Power consumption fluctuation reason positioning method and system based on multi-factor analysis

Abstract

The invention discloses a power consumption fluctuation reason positioning method and system based on multi-factor analysis, and belongs to the technical field of power system monitoring and analysis, and the method comprises the steps: collecting influence factor data related to power consumption fluctuation in a certain region; after cleaning the collected influence factor data, constructing a data set corresponding to the influence factor index system, and then carrying out missing value repair and standardization processing on the data set; constructing a power consumption fluctuation factor decomposition model according to the processed influence factor data set; calculating the contribution rate of each influence factor to the power consumption fluctuation based on the constructed power consumption fluctuation factor decomposition model; sorting the calculated contribution rates of the influence factors from high to low, and setting a contribution rate threshold to screen out the influence factors which have obvious influence on the total fluctuation; and according to sorting of the screened contribution rates, key influence factors influencing power consumption fluctuation can be positioned.

Description

Technical Field

[0001] The present invention belongs to the technical field of power system monitoring and analysis, and in particular relates to a method and system for locating the cause of power consumption fluctuation based on multi-factor analysis. Background Art

[0002] With the rapid development of social economy and the continuous increase in energy demand, the power system is facing increasingly complex load fluctuations and management challenges. The power consumption of various electricity users (such as industry, commerce and residents) is affected by various factors such as seasonal changes, meteorological conditions, macroeconomic situation, policy adjustments, etc., resulting in obvious fluctuations in power consumption. This fluctuation not only directly affects the economy and stability of the power system, but also puts higher requirements on power dispatching, load forecasting and energy management.

[0003] However, existing methods for analyzing electricity fluctuations are usually limited to a single factor or a simple model, making it difficult to comprehensively and systematically identify the combined impact of multiple factors on electricity consumption. In particular, many methods are unable to consider the cross-effects of complex factors such as meteorological conditions, economic conditions, and policies in the same model, resulting in one-sided analysis results and difficulty in accurately locating the dominant factors. At the same time, traditional offline analysis methods have a slow response speed and are unable to meet the needs of rapid positioning and real-time monitoring of electricity consumption fluctuations, thereby limiting the flexibility and response capabilities of the power system. In addition, for factors such as seasonality, meteorological conditions, industrial restructuring, and policy changes, existing technologies lack scientific quantitative methods, making it difficult to effectively combine the fluctuation characteristics of each factor and its actual impact on electricity consumption, resulting in low analysis accuracy.

[0004] The prior art document CN118014604A discloses a quantitative evaluation method for electricity fluctuation factors based on causal test and regression model, including: obtaining daily actual electricity consumption and monthly actual electricity consumption in historical months; constructing daily and monthly electricity consumption influencing factor systems respectively; constructing daily electricity consumption basic model to obtain the influence ratio of ecological factors on daily actual electricity consumption; constructing daily electricity consumption influencing model to obtain the influence ratio of social factors or political factors on daily actual electricity consumption; constructing monthly electricity consumption influencing model to obtain the influence ratio of industry development factors or economic factors on monthly actual electricity consumption; predicting the contribution rate of each influencing factor to the monthly electricity consumption of the current month and ranking them according to the influence ratio of each influencing factor on the actual electricity consumption, so as to output the quantitative evaluation result of each influencing factor on the monthly electricity consumption fluctuation of the current month. Its shortcomings are that it fails to comprehensively analyze the cross-effect of multiple factors, mainly focuses on the static quantification of a single dimension, lacks in-depth research on the dynamic relationship and hysteresis effect between complex factors, and at the same time, its analysis method is mostly an offline model with slow response speed, which is difficult to meet the real-time monitoring and rapid positioning requirements of electricity fluctuation. In addition, its dynamic adaptability to factors such as seasonal changes, meteorological conditions, and policy adjustments is poor, which limits its power regulation capabilities and practical application value in responding to emergencies. Summary of the invention

[0005] In order to solve the deficiencies in the prior art, the present invention provides a method and system for locating the cause of power consumption fluctuation based on multi-factor analysis.

[0006] The present invention adopts the following technical solution.

[0007] A first aspect of the present invention provides a method for locating the cause of power consumption fluctuation based on multi-factor analysis, comprising the following steps:

[0008] Collect data on factors affecting electricity fluctuations in a certain region, including seasonal and meteorological factors, macroeconomic factors, policy adjustment factors, and electricity price factors. Each factor contains multiple indicators.

[0009] After cleaning the collected influencing factor data, we construct an influencing factor data set, and then perform missing value repair and normalization on the data set;

[0010] According to the processed influencing factor data set, the total power consumption fluctuation is decomposed into the superposition effect of multiple influencing factors, and the power consumption fluctuation factor decomposition model is constructed;

[0011] Based on the decomposition model of electricity consumption fluctuation factors, calculate the contribution rate of each influencing factor to electricity consumption fluctuation;

[0012] Sort the calculated contribution rates of various influencing factors from high to low, and set contribution rate thresholds to filter the contribution rate rankings;

[0013] According to the contribution rate ranking after screening, the key factors affecting the fluctuation of electricity consumption can be located.

[0014] Optionally, each of the collected influencing factors includes multiple indicators, among which policy adjustment factors include energy conservation and emission reduction and industrial structure adjustment indicators, macroeconomic factors include GDP growth rate and industrial added value indicators, seasonal and meteorological factors include temperature, humidity, climate and seasonal indicators, and electricity price factors include peak and valley electricity prices, real-time electricity prices and residential and industrial electricity price indicators.

[0015] Optionally, the missing value repair and normalization processing of the influencing factor data set is specifically as follows:

[0016] Use regression interpolation to interpolate the collected data and build a regression model as follows:

[0017] Y=β0+β1X1+β2X2+···+β p X p +ε

[0018] Among them, Y is the interpolation data to be repaired, X1, X2…X p is time, β0, β1, …β p is the model parameter, ε is the error term;

[0019] After interpolation, the mean square error is used to measure the accuracy of the interpolation model. The calculation formula of the mean square error is:

[0020]

[0021] Where n is the total number of observations, y i is the ith observation, is the interpolated value of the ith observation;

[0022] The accuracy of the interpolation process requires that the MSE is less than the set threshold and the goodness of fit of the regression model R 2 Not less than 0.9;

[0023] According to the above standards, the model parameters are continuously adjusted until the accuracy requirements are met, and finally the optimized interpolation model is used to interpolate the missing values ​​in the influencing factor data set;

[0024] The influencing factor data set after interpolation is standardized based on the Min-Max standardization method. The formula is as follows:

[0025]

[0026] Where x is the original data value, min(x) is the smallest value in the data set, max(x) is the largest value in the data set, and x′ is the standardized data value, and the range of x′ is [0,1].

[0027] Optionally, the constructed power consumption fluctuation factor decomposition model is W, and its expression is:

[0028] W=W po +W e +W m +W pr +ε

[0029] Among them, W po represents the impact of policy adjustment factors on electricity consumption, W e Represents the impact of macroeconomic factors on electricity consumption, W m Indicates the impact of seasonal and meteorological factors on electricity consumption, W pr represents the impact of electricity price factors on electricity consumption, and ε is the error term, which is used to describe the impact of other included factors.

[0030] Optionally, the policy adjustment factor W po It means that when the policy is adjusted, it will affect the electricity demand. The model is constructed as follows:

[0031] W po =α po ·P t +∈ po

[0032] where α po represents the influence coefficient of policy adjustment factors on electricity consumption, ∈ po Represents the random error of this part, P t It indicates the electricity demand under normal conditions without the influence of policy factors.

[0033] Optionally, the macroeconomic factor W e It means that macroeconomic conditions directly affect the level of industry activities, which in turn affects electricity demand. The model is constructed as follows:

[0034] W e =β gdp GDP t +β iv IV t +∈ e

[0035] Among them, β gdp and β iv Respectively represent the impact coefficients of GDP and industrial added value on electricity consumption, ∈ e is a random error, GDP t represents the base electricity consumption without the influence of GDP factors, IVt It represents the benchmark electricity consumption without the influence of industrial value-added factors.

[0036] Optionally, the season and meteorological factors W m It refers to the impact of meteorological factors on seasonal electricity demand. The model is constructed as follows:

[0037] W m =γ temp Temp t +γ humidity Humidity t +∈ m

[0038] Among them, γ temp and γ humidity Respectively represent the influence coefficients of temperature and humidity on power consumption, ∈ m Indicates the random error of meteorological factors, Temp t Indicates the benchmark power consumption at an appropriate temperature. Humidity t Indicates the baseline power consumption under appropriate humidity.

[0039] Optionally, the electricity price factor W pr It means that electricity price fluctuations will affect the electricity consumption strategy of enterprises. The model is constructed as follows:

[0040] W pr =δ price Price t +δ price_diff Price_Diff+∈ pr

[0041] Among them, δ price Indicates the influence coefficient of electricity price fluctuation on electricity consumption, δ price_diff Indicates the impact coefficient of electricity price fluctuation on electricity consumption, Price t represents the benchmark electricity consumption before the electricity price fluctuation, and Price_Diff represents the benchmark electricity consumption after the electricity price fluctuation, ∈ pr Represents the random error of electricity price fluctuation.

[0042] Optionally, the calculation of the contribution rate of each influencing factor to the power consumption fluctuation based on the power consumption fluctuation factor decomposition model specifically includes:

[0043] The contribution rates of policy adjustment factors, macroeconomic factors, seasonal and meteorological factors, and electricity price factors to electricity consumption fluctuations are calculated respectively. The contribution rate of each influencing factor to electricity consumption fluctuations is calculated using the following calculation formula:

[0044]

[0045] Where ηi Indicates the contribution rate of the i-th factor, ΔW i It represents the power consumption fluctuation value caused by this factor, and ΔW represents the total power consumption fluctuation value.

[0046] The second aspect of the present invention provides a system for locating the cause of power consumption fluctuation based on multi-factor analysis, which is used to implement the method for locating the cause of power consumption fluctuation based on multi-factor analysis described in the first aspect of the present invention, and the system includes:

[0047] Data collection module, data processing module, decomposition model building module, contribution rate calculation model, contribution rate ranking module and analysis module;

[0048] The data collection module is used to collect influencing factor data, the data processing module is used to process the influencing factor data collected by the data collection module and construct an influencing factor data set, the decomposition model construction module constructs an electricity consumption fluctuation factor decomposition model based on the influencing factor data set, and the contribution rate calculation model calculates the contribution rate of each influencing factor to the electricity consumption fluctuation according to the electricity consumption fluctuation factor decomposition model; the contribution rate sorting module is used to sort the calculated contribution rates of each influencing factor from high to low, and set the contribution rate threshold to screen out the influencing factors that have a significant impact on the total fluctuation; the analysis module analyzes the causes of electricity consumption fluctuations according to the filtered contribution rate sorting.

[0049] The beneficial effect of the present invention is that, compared with the prior art, it solves the limitations of single factor analysis and static models, and proposes a method for evaluating electricity consumption fluctuations that can comprehensively quantify the interaction of multiple factors and hysteresis effects. By introducing a dynamic hysteresis model, the present invention can not only quantify the direct impact of factors such as policy adjustments, economic conditions, and meteorological conditions, but also evaluate the hysteresis effects of these factors in the time dimension, greatly improving the depth and scientificity of the analysis. In addition, the present invention constructs an indicator system with high scalability, which can flexibly add emerging factors such as new energy access and market behavior to meet the needs of the continuous development of complex power systems. Compared with the problem that the prior art relies on offline analysis and has a slow response speed, the present invention combines real-time data processing with rapid modeling technology, can quickly locate the main cause of electricity consumption fluctuations, and realize real-time monitoring and rapid response. Compared with the prior art, the present invention has significantly improved in terms of analysis accuracy, real-time performance, dynamic adaptability, and practical application value, providing technical guarantees for the flexible management and long-term development of modern power systems, and has broad promotion prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is an overall flow chart of the method of the present invention. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the scope of protection of the present invention.

[0052] Driven by energy conservation and emission reduction policies, power control departments and power users are in urgent need of a technology that can quickly and accurately analyze the causes of power consumption fluctuations to achieve scheduling optimization and energy-saving decisions. Therefore, the present invention proposes a comprehensive method and system for locating the causes of power consumption fluctuations. By constructing indicators for multiple factors such as seasonality, meteorological conditions, economic situation, and policy adjustments, a fluctuation cause analysis model is established to quantify the contribution rate of each factor to power consumption changes, so that the main cause of the fluctuation can be quickly identified.

[0053] The purpose of the present invention is to overcome the shortcomings and deficiencies of the prior art and to provide a method and system for locating the cause of power consumption fluctuation based on multi-factor analysis to improve the stability of the power system.

[0054] To achieve the above object, the present invention provides a method for locating the cause of power consumption fluctuation based on multi-factor analysis in embodiment 1, and the method comprises the following steps:

[0055] Step S1: Collect data on influencing factors related to power fluctuations in a certain region, including seasonal and meteorological factors, macroeconomic factors, policy adjustment factors, and electricity price factors, etc. Each influencing factor includes multiple indicators suitable for analysis;

[0056] In a further preferred but non-limiting embodiment, in step S1, each influencing factor collected includes multiple indicators, specifically:

[0057] Influencing factors index Policy adjustment factors Energy conservation and emission reduction, industrial structure adjustment, etc. Macroeconomic factors GDP growth rate, industrial added value, etc. Seasonal and meteorological factors Temperature, humidity, climate, season, etc. Electricity price factors Peak and valley electricity prices, real-time electricity prices, residential and industrial electricity prices, etc.

[0058] In a further preferred but non-limiting embodiment, the influencing factors collected in step S1 may also include international energy price factors and population growth factors.

[0059] Step S2: After cleaning the influencing factor data collected in step S1, construct a data set corresponding to the influencing factor indicator system, and then repair missing values ​​and normalize the data set to ensure that data of different dimensions can be used for unified analysis.

[0060] In a further preferred but non-limiting embodiment, in step S2, the missing value repair and normalization processing of the influencing factor data set specifically includes:

[0061] The collected data were interpolated using the regression interpolation method, and the regression model was established as follows:

[0062] Y=β0+β1X1+β2X2+···+β p X p +ε (1)

[0063] Among them, Y is the interpolation data to be repaired, X1, X2…X p is time, β0, β1, …β p are model parameters and ε is the error term.

[0064] After interpolation, the mean square error is used to measure the accuracy of the interpolation model. The calculation formula of the mean square error is:

[0065]

[0066] Where n is the total number of observations, y i is the ith observation, is the interpolated value of the ith observation.

[0067] The accuracy of the interpolation process requires that the MSE is less than the set threshold (such as MSE ≤ 0.01), and the goodness of fit of the regression model R 2 It should not be less than 0.9 to ensure that the model can accurately explain the data characteristics. According to the above standards, the model parameters are continuously adjusted until the accuracy requirements are met, and finally the optimized interpolation model is used to interpolate the missing values ​​in the influencing factor data set.

[0068] The data set is standardized based on the Min-Max standardization method to ensure that the data sets of influencing factors of different dimensions are suitable for unified analysis. The specific formula is as follows:

[0069]

[0070] Where x is the original data value, min(x) is the smallest value in the data set, max(x) is the largest value in the data set, and x′ is the standardized data value, and the range of x′ is [0,1].

[0071] Step S3: Based on the influencing factor data set processed in step S2, the total power consumption fluctuation is decomposed into the superposition effect of multiple influencing factors, and a power consumption fluctuation factor decomposition model is constructed.

[0072] In a further preferred but non-limiting embodiment, in step S3, the power consumption fluctuation factor decomposition model constructed is W, and its expression formula is:

[0073] W=W po +W e +Wm +W pr +ε

[0074] Among them, W po represents the impact of policy adjustment factors on electricity consumption, W e Represents the impact of macroeconomic factors on electricity consumption, W m Indicates the impact of seasonal and meteorological factors on electricity consumption, W pr represents the impact of electricity price factors on electricity consumption, ε is the error term, which is used to describe the impact of other included factors;

[0075] Specifically, the policy adjustment factor W po It means that when policies are adjusted, electricity demand will be affected, especially the electricity demand of key industries will be significantly affected. The following model is constructed:

[0076] W po =α po ·P t +∈ po

[0077] where α po represents the influence coefficient of policy adjustment factors on electricity consumption, ∈ po Represents the random error of this part, P t It indicates the electricity demand under normal conditions without the influence of policy factors.

[0078] Macroeconomic factors e It means that macroeconomic conditions such as GDP growth and fixed asset investment directly affect the level of industry activities, which in turn affects electricity demand. The following model can be constructed:

[0079] W e =β gdp GDP t +β iv IV t +∈ e

[0080] Among them, β gdp and β iv Respectively represent the impact coefficients of GDP and industrial added value on electricity consumption, ∈ e is a random error, GDP t represents the base electricity consumption without the influence of GDP factors, IV t It represents the benchmark electricity consumption without the influence of industrial value-added factors.

[0081] Seasonal and meteorological factors m It means that meteorological factors (such as temperature, humidity, wind speed, etc.) have a greater impact on seasonal electricity demand (such as air conditioning, electric heating, etc.). The following model can be constructed:

[0082] Wm =γ temp Temp t +γ humidity Humidity t +∈ m

[0083] Among them, γ temp and β humidity Respectively represent the influence coefficients of temperature and humidity on power consumption, ∈ m Indicates the random error of meteorological factors, Temp t Indicates the benchmark power consumption at an appropriate temperature. Humidity t Indicates the benchmark power consumption under suitable humidity;

[0084] Exemplarily, the suitable temperature includes 20° C., and the suitable humidity includes 45%.

[0085] Electricity price factor W pr It means that electricity price fluctuations will affect the electricity consumption strategy of enterprises, especially when the difference between peak and valley electricity prices is significant, the electricity consumption may be adjusted. The following model can be constructed:

[0086] W pr =δ price Price t +δ price_diff Price_Diff+∈ pr

[0087] Among them, δ price Indicates the influence coefficient of electricity price fluctuation on electricity consumption, δ price_diff Indicates the impact coefficient of electricity price fluctuation on electricity consumption, Price t represents the benchmark electricity consumption before the electricity price fluctuation, and Price_Diff represents the benchmark electricity consumption after the electricity price fluctuation, ∈ pr Represents the random error of electricity price fluctuation.

[0088] Step S4: According to the power consumption fluctuation factor decomposition model constructed in step S3, the contribution rate of each influencing factor to the power consumption fluctuation is calculated.

[0089] In a further preferred but non-limiting embodiment, step S4 specifically comprises:

[0090] The contribution rates of policy adjustment factors, macroeconomic factors, seasonal and meteorological factors, and electricity price factors to electricity consumption fluctuations are calculated respectively. The contribution rate of each influencing factor to electricity consumption fluctuations is calculated using the following formula:

[0091]

[0092] where η iIndicates the contribution rate of the i-th factor, ΔW i It represents the power consumption fluctuation value caused by this factor, and ΔW represents the total power consumption fluctuation value.

[0093] Step S5: sort the contribution rates of the various influencing factors calculated in step S4 from high to low, and set a contribution rate threshold to screen out the influencing factors that have a significant impact on the total fluctuation.

[0094] In a further preferred but non-limiting embodiment, the specific rule for setting the contribution rate threshold in step S5 is:

[0095] Arrange the contribution rate of each factor in descending order and calculate the cumulative contribution rate. When the cumulative contribution rate reaches 80%, select the current and previous factors as the main influencing factors, and the remaining factors are regarded as secondary factors or ignored. If the actual business needs emphasize higher coverage, the threshold can be raised to 90%; if it is necessary to focus on fewer core factors, it can be reduced to 70%.

[0096] Step S6: According to the contribution rate ranking after screening in step S5, the key influencing factors affecting the fluctuation of electricity consumption can be located. Then, combined with the actual situation of regional or industrial users, the causes of electricity consumption fluctuations can be determined based on the key influencing factors, and their specific impact mechanism on electricity consumption fluctuations can be explored.

[0097] In a further preferred but non-limiting embodiment, in step S6, a greater contribution rate of the influencing factor indicates a more significant degree of influence.

[0098] For example, a user in the steel industry in the region was directly shut down due to the environmental protection and production restriction policy. Combined with the actual situation, the key factor of this electricity consumption fluctuation can be directly identified as a policy factor.

[0099] In Example 2, the present invention provides a system for locating the cause of power consumption fluctuation based on multi-factor analysis, which is used to implement the method for locating the cause of power consumption fluctuation based on multi-factor analysis described in Example 1, and the system includes:

[0100] Data collection module, data processing module, decomposition model building module, contribution rate calculation model, contribution rate ranking module and analysis module;

[0101] The data collection module is used to collect influencing factor data, the data processing module is used to process the influencing factor data collected by the data collection module and construct an influencing factor data set, the decomposition model construction module constructs an electricity consumption fluctuation factor decomposition model based on the influencing factor data set, and the contribution rate calculation model calculates the contribution rate of each influencing factor to the electricity consumption fluctuation according to the electricity consumption fluctuation factor decomposition model; the contribution rate sorting module is used to sort the calculated contribution rates of each influencing factor from high to low, and set the contribution rate threshold to screen out the influencing factors that have a significant impact on the total fluctuation; the analysis module analyzes the causes of electricity consumption fluctuations according to the filtered contribution rate sorting.

[0102] The present invention constructs a multi-factor cross-analysis model, quantifies the synergistic effects of multi-dimensional factors such as policies, economy, and meteorology, and their dynamic hysteresis effects, and combines real-time data processing and rapid modeling technology to achieve accurate positioning of the main causes of electricity fluctuations. The present invention further uses a highly extensible indicator system and dynamic model to adapt to the needs of complex power scenarios, such as new energy access or distributed power markets, to provide a scientific basis for scheduling and optimization and energy-saving and emission reduction decisions. Based on this significant difference, the technical effects actually achieved by the present invention include at least: real-time and rapid identification of the main causes of electricity fluctuations, improving the stability and regulation efficiency of the power system, supporting the precise implementation of energy-saving and emission reduction measures, and significantly improving the analysis accuracy and applicability, providing an innovative solution for the flexible management of complex power systems.

[0103] 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 above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for locating the cause of power consumption fluctuation based on multi-factor analysis, characterized in that: The steps include: Collect data on factors affecting electricity fluctuations in a certain region, including seasonal and meteorological factors, macroeconomic factors, policy adjustment factors, and electricity price factors. Each factor contains multiple indicators. After cleaning the collected influencing factor data, we construct an influencing factor data set, and then perform missing value repair and normalization on the data set; According to the processed influencing factor data set, the total power consumption fluctuation is decomposed into the superposition effect of multiple influencing factors, and the power consumption fluctuation factor decomposition model is constructed; Based on the decomposition model of electricity consumption fluctuation factors, calculate the contribution rate of each influencing factor to electricity consumption fluctuation; Sort the calculated contribution rates of various influencing factors from high to low, and set contribution rate thresholds to filter the contribution rate rankings; According to the contribution rate ranking after screening, the key factors affecting the fluctuation of electricity consumption can be located.

2. The method for locating the cause of power consumption fluctuation based on multi-factor analysis according to claim 1 is characterized in that: Each of the collected influencing factors includes multiple indicators, among which policy adjustment factors include energy conservation and emission reduction and industrial structure adjustment indicators, macroeconomic factors include GDP growth rate and industrial added value indicators, seasonal and meteorological factors include temperature, humidity, climate and seasonal indicators, and electricity price factors include peak and valley electricity prices, real-time electricity prices and residential and industrial electricity price indicators.

3. The method for locating the cause of power consumption fluctuation based on multi-factor analysis according to claim 1 is characterized in that: The missing value repair and normalization processing of the influencing factor data set are specifically as follows: Use regression interpolation to interpolate the collected data and build a regression model as follows: Y=β0+β1X1+β2X2+…+β p X p +e Among them, Y is the interpolation data to be repaired, X1, X2...X p is time, β0, β1, ...β p is the model parameter, ε is the error term; After interpolation, the mean square error is used to measure the accuracy of the interpolation model. The calculation formula of the mean square error is: Where n is the total number of observations, y i is the ith observation, is the interpolated value of the ith observation; The accuracy of the interpolation process requires that the MSE is less than the set threshold and the goodness of fit of the regression model R 2 Not less than 0.9; According to the above standards, the model parameters are continuously adjusted until the accuracy requirements are met, and finally the optimized interpolation model is used to interpolate the missing values ​​in the influencing factor data set; The influencing factor data set after interpolation is standardized based on the Min-Max standardization method. The formula is as follows: Where x is the original data value, min(x) is the smallest value in the data set, max(x) is the largest value in the data set, and x′ is the standardized data value, and the range of x′ is [0,1].

4. The method for locating the cause of power consumption fluctuation based on multi-factor analysis according to claim 1 is characterized in that: The constructed power consumption fluctuation factor decomposition model is W, and its expression is: W=W po +W e +W m +W pr +ε Among them, W po represents the impact of policy adjustment factors on electricity consumption, W e Represents the impact of macroeconomic factors on electricity consumption, W m Indicates the impact of seasonal and meteorological factors on electricity consumption, W pr represents the impact of electricity price factors on electricity consumption, and ε is the error term, which is used to describe the impact of other included factors.

5. The method for locating the cause of power consumption fluctuation based on multi-factor analysis according to claim 4 is characterized in that: The policy adjustment factor W po It means that when the policy is adjusted, it will affect the electricity demand. The model is constructed as follows: W po =a po ·P t +∈ po where α po represents the influence coefficient of policy adjustment factors on electricity consumption, ∈ po Represents the random error of this part, P t It indicates the electricity demand under normal conditions without the influence of policy factors.

6. A method for locating the cause of power consumption fluctuation based on multi-factor analysis according to claim 2 or 4, characterized in that: The macroeconomic factors e It means that macroeconomic conditions directly affect the level of industry activities, which in turn affects electricity demand. The model is constructed as follows: W e =b gdp ·GDP t +b iv ·IV t +∈ e Among them, β gdp and β iv Respectively represent the impact coefficients of GDP and industrial added value on electricity consumption, ∈ e is a random error, GDP t represents the base electricity consumption without the influence of GDP factors, IV t It represents the benchmark electricity consumption without the influence of industrial value-added factors.

7. A method for locating the cause of power consumption fluctuation based on multi-factor analysis according to claim 2 or 4, characterized in that: The seasonal and meteorological factors m It refers to the impact of meteorological factors on seasonal electricity demand. The model is constructed as follows: W m =c temp ·Temp t +g humidity ·Humidity t +∈ m Among them, γ temp and γ humidity Respectively represent the influence coefficients of temperature and humidity on power consumption, ∈ m Indicates the random error of meteorological factors, Temp t Indicates the benchmark power consumption at an appropriate temperature. Humidity t Indicates the baseline power consumption under suitable humidity.

8. The method for locating the cause of power consumption fluctuation based on multi-factor analysis according to claim 4 is characterized in that: The electricity price factor W pr It means that electricity price fluctuations will affect the electricity consumption strategy of enterprises. The model is constructed as follows: W pr =δ price ·Price t +δ price_diff ·Price_Diff+∈ pr Among them, δ price Indicates the influence coefficient of electricity price fluctuation on electricity consumption, δ price_diff Indicates the impact coefficient of electricity price fluctuation on electricity consumption, Price t represents the benchmark electricity consumption before the electricity price fluctuation, and Price_Diff represents the benchmark electricity consumption after the electricity price fluctuation, ∈ pr Represents the random error of electricity price fluctuation.

9. The method for locating the cause of power consumption fluctuation based on multi-factor analysis according to claim 1 is characterized in that: The calculation of the contribution rate of each influencing factor to the power consumption fluctuation based on the power consumption fluctuation factor decomposition model specifically includes: The contribution rates of policy adjustment factors, macroeconomic factors, seasonal and meteorological factors, and electricity price factors to electricity consumption fluctuations are calculated respectively. The contribution rate of each influencing factor to electricity consumption fluctuations is calculated using the following calculation formula: Where η i Indicates the contribution rate of the i-th factor, ΔW i It represents the power consumption fluctuation value caused by this factor, and ΔW represents the total power consumption fluctuation value.

10. A system for locating the cause of power consumption fluctuation based on multi-factor analysis, used to implement a method for locating the cause of power consumption fluctuation based on multi-factor analysis as claimed in any one of claims 1 to 9, characterized in that: The system includes: Data collection module, data processing module, decomposition model building module, contribution rate calculation model, contribution rate ranking module and analysis module; The data collection module is used to collect influencing factor data, the data processing module is used to process the influencing factor data collected by the data collection module and construct an influencing factor data set, the decomposition model construction module constructs an electricity consumption fluctuation factor decomposition model based on the influencing factor data set, and the contribution rate calculation model calculates the contribution rate of each influencing factor to the electricity consumption fluctuation according to the electricity consumption fluctuation factor decomposition model; the contribution rate sorting module is used to sort the calculated contribution rates of each influencing factor from high to low, and set the contribution rate threshold to screen out the influencing factors that have a significant impact on the total fluctuation; the analysis module analyzes the causes of electricity consumption fluctuations according to the filtered contribution rate sorting.

Images