Regional electricity consumption prediction method based on industry division

Abstract

The invention provides a regional electricity consumption prediction method based on industry division, and belongs to the technical field of energy management, and the method comprises the steps: determining a prediction time range, collecting regional historical electricity consumption data, dividing each industry, calculating the total electricity consumption proportion of each industry, and obtaining a regional industry structure. Evaluating the change condition of the industrial structure in combination with related policies, judging the first influence of the change condition on the power demand, carrying out the first prediction of the regional power consumption based on the influence and the historical power consumption proportion, collecting the data of external factors influencing the power demand, and quantifying the second influence of the factors on the power demand; the first prediction result and the second prediction result are integrated to form a comprehensive prediction model, so that regional power consumption prediction based on industry types is realized, changes of an industrial structure and an external environment are quickly responded, the flexibility and the response speed of a power system are improved, and the power consumption prediction efficiency is improved. And the configuration and scheduling of power resources are optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of energy management, and in particular to a method for predicting regional electricity consumption based on industry division. Background Art

[0002] Different industries have different electricity consumption characteristics. For example, the electricity consumption of the manufacturing industry may be significantly different from that of the service industry. The impact of industrial structure adjustment: The adjustment of industrial structure may lead to an increase in electricity consumption in some industries and a decrease in electricity consumption in other industries. Accurately predicting regional electricity consumption requires consideration of multiple factors, including industry characteristics, external influencing factors, policy orientation, etc. Current technologies include smart grid technology, smart meters, distributed generation, energy storage systems, etc., but they cannot provide high-precision real-time predictions and are difficult to meet the requirements of modern power systems for flexibility and response speed.

[0003] Therefore, the present invention provides a method for predicting regional electricity consumption based on industry division. Summary of the invention

[0004] The present invention provides a method for predicting regional electricity consumption based on industry division, which clarifies the prediction time range, collects and analyzes historical electricity consumption data in the region, classifies each industry and determines its electricity consumption ratio, evaluates the impact of its changes on electricity demand in combination with industrial structure and policy, collects external factor data and quantifies its impact, performs a first and a second prediction respectively, and finally integrates the results to form a comprehensive prediction model, accurately predicts regional electricity consumption, significantly improves the accuracy of regional electricity consumption prediction, quickly responds to changes in industrial structure and external environment, improves the flexibility and response speed of the power system, optimizes the allocation and scheduling of power resources, improves electricity efficiency, and reduces waste.

[0005] The present invention provides a method for predicting regional electricity consumption based on industry division, comprising:

[0006] Step 1: clarify the forecast time range, collect historical electricity consumption data in the region based on the forecast time range, divide the industries in the region, and determine the historical total electricity consumption proportion of each industry based on the historical electricity consumption data;

[0007] Step 2: Obtain the industrial structure in the region, and evaluate the changes in the industrial structure in combination with relevant regional policies, determine the first impact of the changes in the industrial structure on electricity demand, and make a first prediction of regional electricity consumption in combination with the first impact and the historical total electricity consumption ratio;

[0008] Step 3: Collect data on external factors that affect electricity demand in the region, quantify the second impact of the external factor data on electricity demand, build an external model based on the second impact, and use the external model to make a second prediction of regional electricity consumption;

[0009] Step 4: Integrate the first prediction result with the second prediction result to form a comprehensive prediction model to achieve prediction of regional electricity consumption based on industry type.

[0010] The present invention provides a regional electricity consumption prediction method based on industry division, clarifies the prediction time range, and collects historical electricity consumption data in the region based on the prediction time range, including:

[0011] The forecast time range is used to clearly forecast the forecast time granularity required, and a minimum time range is defined according to the forecast time range and the forecast time granularity;

[0012] Confirming first historical power usage data in an internal data source, and at the same time, confirming second historical power usage data in an external data source;

[0013] Check whether the first historical electricity consumption data and the second historical electricity consumption data cover the minimum time range, and at the same time determine whether the historical time granularity corresponding to the first historical electricity consumption data and the second historical electricity consumption data meets the requirements corresponding to the prediction time granularity to obtain a first judgment result; based on the inspection result and the first judgment result, merge the internal data source and the external data source to obtain a merged source, and determine the first effective time range in combination with the coverage and quality of the merged source and the minimum time range; and obtain the historical electricity consumption data of the merged source in the area according to the first effective time range.

[0014] The present invention provides a regional electricity consumption prediction method based on industry division, which divides various industries in a region and determines the historical total electricity consumption proportion of each industry based on the historical electricity consumption data, including:

[0015] Obtaining industry distribution information within the region, dividing the main levels of the industries within the region according to the industry distribution information, and determining the types of industries within the region according to the main levels and various characteristics within the region;

[0016] Match the historical electricity consumption data obtained from the merged source with the industry categories one by one, and add a corresponding industry label to each piece of historical electricity consumption data;

[0017] The first total electricity consumption of each industry within the first valid time range is calculated based on the industry tags, and the second total electricity consumption of all industries within the first valid time range is calculated. The historical total electricity consumption ratio of each industry is determined according to the first total electricity consumption and the second total electricity consumption.

[0018] The present invention provides a regional electricity consumption prediction method based on industry division, confirming first historical electricity consumption data in an internal data source, and at the same time confirming second historical electricity consumption data in an external data source, including:

[0019] Determine a data type of power demand, and based on the data type of power demand, determine an external source that can provide historical power consumption data of a corresponding type to constitute an external data source;

[0020] An internal system storing historical data related to power demand is identified and confirmed, and an internal data source is formed based on the internal system.

[0021] The present invention provides a regional electricity consumption prediction method based on industry division, obtains the industrial structure in the region, evaluates the change of the industrial structure in combination with relevant regional policies, determines the first impact of the change of the industrial structure on the power demand, and performs a first prediction of the regional electricity consumption in combination with the first impact and the historical total electricity consumption ratio, including:

[0022] Taking the first effective time range as a base, determining a second effective time range, acquiring industrial structure data according to the second effective time range, identifying trends and patterns of industry changes based on the industrial structure data, and determining the industrial structure of the region within the second effective time range;

[0023] Evaluate the changes in the industrial structure based on the trends and patterns of the changes in the industry and the relevant policies within the second effective time frame, and determine the first relationship between different types of industries and the second relationship between the policies and the changes in different types of industries based on the evaluation results;

[0024] Identify the electricity consumption patterns of all industries according to the first and second correlations, evaluate the circuit demand of the corresponding industry by combining the historical electricity consumption data of each industry with the electricity consumption pattern, and obtain the first impact of industrial structure changes on electricity demand;

[0025] Based on the historical total electricity consumption ratio in the region, a prediction reference point is determined, and a first prediction of regional electricity consumption is made in combination with the first impact.

[0026] The present invention provides a method for predicting regional electricity consumption based on industry division, which collects external factor data affecting electricity demand in a region, quantifies the second impact of the external factor data on electricity demand, builds an external model based on the second impact, and uses the external model to perform a second prediction on regional electricity consumption, including:

[0027] Determine external factors that may affect electricity demand in the region, perform correlation analysis between the external factors and electricity demand, and determine the secondary impact of different external factors on electricity demand;

[0028] Determine the regression coefficient of each external influencing factor according to the external factors, power demand and the second impact, evaluate the regression coefficient to obtain a coefficient evaluation result, and determine the degree of influence of each external factor on power demand according to the coefficient evaluation result;

[0029] An external model is constructed based on the impact degree and the second impact, and a second prediction of regional electricity consumption is made using the external model.

[0030] The present invention provides a method for predicting regional electricity consumption based on industry division, constructing an external model based on the impact degree and the second impact, and using the external model to perform a second prediction on regional electricity consumption, including:

[0031]

[0032] in, represents the predicted value of electricity demand; β0 represents the predicted value of electricity demand when the external factor is zero; Indicates when X j The change in the power demand forecast value when one unit value is added; X j represents the actual observed value of the jth external factor; γ j Indicates when X j The influence of the square term on electricity demand; jk represents the overall impact of the interaction between the jth external factor and the kth external factor on electricity demand; (X j ×X k ) represents the impact of the jth external factor and the kth external factor on the power demand when they interact, and k≠j; ω j represents the weight of the jth external factor; represents the intensity of the impact of the jth external factor on electricity demand; u represents the contribution value of unobserved factors to electricity demand; η represents the gap between the model prediction and the actual value; p represents the total number of external factors.

[0033] The present invention provides a method for predicting regional electricity consumption based on industry classification, which integrates a first prediction result and a second prediction result to form a comprehensive prediction model, and realizes the prediction of regional electricity consumption based on industry type, including:

[0034] Calculating the error correlation value between the first prediction result, the second prediction result and the actual power demand as the first judgment condition;

[0035] Calculating the determination coefficient between the first prediction result, the second prediction result and the actual power demand as the second judgment condition;

[0036] The accuracy of the comprehensive prediction model is judged based on the first judgment condition and the second judgment condition to obtain a second judgment result, and the regional electricity consumption of different industries in the region is output based on the second judgment result.

[0037] Compared with the prior art, the beneficial effects of the present application are as follows: by clarifying the prediction time range, collecting and analyzing historical electricity consumption data in the region, classifying each industry and determining its electricity consumption ratio, combining the industrial structure and policy to evaluate the impact of its changes on electricity demand, collecting external factor data and quantifying its impact, making the first and second predictions respectively, and finally integrating the results to form a comprehensive prediction model, accurately predicting regional electricity consumption, significantly improving the accuracy of regional electricity consumption prediction, quickly responding to changes in industrial structure and external environment, improving the flexibility and response speed of the power system, optimizing the allocation and scheduling of power resources, improving electricity efficiency, and reducing waste.

[0038] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0039] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0041] Figure 1 It is a flow chart of a method for predicting regional electricity consumption based on industry division provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0043] Embodiment 1:

[0044] The embodiment of the present invention provides a method for predicting regional electricity consumption based on industry division, such as Figure 1 As shown, including:

[0045] Step 1: clarify the forecast time range, collect historical electricity consumption data in the region based on the forecast time range, divide the industries in the region, and determine the historical total electricity consumption proportion of each industry based on the historical electricity consumption data;

[0046] Step 2: Obtain the industrial structure in the region, and evaluate the changes in the industrial structure in combination with relevant regional policies, determine the first impact of the changes in the industrial structure on electricity demand, and make a first prediction of regional electricity consumption in combination with the first impact and the historical total electricity consumption ratio;

[0047] Step 3: Collect data on external factors that affect electricity demand in the region, quantify the second impact of the external factor data on electricity demand, build an external model based on the second impact, and use the external model to make a second prediction of regional electricity consumption;

[0048] Step 4: Integrate the first prediction result with the second prediction result to form a comprehensive prediction model to achieve prediction of regional electricity consumption based on industry type.

[0049] In this embodiment, the division process is to collect relevant data of various industries in the region, such as industrial structure and electricity consumption characteristics; then analyze the industry distribution, identify the main and secondary industries and determine the hierarchy, such as the first, second and third industries, evaluate the industry characteristics based on multiple characteristics, refine the industry types, match the historical electricity consumption data of the merged source with the industry label, calculate the total electricity consumption of each industry within the effective time range, and determine the proportion of each industry in the total historical electricity consumption, so as to provide a basis for subsequent electricity demand forecasting.

[0050] In this embodiment, the historical electricity consumption ratio refers to the ratio of electricity consumption of each industry to the total electricity consumption of the region within a specific time range. The historical electricity consumption ratio of each industry is obtained by calculating the electricity consumption of each industry and comparing it with the total electricity consumption, so as to understand the electricity consumption characteristics and relative importance of different industries.

[0051] In this embodiment, the first prediction result is a regional electricity consumption prediction value obtained based on the above analysis, especially in combination with the impact of changes in industrial structure.

[0052] In this example, the first impact refers to the direct impact of industrial structure changes on electricity demand, reflecting the changes in electricity demand of different types of industries during the structural adjustment process, such as the effect of an increase in electricity consumption in a certain industry leading to an increase or decrease in overall electricity demand.

[0053] In this embodiment, external factors refer to factors that may affect electricity demand, including: economic factors: such as GDP growth rate, industrial output value, employment rate and other indicators reflecting economic activities; climate factors: such as temperature, precipitation, humidity and other meteorological data that have a greater impact on electricity demand; population factors: such as population density, residential electricity consumption characteristics, population growth rate, etc.; policy factors: such as energy policy, environmental protection regulations, subsidy policy and other external policies related to the power industry; technological changes: such as new impacts on electricity demand from distributed generation, smart grid and energy storage technology; market factors: such as electricity price fluctuations, market supply and demand relationship

[0054] In this embodiment, the second impact refers to analyzing the impact of changes in external factors on electricity demand based on the relationship between these external factors and electricity demand. Specifically, it is manifested as the specific impact of changes in various external factors (such as economic growth or temperature changes) on changes in electricity demand. For example, rising temperatures in summer may lead to increased use of air conditioners, thereby increasing electricity demand.

[0055] In this embodiment, the second prediction result is a predicted value of regional electricity consumption obtained by combining an external model. In this model, the influence degree of each external factor determined based on the above analysis is calculated in combination with the actual observed value of the external factor.

[0056] In this embodiment, the accuracy of the comprehensive prediction model is evaluated by calculating the error correlation value and determination coefficient between the first prediction result, the second prediction result and the actual electricity demand, and the regional electricity consumption forecast for different industries in the region is output based on the judgment results of the error and the determination coefficient.

[0057] In this embodiment, the integration process of the first prediction result and the second prediction result is to obtain the regional electricity consumption forecast data of various industries from the first prediction result and the second prediction result, and to standardize (such as z-score standardization) or normalize the prediction values ​​obtained by different methods as needed to make them comparable, and to weight and integrate the first prediction result and the second prediction result according to the evaluation results of the model (such as determination coefficient, error correlation value, etc.) to form a more comprehensive prediction value. For example, a linear weighted algorithm can be used: comprehensive prediction value = w1×first prediction result + w2×second prediction result, where w1 and w2 are weights, which need to be adjusted according to the model performance, and the comprehensive prediction result obtained after integration.

[0058] In this embodiment, the input data of the comprehensive prediction model include: historical electricity consumption data in the region, industry classification data, historical data of external factors, policy change data, etc. Output data include: comprehensive prediction of regional electricity consumption, electricity demand prediction classified by different industries, training set: data set used to train the comprehensive prediction model, including historical electricity consumption data in the region, external influencing factors, industry characteristics, etc. The training set should cover different time periods and diverse industry characteristics to ensure the generalization ability of the model.

[0059] In this embodiment, the input data of the external model mainly includes the external factors affecting the electricity demand in the region, such as economic indicators (GDP), meteorological data (temperature, precipitation, etc.), policy and regulatory changes, population trends, etc. Output data: The regional electricity consumption predicted by the external model is usually based on the changes in electricity demand under the influence of external factors. Training set: The data set used to train the external model should contain external factor data and corresponding actual electricity demand data. It is necessary to ensure that the data is representative and can effectively reflect the relationship between different industries and external factors.

[0060] The working principle and beneficial effects of the above technical solution are: by clarifying the prediction time range, collecting and analyzing historical electricity consumption data in the region, classifying each industry and determining its electricity consumption ratio, combining industrial structure and policy to evaluate the impact of its changes on electricity demand, collecting external factor data and quantifying its impact, making the first and second predictions respectively, and finally integrating the results to form a comprehensive prediction model, accurately predicting regional electricity consumption, significantly improving the accuracy of regional electricity consumption prediction, quickly responding to changes in industrial structure and external environment, improving the flexibility and response speed of the power system, optimizing the allocation and scheduling of power resources, improving electricity efficiency, and reducing waste.

[0061] Embodiment 2:

[0062] The embodiment of the present invention provides a method for predicting regional electricity consumption based on industry division, which clarifies the prediction time range and collects historical electricity consumption data in the region based on the prediction time range, including:

[0063] The forecast time range is used to clearly forecast the forecast time granularity required, and a minimum time range is defined according to the forecast time range and the forecast time granularity;

[0064] Confirming first historical power usage data in an internal data source, and at the same time, confirming second historical power usage data in an external data source;

[0065] Check whether the first historical electricity consumption data and the second historical electricity consumption data cover the minimum time range, and at the same time determine whether the historical time granularity corresponding to the first historical electricity consumption data and the second historical electricity consumption data meets the requirements corresponding to the prediction time granularity to obtain a first judgment result; based on the inspection result and the first judgment result, merge the internal data source and the external data source to obtain a merged source, and determine the first effective time range in combination with the coverage and quality of the merged source and the minimum time range; and obtain the historical electricity consumption data of the merged source in the area according to the first effective time range.

[0066] In this embodiment, the prediction time granularity refers to the smallest time unit used in the prediction process. For example, the time granularity can be hourly, daily, weekly, monthly or annual, etc., which determines the frequency and resolution of the prediction results.

[0067] In this embodiment, the minimum time range refers to the minimum time period covered by the data required for prediction, which is determined based on the prediction time range and the prediction time granularity. For example, if the prediction time range is one year and the time granularity is monthly, then the minimum time range may be one month.

[0068] In this embodiment, the first historical electricity consumption data refers to the historical electricity consumption records obtained from the internal data source, which are usually stored in the internal system, such as the enterprise's energy management system (EMS) or other internal databases, and contain the electricity consumption in a certain period of time in the past; the second historical electricity consumption data refers to the historical electricity consumption records obtained from the external data source, including public power companies, market analysis companies or other external databases, which are used to supplement or verify the information in the internal data source; the difference lies in the different sources of data: the first historical electricity consumption data comes from the internal system, which is usually more detailed and more accurate; the second historical electricity consumption data comes from an external source, which can provide a wider range of background information and supplementary data, and enhance the coverage and accuracy of the data.

[0069] In this embodiment, the first judgment result refers to the conclusion determined after the inspection, which determines whether the first historical electricity consumption data and the second historical electricity consumption data cover the minimum time range, and whether the time granularity of these data meets the requirements of the predicted time granularity. If both meet the conditions, subsequent data merging and analysis can be performed.

[0070] In this embodiment, the merged source refers to a data set obtained by merging the first historical electricity usage data and the second historical electricity usage data that meet the requirements. The data set integrates data from internal and external sources to provide a complete and consistent historical electricity usage record for further analysis and prediction.

[0071] In this embodiment, the first effective time range is determined based on the data coverage and quality of the merged source, which represents the effective time range of the data. The corresponding steps are to determine the time span of the merged source data set to obtain its coverage in time, check the integrity and accuracy of the merged source data set, filter out missing or abnormal data, ensure that the data in the merged source covers the minimum time range and meets the requirements of the prediction time granularity, and combine the above evaluation results to determine the first effective time range that can be used for model analysis.

[0072] The working principle and beneficial effects of the above technical solution are: based on a clear prediction time range and prediction time granularity, define the minimum time range, confirm and check that the historical electricity consumption data of internal and external data sources cover the minimum time range and meet the prediction granularity requirements, merge them into a merged source, and determine the first effective time range based on the data coverage and quality, obtain the historical electricity consumption data in the area from the merged source, provide more comprehensive and consistent electricity consumption data, overcome the problems of data dispersion and fragmentation, help cope with complex changes in industrial structure, and improve the flexibility and rapid response capabilities of the power system.

[0073] Embodiment 3:

[0074] The embodiment of the present invention provides a method for predicting regional electricity consumption based on industry division, which divides various industries in a region and determines the historical total electricity consumption proportion of each industry based on the historical electricity consumption data, including:

[0075] Obtaining industry distribution information within the region, dividing the main levels of the industries within the region according to the industry distribution information, and determining the types of industries within the region according to the main levels and various characteristics within the region;

[0076] Match the historical electricity consumption data obtained from the merged source with the industry categories one by one, and add a corresponding industry label to each piece of historical electricity consumption data;

[0077] The first total electricity consumption of each industry within the first valid time range is calculated based on the industry tags, and the second total electricity consumption of all industries within the first valid time range is calculated. The historical total electricity consumption ratio of each industry is determined according to the first total electricity consumption and the second total electricity consumption.

[0078] In this embodiment, industry distribution information refers to the distribution status of different industries in a specific area, including the quantity, proportion, distribution characteristics of each industry and the relative scale between industries, which is usually obtained through market research, statistical data or industry reports.

[0079] In this embodiment, the main level refers to a method of classifying and stratifying different industries according to industry distribution information, which is divided into primary industries (such as agriculture and mining), secondary industries (such as manufacturing and construction) and advanced industries (such as services and high-tech industries).

[0080] In this embodiment, the multi-faceted characteristics refer to multiple factors that affect the types of industries and their electricity demand in the region. These characteristics may include economic development level, population density, geographical location, energy policy, market demand, technology level, etc.

[0081] In this embodiment, the one-to-one correspondence process is the process of matching historical electricity consumption data with identified industry categories, including: obtaining a list of historical electricity consumption data and industry categories from a merged source, adding corresponding industry tags (such as "manufacturing", "service industry", etc.) to each historical electricity consumption record based on the source of the historical data or other contextual information, and organizing the historical electricity consumption data with industry tags into a structured data set.

[0082] The working principle and beneficial effects of the above technical solution are: divide the industry level and type according to the industry distribution information in the region, correspond the historical electricity consumption data of the merged source with the industry type and add industry labels, calculate the total electricity consumption of each industry within the first effective time range, and then determine the historical total electricity consumption ratio of each industry, improve the data accuracy of the prediction model, accurately reflect the electricity consumption characteristics of each industry, and enhance the adaptability to various complex factors.

[0083] Embodiment 4:

[0084] The embodiment of the present invention provides a method for predicting regional electricity consumption based on industry division, confirming first historical electricity consumption data in an internal data source, and at the same time confirming second historical electricity consumption data in an external data source, including:

[0085] Determine a data type of power demand, and based on the data type of power demand, determine an external source that can provide historical power consumption data of a corresponding type to constitute an external data source;

[0086] An internal system storing historical data related to power demand is identified and confirmed, and an internal data source is formed based on the internal system.

[0087] In this embodiment, the power demand data types include regional power consumption, industry power consumption, time series data, load curves, and customer power consumption patterns.

[0088] In this embodiment, external sources include power companies, statistical yearbooks, market research institutions, environmental monitoring institutions, and industry associations; internal systems include energy management systems, customer relationship management systems, financial systems, monitoring systems, and data warehouses.

[0089] In this embodiment, the relevant historical data includes electricity consumption history, load history data, economic activity data, weather data, policy and regulatory change records, and user behavior data.

[0090] The working principle and beneficial effects of the above technical solution are: by determining the type of power demand data, then identifying the external source that can provide historical power consumption data to form an external data source, and at the same time, confirming the internal system that stores historical data related to power demand to form an internal data source. Integrating internal and external data sources for subsequent power consumption analysis and forecasting will help to quickly adapt to market and policy changes, improve the system's response speed and flexibility, and improve the overall efficiency of the power system.

[0091] Embodiment 5:

[0092] The embodiment of the present invention provides a method for predicting regional electricity consumption based on industry division, obtains the industrial structure in the region, evaluates the change of the industrial structure in combination with relevant regional policies, determines the first impact of the change of the industrial structure on the power demand, and performs a first prediction of the regional electricity consumption in combination with the first impact and the historical total electricity consumption ratio, including:

[0093] Taking the first effective time range as a base, determining a second effective time range, acquiring industrial structure data according to the second effective time range, identifying trends and patterns of industry changes based on the industrial structure data, and determining the industrial structure of the region within the second effective time range;

[0094] Evaluate the changes in the industrial structure based on the trends and patterns of the changes in the industry and the relevant policies within the second effective time frame, and determine the first relationship between different types of industries and the second relationship between the policies and the changes in different types of industries based on the evaluation results;

[0095] Identify the electricity consumption patterns of all industries according to the first and second correlations, evaluate the circuit demand of the corresponding industry by combining the historical electricity consumption data of each industry with the electricity consumption pattern, and obtain the first impact of industrial structure changes on electricity demand;

[0096] Based on the historical total electricity consumption ratio in the region, a prediction reference point is determined, and a first prediction of regional electricity consumption is made in combination with the first impact.

[0097] In this embodiment, the second effective time range is based on the first effective time range. According to different analysis purposes and data availability, the time period that can provide industrial structure data is checked to ensure that there is sufficient data in the selected time period to support the analysis, and the time window for changes in policies, economy and market activities is determined to more accurately capture relevant industrial changes and impacts. Combined with the above judgments, the second effective time range that is most suitable for studying industrial structure changes is confirmed.

[0098] In this embodiment, the industrial structure data generally include: output value of each industry, industry classification, number of employees in the industry, industry profit margin, and investment data.

[0099] In this embodiment, industry change trends and patterns refer to: changes in economic data: the growth, decline or stability trend of an industry in a specific time period, electricity consumption characteristics: the trend of electricity consumption in various industries changing over time (such as seasonality, cyclicality), technology and policy impacts: how the application of new technologies and the implementation of environmental policies affect the electricity demand of different industries, and market share changes: the increase or decrease in market share of certain industries reflects the market competition situation and changes in demand.

[0100] In this embodiment, the step of determining the industrial structure within the second effective time range is as follows: obtaining data related to the industrial structure within the second effective time range, performing trend analysis on the above data, identifying major changes, combining policy changes within the time period, analyzing their impact on the industrial structure, and comprehensively analyzing the results to obtain the characteristics of the industrial structure within the time range.

[0101] In this embodiment, the first relationship is the mutual influence between different types of industries, for example, the electricity demand between the manufacturing industry and the service industry is related to each other and how they influence each other; the second relationship is the impact between policy changes and different industries, for example, a certain policy may promote the development of the green energy industry, but have the opposite impact on traditional energy companies.

[0102] In this embodiment, the process of identifying electricity usage patterns is to collect historical electricity usage data from various industries, use statistical analysis or machine learning methods to identify electricity usage patterns (such as peak electricity usage periods, average load, etc.), compare data from different periods, find changes in electricity usage patterns, and establish corresponding electricity usage models based on the identified patterns for subsequent analysis and prediction.

[0103] In this embodiment, the prediction reference point refers to the historical electricity consumption data or proportion referenced when making electricity demand predictions, and is usually calculated based on the historical total electricity consumption within a specific time range so as to serve as a starting point for a new prediction.

[0104] In this embodiment, the first prediction result is a regional electricity consumption prediction value obtained based on the above analysis, especially in combination with the impact of changes in industrial structure.

[0105] In this embodiment, the first impact refers to the direct impact of industrial structure changes on electricity demand, reflecting the changes in electricity demand of different types of industries during the structural adjustment process, such as the effect of an increase in electricity consumption in a certain industry leading to an increase or decrease in overall electricity demand.

[0106] The working principle and beneficial effects of the above technical solution are: by determining the first effective time range and the second effective time range, obtaining industrial structure data, identifying industry change trends and patterns, combining policies to evaluate industrial structure changes, determining the relationships between industries and the relationship between policies and industry changes, identifying industry electricity consumption patterns based on these relationships, and combining historical electricity consumption data to evaluate industry electricity demand, and deriving the impact of industrial structure changes on electricity demand, so as to determine the forecast benchmark point for regional electricity consumption forecasts, thereby improving the power system's response speed and flexibility to market and policy adjustments, and helping to rationally allocate power resources and reduce energy waste.

[0107] Embodiment 6:

[0108] The embodiment of the present invention provides a method for predicting regional electricity consumption based on industry division, which collects external factor data affecting electricity demand in a region, quantifies the second impact of the external factor data on electricity demand, builds an external model based on the second impact, and uses the external model to perform a second prediction on regional electricity consumption, including:

[0109] Determine external factors that may affect electricity demand in the region, perform correlation analysis between the external factors and electricity demand, and determine the secondary impact of different external factors on electricity demand;

[0110] Determine the regression coefficient of each external influencing factor according to the external factors, power demand and the second impact, evaluate the regression coefficient to obtain a coefficient evaluation result, and determine the degree of influence of each external factor on power demand according to the coefficient evaluation result;

[0111] An external model is constructed based on the impact degree and the second impact, and a second prediction of regional electricity consumption is made using the external model.

[0112] In this embodiment, external factors refer to factors that may affect electricity demand, including: economic factors: such as GDP growth rate, industrial output value, employment rate and other indicators reflecting economic activities; climate factors: such as temperature, precipitation, humidity and other meteorological data that have a greater impact on electricity demand; population factors: such as population density, residential electricity consumption characteristics, population growth rate, etc.; policy factors: such as energy policies, environmental protection regulations, subsidy policies and other external policies involving the power industry; technological changes: such as new impacts on electricity demand such as distributed power generation, smart grids and energy storage technologies; market factors: such as electricity price fluctuations, market supply and demand relationships, etc.

[0113] In this embodiment, the process of correlation analysis is to collect historical data on external factors and electricity demand, including the values ​​of each external factor and the corresponding electricity demand records, use correlation analysis methods (such as Pearson correlation coefficient, Spearman rank correlation, etc.) to calculate the correlation between external factors and electricity demand, build a multivariate linear regression model, and analyze the comprehensive impact of external factors on electricity demand; the results of correlation analysis: through correlation analysis, the correlation strength between different external factors and electricity demand can be determined, and high correlation factors are considered to have a significant impact on electricity demand, while low correlation factors may not be the main factors in the impact.

[0114] In this embodiment, the second impact refers to analyzing the impact of changes in external factors on electricity demand based on the relationship between these external factors and electricity demand. Specifically, it is manifested as the specific impact of changes in various external factors (such as economic growth or temperature changes) on changes in electricity demand. For example, rising temperatures in summer may lead to increased use of air conditioners, thereby increasing electricity demand.

[0115] In this embodiment, the second prediction result is a predicted value of regional electricity consumption obtained by combining an external model. In this model, the influence degree of each external factor determined based on the above analysis is calculated in combination with the actual observed value of the external factor.

[0116] The working principle and beneficial effects of the above technical solution are: by determining the external factors affecting electricity demand and conducting correlation analysis to evaluate their impact on electricity demand, determining the regression coefficients of each external factor through regression analysis, and evaluating the accuracy and impact of these coefficients, building an external model based on the impact level and the second impact of the external factors, and using the model to predict regional electricity consumption, thereby improving the realistic adaptability of the prediction model, providing a scientific basis for electricity management and scheduling, and supporting the optimal configuration of the regional power system.

[0117] Embodiment 7:

[0118] The embodiment of the present invention provides a method for predicting regional electricity consumption based on industry division, constructing an external model based on the impact degree and the second impact, and using the external model to perform a second prediction on regional electricity consumption, including:

[0119]

[0120] in, represents the predicted value of electricity demand; β0 represents the predicted value of electricity demand when the external factor is zero; Indicates when X j The change in the power demand forecast value when one unit value is added; X j represents the actual observed value of the jth external factor; γj Indicates when X j The influence of the square term on electricity demand; jk represents the overall impact of the interaction between the jth external factor and the kth external factor on electricity demand; (X j ×X k ) represents the impact of the jth external factor and the kth external factor on the power demand when they interact, and k≠j; ω j represents the weight of the jth external factor; represents the intensity of the impact of the jth external factor on electricity demand; u represents the contribution value of unobserved factors to electricity demand; η represents the gap between the model prediction and the actual value; p represents the total number of external factors.

[0121] The working principle and beneficial effects of the above technical solution are: through the regression model, the impact of external factors on electricity demand is quantified, the benchmark electricity demand value and the weight of external factors are set, the individual impact, square term impact and interaction effect of each external factor are considered, and a prediction formula is constructed. The actual observed values ​​of external factors are combined to calculate the power demand forecast value, and the impact of unobserved factors on electricity demand is evaluated at the same time, which significantly improves the accuracy of power demand forecasting and meets the needs of modern power systems for high-precision real-time forecasting.

[0122] Embodiment 8:

[0123] The embodiment of the present invention provides a method for predicting regional electricity consumption based on industry classification, which integrates a first prediction result and a second prediction result to form a comprehensive prediction model, and realizes prediction of regional electricity consumption based on industry type, including:

[0124] Calculating the error correlation value between the first prediction result, the second prediction result and the actual power demand as the first judgment condition;

[0125] Calculating the determination coefficient between the first prediction result, the second prediction result and the actual power demand as the second judgment condition;

[0126] The accuracy of the comprehensive prediction model is judged based on the first judgment condition and the second judgment condition to obtain a second judgment result, and the regional electricity consumption of different industries in the region is output based on the second judgment result.

[0127] In this embodiment, the error correlation value is a statistical indicator for measuring the difference between the prediction result and the actual power demand, which is usually achieved by calculating the difference between the prediction value and the actual value, including: root mean square error, mean absolute error and correlation coefficient.

[0128] In this embodiment, the determination coefficient is a statistical indicator that measures the model's ability to explain data variation, indicating the proportion of actual electricity demand that can be explained by the predicted results, and its value is between 0 and 1: determination coefficient = 1: indicating that the model can perfectly explain all data variation (perfect fit), determination coefficient = 0: indicating that the model cannot explain any data variation (invalid model), a high determination coefficient value indicates that the model fits the data variation well, otherwise it indicates that the model effect is poor.

[0129] In this embodiment, the second judgment result is a comprehensive evaluation based on the calculated first judgment condition (error correlation value) and the second judgment condition (determination coefficient), which is used to judge the accuracy of the comprehensive prediction model: Custom standards: Determine whether the model meets the accuracy requirements based on a series of pre-set standards. For example, a threshold is set. If the determination coefficient is higher than a certain value and the error is lower than a certain level, the model can be considered accurate. If both indicate that the model has a high accuracy, then the second judgment result is considered to be valid. Otherwise, it is marked as an invalid model or a model that needs to be improved.

[0130] The working principle and beneficial effects of the above technical solution are: by calculating the error correlation value and determination coefficient between the first prediction result, the second prediction result and the actual electricity demand, the accuracy of the comprehensive prediction model is evaluated, and based on the judgment results of the error and the determination coefficient, the regional electricity consumption forecast of different industries in the region is output, the performance of the model and its error sources are accurately understood, energy waste is reduced, and the overall efficiency and flexibility of the power system are improved.

[0131] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A regional electricity consumption prediction method based on industry division, characterized in that: include: Step 1: clarify the forecast time range, collect historical electricity consumption data in the region based on the forecast time range, divide the industries in the region, and determine the historical total electricity consumption proportion of each industry based on the historical electricity consumption data; Step 2: Obtain the industrial structure in the region, and evaluate the changes in the industrial structure in combination with relevant regional policies, determine the first impact of the changes in the industrial structure on electricity demand, and make a first prediction of regional electricity consumption in combination with the first impact and the historical total electricity consumption ratio; Step 3: Collect data on external factors that affect electricity demand in the region, quantify the second impact of the external factor data on electricity demand, build an external model based on the second impact, and use the external model to make a second prediction of regional electricity consumption; Step 4: Integrate the first prediction result with the second prediction result to form a comprehensive prediction model to achieve prediction of regional electricity consumption based on industry type.

2. The method for predicting regional electricity consumption based on industry division according to claim 1 is characterized in that: Specify the forecast time range and collect historical electricity consumption data in the region based on the forecast time range, including: The forecast time range is used to clearly forecast the forecast time granularity required, and a minimum time range is defined according to the forecast time range and the forecast time granularity; Confirming first historical power usage data in an internal data source, and at the same time, confirming second historical power usage data in an external data source; Check whether the first historical electricity consumption data and the second historical electricity consumption data cover the minimum time range, and at the same time determine whether the historical time granularity corresponding to the first historical electricity consumption data and the second historical electricity consumption data meets the requirements corresponding to the prediction time granularity to obtain a first judgment result; based on the inspection result and the first judgment result, merge the internal data source and the external data source to obtain a merged source, and determine the first effective time range in combination with the coverage and quality of the merged source and the minimum time range; and obtain the historical electricity consumption data of the merged source in the area according to the first effective time range.

3. The method for predicting regional electricity consumption based on industry division according to claim 2 is characterized in that: The industries in the region are divided, and the proportion of historical total electricity consumption of each industry is determined based on the historical electricity consumption data, including: Obtaining industry distribution information within the region, dividing the main levels of the industries within the region according to the industry distribution information, and determining the types of industries within the region according to the main levels and various characteristics within the region; Match the historical electricity consumption data obtained from the merged source with the industry categories one by one, and add a corresponding industry label to each piece of historical electricity consumption data; The first total electricity consumption of each industry within the first valid time range is calculated based on the industry tags, and the second total electricity consumption of all industries within the first valid time range is calculated. The historical total electricity consumption ratio of each industry is determined according to the first total electricity consumption and the second total electricity consumption.

4. The method for predicting regional electricity consumption based on industry division according to claim 2 is characterized in that: Confirming the first historical power usage data in the internal data source and confirming the second historical power usage data in the external data source, including: Determine a data type of power demand, and based on the data type of power demand, determine an external source that can provide historical power consumption data of a corresponding type to constitute an external data source; An internal system storing historical data related to power demand is identified and confirmed, and an internal data source is formed based on the internal system.

5. The method for predicting regional electricity consumption based on industry division according to claim 2 is characterized in that: Obtain the industrial structure in the region, and evaluate the changes in the industrial structure in combination with relevant regional policies, determine the first impact of the changes in the industrial structure on electricity demand, and make a first prediction of regional electricity consumption in combination with the first impact and the historical total electricity consumption ratio, including: Taking the first effective time range as a base, determining a second effective time range, acquiring industrial structure data according to the second effective time range, identifying trends and patterns of industry changes based on the industrial structure data, and determining the industrial structure of the region within the second effective time range; Evaluate the changes in the industrial structure based on the trends and patterns of the changes in the industry and the relevant policies within the second effective time frame, and determine the first relationship between different types of industries and the second relationship between the policies and the changes in different types of industries based on the evaluation results; Identify the electricity consumption patterns of all industries according to the first and second correlations, evaluate the circuit demand of the corresponding industry by combining the historical electricity consumption data of each industry with the electricity consumption pattern, and obtain the first impact of industrial structure changes on electricity demand; Based on the historical total electricity consumption ratio in the region, a prediction reference point is determined, and a first prediction of regional electricity consumption is made in combination with the first impact.

6. The method for predicting regional electricity consumption based on industry division according to claim 1 is characterized in that: Collecting data of external factors affecting power demand in the region, quantifying the second impact of the external factor data on power demand, building an external model based on the second impact, and using the external model to make a second prediction of regional power consumption, including: Determine external factors that may affect electricity demand in the region, perform correlation analysis between the external factors and electricity demand, and determine the secondary impact of different external factors on electricity demand; Determine the regression coefficient of each external influencing factor according to the external factors, power demand and the second impact, evaluate the regression coefficient to obtain a coefficient evaluation result, and determine the degree of influence of each external factor on power demand according to the coefficient evaluation result; An external model is constructed based on the impact degree and the second impact, and a second prediction of regional electricity consumption is made using the external model.

7. The method for predicting regional electricity consumption based on industry division according to claim 6 is characterized in that: Building an external model based on the impact degree and the second impact, and using the external model to make a second prediction of regional electricity consumption, including: in, represents the predicted value of electricity demand; β0 represents the predicted value of electricity demand when the external factor is zero; Indicates when X j The change in the power demand forecast value when one unit value is added; X j represents the actual observed value of the jth external factor; γ j Indicates when X j The influence of the square term on electricity demand; jk represents the overall impact of the interaction between the jth external factor and the kth external factor on electricity demand; (X j ×X k ) represents the impact of the jth external factor and the kth external factor on the power demand when they interact, and k≠j; ω j represents the weight of the jth external factor; represents the intensity of the impact of the jth external factor on electricity demand; u represents the contribution value of unobserved factors to electricity demand; η represents the gap between the model prediction and the actual value; p represents the total number of external factors.

8. The method for predicting regional electricity consumption based on industry division according to claim 1 is characterized in that: The first prediction result and the second prediction result are integrated to form a comprehensive prediction model to realize the prediction of regional electricity consumption based on industry types, including: Calculating the error correlation value between the first prediction result, the second prediction result and the actual power demand as the first judgment condition; Calculating the determination coefficient between the first prediction result, the second prediction result and the actual power demand as the second judgment condition; The accuracy of the comprehensive prediction model is judged based on the first judgment condition and the second judgment condition to obtain a second judgment result, and the regional electricity consumption of different industries in the region is output based on the second judgment result.

Images