Power grid short-term load prediction method and system, computer equipment and storage medium

Abstract

A power grid short-term load prediction method comprises the following steps: determining a prediction target time period; and a pre-trained prediction model is utilized to predict the short-term load of the power grid in the prediction target time period, and the prediction model is # imgabs0 #. The invention provides a short-term load prediction method and system for the power grid, computer equipment and a storage medium. The influence of the long-term trend, seasonal factors, special events such as holidays and festivals of the power grid load, weather changes and the like on the power grid load is comprehensively considered, and a load prediction model is established to accurately predict the future load of the power grid.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid management, and in particular to a method, system, computer equipment and storage medium for short-term load forecasting of a power grid. Background Art

[0002] At present, the installed capacity of non-fossil energy power generation in my country's power system is gradually increasing. The continuous increase in the proportion of new energy power sources has reduced the dependence on fossil energy, but it has also brought new problems. The main problem is that the intermittent, random and volatile characteristics of new energy power sources make the regulation of the power grid system more difficult, and will also quickly consume the flexible regulation resources of the power system. In particular, for some large-scale new energy bases with weak grids and lack of synchronous power support, there is a general problem of insufficient system support capacity, which affects the safe and reliable transmission of new energy. In addition, although the overall utilization rate of new energy in the country has remained at a high level in recent years, the absorption foundation is not yet solid, and the problem of wind and solar abandonment in some areas and at some times is still prominent. In the future, the large-scale and high-proportion development of new energy requires a rapid improvement in the system regulation capacity, but the construction of regulatory resources faces many constraints, and the risk of efficient absorption of regional new energy increases, which restricts the efficient use of new energy. The "double high" characteristics of high-proportion renewable energy and high-proportion power electronic equipment are becoming increasingly prominent, and safe and stable operation faces great risks and challenges. Compared with the traditional power system dominated by synchronous generators, the "double high" power system has obvious characteristics such as low inertia, low damping, and weak voltage support. In addition, my country's power grid presents a grid form with strong coupling between AC and DC transmission and reception ends and complex voltage levels. It is difficult to coordinate between the transmission and reception end grids and between high and low voltage grids, and a chain reaction is easily triggered after a fault. These challenges will have a profound impact on the dispatching and operation of the power system.

[0003] Among them, power load forecasting is an important topic in the power dispatching and operation process of the power system. The power forecasting method that accurately predicts and responds quickly can greatly improve the safety, stability and economy of the power system operation. In the prior art, the power grid load forecasting is mainly carried out periodically through historical data. However, because the power grid load is affected by many factors, such as power consumption scenarios, meteorological factors, etc., for large power consumption areas, due to the large number of power consumption scenarios and the possibility of multiple meteorological factors in the area, the power grid load forecasting of the entire power consumption area is affected. First, the meteorological conditions in large power consumption areas are complex and changeable, and there may be significant differences in different meteorological modes, resulting in the meteorological condition input of the power grid load model cannot take into account various meteorological conditions. Secondly, the power consumption scenarios involved in large power consumption areas are diverse, such as industrial power consumption, residential power consumption and commercial power consumption. The changing trends of different power consumption scenarios are inconsistent, and the power grid load model is difficult to construct, which ultimately affects the accuracy of power load forecasting in large power consumption areas. These complex factors lead to limited accuracy of the prediction results of existing prediction methods. Summary of the invention

[0004] In order to address the deficiencies in the prior art, the present invention provides a method, system, computer equipment and storage medium for short-term load forecasting of a power grid, which comprehensively considers the impact of long-term trends in power grid load, seasonal factors, special events such as holidays and weather changes on power grid load, and establishes a load forecasting model to accurately predict the future load of the power grid.

[0005] In order to achieve the above object, the specific scheme adopted by the present invention is:

[0006] A method for short-term load forecasting of a power grid comprises the following steps: Determine the forecast target period; The pre-trained prediction model is used to predict the short-term load of the power grid in the prediction target period. The prediction model is: Among them, α is the weight coefficient, L 近期 is the recent load of the power grid, L 去年,同日 is the grid load corresponding to the forecast target period, G is the average annual growth rate of load, a T and a H are the temperature influence coefficient and humidity influence coefficient respectively, T and H are the reference temperature and reference humidity of the forecast target period respectively, c is the holiday influence coefficient, D is the holiday factor, e is the weekend influence coefficient, E is the weekend factor, f is the influence coefficient when holidays and weekends overlap, F is the holiday and weekend overlap factor, ΔL 大工业 It is the expected load change reported in advance by large industrial users.

[0007] As a further optimization of the above-mentioned short-term load forecasting method for a power grid: the method for training the forecasting model includes: obtaining the expected load change of large industrial users and comprehensive historical data within a preset historical period, the comprehensive historical data including load historical data and meteorological historical data; Screening the comprehensive historical data to obtain the associated basic load data and basic meteorological data; Generate basic load curve after clustering basic load data; Dynamically obtain the elastic load data of the power grid, and calculate the elastic load average value and elastic load time index within a preset sample period based on the elastic load data; A sample set is formed based on the expected load change, basic load curve, elastic load average value over the same period and elastic load time index; the prediction model is iteratively trained using the sample set.

[0008] As a further optimization of the above-mentioned short-term load forecasting method for power grid, a method for screening comprehensive historical data to obtain associated basic load data and basic meteorological data includes: Extracting multiple load segments from the load history data and multiple meteorological segments from the meteorological history data, wherein the duration of the load segments and the meteorological segments both reach a preset duration threshold; Calculate the time similarity between the load segment and the meteorological segment, and if the time similarity reaches a preset similarity threshold, the meteorological segment and the load segment are recorded as being associated; All related load segments and meteorological segments are collated to generate basic load data and basic meteorological data.

[0009] As a further optimization of the above-mentioned short-term load forecasting method for power grid, the method of clustering basic load data includes: Determine the load level corresponding to the load segment and determine the meteorological classification corresponding to the meteorological segment; All load segments of the same load level are grouped together.

[0010] As a further optimization of the above-mentioned short-term load forecasting method for power grid, a method for calculating the elastic load average value and the elastic load time index within a preset sample period based on the elastic load data includes: Segmenting the dynamically acquired elastic load data based on a preset sample period; Calculate the contemporaneous average of the elastic load data within each sample period; Calculate the average total value based on all the average values ​​of the same period; The ratio of the average value during the same period to the average total value is calculated as the elastic load time index.

[0011] As a further optimization of the above-mentioned short-term load forecasting method for power grid, the comprehensive historical data within a preset historical period is directly obtained from the automatic dispatching system of the power grid.

[0012] As a further optimization of the above-mentioned short-term load forecasting method for power grid: the elastic load data includes electric vehicle charging and discharging load data and energy storage system load data.

[0013] The power grid short-term load forecasting system includes: A data acquisition device for obtaining the estimated load change of large industrial users and comprehensive historical data within a preset historical period; a data management device for forming a sample set; A prediction device, used to predict the short-term load of the power grid in the prediction target period by using a pre-trained prediction model;

[0014] The data output device is used to output the prediction result of the prediction device.

[0015] Computer equipment, including: a memory storing a computer program; The processor is used to read and execute the computer program to implement the above-mentioned method for short-term load forecasting of a power grid.

[0016] The storage medium is used to store a computer program, which implements the above-mentioned short-term load forecasting method for a power grid when executed.

[0017] Beneficial effects: The present invention comprehensively considers the impact of long-term trends in power grid load, seasonal factors, special events such as holidays, and weather changes on power grid load, establishes a load forecasting model, and optimizes the model parameters through data. The trained model can accurately predict the future load of the power grid based on the historical power grid load and the recent average load, thereby improving the accuracy of power grid load forecasting and providing better guidance for dispatchers to carry out real-time balanced dispatching of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a flow chart of the method of the present invention; Figure 2 It is a diagram of experimental results in a specific implementation manner. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] like Figure 1 As shown, the present invention first provides a method for short-term load forecasting of a power grid, including S1 to S2.

[0021] S1. Determine the forecast target period. The length of the target period can be determined according to actual needs, for example, it can be determined to be one day, one week, or multiple days.

[0022] S2. Use the pre-trained prediction model to predict the short-term load of the power grid in the prediction target period. The prediction model is: Among them, α is the weight coefficient, L 近期 is the recent load of the power grid, L 去年,同日 is the grid load corresponding to the forecast target period, G is the average annual growth rate of load, a T and a Hare the temperature influence coefficient and humidity influence coefficient respectively, T and H are the reference temperature and reference humidity of the forecast target period respectively, c is the holiday influence coefficient, D is the holiday factor, e is the weekend influence coefficient, E is the weekend factor, f is the influence coefficient when holidays and weekends overlap, F is the holiday and weekend overlap factor, ΔL 大工业 It is the expected load change reported in advance by large industrial users. This value is determined according to the expected activities or operational changes of large industrial electricity users and is used to adjust the overall load forecast in real time. Large industrial electricity users are electricity users whose electricity load accounts for more than 3%.

[0023] More specifically, for the holiday factor D, when a certain day is a holiday, the value of D is 1, otherwise the value is 0, and holidays here refer to statutory holidays; for the weekend factor E, when a certain day is a weekend, the value of E is 1, otherwise the value is 0; for the holiday and weekend overlap factor F, when a certain day is both a holiday and a weekend, the value of F is 1, otherwise the value is 0.

[0024] The method of training the prediction model includes T1 to T6.

[0025] T1. Obtain the expected load change of large industrial users and the comprehensive historical data within the preset historical period. The comprehensive historical data includes load historical data and meteorological historical data. Among them, the expected load change of large industrial users is reported by large industrial users according to their own conditions. The comprehensive historical data within the preset historical period can be directly obtained from the distribution automation system of the power grid. The distribution automation system is a mature existing technology and will not be described here. The preset historical period can be determined according to actual needs, for example, it can be determined as three consecutive years.

[0026] T2. Screening the comprehensive historical data to obtain the associated basic load data and basic meteorological data. The method of screening the comprehensive historical data to obtain the associated basic load data and basic meteorological data includes T21 to T23.

[0027] T21. Extract multiple load segments from the load history data and multiple meteorological segments from the meteorological history data, and the duration of the load segments and the meteorological segments reaches the preset duration threshold. Furthermore, within the same load segment, the difference between the total load of the power grid at any two moments is less than the preset difference threshold, so as to ensure that the load state of the power grid remains stable within the same load segment; in the same meteorological segment, the meteorological state also remains stable, where the meteorological state may include rainfall state, temperature state, humidity state, etc. By making the duration of the load segment and the meteorological segment reach the preset duration threshold, interference caused by local abnormal data can be avoided.

[0028] T22. Calculate the time similarity of the load segment and the meteorological segment. If the time similarity reaches a preset similarity threshold, the meteorological segment and the load segment are recorded as being associated. Specifically, if the time similarity of the load segment and the meteorological segment reaches a preset similarity threshold, it means that the grid load recorded by the load segment is related to the meteorological state, that is, the meteorological state may have an impact on the grid load, so the meteorological segment and the load segment can be recorded as being associated.

[0029] T23. Arrange all related load segments and meteorological segments to generate basic load data and basic meteorological data.

[0030] T3, clustering the basic load data to generate a basic load curve. The method of clustering the basic load data includes T31 to T32.

[0031] T31. Determine the load level corresponding to the load segment and the meteorological classification corresponding to the meteorological segment. The load level can be high load, medium load, low load, etc., and the meteorological classification can be sunny, cloudy, rainy, etc., which can be selected according to the actual situation.

[0032] T32. Group all load segments of the same load level into one category.

[0033] T4. Dynamically obtain the elastic load data of the power grid, and calculate the elastic load average value and elastic load time index within the preset sample period based on the elastic load data. Specifically, the comprehensive historical data within the preset historical period is directly obtained from the automatic dispatching system of the power grid. The elastic load data includes the charging and discharging load data of electric vehicles and the load data of the energy storage system. The electric vehicles contained in the power grid participate in V2G dispatching, that is, electric vehicles are given priority to participate in dispatching during peak loads, and they are discharged to the power grid as energy storage devices, while they are still charged for electrical equipment to supplement their own power when the system load is low; when the distributed energy storage system is connected to the power grid, it can store and release energy according to the existing load changes of the system. When the power grid load is low, the distributed energy storage can be used as a load to supplement its own power to fill the trough; and when the power grid load is at a peak, the energy storage can act as a power source to release electric energy to eliminate the peak, achieving the effect of peak shaving and valley filling.

[0034] The method of calculating the elastic load contemporaneous average value and the elastic load time index within a preset sample period based on the elastic load data includes T41 to T44.

[0035] T41. Segment the dynamically acquired elastic load data based on a preset sample period. The sample period can be set to one week or one month.

[0036] T42. Calculate the contemporaneous average of the elastic load data within each sample period.

[0037] T43. Calculate the average total value based on all the average values ​​of the same period.

[0038] T44. Calculate the ratio of the average value during the same period to the average total value as the elastic load time index.

[0039] T5. A sample set is formed based on the expected load change, the basic load curve, the average value of the elastic load over the same period, and the elastic load time index. In one embodiment of the present invention, two thousand sets of data are collected to form a sample set, and the sample set is divided into a training set and a validation set in a ratio of 8:2, wherein the training set is used to train the model, and the validation set is used to verify the effect of the trained model.

[0040] T6. Use the sample set to iteratively train the prediction model.

[0041] In order to verify the effect of the present invention, the following experiment was conducted based on data actually obtained from a certain power grid.

[0042] Table 1 shows the load forecast results and actual load comparison of a certain power grid on October 29, 2024, in MW. 近期 Take the average load of the power grid from October 21 to 27, 2024, L 去年,同日 Take the grid load on October 29, 2023, G is the average annual load growth of 17.55% calculated by the maximum load growth of the entire grid from 2015 to 2024, ΔL 大工业 Based on the estimated load changes reported by large industrial users of the power grid, -140MW is taken, and the resulting comprehensive load forecast data for December 29, 2023 will have a deviation of less than 3% from the actual load of the power grid.

[0043] Table 1 Experimental data table

[0044] Further, such as Figure 2 As shown, it can be seen that the load data predicted by the present invention has a high degree of fit with the actual load data, and the present invention can accurately predict the power grid load.

[0045] The present invention further provides a short-term load forecasting system for a power grid, comprising a data acquisition device, a data management device, a forecasting device and a data output device.

[0046] The data acquisition device is used to obtain the estimated load change of large industrial users and the comprehensive historical data within a preset historical period. The data acquisition device can be directly connected to the automatic dispatching system of the power grid and directly obtain data from the automatic dispatching system.

[0047] A data management device is used to assemble a sample set.

[0048] The prediction device is used to predict the short-term load of the power grid in the prediction target period by using a pre-trained prediction model.

[0049] The data output device is used to output the prediction result of the prediction device.

[0050] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, the specific working process of the system and device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here. In the several embodiments provided in the present disclosure, it should be understood that the disclosed system, device and method can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0051] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0052] The present invention also provides a computer device, including a memory and a processor.

[0053] A memory storing a computer program.

[0054] The processor is used to read and execute the computer program to implement the above-mentioned method for short-term load forecasting of a power grid.

[0055] The present invention finally provides a storage medium for storing a computer program, and the computer program implements the above-mentioned method for short-term load forecasting of a power grid when executed.

[0056] Based on the storage medium, the technical solution of the present disclosure, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present disclosure. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.

[0057] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for short-term load forecasting of a power grid, characterized in that: The steps include: Determine the forecast target period; The pre-trained prediction model is used to predict the short-term load of the power grid in the prediction target period. The prediction model is: Among them, α is the weight coefficient, L 近期 is the recent load of the power grid, L 去年,同日 is the grid load corresponding to the forecast target period, G is the average annual growth rate of load, a T and a H are the temperature influence coefficient and humidity influence coefficient respectively, T and H are the reference temperature and reference humidity of the forecast target period respectively, c is the holiday influence coefficient, D is the holiday factor, e is the weekend influence coefficient, E is the weekend factor, f is the influence coefficient when holidays and weekends overlap, F is the holiday and weekend overlap factor, ΔL 大工业 It is the expected load change reported in advance by large industrial users.

2. A method for short-term load forecasting of a power grid as claimed in claim 1, characterized in that: The method of training the prediction model comprises: Obtain the expected load change of large industrial users and comprehensive historical data within a preset historical period, the comprehensive historical data includes load historical data and meteorological historical data; Screening the comprehensive historical data to obtain the associated basic load data and basic meteorological data; Generate basic load curve after clustering basic load data; Dynamically obtain the elastic load data of the power grid, and calculate the elastic load average value and elastic load time index within a preset sample period based on the elastic load data; A sample set is formed based on the expected load change, basic load curve, elastic load average value over the same period and elastic load time index; the prediction model is iteratively trained using the sample set.

3. A method for short-term load forecasting of a power grid as claimed in claim 2, characterized in that: The method of screening the comprehensive historical data to obtain the associated basic load data and basic meteorological data includes: Extracting multiple load segments from the load history data and multiple meteorological segments from the meteorological history data, wherein the duration of the load segments and the meteorological segments both reach a preset duration threshold; Calculate the time similarity between the load segment and the meteorological segment, and if the time similarity reaches a preset similarity threshold, the meteorological segment and the load segment are recorded as being associated; All related load segments and meteorological segments are collated to generate basic load data and basic meteorological data.

4. A method for short-term load forecasting of a power grid as claimed in claim 3, characterized in that: Methods for clustering base load data include: Determine the load level corresponding to the load segment and determine the meteorological classification corresponding to the meteorological segment; All load segments of the same load level are grouped together.

5. A method for short-term load forecasting of a power grid as claimed in claim 2, characterized in that: The method for calculating the elastic load average value and the elastic load time index within a preset sample period based on the elastic load data includes: Segmenting the dynamically acquired elastic load data based on a preset sample period; Calculate the contemporaneous average of the elastic load data within each sample period; Calculate the average total value based on all the average values ​​of the same period; The ratio of the average value during the same period to the average total value is calculated as the elastic load time index.

6. A method for short-term load forecasting of a power grid as claimed in claim 2, characterized in that: The comprehensive historical data within a preset historical period is directly obtained from the power grid's automated dispatching system.

7. A method for short-term load forecasting of a power grid as claimed in claim 2, characterized in that: Elastic load data includes electric vehicle charging and discharging load data and energy storage system load data.

8. The power grid short-term load forecasting system is characterized by: include: Data acquisition device, used to obtain the estimated load change of large industrial users and comprehensive historical data within a preset historical period; A data management device for forming a sample set; A prediction device, used to predict the short-term load of the power grid in the prediction target period by using a pre-trained prediction model; The data output device is used to output the prediction result of the prediction device.

9. Computer device, characterized in that include: a memory storing a computer program; A processor is used to read and execute the computer program to implement a short-term load forecasting method for a power grid as described in any one of claims 1 to 7.

10. A storage medium, characterized in that Used to store a computer program, which, when executed, implements a power grid short-term load forecasting method as described in any one of claims 1-7.

Images