Whole society power consumption load prediction method and system considering industry power consumption characteristics

Abstract

The invention discloses a whole-society power consumption load prediction method and system considering industry power consumption characteristics. The method comprises the following steps: acquiring power consumption load data of a plurality of industries subordinate to a specific area and corresponding air temperature data and day type data; for each industry, calculating the correlation between the electrical load data and the air temperature data, and the correlation between the electrical load data and the day type data; according to the two calculated correlations, determining influence factors of the electrical load of each industry; and according to the electrical load data of the plurality of industries subordinate to the specific region and the influence factors of the electrical load of each industry, predicting the electrical load of the whole society of the specific region. And the accuracy of the whole society electrical load prediction is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electricity load forecasting, and in particular to a method and system for forecasting the electricity load of the entire society taking into account the electricity consumption characteristics of various industries. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] By accurately predicting the electricity load of the entire society, it is possible to effectively regulate the load of the power system.

[0004] Currently, when predicting the electricity load of the whole society, only the historical electricity load data of the whole society is taken into consideration, and the impact of the electricity load data of various industries on the electricity load forecast of the whole society is not considered, and the differences in influencing factors in different industries are not considered, which leads to inaccurate results in the prediction of the electricity load of the whole society. Summary of the invention

[0005] In order to solve the above problems, the present invention proposes a method and system for predicting the electricity load of the whole society taking into account the electricity consumption characteristics of the industry, thereby improving the accuracy of the prediction of the electricity load of the whole society.

[0006] To achieve the above object, the present invention adopts the following technical solution:

[0007] First, a method for predicting the electricity load of the whole society considering the electricity consumption characteristics of various industries is proposed, including:

[0008] Obtain electricity load data for multiple industries in a specific area and the corresponding temperature data and day type data;

[0009] For each industry, calculate the correlation between its electricity load data and temperature data, and the correlation between its electricity load data and daily type data;

[0010] Based on the two calculated correlations, the influencing factors of the electricity load of each industry are determined;

[0011] Based on the electricity load data of multiple industries in a specific area and the influencing factors of the electricity load of each industry, the total social electricity load of the specific area is predicted.

[0012] Further, the Spearman correlation coefficient between the power load data and the temperature data is calculated as the correlation between the power load data and the temperature data;

[0013] or,

[0014] Construct a violin plot of the electricity load data corresponding to each industry and each type of daily data;

[0015] According to the violin plot, the difference of power load data under different daily data types of each industry is determined, and the difference is used as the correlation between the power load data of the industry and the daily data type;

[0016] or,

[0017] The differences in electricity load data under different types of daily data in different industries include the differences in the median of the violin plots of different types of daily data in different industries and the differences in the density width of the violin plots of different types of daily data in different industries.

[0018] Further, for each industry, temperature data and / or day type data with a correlation greater than a set threshold are selected as influencing factors of the electricity load of the industry;

[0019] or,

[0020] For each industry, when one or both of the median difference in the violin plot of different day-type data of the industry and the density width difference in the violin plot of different day-type data of the industry are greater than the set threshold, it is determined that the difference in electricity load data under different day-type data of the industry is greater than the set threshold.

[0021] Furthermore, the day type data includes working day data and rest day data;

[0022] or,

[0023] The temperature data includes spring temperature data, summer temperature data, autumn temperature data and winter temperature data.

[0024] Furthermore, the whole society power load forecasting model is used to process the power load data of multiple industries under a specific area and the influencing factors of the power load of each industry to obtain the whole society power load forecasting result of the specific area; wherein, the process of using the whole society power load forecasting model to process the power load data of multiple industries under a specific area and the influencing factors of the power load of each industry includes:

[0025] For each industry under a specific region, the power load data and the factors affecting the power load of the industry are used as the data of the industry;

[0026] The social electricity load forecasting model extracts the data characteristics of each industry data, and predicts the social electricity load based on the data characteristics of each industry data to obtain the social electricity load forecast results predicted by each industry; the social electricity load forecast results predicted by all industries are weighted and integrated to obtain the final social electricity load forecast results for a specific area.

[0027] Furthermore, the data features of each industry data are extracted through convolutional neural networks;

[0028] Or, through BiLSTM, the electricity load of the whole society is predicted according to the data characteristics of each industry data, and the prediction results of the electricity load of the whole society predicted by each industry are obtained;

[0029] Alternatively, an attention mechanism is used to perform weighted fusion on the social electricity load forecast results predicted by all industries to obtain the final social electricity load forecast results for a specific area.

[0030] Secondly, a social electricity load forecasting system considering the electricity consumption characteristics of various industries is proposed, including:

[0031] A data acquisition module is used to obtain the power load data of multiple industries in a specific area and the corresponding temperature data and day type data;

[0032] A correlation calculation module is used to calculate the correlation between the power load data and the temperature data, and the correlation between the power load data and the daily type data for each industry;

[0033] An influencing factor determination module is used to determine the influencing factors of the power load of each industry based on the two calculated correlations;

[0034] The module for predicting the total social electricity load is used to predict the total social electricity load in a specific area based on the electricity load data of multiple industries in the specific area and the influencing factors of the electricity load of each industry.

[0035] In a third aspect, a computer device is provided, the device comprising:

[0036] a processor adapted to execute a computer program;

[0037] A computer-readable storage medium having a computer program stored therein, wherein when the computer program is executed by the processor, the method for predicting the electricity load of the whole society taking into account the electricity consumption characteristics of the industry proposed in the first aspect is implemented.

[0038] In a fourth aspect, a computer-readable storage medium is proposed, wherein the computer-readable storage medium stores a computer program, and the computer program is suitable for being loaded by a processor and executing the method for predicting the electricity load of the whole society considering the electricity consumption characteristics of the industry proposed in the first aspect.

[0039] In a fifth aspect, a computer program product is proposed, which includes a computer program. When the computer program is executed by a processor, it implements the method for predicting the electricity load of the whole society taking into account the electricity consumption characteristics of the industry proposed in the first aspect.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] The present invention proposes a method and system for predicting the total social electricity load that takes into account the electricity consumption characteristics of different industries. When predicting the total social electricity load of a region, the method fully considers the electricity load data of multiple industries in the region, and selects several influencing factors that are most relevant to the electricity load data of each industry for each industry. The total social electricity load of the region is predicted based on the electricity load data of multiple industries in the region and the influencing factors of the electricity load of each industry selected, thereby improving the accuracy of the prediction of the total social electricity load.

[0042] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The drawings in the specification, which constitute a part of the present application, are used to provide further understanding of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute improper limitations on the present application.

[0044] Figure 1 A flow chart of a method for predicting the electricity load of the whole society taking into account the electricity consumption characteristics of various industries disclosed in the embodiment;

[0045] Figure 2 A schematic diagram of a violin disclosed in an embodiment;

[0046] Figure 3 The comprehensive electricity consumption evaluation index of each industry disclosed in the embodiment;

[0047] Figure 4 The correlation between various industries and temperature disclosed in the embodiment;

[0048] Figure 5 Violin plots between industries and day types disclosed in the embodiment;

[0049] Figure 6 It is the difference between each industry and day type disclosed in the embodiment. DETAILED DESCRIPTION

[0050] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0051] It should be noted that the following detailed descriptions are illustrative and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0052] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0053] Example 1

[0054] As the electricity consumption characteristics of different industries gradually change, it in turn affects the electricity load of the whole society, increasing the difficulty of predicting the electricity load of the whole society. Effective and accurate analysis of electricity consumption characteristics can explore electricity consumption patterns and help improve the accuracy of load forecasting.

[0055] In this context, the patent with publication number CN118316033A proposed a power load forecasting method based on CNN-GRU and ARIMA models, which extracts power load data and performs data preprocessing, and then builds GRU-CNN model and ARIMA model. Finally, the model is trained and the prediction results are output. This helps to more accurately extract the time domain and frequency domain features of power load data and solve the shortcomings of traditional load forecasting. On the other hand, the patent with publication number CN118316019A proposed a short-term load forecasting method based on TCN-BiLSTM-PSO combined model, which decomposes historical load data to obtain the input of TCN model and BiLSTM model, and then accurately predicts the power load.

[0056] However, when analyzing and predicting the data related to the electricity load of the whole society, the above two methods did not conduct a qualitative analysis of the electricity consumption characteristics of different industries, nor did they fully consider the possible differences in the influencing factors of different industries. In addition, there are still deficiencies in extracting and analyzing power load data, which ultimately leads to inaccurate predictions of the electricity load of the whole society.

[0057] To this end, in order to improve the accuracy of the prediction of the total social electricity load, this embodiment focuses on the industry electricity consumption characteristics, and proposes a method for predicting the total social electricity load taking into account the industry electricity consumption characteristics. First, the influencing factors of the electricity load of each industry in the region are screened. Then, based on the electricity load data of each industry in the region and the influencing factors screened out by each industry, the total social electricity load of the region is predicted, thereby improving the accuracy of the prediction of the total social electricity load.

[0058] The method for predicting the electricity load of the whole society taking into account the electricity consumption characteristics of various industries disclosed in this embodiment is described in detail.

[0059] The method for predicting the electricity load of the whole society considering the electricity consumption characteristics of the industry disclosed in this embodiment is as follows: Figure 1-Figure 6As shown, including:

[0060] Obtain electricity load data for multiple industries in a specific area and the corresponding temperature data and day type data;

[0061] For each industry, calculate the correlation between its electricity load data and temperature data, and the correlation between its electricity load data and daily type data;

[0062] Based on the two calculated correlations, the influencing factors of the electricity load of each industry are determined;

[0063] Based on the electricity load data of multiple industries in a specific area and the influencing factors of the electricity load of each industry, the total social electricity load of the specific area is predicted.

[0064] Among them, the day type data includes working day data and rest day data.

[0065] The temperature data includes spring temperature data, summer temperature data, autumn temperature data and winter temperature data.

[0066] Before screening the influencing factors of the power load of each industry in a specific area, this embodiment first obtains the power load data of each industry in the specific area, and determines whether there are differences in the power consumption characteristics of each industry by performing qualitative and quantitative analysis on the power load data of each industry.

[0067] The process of qualitative and quantitative analysis of electricity load data for various industries includes:

[0068] S111: Clustering the power load data of each industry in a specific area according to time to obtain a power load curve of each industry in the specific area.

[0069] The electricity demand of an industry is affected by external factors such as policies and supply fluctuations in the industrial chain on the one hand, and by internal factors such as production process flow and technological innovation on the other hand. Therefore, the electricity load curves of different industries on the same day may show significant differences. This embodiment uses the elbow rule to determine the number of cluster centers of the electricity load data of each industry, and then uses the K-means clustering algorithm (K-means) to perform cluster analysis on the electricity load data of each industry to obtain the electricity load curve of each industry.

[0070] S121: Determine the comprehensive electricity consumption evaluation index of each industry according to the electricity consumption load curve of each industry, wherein the comprehensive electricity consumption evaluation index of each industry is the weighted average of the fluctuation degree, impact degree, daily peak-to-valley difference rate and daily load rate of the electricity consumption load curve of each industry.

[0071] This embodiment uses the entropy weight method to analyze the fluctuation, impact, daily peak-to-valley difference rate, and daily load rate of the electricity load curve of each industry to obtain comprehensive electricity consumption evaluation indicators for each industry to quantify the industry's electricity consumption characteristics.

[0072] Fluctuation is used to characterize the discrete degree of industrial power load, and its magnitude is positively correlated with the load fluctuation degree, that is, the smaller the fluctuation degree, the smaller the load fluctuation degree, and the more stable the load power consumption. The fluctuation formula is as follows:

[0073]

[0074] In the formula, S1 represents the fluctuation, P′ represents the normalized afterload data, and P′ mean represents the normalized mean value, and N represents the number of sampling points.

[0075] Impact load refers to a load that is periodic or non-periodic and has a large range of changes. Its value can reach several times the average load. This type of load usually lasts for a short time, but the peak value is very high. It is an important law of electricity consumption. The impact degree formula is as follows:

[0076]

[0077] In the formula, S2 represents the impact degree, P t Indicates load curve data, P t smooth represents the load curve data after smoothing, and N represents the number of sampling points. To ensure the consistency of data ranges in various industries, S2 can be normalized for comparison.

[0078] The daily peak-to-valley difference rate refers to the ratio of the daily peak-to-valley difference of the power system to the maximum load within a day. It measures the magnitude of the load fluctuation of the power system within a day from the perspective of extreme values ​​and is an important power system indicator. The formula for the daily peak-to-valley difference rate is as follows:

[0079]

[0080] In the formula, S3 represents the daily peak-to-valley difference rate, P max Indicates the maximum value of the daily load curve, P min Indicates the minimum value of the daily load curve.

[0081] The daily load rate refers to the ratio of the average load to the maximum load within a day, and is an important indicator of the power system. The daily load rate formula is as follows:

[0082]

[0083] Where S4 represents the daily load rate, P mean Represents the average value of daily load curve, P max Indicates the maximum value of the daily load curve.

[0084] Among the above four indicators, the greater the volatility, impact, and daily peak-to-valley difference, the more unstable the industry power load curve; the greater the daily load rate, the more stable the industry power load curve. In order to comprehensively measure the degree of representation of the industry power load curve by each indicator, the entropy weight method is introduced to calculate the weight of each indicator and obtain the comprehensive power consumption evaluation index of each industry.

[0085] The entropy weight method is an objective weighting method used to determine the weight of each indicator in a multi-indicator comprehensive evaluation system. It is based on the concept of information entropy and measures the relative importance of each indicator in the evaluation system by calculating the information entropy of each indicator. The smaller the information entropy, the greater the variability of the indicator and the more information it provides, so it will be given a greater weight in the comprehensive evaluation. The basic process of the entropy weight method is as follows:

[0086] (1) Data standardization

[0087] In order to eliminate the possible differences in the dimensions and numerical ranges of each indicator, each indicator is standardized to ensure the consistency of the magnitude of each indicator.

[0088] (2) Calculation of index weight

[0089] For each indicator, the weight calculation formula is as follows:

[0090]

[0091] In the formula, r ij is the specific gravity, x ij is the value of the jth evaluation object on the ith indicator, and m is the total number of evaluation objects.

[0092] (3) Calculate the index entropy value

[0093] The entropy value of each indicator is calculated using the proportion. The calculation formula is as follows:

[0094]

[0095] e i is the entropy value; if r ij =0, then r ij ln ij =0.

[0096] (4) Calculate the coefficient of variation of the indicator

[0097] Using the coefficient of variation g i Reflecting the effectiveness of the indicator, the calculation formula is as follows:

[0098]

[0099] (5) Determine the weight of the indicator

[0100] Calculate the weight w of each indicator according to the coefficient of difference i , the calculation formula is as follows:

[0101]

[0102] In the formula, n represents the total number of indicators.

[0103] The comprehensive electricity consumption evaluation index is the weighted average of fluctuation, impact, daily peak-to-valley difference rate and daily load rate. The calculation formula is as follows:

[0104]

[0105] The power load data used are the power load of the whole society in a certain area in 2022 and the data of the eleven subordinate industries and the power load of residents. For the convenience of expression, they are collectively referred to as the twelve major industries in the following text. The data time scale is 15 minutes, that is, one data point is collected every 15 minutes; the temperature data is the temperature data at 24 o'clock every day in the same period of the region. The specific industry names of the power load data are shown in Table 1. The linear interpolation method is used to fill the missing data in the data set, and the average load in each hour is calculated, forming a total of 8760*11 load data for each category.

[0106] Table 1 Dataset category names

[0107]

[0108]

[0109] In addition, in order to accelerate the convergence of the model and improve the generalization ability, the obtained power load data and temperature data are normalized. The normalization method used is Min-Max Normalization, and the formula is as follows:

[0110]

[0111] Where, is the normalized data, P is the original power load data, and P min is the minimum value of the original power load data, P max It is the maximum value of the original power load data.

[0112] Analysis of industry electricity consumption characteristics: (1) Qualitative analysis results. Based on the above qualitative analysis principles, the elbow rule was used to analyze the electricity load data of the whole society and the twelve major industries to determine the optimal number of clustering centers for each industry. The results are shown in Table 2.

[0113] Table 2 Results of the elbow rule for the whole society and the twelve major industries

[0114]

[0115]

[0116] According to the results of the elbow rule, the K-means clustering method is used to cluster the power load data of each industry, and the clustering results of the power load data of each industry are obtained, that is, the power load data curve of each industry. The overall social power load curve gradually rises around 4 o'clock, reaches the first peak around 10 o'clock, and then falls back slightly, reaching the second peak around 16 o'clock, and then shows a downward trend, but there will be a rebound at 19 o'clock.

[0117] The electricity load curves of agriculture, forestry, animal husbandry and fishery mostly reach the peak value of the day at 6 o'clock and drop to the valley value of the day at 18 o'clock; the changes in the construction industry are similar to those of the industry, so I will not repeat them; the changes in the transportation / warehousing and postal industry, information transmission / software and information technology services, wholesale and retail industries, real estate industries, leasing and business services industries are similar, reaching the peak value around 10 o'clock, and then remaining stable for a long period of time, with a slight decline around 14 o'clock and further decline around 18 o'clock; the accommodation and catering industry has been introduced in the previous article and will not be repeated; the changes in the financial industry, public services and management organizations are similar, reaching the peak value around 10 o'clock and then gradually declining, accompanied by several small rebounds in the middle; the electricity consumption of urban and rural residents shows a downward trend after 0 o'clock, gradually recovers around 5 o'clock, usually reaches the morning peak at 6 o'clock, then declines slightly, reaches a peak again at 10 o'clock, rises sharply after 15 o'clock, reaches the peak value of the day at 20 o'clock, and then begins to decline sharply. It can be seen that the electricity load curves of the twelve major industries change differently, and there are large differences from the electricity load curves of the whole society in certain periods. Therefore, when predicting the electricity load curve for the whole society, the electricity consumption characteristics of each industry need to be considered.

[0118] (2) Quantitative analysis results

[0119] According to the above quantitative analysis principle, the fluctuation, impact, daily peak-to-valley difference and daily load rate of daily electricity load data of the whole society and the twelve major industries are calculated respectively, and the entropy weight method is used to determine the weight of each sub-indicator in the comprehensive electricity consumption evaluation index. The average values ​​and weight results of fluctuation, impact, daily peak-to-valley difference and daily load rate are shown in Table 3.

[0120] Table 3 Average values ​​and weights of indicators for the whole society and the twelve major industries

[0121]

[0122] According to Table 3, the average values ​​of comprehensive electricity consumption evaluation indicators for the whole society and the twelve major industries are as follows: Figure 3 As shown, Figure 3 The industry categories are shown in Table 1.

[0123] Depend on Figure 3 It can be seen that the comprehensive electricity consumption evaluation index of the whole society is 0.476, while the comprehensive electricity consumption evaluation index of most industries is around 0.4. Only the comprehensive electricity consumption evaluation index of industry, construction and urban and rural residents is close to 0.476, and the transportation / warehousing and postal industries even exceed 0.5. Therefore, the comprehensive electricity consumption evaluation index results of different industries vary greatly, indicating that there are large differences in the electricity consumption characteristics of various industries. Therefore, when predicting the electricity load curve of the whole society, it is necessary to consider the electricity consumption characteristics of the industry.

[0124] This embodiment performs qualitative and quantitative analysis on the industry power load data, determines that when predicting the power load of the whole society, the power consumption characteristics of each industry should be considered, and then screens the factors affecting the power load of each industry.

[0125] As an important factor affecting the power load, the impact of temperature on the power load is mainly reflected in the air conditioning cooling load in summer and the electric heating load in winter. Although the temperature has an important impact on the power load, it does not affect the power load all year round. At present, the impact of temperature on the power load is mainly analyzed by the size of the Pearson correlation coefficient (Pearson correlation coefficient), and there are few studies on the nonlinear relationship between temperature and power load. On the other hand, a major premise of the Pearson correlation coefficient is that the data conforms to the normal distribution assumption, but the temperature and power load data do not conform to the normal distribution assumption. Compared with the Pearson correlation coefficient, the Spearman correlation coefficient can not only analyze the nonlinear relationship between sequences, but also does not require the data to obey the normal distribution. Therefore, this embodiment selects the Spearman correlation coefficient to analyze the correlation between the temperature data and the industry power load data, that is, to calculate the Spearman correlation coefficient between the power load data and the temperature data as the correlation between the power load data and the temperature data. The calculation formula of the Spearman correlation coefficient is as follows:

[0126]

[0127] In the formula, n is the number of samples, which indicates the number of observation logarithms involved in calculating the correlation, that is, the length of the temperature data corresponding to the load data, and d i is the rank difference of the i-th pair of observations, that is, X i and Y i The difference in the sequence numbers of the i-th samples in the two columns after they are arranged in descending order.

[0128] By calculating the correlation coefficient between the temperature data and the power load data of each season, we can more clearly understand the impact of temperature on power load. The months included in different seasons are as follows: spring is from March to May, summer is from June to August, autumn is from September to November, and winter is from January to February and December.

[0129] The Spearman correlation coefficients of the daily electricity load curves of the whole society and the twelve major industries and the temperature data in different seasons are calculated respectively. The results are as follows Figure 4 shown.

[0130] Depend on Figure 4 It can be seen that agriculture, forestry, animal husbandry and fishery have a strong correlation with temperature in spring, summer and autumn; except for industry in summer, other industries have a strong correlation with temperature; transportation / warehousing and postal industries, information transmission / software and information technology services, wholesale and retail industries, accommodation and catering industries, finance, real estate industries, leasing and business services industries also have a strong correlation with temperature in autumn; the correlation between various industries and temperature in winter is relatively low, mainly because the winter climate in the selected areas is relatively high and the heating demand is low.

[0131] This embodiment constructs a violin plot of the power load data corresponding to each type of daily data in each industry;

[0132] According to the violin plot, the difference of electricity load data under different day type data of each industry is determined, and the difference is used as the correlation between industry electricity load data and day type data.

[0133] The differences in electricity load data under different types of daily data in different industries include the differences in the median of the violin plots of different types of daily data in different industries and the differences in the density width of the violin plots of different types of daily data in different industries.

[0134] Among them, the day type data includes working day data and rest day data; the impact of day type on load is usually reflected in the difference between working days and rest days. Usually, the power load on working days is higher than that on rest days due to the daily production, office and other work. Since the day type data is discrete data, the power load data is time series data, and the correlation coefficient is mainly used to analyze the correlation between two continuous sequences, when the correlation coefficient is used to analyze the correlation between day type and load, the correlation coefficient may be very low. Therefore, it is considered to use visualization tools to analyze the patterns and differences of power load under different day types.

[0135] like Figure 2As shown in the figure, the violin plot is a data visualization method that combines the advantages of the box plot and the density plot. It can display a large amount of data distribution characteristics in a limited space and is very suitable for comparing and analyzing the distribution of data of different categories. The violin plot usually consists of two structures: the box plot part and the density curve. The box plot is located in the center of the violin plot, showing the median and upper and lower quartiles of the data, as well as possible outliers, usually represented by points; the density curve is located on the outer edge of the box plot, and the width of the curve represents the density estimate of the data points. In the "neck" of the violin plot (the area of ​​the box plot), the density curve is narrower due to the small number of data points. In the "body" part, the density of data points is higher and the density curve becomes wider.

[0136] Analysis of the impact of daily types, considering that the magnitude of electricity load data of various industries varies greatly, after normalizing the electricity load data of various industries, a violin diagram of the electricity load of the whole society and the twelve major industries is drawn, as shown in Figure 5 As shown in the figure, according to the violin plot of the electricity load of the whole society and the twelve major industries, the correlation between the electricity load of the whole society and the twelve major industries and the daily type data is determined. By comparing the median and density map width of the violin plot of the electricity load data corresponding to the working day data and the violin plot of the electricity load data corresponding to the rest day data, the difference in electricity load under the two different daily type data can be analyzed, thereby reflecting the impact of the daily type data on the electricity load. By analyzing the correlation between each industry and the influencing factors, the violin plot visualization analysis is introduced and the difference quantification index is constructed to screen the main influencing factors of each industry, such as Figure 6 Shown are the median differences and density map width differences of electricity load data under different day types of data for each industry.

[0137] The differences in electricity load data under different types of daily data in different industries include the differences in the median of the violin plots of different types of daily data in different industries and the differences in the density width of the violin plots of different types of daily data in different industries.

[0138] Among them, the median difference of the violin plots of different day types of data in the industry is calculated based on the median of the violin plots of the electricity load data on working days and the median of the violin plots of the electricity load data on rest days in the industry.

[0139] The difference in the density width of the violin plots of different types of daily data in an industry is calculated based on the density width of the violin plots of the applied electricity load data on working days and the density width of the violin plots of the applied electricity load data on rest days in the industry.

[0140] The calculation formula for the median difference of the violin plot of different daily data types in the industry and the density width difference of the violin plot of different daily data types in the industry is:

[0141]

[0142] Where D represents the median difference of the violin plots of different types of daily data or the density width difference of the violin plots of different types of daily data in different industries, P mid,0 The median of the violin plot of the weekday versus electricity load data or the density width of the violin plot of the weekday versus electricity load data, P mid1 Represents the median of the violin plot of rest days versus applied electricity load data or the width of the violin density plot of rest days versus applied electricity load data.

[0143] Depend on Figure 5 and Figure 6 It can be seen that the load distribution differences between working days and rest days are relatively small for agriculture / forestry / animal husbandry / fishery and wholesale and retail industries, i.e., categories 1 and 6, while the load distribution differences between different types of days for other industries are relatively large and significant. Therefore, when forecasting the electricity load of the whole society, it is necessary to consider the electricity consumption characteristics of each industry under different types of day data.

[0144] In this embodiment, for each industry, temperature data and / or day type data with a correlation greater than a set threshold are selected as influencing factors of the power load of the industry.

[0145] For each industry, when one or both of the median difference of the violin plot of different day-type data in the industry and the density width difference of the violin plot of different day-type data in the industry are greater than the set threshold, it is determined that the difference of the power load data under different day-type data in the industry is greater than the set threshold, and the day-type data is used as the influencing factor of the power load of the industry.

[0146] According to the above analysis results, the influencing factors that need to be considered in the twelve major industries are screened out and shown in Table 4, where "yes" means that the influencing factor is taken into consideration. When the social electricity load forecast is carried out subsequently, the influencing factor is used as the input data of the social electricity load forecast model, so that the influencing factor is involved in the social electricity load forecast. "no" means that the influencing factor is not considered. When the social electricity load forecast is carried out subsequently, the influencing factor data is not obtained, the input data of the social electricity load forecast model does not contain the influencing factor, and the influencing factor is not involved in the social electricity load forecast.

[0147] Table 4 Influencing factors of the twelve major industries

[0148]

[0149]

[0150] After screening out the influencing factors of the power load of each industry, this embodiment uses the whole society power load forecasting model to process the power load data of multiple industries under a specific area and the influencing factors of the power load of each industry to obtain the whole society power load forecasting result of the specific area; wherein, using the whole society power load forecasting model to process the power load data of multiple industries under a specific area and the influencing factors of the power load of each industry includes:

[0151] For each industry under a specific region, the power load data and the factors affecting the power load of the industry are used as the data of the industry;

[0152] The social electricity load forecasting model extracts the data characteristics of each industry data, and predicts the social electricity load based on the data characteristics of each industry data to obtain the social electricity load forecast results predicted by each industry; the social electricity load forecast results predicted by all industries are weighted and integrated to obtain the final social electricity load forecast results for a specific area.

[0153] This embodiment extracts the data features of each industry data through a convolutional neural network (CNN); predicts the total social electricity load according to the data features of each industry data through a bidirectional long short-term memory network (BiLSTM), and obtains the total social electricity load prediction results predicted for each industry; and uses an attention mechanism to weightedly fuse the total social electricity load prediction results corresponding to all industries to obtain the final total social electricity load prediction results for a specific area.

[0154] This embodiment regards the electricity consumption of the whole society as a special industry, and includes the electricity consumption data of the whole society in the data of multiple industries under the specific area obtained.

[0155] The method for predicting the total social electricity load taking into account the electricity consumption characteristics of different industries disclosed in this embodiment fully considers the electricity load data of multiple industries in the region when predicting the total social electricity load of a region, and selects several influencing factors that are most relevant to the electricity load data of each industry for each industry; the total social electricity load of the region is predicted by using the electricity load data of multiple industries in the region and the influencing factors of the electricity load of each industry selected, thereby improving the accuracy of the prediction of the total social electricity load.

[0156] Example 2

[0157] In this embodiment, a system for predicting the electricity load of the whole society taking into account the electricity consumption characteristics of various industries is disclosed, including:

[0158] A data acquisition module is used to obtain the power load data of multiple industries in a specific area and the corresponding temperature data and day type data;

[0159] A correlation calculation module is used to calculate the correlation between the power load data and the temperature data, and the correlation between the power load data and the daily type data for each industry;

[0160] An influencing factor determination module is used to determine the influencing factors of the power load of each industry based on the two calculated correlations;

[0161] The module for predicting the total social electricity load is used to predict the total social electricity load in a specific area based on the electricity load data of multiple industries in the specific area and the influencing factors of the electricity load of each industry.

[0162] The present invention also discloses a computer device, which includes:

[0163] a processor adapted to execute a computer program;

[0164] A computer-readable storage medium having a computer program stored therein, wherein when the computer program is executed by the processor, the method for predicting the electricity load of the whole society taking into account the electricity consumption characteristics of the industry disclosed in Example 1 is implemented.

[0165] The present invention also discloses a computer-readable storage medium storing a computer program, wherein the computer program is suitable for being loaded by a processor and executing the method for predicting the electricity load of the whole society taking into account the electricity consumption characteristics of the industry disclosed in Example 1.

[0166] The present invention also discloses a computer program product, which includes a computer program. When the computer program is executed by a processor, it implements the method for predicting the electricity load of the whole society taking into account the electricity consumption characteristics of the industry disclosed in Example 1.

[0167] The method disclosed in Example 1 can be directly embodied as a hardware processor, or can be executed by a combination of hardware and software modules in the processor. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware. To avoid repetition, it will not be described in detail here.

[0168] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with this embodiment can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0169] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. A method for predicting the electricity load of the whole society considering the electricity consumption characteristics of various industries, characterized in that: include: Obtain electricity load data for multiple industries in a specific area and the corresponding temperature data and day type data; For each industry, calculate the correlation between its electricity load data and temperature data, and the correlation between its electricity load data and daily type data; Based on the two calculated correlations, the influencing factors of the electricity load of each industry are determined; Based on the electricity load data of multiple industries in a specific area and the influencing factors of the electricity load of each industry, the total social electricity load of the specific area is predicted.

2. The method for predicting the electricity load of the whole society considering the electricity consumption characteristics of the industry as claimed in claim 1, characterized in that: The Spearman correlation coefficient between the power load data and the temperature data is calculated as the correlation between the power load data and the temperature data; or, Construct a violin plot of the electricity load data corresponding to each industry and each type of daily data; According to the violin plot, the difference of power load data under different daily data types of each industry is determined, and the difference is used as the correlation between the power load data of the industry and the daily data type; or, The differences in electricity load data under different types of daily data in different industries include the differences in the median of the violin plots of different types of daily data in different industries and the differences in the density width of the violin plots of different types of daily data in different industries.

3. The method for predicting the electricity load of the whole society considering the electricity consumption characteristics of the industry as claimed in claim 2, characterized in that: For each industry, select temperature data and / or day type data with correlation greater than a set threshold as the influencing factors of the industry's electricity load; or, For each industry, when one or both of the median difference in the violin plot of different day-type data of the industry and the density width difference in the violin plot of different day-type data of the industry are greater than the set threshold, it is determined that the difference in electricity load data under different day-type data of the industry is greater than the set threshold.

4. The method for predicting the electricity load of the whole society considering the electricity consumption characteristics of the industry as claimed in claim 1, characterized in that: Day type data includes working day data and rest day data; or, The temperature data includes spring temperature data, summer temperature data, autumn temperature data and winter temperature data.

5. The method for predicting the electricity load of the whole society considering the electricity consumption characteristics of various industries as claimed in claim 1, characterized in that: The whole society power load forecasting model is adopted to process the power load data of multiple industries under a specific area and the influencing factors of the power load of each industry to obtain the whole society power load forecasting result of the specific area; wherein, the process of using the whole society power load forecasting model to process the power load data of multiple industries under a specific area and the influencing factors of the power load of each industry includes: For each industry under a specific region, the power load data and the factors affecting the power load of the industry are used as the data of the industry; The social electricity load forecasting model extracts the data characteristics of each industry data, and predicts the social electricity load based on the data characteristics of each industry data to obtain the social electricity load forecast results predicted by each industry; the social electricity load forecast results predicted by all industries are weighted and integrated to obtain the final social electricity load forecast results for a specific area.

6. The method for predicting the electricity load of the whole society considering the electricity consumption characteristics of the industry as claimed in claim 5, characterized in that: Extract the data features of each industry data through convolutional neural network; or, Through BiLSTM, the electricity load of the whole society is predicted according to the data characteristics of each industry data, and the prediction results of the electricity load of the whole society predicted by each industry are obtained; or, The attention mechanism is used to perform weighted fusion on the social electricity load forecast results predicted by all industries to obtain the final social electricity load forecast results for a specific area.

7. The whole society electricity load forecasting system considering the industry electricity consumption characteristics is characterized by: include: A data acquisition module is used to obtain the power load data of multiple industries in a specific area and the corresponding temperature data and day type data; A correlation calculation module is used to calculate the correlation between the power load data and the temperature data, and the correlation between the power load data and the daily type data for each industry; An influencing factor determination module is used to determine the influencing factors of the power load of each industry based on the two calculated correlations; The module for predicting the total social electricity load is used to predict the total social electricity load in a specific area based on the electricity load data of multiple industries in the specific area and the influencing factors of the electricity load of each industry.

8. An electronic device, characterized in that: The device comprises: a processor adapted to execute a computer program; A computer-readable storage medium having a computer program stored therein, wherein when the computer program is executed by the processor, the method for predicting the electricity load of the whole society taking into account the electricity consumption characteristics of the industry as described in any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is suitable for being loaded by a processor and executing the method for predicting the electricity load of the whole society taking into account the electricity consumption characteristics of the industry as described in any one of claims 1-6.

10. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, it implements the method for predicting the electricity load of the whole society taking into account the electricity consumption characteristics of the industry as described in any one of claims 1-6.

Images