CN116050482A - Rural area electric energy substitution potential prediction method - Google Patents
Rural area electric energy substitution potential prediction method Download PDFInfo
- Publication number
- CN116050482A CN116050482A CN202211517587.4A CN202211517587A CN116050482A CN 116050482 A CN116050482 A CN 116050482A CN 202211517587 A CN202211517587 A CN 202211517587A CN 116050482 A CN116050482 A CN 116050482A
- Authority
- CN
- China
- Prior art keywords
- rural
- electric energy
- year
- substitution
- energy consumption
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention belongs to the technical field of rural area electric energy substitution potential evaluation, and particularly relates to a rural area electric energy substitution potential prediction method; the method comprises the following steps: step 1, calculating the electric energy substitution quantity; step 2, determining influencing factors of electric energy replacing electric quantity; step 3, predicting and modeling the electric energy substitution quantity by adopting a multi-element nonlinear regression curve; step 4, building a neural network prediction model; the invention relates to a rural area electric energy substitution potential prediction method, which provides an intelligent prediction mixed model based on a multi-element nonlinear regression and a neural network, and predicts and quantitatively analyzes the rural area electric energy substitution potential; according to the invention, the residual error between the nonlinear fitting value and the actual value is trained through the wavelet neural network, the result is corrected, the fitting precision is further improved, and the obtained predicted value of the electric energy substitution quantity is the most accurate.
Description
Technical Field
The invention belongs to the technical field of rural area electric energy substitution potential evaluation, and particularly relates to a rural area electric energy substitution potential prediction method.
Background
Energy transformation is continuously carried out in various industries so as to promote the development of green low carbon in China; the rural power grid is used as a hub for connecting the two sides of rural energy supply and demand, and is a head-row soldier for constructing a low-carbon rural area and assisting carbon emission reduction in a rural area; the special committee for climate change of united states and mansion in 2019 counts the emission of agricultural carbon dioxide, and the emission of greenhouse gases in rural areas of China reaches about 15% of the total national emission, wherein rural energy activities account for 60% of the total emission in rural areas; for a long time, the rural environment protection consciousness and the emission reduction capability of China are weak, the energy supply depends on fossil energy mainly comprising coal, straw and the like, and the greenhouse gas emission is serious; for a long time, agricultural rural areas have no clear carbon emission reduction requirements and constraint indexes, and the electric energy substitution potential evaluation of rural areas is less in research, so that the electric energy substitution potential of rural areas is required to be quantitatively and deeply researched to exert the carbon emission reduction effect.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a rural area electric energy substitution potential prediction method.
The purpose of the invention is realized in the following way: a method for predicting the potential of electric energy substitution in rural areas comprises the following steps:
step 1, calculating the electric energy substitution quantity;
step 3, predicting and modeling the electric energy substitution quantity by adopting a multi-element nonlinear regression curve;
and 4, building a neural network prediction model.
The step 1 of calculating the electric energy substitution amount specifically includes:
taking the t th year as a reference year, the actual electric energy consumption of the reference year is E (t), and the total energy consumption is D sum (t) when the terminal energy pattern is maintained at the reference year level, the specific gravity of the terminal energy occupied by the electric energy is the same as that of the reference year, and the increase of the electric energy consumption in the t-year compared with the electric energy consumption in the t+1th year is defined as the electric energy substitution D in the t+1th year sub (t+1) calculated as:
wherein D is sub (t+1) is an electric energy substitution amount; e (t) is the actual electric energy consumption of the t th year; d (D) sum And (t) is the total energy consumption in the t th year.
The determining the influencing factors of the electric energy replacing electric quantity in the step 2 specifically comprises the following steps:
5 key factors affecting the replacement of electric energy are selected: the total rural production value, the rural population growth rate, the total rural energy consumption, the rural electric energy consumption ratio and the rural newly-built electric power investment proportion are quantized to influence the electric energy substitution process;
A. the total production value of the rural area is selected to reflect the influence of the rural area economy on the electric energy substitution:
G M,t ＝k 1 ·G sum,t ·h (2)
wherein k is 1 The elastic coefficient representing rural economic development; g M,t Representing the total production value of rural areas in the t th year of China; g sum,t The total production value is produced in China in the t th year; h is the production contribution rate of rural areas in China;
B. selecting rural population growth rate to reflect influence of rural area population growth on electric energy substitution:
wherein k is 2 An elastic coefficient representing an increase in rural population; g R,t Representing the population growth rate of the t year in rural areas; r is R 0,t Is the population of the first year of the t year; r is R 1,t Population at the end of the t year;
population for the year t;
C. the rural energy consumption type is mainly coal and biomass energy, the electric power ratio is generally lower, and the saturation growth characteristics of terminal energy consumption are described as follows:
G sum,t ＝k 3 ·E sum,t (4)
wherein k is 3 The elastic coefficient of the total energy consumption of the rural area is represented; e (E) sum,t The total energy consumption amount of the t-th year in rural areas;
D. the proportion of rural electric energy consumption to the total rural energy consumption is an important factor affecting the electric energy substitution, and is as follows:
wherein k is 4 The elastic coefficient of the rural electric energy consumption ratio is represented; g E,t Representing the power consumption ratio of the t-th year in rural areas of China; e (E) e,t The power consumption of the rural area in the t year; e (E) sum,t The total energy consumption amount of the t-th year in rural areas;
E. quantifying government investment in electricity replacement development based on the ratio of new electricity capital investment to new energy capital (electricity, coal, oil, and gas) investment from a macroscopic perspective:
wherein k is 5 The elasticity coefficient of the rural newly-built power investment proportion is represented; g Z,t Representing the effect of the year t policy on power replacement; f (F) e,t Representing the newly built fixed asset investment of the electric power in the t th year; f (F) coal,t Representing the newly built fixed asset investment of the coal in the t th year; f (F) oil,t Representing the newly built fixed asset investment of petroleum in the t th year; f (F) gas,t And represents the newly built fixed asset investment of natural gas in the t th year.
The step 3 of predicting and modeling the electric energy substitution quantity by adopting a multi-element nonlinear regression curve specifically comprises the following steps:
let the total value of rural production be x 1 The population in rural area is x 2 The total energy consumption in rural areas is x 3 The rural power consumption duty ratio is x 4 The rural newly-built power investment ratio is x 5 Through historical data correlation analysis, the approximate linear relation of the total rural production value, the rural population, the total rural energy consumption and the electric energy substitution amount, the rural electric energy consumption ratio, the approximate quadratic relation of the rural newly-built electric power investment ratio and the electric energy substitution amount can be known, so that a polynomial fitting model is established as follows:
where a, b, c, d, e, f, g, h is the model parameter.
The step 4 of establishing a neural network prediction model specifically comprises the following steps:
predicting 5 independent variables of a rural production total value, a rural population growth rate, a rural energy consumption total amount, a rural electric energy consumption ratio and a rural newly-built electric power investment ratio by adopting a neural network, and then carrying the 5 independent variables into a (7) so as to obtain a fitting predicted value of the electric energy substitution; setting the training step length of the BP neural network to 5000, setting an error cut-off target to 0.001, setting the learning rate to 0.1, adopting sigmoid type functions for the hidden layer and the output layer excitation functions, and repeatedly training the network by using a self-variable numerical sequence until training meets the error cut-off target;
actual value D of electric energy substitution quantity in the t th year sub,t Substitution D 'obtained by fitting' sub,t The difference is the t-th year residual error r ε,t ：
r ε,t ＝D sub,t -D' sub,t (8)
Setting the node number at the input side of the wavelet neural network as 10, setting the node number at the hidden layer as 16, setting the output layer as 1, taking the first 10 data in the residual sequence as network input, taking the 11 th data as expected repeated training network until the output error training meets the error cut-off target, and obtaining the residual error after the wavelet neural network correction
Will fit the predicted substitution D' sub,t And residual sequence correction result->
By combining, a more accurate predicted value of the electric energy substitution quantity can be obtained>
The invention has the beneficial effects that: according to the rural area electric energy substitution potential prediction method, the electric energy substitution quantity is calculated through the step 1; step 2, determining influencing factors of electric energy instead of electric quantity; step 3, predicting and modeling the electric energy substitution quantity by adopting a multi-element nonlinear regression curve; step 4, building a neural network prediction model; the intelligent prediction mixed model based on the multi-element nonlinear regression and the neural network is provided, and the prediction and quantitative analysis can be carried out on the electric energy substitution potential of rural areas.
Drawings
Fig. 1 is a schematic flow chart of a method for predicting electric energy substitution potential in rural areas.
Fig. 2 is a schematic diagram of a power substitution prediction model structure of a power substitution potential prediction method in rural areas.
FIG. 3 is a schematic diagram of a comparison of a predicted value and an actual value of a total rural production value in a rural area.
FIG. 4 is a schematic diagram of a comparison of a predicted value and an actual value of a rural population in a rural area.
FIG. 5 is a schematic diagram of a comparison of predicted and actual rural energy consumption values in rural areas.
Fig. 6 is a schematic diagram of comparison between a predicted value and an actual value of rural power consumption in a rural area.
Fig. 7 is a schematic diagram of comparison between a predicted value and an actual value of a rural newly-built power investment ratio in a rural area.
FIG. 8 is a graph showing comparison of power substitution predictions for different algorithms.
Detailed Description
The invention is further described below with reference to the accompanying drawings.
As shown in fig. 1, a method for predicting the potential of electric energy substitution in rural areas comprises the following steps:
step 1, calculating the electric energy substitution quantity;
step 3, predicting and modeling the electric energy substitution quantity by adopting a multi-element nonlinear regression curve;
and 4, building a neural network prediction model.
The step 1 of calculating the electric energy substitution amount specifically includes:
taking the t th year as a reference year, the actual electric energy consumption of the reference year is E (t), and the total energy consumption is D sum (t) when the terminal energy pattern is maintained at the reference year level, the specific gravity of the terminal energy occupied by the electric energy is the same as that of the reference year, and the increase of the electric energy consumption in the t-year compared with the electric energy consumption in the t+1th year is defined as the electric energy substitution D in the t+1th year sub (t+1) calculated as:
wherein D is sub (t+1) is an electric energy substitution amount; e (t) is the actual electric energy consumption of the t th year; d (D) sum And (t) is the total energy consumption in the t th year.
The determining the influencing factors of the electric energy replacing electric quantity in the step 2 specifically comprises the following steps:
the electric energy replacement electric quantity is mainly influenced by economy, population growth, technology, energy policy and the like, and 5 key factors influencing the electric energy replacement are selected: the total rural production value, the rural population growth rate, the total rural energy consumption, the rural electric energy consumption ratio and the rural newly-built electric power investment proportion are quantized to influence the electric energy substitution process; considering different influencing factors of electric energy substitution in different periods, the invention introduces the elasticity coefficient to quantify the influence degree of different factors on the electric energy substitution so as to more accurately predict the electric energy substitution potential. In the invention, the elastic coefficients are all taken as 1, assuming that 4 factors have no obvious difference in the influence on the potential of electric energy substitution.
A. The power demand and the economic growth of China have intrinsic performance, the economic growth condition has great influence on the power demand of the terminal, and the overall economic development condition of one region has influence on the power consumption and other energy consumption; the total production value of the rural area is selected to reflect the influence of the rural area economy on the electric energy substitution:
G M,t ＝k 1 ·G sum,t ·h (2)
wherein k is 1 The elastic coefficient representing rural economic development; g M,t Representing the total production value of rural areas in the t th year of China; g sum,t The total production value is produced in China in the t th year; h is the production contribution rate of rural areas in China;
B. the electric energy replacing electric quantity has stronger correlation with population factors, population influences the consumption condition of terminal energy sources including electric energy, the faster the population grows, the faster the electric energy consumption rate can be caused, and the selection preference of rural population to different energy sources such as electric energy, coal, petroleum and the like determines the consumption structure of the terminal energy sources to a certain extent; selecting rural population growth rate to reflect influence of rural area population growth on electric energy substitution:
wherein k is 2 An elastic coefficient representing an increase in rural population; g R,t Representing the population growth rate of the t year in rural areas; r is R 0,t Is the population of the first year of the t year; r is R 1,t Population at the end of the t year;
population for the year t;
C. the rural energy supply and demand form a multiple trend mainly by burning straws, trees and the like, the renewable energy source such as photovoltaic and the like develops faster, the low-carbon energy consumption trend is accelerated, and the use of petroleum, electric power and liquefied petroleum gas is increased in the 80 s; the rural energy consumption type is mainly coal and biomass energy, the electric power ratio is generally lower, and the saturation growth characteristics of terminal energy consumption are described as follows:
G sum,t ＝k 3 ·E sum,t (4)
wherein k is 3 The elastic coefficient of the total energy consumption of the rural area is represented; e (E) sum,t The total energy consumption amount of the t-th year in rural areas;
D. whether the electric energy replaces other energy sources in the terminal energy consumption link or not can be reflected by the proportion of the electric energy accounting for the terminal energy consumption in a macroscopic manner; the proportion of rural electric energy consumption to the total rural energy consumption is an important factor affecting the electric energy substitution, and is as follows:
wherein k is 4 The elastic coefficient of the rural electric energy consumption ratio is represented; g E,t Representing the power consumption ratio of the t-th year in rural areas of China; e (E) e,t The power consumption of the rural area in the t year; e (E) sum,t The total energy consumption amount of the t-th year in rural areas;
E. quantifying government investment in electricity replacement development based on the ratio of new electricity capital investment to new energy capital (electricity, coal, oil, and gas) investment from a macroscopic perspective:
wherein k is 5 The elasticity coefficient of the rural newly-built power investment proportion is represented; g Z,t Representing the effect of the year t policy on power replacement; f (F) e,t Representing the newly built fixed asset investment of the electric power in the t th year; f (F) coal,t Representing the newly built fixed asset investment of the coal in the t th year; f (F) oil,t Representing the newly built fixed asset investment of petroleum in the t th year; f (F) gas,t And represents the newly built fixed asset investment of natural gas in the t th year.
The increasing trend of the electric energy substitution is restricted by the common influence of the total rural production value, the rural population, the total rural energy consumption, the rural electric energy consumption ratio and the rural newly-built electric power investment ratio, and the total rural energy consumption ratio and the rural newly-built electric power investment ratio are not simple linear regression relations, but exhibit nonlinear growth characteristics, so that the invention firstly adopts a multiple nonlinear regression curve to carry out predictive modeling on the electric energy substitution.
The step 3 of predicting and modeling the electric energy substitution quantity by adopting a multi-element nonlinear regression curve specifically comprises the following steps:
let the total value of rural production be x 1 The population in rural area is x 2 The total energy consumption in rural areas is x 3 The rural power consumption duty ratio is x 4 The rural newly-built power investment ratio is x 5 Through historical data correlation analysis, the approximate linear relation of the total rural production value, the rural population, the total rural energy consumption and the electric energy substitution amount, the rural electric energy consumption ratio, the approximate quadratic relation of the rural newly-built electric power investment ratio and the electric energy substitution amount can be known, so that a polynomial fitting model is established as follows:
where a, b, c, d, e, f, g, h is the model parameter.
The step 4 of establishing a neural network prediction model specifically comprises the following steps:
after obtaining the fitting polynomial (7) of the electric energy substitution quantity, in order to predict the future electric energy substitution potential, the values of 5 independent variables, namely, the total rural production value, the rural population, the total rural energy consumption, the rural electric energy consumption ratio and the rural newly-built electric power investment ratio, are required to be predicted; the neural network algorithm is an algorithm for simulating the human brain by constructing neurons, has stronger learning capacity and adaptability, and can take various factors influencing a prediction result into consideration by training, so that the neural network algorithm has higher prediction precision; the invention adopts the neural network to predict 5 independent variables of the total rural production value, the rural population growth rate, the total rural energy consumption, the rural electric energy consumption ratio and the rural newly-built electric power investment ratio, and then brings the 5 independent variables into the (7) so as to obtain a fitting predicted value of the electric energy substitution; setting the training step length of the BP neural network to 5000, setting an error cut-off target to 0.001, setting the learning rate to 0.1, adopting sigmoid type functions for the hidden layer and the output layer excitation functions, and repeatedly training the network by using a self-variable numerical sequence until training meets the error cut-off target;
actual value D of electric energy substitution quantity in the t th year sub,t Substitution D 'obtained by fitting' sub,t The difference is the t-th year residual error r ε,t ：
r ε,t ＝D sub,t -D' sub,t (8)
Because the fitting residual has extremely strong randomness and uncertainty, the change trend of the fitting residual is difficult to predict, the hidden layer of the wavelet neural network takes a wavelet function as a transfer function, and the time-frequency localization characteristic of wavelet analysis is combined with the self-adaptive learning capability of the neural network, so that the residual sequence can be predicted well.
Setting the node number at the input side of the wavelet neural network as 10, setting the node number at the hidden layer as 16, setting the output layer as 1, taking the first 10 data in the residual sequence as network input, taking the 11 th data as expected repeated training network until the output error training meets the error cut-off target, and obtaining the residual error after the wavelet neural network correction
Will fit the predicted substitution D' sub,t And residual sequence correction result->
By combining, a more accurate predicted value of the electric energy substitution quantity can be obtained>
/>
In summary, the structure of the rural area electric energy substitution potential overall prediction model based on the nonlinear regression-neural network provided by the invention is shown in figure 2,
step S1: inputting relevant data of economy, population, energy consumption and technical policy in rural areas, and calculating according to formulas (1) to (5) to obtain an original sequence of 5 influencing factors about the electric energy substitution quantity, the total rural production value, the total rural population, the total rural energy consumption, the rural electric energy consumption ratio and the rural newly-built electric power investment ratio;
step S2: using four influencing factors as independent variables and using electric energy substitution quantity as dependent variables, and adopting a multiple nonlinear regression fitting algorithm to perform fitting to obtain a fitting polynomial y of the electric energy substitution quantity sub ；
Step S3: predicting the total rural production value, rural population growth rate, rural power consumption ratio and policy-supported value of the next t years through BP neural network to obtain an independent variable predicted value X t Substituting the fitting polynomial y sub Fitting predicted value D of the electric energy substitution quantity in the t-th year can be obtained s ' ub,t ；
Step S4: let t-th year fit prediction value D s ' ub,t And the actual value D of the electric energy substitution quantity sub,t The difference is called the prediction residual r of the t-th year ε,t Prediction residual error r based on wavelet neural network ε,t Training and correcting to obtain
Finally, the predicted value D will be fitted s ' ub,t And residual correction result->
Adding to obtain more accurate predicted value of the electric energy substitution quantity +.>
The analysis is performed as a specific example below.
(1) Data processing
In order to verify the effectiveness of the electric energy substitution prediction model, taking the development of electric energy substitution in fifteen years in China as an example, relevant data of economy, population, energy consumption and technical policy in rural areas in 2005-2020 are selected, as shown in table 1.
Table 1 rural area related data of approximately fifteen years
Note that: the related data is derived from Chinese statistical yearbook.
In order to obtain the rural electric energy consumption ratio, the total consumption amount of the terminal energy and the electric energy consumption amount are converted into unified dimensions, the electric energy consumption amount is converted into standard coal by adopting a thermal equivalent method, and the ratio of the rural electric energy consumption amount to the total consumption amount of the rural energy can be obtained through the formula (4) according to the fact that the heat value of 1kWh electric energy is equal to that of 0.123kg standard coal.
Meanwhile, the electric energy substitution amount in rural areas every year can be obtained by calculation according to the data of the table 1 and the electric energy substitution amount calculation formula (1) as shown in the table:
(2) Prediction results and analysis
Training by taking 2005-2015 historical data as a sample, respectively predicting 2016-2020 related data and electric energy substitution by using a neural network and multiple nonlinear fitting, and comparing and analyzing with a true value; the comparison of the predicted value and the actual value of each variable in rural areas is shown in figures 3-7.
Based on 2005-2015 electric energy substitution in tables 1 and 2 and the numerical relation between the electric energy substitution and the rural production total value, the rural population, the rural energy consumption total amount, the rural electric energy consumption duty ratio and the rural newly-built electric power investment ratio, a polynomial nonlinear model is adopted for fitting, and the preliminary fitting predicted value of the electric energy substitution can be obtained by combining the predictions of the values of 5 main influencing factors. And then, self-learning the fitting error by adopting a wavelet neural network, so that the predicted value of the electric energy substitution quantity can be obtained.
The prediction model, 3 models which are directly predicted by using a nonlinear algorithm and directly predicted by using a BP neural network, and the prediction results of the electric energy substitution quantity in 2016-2020 in rural areas of China are shown in figure 8, and as seen from figure 8, the three algorithms can be used for predicting the increasing trend of the electric energy substitution quantity well. The BP neural network algorithm is adopted to predict the electric energy substitution quantity to be superior to a nonlinear fitting method. This is because the neural network algorithm has a strong learning ability and generalization ability, and various factors affecting the prediction result can be taken into consideration in the training process, so that the neural network algorithm has a high prediction accuracy. While nonlinear regression fitting, although less accurate in prediction, can better characterize the trend of increasing energy substitution. The hybrid algorithm provided by the invention combines the advantages of the two algorithms, the residual error of the nonlinear fitting value and the actual value is trained through the wavelet neural network, the result is corrected, the fitting precision is further improved, and the obtained predicted value of the electric energy substitution quantity is the most accurate.
In summary, according to the rural area electric energy substitution potential prediction method, the electric energy substitution quantity is calculated through the step 1; step 2, determining influencing factors of electric energy instead of electric quantity; step 3, predicting and modeling the electric energy substitution quantity by adopting a multi-element nonlinear regression curve; step 4, building a neural network prediction model; the intelligent prediction mixed model based on the multi-element nonlinear regression and the neural network is provided, and the prediction and quantitative analysis can be carried out on the electric energy substitution potential of rural areas.
Claims (5)
1. The method for predicting the electric energy substitution potential in rural areas is characterized by comprising the following steps of:
step 1, calculating the electric energy substitution quantity;
step 2, determining influencing factors of electric energy replacing electric quantity;
step 3, predicting and modeling the electric energy substitution quantity by adopting a multi-element nonlinear regression curve;
and 4, building a neural network prediction model.
2. The method for predicting the potential of power substitution in rural areas according to claim 1, wherein the calculating of the power substitution amount in the step 1 specifically comprises:
taking the t th year as a reference year, the actual electric energy consumption of the reference year is E (t), and the total energy consumption is D sum (t) when the terminal energy pattern is maintained at the reference year level, the specific gravity of the terminal energy occupied by the electric energy is the same as that of the reference year, and the increase of the electric energy consumption in the t-year compared with the electric energy consumption in the t+1th year is defined as the electric energy substitution D in the t+1th year sub (t+1) by the following formulaThe method comprises the following steps:
wherein D is sub (t+1) is an electric energy substitution amount; e (t) is the actual electric energy consumption of the t th year; d (D) sum And (t) is the total energy consumption in the t th year.
3. The method for predicting power replacement potential in rural areas according to claim 1, wherein determining the influencing factors of the power replacement power in step 2 specifically comprises:
5 key factors affecting the replacement of electric energy are selected: the total rural production value, the rural population growth rate, the total rural energy consumption, the rural electric energy consumption ratio and the rural newly-built electric power investment proportion are quantized to influence the electric energy substitution process;
A. the total production value of the rural area is selected to reflect the influence of the rural area economy on the electric energy substitution:
G M,t ＝k 1 ·G sum,t ·h (2)
wherein k is 1 The elastic coefficient representing rural economic development; g M,t Representing the total production value of rural areas in the t th year of China; g sum,t The total production value is produced in China in the t th year; h is the production contribution rate of rural areas in China;
B. selecting rural population growth rate to reflect influence of rural area population growth on electric energy substitution:
wherein k is 2 An elastic coefficient representing an increase in rural population; g R,t Representing the population growth rate of the t year in rural areas; r is R 0,t Is the population of the first year of the t year; r is R 1,t Population at the end of the t year;
population for the year t;
C. the rural energy consumption type is mainly coal and biomass energy, the electric power ratio is generally lower, and the saturation growth characteristics of terminal energy consumption are described as follows:
G sum,t ＝k 3 ·E sum,t (4)
wherein k is 3 The elastic coefficient of the total energy consumption of the rural area is represented; e (E) sum,t The total energy consumption amount of the t-th year in rural areas;
D. the proportion of rural electric energy consumption to the total rural energy consumption is an important factor affecting the electric energy substitution, and is as follows:
wherein k is 4 The elastic coefficient of the rural electric energy consumption ratio is represented; g E,t Representing the power consumption ratio of the t-th year in rural areas of China; e (E) e,t The power consumption of the rural area in the t year; e (E) sum,t The total energy consumption amount of the t-th year in rural areas;
E. quantifying government investment in electricity replacement development based on the ratio of new electricity capital investment to new energy capital (electricity, coal, oil, and gas) investment from a macroscopic perspective:
wherein k is 5 The elasticity coefficient of the rural newly-built power investment proportion is represented; g Z,t Representing the effect of the year t policy on power replacement; f (F) e,t Representing the newly built fixed asset investment of the electric power in the t th year; f (F) coal,t Representing the newly built fixed asset investment of the coal in the t th year; f (F) oil,t Representing the newly built fixed asset investment of petroleum in the t th year; f (F) gas,t And represents the newly built fixed asset investment of natural gas in the t th year.
4. The method for predicting the potential of electric energy substitution in rural areas according to claim 1, wherein the step 3 of predicting and modeling the electric energy substitution amount by using a multiple nonlinear regression curve specifically comprises:
let the total value of rural production be x 1 The population in rural area is x 2 The total energy consumption in rural areas is x 3 The rural power consumption duty ratio is x 4 The rural newly-built power investment ratio is x 5 Through historical data correlation analysis, the approximate linear relation of the total rural production value, the rural population, the total rural energy consumption and the electric energy substitution amount, the rural electric energy consumption ratio, the approximate quadratic relation of the rural newly-built electric power investment ratio and the electric energy substitution amount can be known, so that a polynomial fitting model is established as follows:
where a, b, c, d, e, f, g, h is the model parameter.
5. The method for predicting power substitution potential in rural areas according to claim 4, wherein the step 4 of establishing a neural network prediction model specifically comprises:
predicting 5 independent variables of a rural production total value, a rural population growth rate, a rural energy consumption total amount, a rural electric energy consumption ratio and a rural newly-built electric power investment ratio by adopting a neural network, and then carrying the 5 independent variables into a (7) so as to obtain a fitting predicted value of the electric energy substitution; setting the training step length of the BP neural network to 5000, setting an error cut-off target to 0.001, setting the learning rate to 0.1, adopting sigmoid type functions for the hidden layer and the output layer excitation functions, and repeatedly training the network by using a self-variable numerical sequence until training meets the error cut-off target;
actual value D of electric energy substitution quantity in the t th year sub,t Substitution D 'obtained by fitting' sub,t The difference is the t-th year residual error r ε,t ：
r ε,t ＝D sub,t -D′ sub,t (8)
Setting the node number at the input side of the wavelet neural network as 10, setting the node number at the hidden layer as 16, setting the output layer as 1, taking the first 10 data in the residual sequence as network input, taking the 11 th data as expected repeated training network until the output error training meets the error cut-off target, and obtaining the residual error after the wavelet neural network correction
Will fit the predicted substitution quantity D s ' ub,t And residual sequence correction result->
By combining, a more accurate predicted value of the electric energy substitution quantity can be obtained>
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211517587.4A CN116050482A (en) | 2022-11-29 | 2022-11-29 | Rural area electric energy substitution potential prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211517587.4A CN116050482A (en) | 2022-11-29 | 2022-11-29 | Rural area electric energy substitution potential prediction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116050482A true CN116050482A (en) | 2023-05-02 |
Family
ID=86115508
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211517587.4A Pending CN116050482A (en) | 2022-11-29 | 2022-11-29 | Rural area electric energy substitution potential prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116050482A (en) |
-
2022
- 2022-11-29 CN CN202211517587.4A patent/CN116050482A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | A seasonal GM (1, 1) model for forecasting the electricity consumption of the primary economic sectors | |
| Luo et al. | A novel Riccati equation grey model and its application in forecasting clean energy | |
| CN112801388B (en) | Power load prediction method and system based on nonlinear time series algorithm | |
| CN115660918A (en) | Industrial park carbon emission calculation and prediction method | |
| CN112650063B (en) | Self-adaptive soft measurement method based on semi-supervised incremental Gaussian mixture regression | |
| CN111600300B (en) | Robust optimal scheduling method considering wind power multivariate correlation ellipsoid set | |
| CN104834975A (en) | Power network load factor prediction method based on intelligent algorithm optimization combination | |
| CN104573857A (en) | Power grid load rate prediction method based on intelligent algorithm optimization and combination | |
| Shi | Forecast of China’s carbon emissions under the background of carbon neutrality | |
| CN111952965B (en) | CCHP system optimized operation method based on predictive control and interval planning | |
| Zhao et al. | Regulation factors driving vegetation changes in China during the past 20 years | |
| Wang et al. | Carbon market price prediction based on sequence decomposition-reconstruction-dimensionality reduction and improved deep learning model | |
| Zheng | Evaluation model of ecological economic benefits based on discrete mathematical algorithm | |
| CN112508231A (en) | Medium-and-long-term power load prediction method and system based on system dynamics | |
| CN116050482A (en) | Rural area electric energy substitution potential prediction method | |
| CN116049938A (en) | Building energy consumption prediction method based on multi-factor index and GM gray system | |
| CN111126827A (en) | Input-output accounting model construction method based on BP artificial neural network | |
| CN114139829A (en) | Power grid load prediction method based on artificial intelligence | |
| CN115511230A (en) | Electric energy substitution potential analysis and prediction method | |
| CN117057516B (en) | Carbon accounting parameter analysis and prediction method based on system dynamics | |
| Zhao et al. | An interpretable ultra-short-term wind power prediction model based on the feature matrix reconstruction through regression trees | |
| Zhou et al. | Study on the Method of Carbon Emission Modelling in China Based on Machine Learning | |
| Nurullayevna et al. | Construction of Low-carbon Economic Development Level Model Based on BP Neural Network Algorithm | |
| CN117634678A (en) | Low-carbon park carbon emission prediction method based on actual operation scene | |
| CN117709524A (en) | Bayesian optimization-based carbon emission prediction method and system for steel industry |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |