long-term asynchronous change differences.

[0057] The first small step is to build short-term asynchronous feature matrices based on the power consumption data of the user end collected at different times.

[0058] Based on the above analysis, the short-term or long-term relative characteristic differences between different electricity consumption data are analyzed by constructing an electricity consumption data analysis matrix. The construction process of the electricity consumption data analysis matrix is: each electricity consumption data sequence collected by the user end is taken as a row element in the matrix, and the standardized processing result of the matrix composed of all electricity consumption data sequences is taken as the electricity consumption data analysis matrix; wherein the electricity consumption data sequence includes electricity data sequence, power data sequence and temperature data sequence.

[0059] Generally, different electricity consumption data have synchronous change characteristics. After the different electricity consumption data are superimposed and summed, there may be significant complex change differences in the period of asynchronous change characteristics. Based on this, the electricity consumption data analysis matrix is ​​superimposed and mapped, and the data is divided based on the superimposed mapping results to analyze the short-term asynchronous change characteristic differences according to the division results. The specific division process is as follows:

[0060] Each column of data in the power consumption data analysis matrix corresponds to the electricity data, power data and temperature data collected at the same time. The mean of each column of data in the power consumption data analysis matrix is ​​used as the superimposed feature data at the corresponding collection time. The sequence composed of all the superimposed feature data corresponding to the power consumption data analysis matrix in ascending time order is used as the superimposed feature data sequence. A sequence segmentation algorithm is used to obtain the division result of the superimposed feature data sequence. The sequence composed of each group of data in the division result is used as a subsequence of the superimposed feature data sequence. Since each element in the superimposed feature data sequence corresponds to a column of data in the application point data analysis matrix, the column vectors corresponding to all the superimposed feature data in each subsequence in the power consumption data analysis matrix constitute a short-term asynchronous feature matrix. The short-term asynchronous feature matrix divides the power consumption data that may have short-term asynchronous characteristics into a group of data for analysis.

[0061] It should be noted that, in the present embodiment, in the division result of the superimposed feature data sequence obtained by using a sequence segmentation algorithm, the sequence segmentation algorithm may be a Variable-length Subsequence Clustering algorithm. As other embodiments, on the basis of achieving the above-mentioned sequence division purpose, the implementer may use other sequence segmentation algorithms for processing.

[0062] The second small step: analyzing the relationship between the modal components of each group of elements in any two rows in the short-term asynchronous characteristic matrix, and determining the first asynchronous characteristic coefficients of the modal components of each group of elements in any two rows.

[0063] Furthermore, the differences in the periodic asynchronous change characteristics between different data in each short-term asynchronous feature matrix are analyzed to calculate the user periodic characteristic coefficient. The user periodic characteristic coefficient reflects the degree to which the short-term electricity consumption data characteristics of the user-side electricity consumption data are affected by the user-side interference factors.

[0064] For each row of elements in the short-term asynchronous feature matrix, the sequence of elements in each row is taken as input, and the empirical mode decomposition algorithm is used to obtain the (The value in this embodiment is 5) modal components and 1 residual component, that is, five modal decomposition processes are performed in this embodiment; the elements of each row are The modal components are sorted in the order of decomposition to correspond to the modal components between elements in different rows. For example, for any two rows of elements, the modal components obtained by any two elements in the same decomposition order are grouped into a group of modal components, which are corresponded in the decomposition order in turn.

[0065] For any two rows of elements in the short-term asynchronous feature matrix, the sequence composed of the elements in one row is used as the first comparison sequence, and the sequence composed of the elements in the other row is used as the second comparison sequence. The calculation result of the correlation coefficient of each group of modal components corresponding to the first comparison sequence and the second comparison sequence is used as the first asynchronous feature coefficient of each group of modal components corresponding to the first comparison sequence and the second comparison sequence. In this embodiment, the correlation coefficient is the Spearman correlation coefficient. The implementer may also use other correlation coefficient calculation methods, which are determined by the implementer according to the actual application scenario;

[0066] It should be understood that the smaller the first asynchronous characteristic coefficient is, the lower the correlation change degree of short-term data between different power consumption data is, and the greater the possibility of presenting local asynchronous characteristics;

[0067] The third small step: take the trend change difference between any two rows of elements in the short-term asynchronous feature matrix as the second asynchronous feature coefficient of the any two rows of elements.

[0068] Specifically, in this embodiment, the residual components of the first comparison sequence and the second comparison sequence are respectively used as inputs. In this embodiment, the Mann-Kendall trend check algorithm is used to obtain the trend statistics of the residual components of the first comparison sequence and the second comparison sequence respectively, and the difference between the trend statistics corresponding to the first comparison sequence and the second comparison sequence is used as the second asynchronous characteristic coefficient. In this embodiment, the difference in trend statistics can be calculated by taking the absolute value of the difference between the trend statistics corresponding to the two sequences;

[0069] It should be understood that the larger the second asynchronous characteristic coefficient is, the more significant the asynchronous deviation characteristics of short-term data between different electricity consumption data are, which indicates that the difference analysis of the overall trend characteristics expressed by the residual components of each comparison sequence is more significant.

[0070] The fourth sub-step: determining the user time period characteristic coefficient between the elements of any two rows based on the first asynchronous characteristic coefficient and the second asynchronous characteristic coefficient.

[0071] The user period characteristic coefficient between any two row elements in the short-term asynchronous characteristic matrix consists of two parts: the first asynchronous characteristic coefficient and the second asynchronous characteristic coefficient, wherein the user period characteristic coefficient is negatively correlated with the first asynchronous characteristic coefficient, and the second asynchronous characteristic coefficient is positively correlated.

[0072] It should be noted that negative correlation is used to characterize that the changing trends between variables are opposite, and the changes in the independent variable and the dependent variable show an opposite relationship, which can be a negative correlation such as a ratio relationship and an inverse relationship; positive correlation means that the changing trends between the variables are the same, and the dependent variable increases (decreases) as the independent variable increases (decreases), which can be a positive proportional relationship, a multiplication relationship, etc., which depends on the actual application scenario.

[0073] Specifically, in one embodiment of the present application, the mean of all first asynchronous characteristic coefficients corresponding to the first comparison sequence and the second comparison sequence is calculated, recorded as the first asynchronous mean, and the ratio of the second asynchronous characteristic coefficient between the first comparison sequence and the second comparison sequence to the first asynchronous mean is used as the user period characteristic coefficient between any two row elements in the short-term asynchronous characteristic matrix.

[0074] It should be understood that the larger the user's time period characteristic coefficient is, the greater the short-term asynchronous deviation of different power consumption data at the user end is.

[0075] See also Figure 2 , the user time period characteristic coefficient construction process diagram is as follows Figure 2 shown.

[0076] Step 3: Determine the load period adjustment coefficient based on the user's period characteristic coefficient and the changes in the user's electricity load.

[0077] Through the above calculation, the user period characteristic coefficients between any two rows of elements in the short-term asynchronous feature matrix are obtained. Further, the user period characteristic coefficients between all any two rows of elements in each short-term asynchronous feature matrix are combined to form the short-term asynchronous period characteristic sequence of each short-term asynchronous feature matrix.

[0078] Since each short-term asynchronous characteristic matrix is ​​the electricity consumption data of the user in different periods, the load period adjustment coefficient of each short-term asynchronous characteristic matrix is ​​constructed by comprehensively comparing the relative differences of asynchronous deviations in different periods. The load period adjustment coefficient reflects the relative asynchronous change differences of electricity consumption data in different periods at the user end. The specific analysis process is as follows:

[0079] For the short-term asynchronous feature sequence of each short-term asynchronous feature matrix, the mean of all elements in the short-term asynchronous feature sequence is taken as the first period contrast coefficient of each short-term asynchronous feature matrix.

[0080] It should be understood that the larger the first period comparison coefficient is, the larger the asynchronous deviation of the period data change of the electricity consumption data in the period is, according to the asynchronous characteristic analysis between different electricity consumption data in each period;

[0081] The difference between the short-term asynchronous feature sequence and the short-term asynchronous feature sequences of other short-term asynchronous feature matrices is taken as the average of all short-term asynchronous feature matrices as the second period comparison coefficient of each short-term asynchronous feature matrix. In this embodiment, the DTW distance is used to measure the difference between the short-term asynchronous feature sequences. As other implementation methods, the implementer can select other sequence distance measurement methods, and this application does not make special restrictions;

[0082] It should be understood that the larger the second time period contrast coefficient is, the larger the relative difference in the asynchronous deviation characteristics of the power consumption data of the user end in different time periods is;

[0083] The load period adjustment coefficient of each short-term asynchronous characteristic matrix is ​​composed of the first period contrast coefficient and the second period contrast coefficient. The load period adjustment coefficient is the positive fusion of the first period contrast coefficient and the second period contrast coefficient.

[0084] Specifically, the forward fusion is the same change trend between variables, and the forward fusion can adopt a positive correlation relationship such as addition and multiplication. In one embodiment of the present application, the product of the first period contrast coefficient and the second period contrast coefficient of each short-term asynchronous characteristic matrix is ​​used as the load period adjustment coefficient of each short-term asynchronous characteristic matrix; in other embodiments, the sum of the first period contrast coefficient and the second period contrast coefficient can be used as the load period adjustment coefficient of each short-term asynchronous characteristic matrix.

[0085] It can be understood that the larger the load period adjustment coefficient is, the greater the degree of difference in the irregular trend characteristics of the user-side electricity consumption data characteristics in this period will be affected by interference factors such as user behavior habits through the comparison of the characteristics of different electricity consumption data in each period and the difference analysis of asynchronous deviations in different periods.

[0086] At this point, the load period adjustment coefficient is obtained.

[0087] See also Figure 3 The schematic diagram of the construction process of the load period adjustment coefficient is as follows: Figure 3 shown.

[0088] Step 4: Forecast the power load at the user end according to the load period adjustment coefficient, and flexibly adjust the smart energy unit at the user end based on the power load forecast results at the user end.

[0089] Generally speaking, if the characteristics of user data in different time periods are greatly affected by the differences in user behavior habits and user power facilities at the user end, the error and processing efficiency of load forecasting analysis using existing forecasting methods are poor, and it is impossible to cope with the impact of interference factors on power consumption data at the user end; therefore, based on the long-term and short-term asynchronous deviation analysis of power consumption data collected at the user end, the Holt-Winters exponential smoothing algorithm is optimized and adjusted to improve the accuracy of load forecasting at the user end, thereby improving the effect of smart energy units for flexible load regulation at the user end. The specific adjustment process is as follows:

[0090] The set of load period adjustment coefficients of all short-term asynchronous characteristic matrices is used as the load characteristic analysis data set, the load characteristic analysis data set is used as input, the Z-score normalization algorithm is used to obtain the normalized results of all elements in the load characteristic analysis data set, the mean of the normalized results is calculated, and the mean is used as the smoothing reference coefficient ; If the long-term or short-term electricity consumption data used for load forecasting presents unstable and irregular change characteristics, the larger the calculated smoothing reference coefficient is, that is, through the comprehensive analysis of the user's electricity consumption, power and temperature, the change characteristics of the user's electricity consumption data used for load forecasting have a greater impact on the forecast results. Therefore, when using the Holt-Winters exponential smoothing algorithm for the user's electricity consumption data characteristics, a smaller weight parameter needs to be set. There are 3 weight parameters in the Holt-Winters exponential smoothing algorithm, which are respectively recorded as , and , and among them , .

[0091] Furthermore, the weight parameter is determined by analyzing the smoothing reference coefficient obtained by the user's power consumption data characteristics. , specific weight parameters The calculation relationship is: , where h is the smoothing reference coefficient.

[0092] Among them, the smoothing reference coefficient The larger the value, the more significant the short-term change characteristics of the user's electricity consumption data are due to the influence of the user's electricity consumption behavior, so a smaller weight parameter is set. , and obtain accurate prediction results.

[0093] The specific load forecasting process is as follows: the input is the power data sequence and the corresponding weight parameters , and , the Holt-Winters exponential smoothing algorithm is used to obtain the predicted power, and the predicted power is the load prediction result of the user end.

[0094] To facilitate analysis, the load forecast results can be normalized and input into the load regulation allocation module. The load regulation allocation module determines whether the regulation threshold is reached based on the received load forecast results. The implementation example of the regulation threshold is 0.75. If the regulation threshold is reached, the load regulation instruction of the current user end is issued. The smart energy unit issues a power control instruction to the device according to the load regulation instruction of the current user end, thereby limiting the operating power of the device to achieve the purpose of load regulation.

[0095] See also Figure 4 The schematic diagram of load change after the user uses the smart energy unit to adjust the load is as follows: Figure 4 As shown, the horizontal axis is time, in hours; the vertical axis is power, and a represents the curve before the user-side composite adjustment, and b represents the curve after the user-side load adjustment.

[0096] See also Figure 5 , Figure 5 Schematic diagram of the structure of the load flexibility adjustment device of the smart energy unit provided in the embodiment of the present application. In this embodiment, each unit included in the terminal is used to execute each step in the embodiment corresponding to the load flexibility adjustment method of the smart energy unit. Figure 5 The load flexible adjustment device includes: energy storage data acquisition module, data feature analysis module, and load adjustment distribution module.

[0097] Energy storage data acquisition module, used to obtain real-time power consumption data at the user end;

[0098] The data characteristic analysis module is used to determine the user's periodic characteristic coefficient based on the periodic variation difference of each power consumption data of the user end; determine the load periodic adjustment coefficient based on the user's periodic characteristic coefficient and the change of the power load at the user end;

[0099] The load regulation and allocation module is used to predict the actual power load at the user end according to the load period adjustment coefficient, and flexibly adjust the smart energy unit at the user end based on the power load prediction results at the user end.

[0100] Based on the same inventive concept as the above method, an embodiment of the present application also provides a load flexible regulation system for a smart energy unit, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, the steps of any one of the above-mentioned load flexible regulation methods for the smart energy unit are implemented.

[0101] It is to be understood that the sequence of the embodiments of the present application described above is for description only and does not represent the advantages and disadvantages of the embodiments. The above describes specific embodiments of the present specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0102] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0103] The above content is only an implementation method of the present application and is not intended to limit the 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 used in other related technical fields, are also included in the protection scope of the present application.

Claims

1. A load flexible adjustment method for a smart energy unit, characterized in that: The method comprises the following steps: Obtain real-time power consumption data from the user end; Determine the user's time characteristic coefficient based on the time-varying differences of each user's electricity consumption data; Determine the load period adjustment coefficient based on the user's period characteristic coefficient and the change of the user's end power load; The power load at the user end is predicted according to the load period adjustment coefficient, and the smart energy unit at the user end is flexibly adjusted based on the power load prediction results at the user end; The process of determining the user's time period characteristic coefficient is as follows: Constructing various short-term asynchronous feature matrices based on various power consumption data collected at different times from the user end; Analyze the relationship between each group of modal components of any two rows of elements in the short-term asynchronous characteristic matrix, and determine the first asynchronous characteristic coefficients of each group of modal components of the any two rows of elements; The trend change difference between any two rows of elements in the short-term asynchronous characteristic matrix is ​​used as the second asynchronous characteristic coefficient of the any two rows of elements; The user time period characteristic coefficient between any two rows of elements is determined based on the first asynchronous characteristic coefficient and the second asynchronous characteristic coefficient, wherein the user time period characteristic coefficient between any two rows of elements is negatively correlated with the mean of the first asynchronous characteristic coefficients of all group modal components of the any two rows of elements, and is positively correlated with the second asynchronous characteristic coefficient.

2. The load flexible adjustment method of the smart energy unit according to claim 1, characterized in that: The construction process of the short-term asynchronous feature matrix is ​​as follows: Arrange all collected data of each type of electricity consumption data in ascending order according to the collection time to form each electricity consumption data sequence, form all electricity consumption data sequences into a matrix and perform standardization processing to obtain an electricity consumption data analysis matrix; The mean value of each column of data in the power consumption data analysis matrix is ​​used as the superposition feature data at the corresponding collection time, and all the superposition feature data corresponding to the power consumption data analysis matrix are combined into a superposition feature data sequence in ascending time order; The sequence segmentation algorithm is used to divide the superimposed feature data sequence into subsequences, and the column vectors corresponding to the power consumption data analysis matrix of all superimposed feature data in each subsequence form a short-term asynchronous feature matrix.

3. The load flexible adjustment method of the smart energy unit according to claim 1, characterized in that: The process of determining the first asynchronous characteristic coefficient is: Multiple modal decompositions are performed on each row element in the short-term asynchronous characteristic matrix. Each row element corresponds to multiple modal components. The modal components of any two row elements at the corresponding decomposition times form a group of modal components. The correlation coefficients of each group of modal components of any two rows of elements in the short-term asynchronous characteristic matrix are analyzed and recorded as the first asynchronous characteristic coefficients of each group of modal components of any two rows of elements.

4. The load flexible adjustment method of the smart energy unit according to claim 3, characterized in that: The process of determining the load period adjustment coefficient is as follows: The user period feature coefficients between all two row elements in each short-term asynchronous feature matrix are used to form a short-term asynchronous period feature sequence of each short-term asynchronous feature matrix; The mean of the periodic characteristic coefficients of all users in the short-term asynchronous feature sequence is taken as the first periodic contrast coefficient of each short-term asynchronous feature matrix; The mean of the difference between the short-term asynchronous feature sequence and other short-term asynchronous feature sequences is used as the second period contrast coefficient of each short-term asynchronous feature matrix; The load periodic adjustment coefficient of each short-term asynchronous characteristic matrix is ​​determined based on the first periodic comparison coefficient and the second periodic comparison coefficient.

5. The load flexible adjustment method of the smart energy unit according to claim 4, characterized in that: The load periodic adjustment coefficient of each short-term asynchronous characteristic matrix is ​​a forward fusion of the first periodic comparison coefficient and the second periodic comparison coefficient of each short-term asynchronous characteristic matrix.

6. The load flexible adjustment method of the smart energy unit according to claim 5, characterized in that: The process of predicting the power load at the user end is as follows: Normalize the load period adjustment coefficients of all short-term asynchronous characteristic matrices, and obtain the average of the normalized results of all load period adjustment coefficients, which is recorded as the smoothing reference coefficient; The weight parameter of the exponential smoothing algorithm is determined based on the smoothing reference coefficient, and the load forecasting result is obtained by combining the exponential smoothing algorithm with the power data collected by the user end.

7. The load flexible adjustment method of the smart energy unit according to claim 6, characterized in that: The calculation formulas for the weight parameters of the exponential smoothing algorithm are: , , , where , and are the three weight parameters of the exponential smoothing algorithm, is the smoothing reference coefficient.

8. A load flexible adjustment device for a smart energy unit, implementing the method according to claim 1, characterized in that: The load flexibility adjustment device comprises: Energy storage data acquisition module, used to obtain real-time power consumption data at the user end; The data characteristic analysis module is used to determine the user's periodic characteristic coefficient based on the periodic variation difference of each power consumption data of the user end; determine the load periodic adjustment coefficient based on the user's periodic characteristic coefficient and the change of the power load at the user end; The load regulation and allocation module is used to predict the power load at the user end according to the load period adjustment coefficient, and flexibly adjust the smart energy unit at the user end based on the power load prediction results at the user end.

9. A load flexible regulation system for a smart energy unit, comprising a memory, a processor, and a computer program stored in the memory and running 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.