Daylight power prediction method and system based on fuzzy membership function

Abstract

The application relates to a daily power consumption prediction method based on a fuzzy membership function, which comprises the following steps: step S1: constructing a time sequence of historical daily power consumption and historical air temperature by taking date as an index; step S2: establishing a fuzzy membership function; step S3: taking the maximum Pearson correlation coefficient as an objective function, and optimizing the parameters of the fuzzy membership function through a meta-heuristic algorithm; step S4: converting the historical air temperature and the predicted air temperature into relevant factors based on the optimized fuzzy membership function, and marking the relevant factors as covariates; and step S5: establishing a time sequence prediction model, training the model by using the historical daily power consumption and the historical covariates, and finally inputting the predicted covariates to calculate the daily power consumption prediction value. The application fully utilizes weather forecast data, and effectively improves the accuracy of daily power consumption prediction of a power grid.

Description

Technical Field

[0001] This invention relates to the field of power grid power prediction technology, specifically to a daily power prediction method and system based on fuzzy membership functions. Background Technology

[0002] Electricity is an indispensable pillar of national economic development and holds a vital strategic position. Electricity consumption, as the most important indicator in the power industry, reflects electricity sales and can be used to assess economic development, possessing immense data mining value. Long-term experience shows that daily electricity consumption data exhibits certain regularities and a strong correlation with temperature. Based on historical daily electricity consumption data and weather forecasts, it is possible to predict daily electricity consumption for a future period. Power grid companies hope to use daily electricity consumption data to plan future power supply to ensure a secure and economical power supply, and they also hope to use this data to analyze economic trends and formulate corresponding policy incentives. Therefore, a method for predicting daily electricity consumption is needed. Summary of the Invention

[0003] In view of this, the purpose of this invention is to provide a method and system for predicting daily electricity consumption based on fuzzy membership functions, which makes full use of weather forecast data and effectively improves the accuracy of daily electricity consumption prediction for the power grid.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] A daily electricity consumption prediction method based on fuzzy membership functions includes the following steps:

[0006] Step S1: Construct a time series of historical daily electricity consumption and historical temperature using the date as the index;

[0007] Step S2: Establish fuzzy membership functions;

[0008] Step S3: Using the maximum Pearson correlation coefficient as the objective function, optimize the parameters of the fuzzy membership function through a metaheuristic algorithm;

[0009] Step S4: Based on the optimized fuzzy membership function, historical and predicted temperatures are transformed into relevant factors and labeled as covariates;

[0010] Step S5: Establish a time series forecasting model, train the model using historical daily electricity consumption and historical covariates, and finally input the forecast covariates to calculate the predicted daily electricity consumption value.

[0011] Furthermore, the fuzzy membership function decomposes the temperature into five intervals and establishes five sets of mapping relationships, specifically including:

[0012] Let the temperature be x, the mapping value be y, and the mapping relationship be f(·), where a, b, c, and d are all parameters of the fuzzy membership function;

[0013] Low-temperature saturation region: Consider x≤a as the low-temperature saturation region, and we have y=f(x)=ba;

[0014] Low temperature influence zone: a < x ≤ b is considered as the low temperature influence zone, and y = f(x) = bx;

[0015] Comfort zone: Consider b < x ≤ c as the comfort zone, and have y = f(x) = 0;

[0016] High-temperature influence zone: Consider c < x ≤ d as the high-temperature influence zone, and we have y = f(x) = xc;

[0017] High-temperature saturation region: Consider x > d as the high-temperature saturation region, and we have y = f(x) = dc.

[0018] Furthermore, the parameters of the fuzzy membership function include four values: a, b, c, and d. The parameters of the fuzzy membership function are optimized as follows:

[0019] (1): Randomly generate N sets of data. Each set of data contains 4 values, and all 4 values ​​are within a preset range. After sorting from smallest to largest, they are assigned the values ​​a, b, c, d respectively.

[0020] (2): For each set of data, determine the fuzzy membership function and calculate the mapping value of historical temperature through the fuzzy membership function;

[0021] (3): Match historical temperature mapping values ​​with historical daily electricity consumption on the same date to generate two sequences, and calculate the Pearson correlation coefficient;

[0022] (4): Repeat (2) and (3) until the Pearson correlation coefficients corresponding to N sets of data are calculated. Find the set of data with the smallest Pearson correlation coefficient and record it as the optimal initial value.

[0023] (5): Using the optimal initial value obtained in (4) as the initial value for iteration, design a fitness function and update the four values ​​a, b, c, d through selection, crossover, and mutation operations until the four values ​​a, b, c, d no longer change.

[0024] Furthermore, the fitness function is specifically as follows:

[0025] Let the input to the fitness function be a vector p, with elements p1, p2, p3, p4. Let a = p1, b = p2, c = p3, and d = p4, and determine the mapping relationship f(·). Let the historical temperature sequence ordered in chronological order be a vector X = {x tThe mapping value y is calculated using the mapping relationship f(·). t Calculate y t The average value is denoted as Let the historical electricity sequence ordered in chronological order be a vector Z = {z} t}, calculate z t The average value is denoted as

[0026] The fitness output r = h(p) is calculated as the fitness function using the following formula:

[0027]

[0028] +10×[max(0,p2-p3)+max(0,p3-p4)]

[0029] Furthermore, the time series has a step size of 1 day and includes 6 sub-items, specifically including:

[0030] Annual trend item: Fit the annual growth rate from historical electricity data with a step size of 1 year;

[0031] Monthly cycle item: Determine the proportion of electricity consumption in each month to the annual electricity consumption in a 12-month cycle;

[0032] Weekly cycle item: Using a 7-day cycle, determine the proportion of electricity consumption from Monday to Sunday in the weekly electricity consumption;

[0033] Statutory holiday revisions: The timeline has been revised according to the statutory holidays and work schedule adjustments issued by the State Council;

[0034] Covariate adjustment term: The influence of other time series that are strongly correlated with electricity consumption on electricity consumption is obtained by fitting the part after adjusting for actual electricity consumption - annual trend term - monthly cycle term - weekly cycle term - statutory holidays term with the covariate;

[0035] White noise term: The remaining residual terms that cannot be fitted by the above five terms satisfy a normal distribution.

[0036] Furthermore, the covariates are the correlation factors obtained after the optimized fuzzy membership function transformation of the highest and lowest temperatures.

[0037] A daily electricity consumption forecasting system based on fuzzy membership functions includes:

[0038] The data preprocessing module is used to collect, clean, query, and group daily electricity and temperature data, and construct time series data.

[0039] The data transformation module is used to convert temperature data in a time series into influencing factors.

[0040] The training module is used to build a time series prediction model and uses the above time series as input to train the model parameters, so as to obtain the expression of electricity consumption with respect to time and covariates;

[0041] The prediction module is used to input the time series to be predicted and the predicted temperature, and converts the predicted temperature into an influence factor according to the same historical degree function. Finally, it calculates the predicted electricity value based on the trained time series prediction model.

[0042] The results display module is used to show historical data and predicted data.

[0043] Furthermore, the data preprocessing module decomposes the provincial power grid electricity according to the power supply unit and the electricity consumption category. The electricity consumption category includes large industrial electricity consumption, non-general industrial electricity consumption, commercial electricity consumption, non-residential lighting electricity consumption, residential electricity consumption and agricultural electricity consumption. The power supply unit is divided according to the county and district, and the highest and lowest temperatures of the corresponding county and district are also connected to form multiple time series.

[0044] Furthermore, the data conversion module uses fuzzy membership functions to divide the highest and lowest temperatures into five zones: low-temperature saturation zone, low-temperature influence zone, comfort zone, high-temperature influence zone, and high-temperature saturation zone. The parameters of the fuzzy membership functions are optimized to convert the highest and lowest temperatures into influencing factors.

[0045] Furthermore, the training module fits expressions for annual trend terms, monthly cycle terms, weekly cycle terms, statutory holiday correction terms, covariate correction terms, and white noise terms based on the time series input data, and combines them to obtain expressions for electricity consumption with respect to time, the influence factor of the highest temperature, and the influence factor of the lowest temperature.

[0046] Compared with the prior art, the present invention has the following advantages:

[0047] 1. This invention uses fuzzy membership functions to transform temperature into an influencing factor, avoiding the difficulties caused to the prediction model by the difference in the influence trends of high and low temperatures on electricity. At the same time, it uses Pearson correlation coefficient and metaheuristic algorithm to determine the parameters of fuzzy membership functions, which can adaptively adjust the parameters according to climate change, thereby improving the accuracy and robustness of the electricity prediction model.

[0048] 2. This invention divides the time series prediction model into six parts: annual trend term, monthly cycle term, weekly cycle term, statutory holiday correction term, covariate correction term, and white noise term. It fully considers the nonlinear characteristics of the daily electricity consumption curve, improves the curve fitting accuracy, and improves the accuracy of electricity consumption prediction.

[0049] 3. This invention fully considers the differences in meteorological conditions between regions and the different correlations between different electricity consumption categories and temperature. Finally, it summarizes the daily electricity consumption forecast data by county and electricity consumption category to restore the daily electricity consumption of the whole province, which can improve the accuracy of electricity consumption forecast. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 A schematic flowchart of an embodiment of the daily electricity consumption prediction method based on fuzzy membership function provided in this application;

[0052] Figure 2 A schematic diagram of a fuzzy membership function in a specific scenario provided for this application;

[0053] Figure 3 A flowchart illustrating the optimization of fuzzy membership function parameters using a metaheuristic algorithm in a specific scenario provided in this application;

[0054] Figure 4 A schematic diagram illustrating the training and prediction of a time series prediction model in a specific scenario provided in this application;

[0055] Figure 5 A schematic diagram of the time series model components in a specific scenario provided for this application;

[0056] Figure 6 A schematic diagram of an embodiment of the daily electricity consumption forecasting system for provincial power grids provided in this application;

[0057] Figure 7 A visualization chart generated by the result display module in a specific scenario provided in this application. Detailed Implementation

[0058] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0059] Please refer to Figure 1 This invention provides a method for predicting daily electricity consumption based on fuzzy membership functions, comprising the following steps:

[0060] Step S1: Construct a time series of historical daily electricity consumption and historical temperature using the date as the index;

[0061] Step S2: Establish fuzzy membership functions;

[0062] Step S3: Using the maximum Pearson correlation coefficient as the objective function, optimize the parameters of the fuzzy membership function through a metaheuristic algorithm;

[0063] Step S4: Based on the optimized fuzzy membership function, historical and predicted temperatures are transformed into relevant factors and labeled as covariates;

[0064] Step S5: Establish a time series forecasting model, train the model using historical daily electricity consumption and historical covariates, and finally input the forecast covariates to calculate the predicted daily electricity consumption value.

[0065] In this embodiment, preferably, such as Figure 2 As shown, the fuzzy membership function can decompose the temperature into five intervals and establish five sets of mapping relationships, as follows:

[0066] Let the temperature be x, the mapping value be y, and the mapping relationship be f(·).

[0067] Low temperature saturation region (e.g.) Figure 2 As shown in Figure 201: x≤a is considered as the low-temperature saturation region, and y=f(x)=ba.

[0068] Low temperature affected areas (such as Figure 2 As shown in Figure 202): a < x ≤ b is considered as the low temperature influence zone, and y = f(x) = bx.

[0069] Comfort zone (e.g.) Figure 2 As shown in Figure 203): b < x ≤ c is considered as the comfort zone, and y = f(x) = 0.

[0070] High-temperature affected areas (such as...) Figure 2 As shown in Figure 204): c < x ≤ d is considered as the high temperature influence zone, and y = f(x) = xc.

[0071] High temperature saturation zone (e.g.) Figure 2 As shown in Figure 205): x > d is considered as the high-temperature saturation region, and y = f(x) = dc.

[0072] Where a, b, c, and d are all parameters of the fuzzy membership function, such as Figure 2 As shown in Figure 206.

[0073] refer to Figure 3 In this embodiment, the optimization process of the fuzzy membership function parameters is as follows:

[0074] Step 301: Randomly generate N sets of data, each set containing 4 values, all of which are between 0 and 40. Sort them in ascending order and assign them the values ​​a, b, c, and d respectively.

[0075] Step 302: For each set of data, determine the fuzzy membership function and calculate the mapping value of historical temperature using the fuzzy membership function.

[0076] Step 303: Match historical temperature mapping values ​​with historical daily electricity consumption for the same date to generate two sequences, and calculate the Pearson correlation coefficient.

[0077] Step 304: Repeat steps 302 and 303 until the Pearson correlation coefficients for N sets of data are calculated. Find the set of data with the smallest Pearson correlation coefficient and record it as the optimal initial value.

[0078] Step 305: Using the optimal initial value obtained in step 304 as the initial value for iteration, design the fitness function, call the genetic algorithm, and update the four values ​​a, b, c, d through selection, crossover, mutation and other operations until the four values ​​a, b, c, d no longer change.

[0079] Let the input of the fitness function be a vector p, with elements p1, p2, p3, p4. Let a = p1, b = p2, c = p3, and d = p4, and determine the mapping relationship f(·). Let the historical temperature sequence ordered in chronological order be a vector X = {x}. t The mapping value y is calculated using the mapping relationship f(·). t Calculate y t Let the average value be denoted as y, and let the historical electricity sequence ordered in chronological order be a vector Z = {z} t}, calculate z t The average value is denoted as z, and the fitness is calculated as the output r = h(p) as the fitness function according to the following formula:

[0080]

[0081] +10×[max(0,p2-p3)+max(0,p3-p4)]

[0082] In this embodiment, reference Figure 4 Construct a time series prediction model, train the model, and use the model to predict electricity consumption, specifically as follows:

[0083] A time series model was constructed using the Prophet model developed by Facebook. In the model training section 401, the variable "ds" was set as the historical date with a step size of 1 day; the variable "y" was set as the historical daily electricity consumption; the variable "add1" was set as the influence factor of the historical daily maximum temperature after transformation by a fuzzy membership function; the variable "add2" was set as the influence factor of the historical daily minimum temperature after transformation by a fuzzy membership function; annual periodicity and weekday periodicity were set; the variables "add1" and "add2" were labeled as covariates; and an additive model was used for fitting.

[0084] In the model prediction section 402, the variable "ds" is set as the prediction date with a step size of 1 day; the variable "add1" is set as the influence factor of the predicted daily maximum temperature after transformation by the same fuzzy membership function as above; the variable "add2" is set as the influence factor of the predicted daily minimum temperature after transformation by the same fuzzy membership function as above; and the daily electricity consumption is predicted using the Prophet model trained above.

[0085] like Figure 5 As shown, the time series forecasting model contains 6 sub-items, specifically including:

[0086] Annual trend item 501: Fits the annual growth rate from historical electricity data with a step size of 1 year.

[0087] Monthly Cycle Item 502: Determine the proportion of electricity consumption in each month to the annual electricity consumption, using a 12-month cycle.

[0088] Weekly Cycle Item 503: Determine the proportion of electricity consumption from Monday to Sunday in the weekly electricity consumption, using a 7-day cycle.

[0089] Statutory Holiday Amendment Item 504: The time sequence is amended according to the statutory holidays and work schedule adjustments issued by the State Council.

[0090] Covariate correction term 505: The influence of other time series that are strongly correlated with electricity consumption on electricity consumption, obtained by fitting the part after actual electricity consumption - annual trend term - monthly cycle term - weekly cycle term - statutory holiday correction term with the covariate.

[0091] White noise term 506: The remaining residual terms that cannot be fitted by the above five terms must satisfy a normal distribution.

[0092] The predicted power consumption is obtained by adding the values ​​of the six components together.

[0093] This invention also provides a daily electricity consumption forecasting system 60 for provincial power grids, such as... Figure 6 As shown, the daily electricity consumption prediction system in this embodiment includes five functional modules: a data preprocessing module 601, a data conversion module 602, a training module 603, a prediction module 604, and a result display module 605. The specific descriptions of these five functional modules are as follows:

[0094] The data preprocessing module 601 is used to collect, clean, query, and group daily electricity and temperature data, and to construct time series data.

[0095] Data conversion module 602 is used to convert temperature in a time series into influencing factors;

[0096] Training module 603 is used to establish a time series prediction model and to train the model parameters by taking the above time series as input, so as to obtain the expression of electricity consumption with respect to time and covariates;

[0097] The prediction module 604 is used to input the time series to be predicted and the predicted temperature, convert the predicted temperature into an influence factor according to the same historical degree function, and finally calculate the predicted power consumption value based on the trained time series prediction model.

[0098] The results display module 605 is used to display historical and forecast data in a chart format. It can display the changing trends of electricity and temperature on the same chart and can distinguish between historical and forecast data.

[0099] The data preprocessing module 601 decomposes the provincial power grid electricity consumption according to the power supply unit and electricity consumption category. The electricity consumption category includes large industrial electricity consumption, non-general industrial electricity consumption, commercial electricity consumption, non-residential lighting electricity consumption, residential electricity consumption, and agricultural electricity consumption. The power supply unit is divided according to county / district, and the highest and lowest temperatures of the corresponding county / district are also accessed to form multiple time series. For a specific time series, it includes four fields: date, highest temperature of the county / district, lowest temperature of the county / district, and electricity consumption of a certain electricity consumption category in the county / district.

[0100] Furthermore, in the data conversion module 602, the highest and lowest temperatures are divided into five intervals using fuzzy membership functions: low temperature saturation zone, low temperature influence zone, comfort zone, high temperature influence zone, and high temperature saturation zone. The parameters of the fuzzy membership function are optimized to convert the highest and lowest temperatures into influence factors.

[0101] Meanwhile, the training module 603 fits expressions for annual trend terms, monthly cycle terms, weekly cycle terms, statutory holiday correction terms, covariate correction terms, and white noise terms based on the time series input data, and combines them to obtain expressions for electricity consumption with respect to time, the highest temperature influence factor, and the lowest temperature influence factor.

[0102] In addition, the prediction module 604 can convert the highest and lowest temperatures into influencing factors based on the weather forecast results of each county and district, and predict the daily electricity consumption of different electricity consumption categories in each county and district.

[0103] Finally, the results display module 605 can utilize Baidu's Echarts technology to generate dynamic visual reports, such as... Figure 7 As shown, the report includes the historical daily electricity consumption, the predicted daily electricity consumption, the daily electricity consumption during the same period, the historical highest temperature, the predicted highest temperature, the highest temperature during the same period, the historical lowest temperature, the predicted lowest temperature, and the lowest temperature during the same period, which can provide a clear and intuitive analysis of the impact of temperature on electricity consumption.

[0104] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the above system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0105] It should be noted that the methods and detailed examples provided in the above embodiments can be combined with the systems provided in the embodiments for mutual reference, and will not be repeated here.

[0106] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0107] In the embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For instance, the division of modules or units described above is merely a logical functional division, and in actual implementation, it can be divided in other ways. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0108] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for predicting daily electricity consumption based on fuzzy membership functions, characterized in that, Includes the following steps: Step S1: Construct a time series of historical daily electricity consumption and historical temperature using the date as the index; Step S2: Establish fuzzy membership functions; Step S3: Using the maximum Pearson correlation coefficient as the objective function, optimize the parameters of the fuzzy membership function through a metaheuristic algorithm; Step S4: Based on the optimized fuzzy membership function, historical and predicted temperatures are transformed into relevant factors and labeled as covariates; Step S5: Establish a time series forecasting model, train the model using historical daily electricity consumption and historical covariates, and finally input the forecast covariates to calculate the predicted daily electricity consumption value. The fuzzy membership function decomposes the temperature into five intervals and establishes five sets of mapping relationships, specifically including: Let the temperature be x, the mapping value be y, and the mapping relationship be f(·), where a, b, c, and d are all parameters of the fuzzy membership function; Low-temperature saturation region: Consider x≤a as the low-temperature saturation region, and we have y=f(x)=ba; Low temperature influence zone: a < x ≤ b is considered as the low temperature influence zone, and y = f(x) = bx; Comfort zone: Consider b < x ≤ c as the comfort zone, and have y = f(x) = 0; High-temperature influence zone: Consider c < x ≤ d as the high-temperature influence zone, and we have y = f(x) = xc; High-temperature saturation region: Consider x > d as the high-temperature saturation region, and we have y = f(x) = dc.

2. The daily electricity consumption prediction method based on fuzzy membership function according to claim 1, characterized in that, The parameters of the fuzzy membership function include four values: a, b, c, and d. The parameters of the fuzzy membership function are optimized as follows: (1): Randomly generate N sets of data. Each set of data contains 4 values, and all 4 values ​​are within a preset range. After sorting from smallest to largest, they are assigned the values ​​a, b, c, d respectively. (2): For each set of data, determine the fuzzy membership function and calculate the mapping value of historical temperature through the fuzzy membership function; (3): Match historical temperature mapping values ​​with historical daily electricity consumption on the same date to generate two sequences, and calculate the Pearson correlation coefficient; (4): Repeat (2) and (3) until the Pearson correlation coefficients corresponding to N sets of data are calculated. Find the set of data with the smallest Pearson correlation coefficient and record it as the optimal initial value. (5): Using the optimal initial value obtained in (4) as the initial value for iteration, design a fitness function and update the four values ​​a, b, c, d through selection, crossover, and mutation operations until the four values ​​a, b, c, d no longer change.

3. The daily electricity consumption prediction method based on fuzzy membership function according to claim 2, characterized in that, The fitness function is specifically as follows: Let the input to the fitness function be a vector p, with elements p1, p2, p3, p4. Let a = p1, b = p2, c = p3, and d = p4, and determine the mapping relationship f(·). Let the historical temperature sequence ordered in chronological order be a vector X = {x t The mapping value y is calculated using the mapping relationship f(·). t Calculate y t The average value is denoted as Let the historical electricity sequence ordered in chronological order be a vector Z = {z} t }, calculate z t The average value is denoted as The fitness output r = h(p) is calculated as the fitness function using the following formula:

4. The daily electricity consumption prediction method based on fuzzy membership function according to claim 1, characterized in that, The time series has a step size of 1 day and includes 6 sub-items, specifically: Annual trend item: Fit the annual growth rate from historical electricity data with a step size of 1 year; Monthly cycle item: Determine the proportion of electricity consumption in each month to the annual electricity consumption in a 12-month cycle; Weekly cycle item: Using a 7-day cycle, determine the proportion of electricity consumption from Monday to Sunday in the weekly electricity consumption; Statutory holiday corrections: The time series is corrected according to statutory holidays and work-off adjustments; Covariate adjustment term: The influence of other time series that are strongly correlated with electricity consumption on electricity consumption is obtained by fitting the part after adjusting for actual electricity consumption - annual trend term - monthly cycle term - weekly cycle term - statutory holidays term with the covariate; White noise term: The remaining residual terms that cannot be fitted by the above five terms satisfy a normal distribution.

5. The daily electricity consumption prediction method based on fuzzy membership function according to claim 4, characterized in that, The covariates are the correlation factors obtained after the optimized fuzzy membership function transformation of the highest and lowest temperatures.

6. A daily electricity consumption forecasting system based on fuzzy membership functions, characterized in that, include: The data preprocessing module is used to collect, clean, query, and group daily electricity and temperature data, and construct time series data. The data transformation module is used to convert temperature data in a time series into influencing factors. The training module is used to build a time series prediction model and uses the above time series as input to train the model parameters, so as to obtain the expression of electricity consumption with respect to time and covariates; The prediction module is used to input the time series to be predicted and the predicted temperature, and converts the predicted temperature into an influence factor according to the same historical degree function. Finally, it calculates the predicted electricity value based on the trained time series prediction model. The results display module is used to display historical data and predicted data; The fuzzy membership function decomposes the temperature into five intervals and establishes five sets of mapping relationships, specifically including: Let the temperature be x, the mapping value be y, and the mapping relationship be f(·), where a, b, c, and d are all parameters of the fuzzy membership function; Low-temperature saturation region: Consider x≤a as the low-temperature saturation region, and we have y=f(x)=ba; Low temperature influence zone: a < x ≤ b is considered as the low temperature influence zone, and y = f(x) = bx; Comfort zone: Consider b < x ≤ c as the comfort zone, and have y = f(x) = 0; High-temperature influence zone: Consider c < x ≤ d as the high-temperature influence zone, and we have y = f(x) = xc; High-temperature saturation region: Consider x > d as the high-temperature saturation region, and we have y = f(x) = dc; The data conversion module uses fuzzy membership functions to divide the highest and lowest temperatures into five zones: low-temperature saturation zone, low-temperature influence zone, comfort zone, high-temperature influence zone, and high-temperature saturation zone. The parameters of the fuzzy membership functions are optimized to convert the highest and lowest temperatures into influencing factors.

7. The daily electricity consumption prediction system based on fuzzy membership function according to claim 6, characterized in that, The data preprocessing module decomposes the provincial power grid electricity according to the power supply unit and electricity consumption category. The electricity consumption category includes large industrial electricity consumption, non-general industrial electricity consumption, commercial electricity consumption, non-residential lighting electricity consumption, residential electricity consumption, and agricultural electricity consumption. The power supply unit is divided according to the county and district, and the highest and lowest temperatures of the corresponding county and district are also connected to form multiple time series.

8. The daily electricity consumption prediction system based on fuzzy membership function according to claim 6, characterized in that, The training module fits expressions for annual trend terms, monthly cycle terms, weekly cycle terms, statutory holiday correction terms, covariate correction terms, and white noise terms based on the time series input data, and combines them to obtain expressions for electricity consumption with respect to time, the influence factor of the highest temperature, and the influence factor of the lowest temperature.

Images