CN113435653B - Method and system for predicting saturated power consumption based on logistic model - Google Patents

Method and system for predicting saturated power consumption based on logistic model Download PDF

Info

Publication number
CN113435653B
CN113435653B CN202110749465.7A CN202110749465A CN113435653B CN 113435653 B CN113435653 B CN 113435653B CN 202110749465 A CN202110749465 A CN 202110749465A CN 113435653 B CN113435653 B CN 113435653B
Authority
CN
China
Prior art keywords
initial
curve
power consumption
logistic
target area
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.)
Active
Application number
CN202110749465.7A
Other languages
Chinese (zh)
Other versions
CN113435653A (en
Inventor
何琳
刘锦明
崔翔
门艳
刘芳
金梦
刘景华
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Power Construction Technology Economic Consulting Center Of China Electricity Council
Economic and Technological Research Institute of State Grid Xinjiang Electric Power Co Ltd
Original Assignee
Power Construction Technology Economic Consulting Center Of China Electricity Council
Economic and Technological Research Institute of State Grid Xinjiang Electric Power Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Power Construction Technology Economic Consulting Center Of China Electricity Council, Economic and Technological Research Institute of State Grid Xinjiang Electric Power Co Ltd filed Critical Power Construction Technology Economic Consulting Center Of China Electricity Council
Priority to CN202110749465.7A priority Critical patent/CN113435653B/en
Publication of CN113435653A publication Critical patent/CN113435653A/en
Application granted granted Critical
Publication of CN113435653B publication Critical patent/CN113435653B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
    • G06F18/24133—Distances to prototypes
    • G06F18/24137—Distances to cluster centroïds
    • G06F18/2414—Smoothing the distance, e.g. radial basis function networks [RBFN]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/20—Ensemble learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Economics (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Artificial Intelligence (AREA)
  • Mathematical Physics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Software Systems (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Health & Medical Sciences (AREA)
  • Operations Research (AREA)
  • Pure & Applied Mathematics (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Medical Informatics (AREA)
  • Game Theory and Decision Science (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Computing Systems (AREA)
  • Development Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Probability & Statistics with Applications (AREA)
  • Quality & Reliability (AREA)
  • Algebra (AREA)

Abstract

The application provides a saturated power consumption prediction method and a saturated power consumption prediction system based on a logistic model, which comprise the following steps: determining an initial logistic fitting curve of the power consumption of the target area according to the historical power consumption data of the target area; calculating to obtain an initial difference curve according to the initial logistic fitting curve and the historical power consumption data of the target area; respectively carrying out normalization processing on the initial logistic fitting curve and the initial difference value curve to obtain a training sample data set; training and learning the training sample data set through a support vector machine to obtain a regression prediction curve; and adding the regression prediction curve and the initial logistic fitted curve to obtain an optimized logistic fitted curve. A traditional preliminary prediction result of the logistic is optimized by introducing a prediction value of the difference value, so that the deviation condition of the actual urban power consumption and the initial logistic fitting curve caused by the influence of multiple factors can be made up, and the defect of inaccurate parameter estimation of a logistic model is made up.

Description

Saturated power consumption prediction method and system based on logistic model
Technical Field
The application belongs to the technical field of electric quantity prediction, and particularly relates to a saturated power consumption prediction method and system based on a logistic model.
Background
Urban electricity consumption generally occurs in three stages, namely a slow development stage, a high-speed growth stage and a low-speed saturation stage. In the initial stage of urban development, due to social and economic reasons, the electricity consumption tends to increase at a low speed, and the stage is a slow development stage; after the society and the economy develop to a certain degree, the electricity consumption can be increased at a high speed, so that the electricity consumption enters a high-speed increasing stage; along with the continuous development of cities, due to the influence of objective factors, the power consumption is stabilized within a certain range and fluctuates in a small amplitude, the increasing speed of the power consumption is gradually reduced or even stops increasing, the stage is a low-speed saturation stage, and generally, when the increasing rate of the power consumption of the cities is lower than 2%, the power demand of the cities can be judged to enter the saturation stage.
The traditional logistic model can fully reflect the long-term development rule of the power consumption and is very suitable for predicting the saturated power consumption, but the urban saturated power consumption is influenced by various factors, so that the prediction result is not ideal. On the other hand, if a combined prediction method is considered, that is, two basic prediction methods are used for combined prediction, the two methods are weighted according to different standards and then summed to obtain a final result, and since the method needs to determine the weight, a more scientific and reasonable weight determination method is difficult to find, and thus the prediction accuracy cannot be effectively improved.
Therefore, how to provide a method capable of ensuring the prediction accuracy of the saturated power consumption is a technical problem which needs to be solved urgently at present.
Disclosure of Invention
In order to solve the technical problems in the prior art, the application provides a method and a system for predicting saturated power consumption based on a logistic model.
In a first aspect, the present application provides a saturated power consumption prediction method based on a logistic model, including: acquiring historical electricity consumption data of a target area; determining an initial logistic fit curve of the power consumption of the target area according to the historical power consumption data of the target area, wherein the independent variable of the initial logistic fit curve is time, and the dependent variable is an initial predicted value of the power consumption of the target area; calculating to obtain an initial difference value curve according to the initial logistic fitting curve and the historical power consumption data of the target area, wherein the independent variable of the initial difference value curve is time, and the dependent variable is an initial difference value, and the initial difference value is the difference value between the historical power consumption of the target area and the initial predicted value of the power consumption; respectively carrying out normalization processing on the initial logistic fitted curve and the initial difference value curve to obtain a training sample data set, wherein the training sample data set comprises sample data of an initial predicted value of the power consumption of the target area and the sample data of the initial difference value; training and learning the training sample data set through a support vector machine to obtain a regression prediction curve, wherein an independent variable of the regression prediction curve is time, a dependent variable of the regression prediction curve is a predicted value of a difference value, and the predicted value of the difference value is the difference value of the historical power consumption and the optimized predicted value of the power consumption of the target area; and adding the regression prediction curve and the initial logistic fitted curve to obtain an optimized logistic fitted curve, wherein the independent variable of the optimized logistic fitted curve is time, and the dependent variable is the optimized predicted value of the power consumption.
Optionally, the method further includes: and predicting the saturated power consumption of the target time period of the target area according to the optimized logistic fitting curve.
Optionally, the determining an initial logistic fit curve of the target area power consumption according to the historical power consumption data of the target area includes: estimating parameters a, b and c in a logistic fitting curve by using a Yule algorithm, wherein the logistic fitting curve satisfies the following first relational expression:
Figure GDA0003789262900000021
wherein t represents time, y t Representing the initial predicted value of the electricity consumption of the target area, and a, b and c represent parameters of a logistic fitting curve; and determining an initial logistic fitting curve of the target area electricity consumption according to the estimated parameters a, b and c.
Optionally, the initial logistic fitted curve is normalized according to a second relational expression, where the second relational expression is:
Figure GDA0003789262900000022
wherein the content of the first and second substances,
Figure GDA0003789262900000023
sample data representing an initial predicted value of a power usage amount of the target area,
Figure GDA0003789262900000024
an initial predicted value representing a used amount of electricity of the target area,
Figure GDA0003789262900000025
and m represents the initial predicted value of the electricity consumption with the change rate within 1% compared with the historical electricity consumption of the target area.
Optionally, the difference curve is normalized according to a third relation, where the third relation is:
Figure GDA0003789262900000026
wherein the content of the first and second substances,
Figure GDA0003789262900000027
sample data representing said initial difference, g t Represents the initial difference, min (g) t ) Represents the minimum value, max (g), of said initial difference t ) Represents the maximum value of the initial difference values.
Optionally, a gaussian radial basis function is selected as a kernel function for training and learning the training sample data set by using a support vector machine, where a formula of the kernel function is:
Figure GDA0003789262900000028
wherein, K (x) i ,x j ) Representing the kernel function, x i And x j And the parameter gamma is set to be 2.8 according to the sample data of the initial predicted value of the electricity consumption of the target area in different time periods.
In a second aspect, the application further provides a saturated power consumption prediction system based on the logistic model, and the acquisition module is used for acquiring historical power consumption data of a target area; the initial logistic fitting curve determining module is used for determining an initial logistic fitting curve of the target area power consumption according to the historical power consumption data of the target area, wherein the independent variable of the initial logistic fitting curve is time, and the dependent variable is an initial predicted value of the target area power consumption; an initial difference curve calculation module, configured to calculate an initial difference curve according to the initial logistic fit curve and the historical power consumption data of the target area, where an independent variable of the initial difference curve is time, and a dependent variable is an initial difference, where the initial difference is a difference between the historical power consumption of the target area and an initial predicted value of the power consumption; a normalization processing module, configured to perform normalization processing on the initial logistic fitted curve and the initial difference curve, respectively, to obtain a training sample data set, where the training sample data set includes sample data of an initial predicted value of power consumption of the target area and sample data of the initial difference; the regression prediction curve calculation module is used for training and learning the training sample data set through a support vector machine to obtain a regression prediction curve, wherein the independent variable of the regression prediction curve is time, the dependent variable of the regression prediction curve is a prediction value of a difference value, and the prediction value of the difference value is the difference value between the historical electricity consumption of the target area and the optimized prediction value of the electricity consumption; and the addition module is used for adding the regression prediction curve and the initial logistic fitted curve to obtain an optimized logistic fitted curve, wherein the independent variable of the optimized logistic fitted curve is time, and the dependent variable is the optimized predicted value of the power consumption.
Optionally, the system further includes a prediction module, configured to predict, according to the optimized logistic fitting curve, a saturated power consumption in a target time period of the target area.
Optionally, the normalization processing module includes a first normalization processing sub-module and a second normalization processing sub-module;
the first normalization processing sub-module is configured to perform normalization processing on the initial logistic fitted curve according to a second relational expression, where the second relational expression is:
Figure GDA0003789262900000031
wherein, the first and the second end of the pipe are connected with each other,
Figure GDA0003789262900000032
sample data representing an initial predicted value of a power usage amount of the target area,
Figure GDA0003789262900000033
an initial predicted value representing a used amount of electricity of the target area,
Figure GDA0003789262900000034
and m represents the initial predicted value of the electricity consumption with the change rate within 1% compared with the historical electricity consumption of the target area.
The second normalization processing sub-module is configured to perform normalization processing on the difference curve according to a third relation, where the third relation is:
Figure GDA0003789262900000035
wherein, the first and the second end of the pipe are connected with each other,
Figure GDA0003789262900000036
sample data representing said initial difference, g t Represents the initial difference, min (g) t ) Represents the minimum value, max (g), of said initial difference t ) Represents the maximum of the initial differences.
Optionally, a gaussian radial basis function is selected as a kernel function for training and learning the training sample data set by using a support vector machine, where a formula of the kernel function is:
Figure GDA0003789262900000041
wherein, K (x) i ,x j ) Representing the kernel function, x i And x j And the parameter gamma is set to be 2.8 according to the sample data of the initial predicted value of the electricity consumption of the target area in different time periods.
In summary, the method and the system for predicting the saturated power consumption based on the logistic model provided by the application introduce the difference g t The prediction value optimizes the initial prediction result of the traditional logistic, so that the deviation condition of the actual urban power consumption and the initial logistic fitting curve caused by the influence of multiple factors can be made up, the defect of inaccurate parameter estimation of a logistic model and the difference value g can be made up t The prediction value is obtained by predicting through a support vector machine regression based on the initial logistic fitting curve, the method improves the condition that the prediction effect of the original single logistic prediction method is poor, the accuracy of the prediction of the saturated power consumption is effectively improved, and the method has good generalization capability, can ensure the balance of power supply and demand, provides important guarantee for improving the living standard of people, and provides powerful support for the healthy development of economy.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
FIG. 1 is a schematic workflow diagram of a saturated power consumption prediction method based on a logistic model according to an embodiment of the present application;
FIG. 2 is a graph showing the prediction results of the total social electricity consumption from 2001 to 2019 in Beijing, using different prediction methods;
FIG. 3 is a graph of absolute percentage error of total social power consumption predictions from 2001 to 2019 in Beijing, using different prediction methods;
FIG. 4 is a block diagram of a saturated power consumption prediction system based on a logistic model according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
To facilitate understanding of the present application, a logistic model will first be described.
The logistic model is a model for mathematically describing an S-shaped curve, and fully shows three stages of electricity consumption, namely a slow development stage, a high-speed growth stage and a low-speed saturation stage. Therefore, the logistic model can be applied to the saturated power consumption prediction. The logistic model satisfies the following first relation (1):
Figure GDA0003789262900000042
for the scenario of power consumption prediction, y is given in the above equation t Representing electricity usage, t representing time, a, b, c representing parameters of the model.
The following describes a saturated power consumption prediction method based on a logistic model provided by the present application with reference to the accompanying drawings.
As shown in fig. 1, the method for predicting saturated power consumption based on a logistic model provided in the present application includes the following steps:
and step 10, obtaining historical electricity consumption data of the target area.
The historical power consumption data is generally annual power consumption data. The target area may be a city, province, or country, etc., which is not limited in this application.
And 20, determining an initial logistic fit curve of the target area power consumption according to the historical power consumption data of the target area, wherein the independent variable of the initial logistic fit curve is time, and the dependent variable is an initial predicted value of the target area power consumption.
Step 20 corresponds to constructing a logistic model that fits a curve to the initial logistic. The parameters a, b, c in the initial logistic fit curve of the power consumption in the target area can be determined by using a conventional parameter estimation method in step 20, for example, yule algorithm, rhodes algorithm, nair algorithm, etc. can be used.
In one implementation, the parameters a, b and c are estimated by a Yule algorithm which is simple and convenient to calculate and easy to operate.
Firstly, a first sub-relation (1-1) is obtained by transforming the first relation (1), wherein the first sub-relation (1-1) is as follows:
Figure GDA0003789262900000051
is provided with
Figure GDA0003789262900000052
γ＝1-e b And β = -c (1-e) b ) Then, the first sub-relation (1-1) is further transformed into a linear equation z t ＝γ+βy t . By using the least square method, the estimated values of the parameters gamma and beta of the linear equation can be obtained, and then the estimated values of b and c can be obtained by calculation according to the estimated values of the parameters gamma and beta
Figure GDA0003789262900000053
And
Figure GDA0003789262900000054
to obtain an estimated value of the parameter a, the first relation (1) is transformed into a second sub-relation (1-2), the second sub-relation (1-2) being:
Figure GDA0003789262900000055
wherein t =1,2, \8230;, n in the second sub-relation (1-2).
Summing t on the left and right in the second sub-relation (1-2) respectively to obtain an estimated value of a
Figure GDA0003789262900000056
Comprises the following steps:
Figure GDA0003789262900000057
according to the parameter estimation value obtained by the calculation
Figure GDA0003789262900000058
And
Figure GDA0003789262900000059
an initial logistic fit curve for target area power usage can be determined
Figure GDA00037892629000000510
Wherein, the first and the second end of the pipe are connected with each other,
Figure GDA00037892629000000511
and representing an initial predicted value of the power consumption of the target area, and t represents time.
The initial predicted value of the power consumption of the target area is further optimized through the following steps 30 to 70, so that a predicted value of the power consumption with higher precision is obtained.
And step 30, calculating to obtain an initial difference value curve according to the initial logistic fitting curve and the historical power consumption data of the target area, wherein the independent variable of the initial difference value curve is time, and the dependent variable is an initial difference value, and the initial difference value is the difference value between the historical power consumption of the target area and the initial predicted value of the power consumption.
It should be appreciated that from the target area historical power usage data, a target area historical power usage curve may be determined that represents the actual power usage y for each time segment of the target area t 。
Further, the real power consumption y can be calculated t Initial predicted value of electricity consumption
Figure GDA0003789262900000068
Obtaining an initial difference curve, wherein the dependent variable in the initial difference curve is the difference g between the initial predicted value of the power consumption and the real power consumption t 。
And step 40, respectively carrying out normalization processing on the initial logistic fitting curve and the initial difference value curve to obtain a training sample data set, wherein the training sample data set comprises sample data of the initial predicted value of the power consumption of the target area and the sample data of the initial difference value.
The method for obtaining the initial predicted value of the power consumption by adopting the regression method of the support vector machine
Figure GDA0003789262900000069
Difference g from the initial t The regression equation between them first needs to be normalized to the initial logistic fit curve and the initial difference curve.
The initial logistic fitting curve can be normalized according to a second relational expression, where the second relational expression is:
Figure GDA0003789262900000061
second relationIn the formula (2), the reaction mixture is,
Figure GDA0003789262900000062
sample data representing an initial predicted value of electricity usage by the target area,
Figure GDA0003789262900000063
an initial predicted value representing a used amount of electricity of the target area,
Figure GDA0003789262900000064
and m represents the initial predicted value of the electricity consumption with the change rate within 1% compared with the historical electricity consumption of the target area.
It should be noted that, in the second relational expression, m should be the maximum value in general, but since the saturated power consumption has an increasing trend, the maximum value cannot be determined, and in order to ensure the rationality and validity of the normalization result and combine the average absolute percentage error condition, the prediction result when the change rate of the initial prediction value compared with the actual power consumption is within 1% is set to be m.
The difference curve may be normalized according to a third relation (3), where the third relation is:
Figure GDA0003789262900000065
in the third relation (3),
Figure GDA0003789262900000066
sample data representing said initial difference, g t Represents the initial difference, min (g) t ) Represents the minimum value, max (g), of said initial difference t ) Represents the maximum of the initial differences.
The training sample data set obtained in the step 40 is a plurality of data groups, and each data group includes initial predicted values and initial difference values which are in one-to-one correspondence. For example, a training sample data setCan be expressed as
Figure GDA0003789262900000067
And 50, training and learning the training sample data set through a support vector machine to obtain a regression prediction curve, wherein the independent variable of the regression prediction curve is time, the dependent variable of the regression prediction curve is a predicted value of a difference value, and the predicted value of the difference value is the difference value between the historical electricity consumption of the target area and the optimized predicted value of the electricity consumption.
Firstly, it should be noted that the support vector machine is an algorithm, and the basic idea of the algorithm is to map an input vector to a high-dimensional feature space through a certain rule, and construct an optimal classification hyperplane in the space, so as to convert a nonlinear problem into a linear problem and solve the linear problem.
Before the training and learning of the training sample data set are carried out by adopting a support vector machine, setting proper support vector machine regression parameters is the basis for ensuring that an accurate regression prediction curve is obtained. In one implementation, support vector machine regression is implemented by using an LIBSVM toolkit in MATLAB, a gaussian Radial Basis Function (RBF) is selected as a kernel function, and the RBF has the advantages of small deviation, radial symmetry, strong generalization capability, good smoothness and the like, wherein a gamma parameter in the kernel function is set to be 2.8, and the value of a loss function p is 0.01.
Based on the calculation principle of the support vector machine
Figure GDA0003789262900000071
Thus the training sample set can be expressed as { (x) i ，y i ) I =1, \8230;, n }, where the regression of the support vector machine is implemented by using LIBSVM toolkit in MATLAB, the basic principle is to find the problem of optimal regression hyperplane, i.e., solving the function y = w · x + b, which can be converted into solving the following quadratic programming problem:
Figure GDA0003789262900000072
Figure GDA0003789262900000073
when solving the quadratic programming problem, introducing a Lagrange multiplier a i ，
Figure GDA0003789262900000074
A dual problem is obtained:
Figure GDA0003789262900000075
Figure GDA0003789262900000076
in the fifth relational expression, K (x) i ，x j ) Representing a kernel function, the kernel function satisfying the following sixth relation (6):
Figure GDA0003789262900000077
therefore, w and b in the fourth relational expression (4) can be solved from the fifth relational expression (5) and the sixth relational expression (6), and the solving function can be determined to be
Figure GDA0003789262900000078
It can be seen that the above solving function is the regression prediction curve of the present application, where the independent variable x is time, and the dependent variable y is the difference g t The predicted value of (g), the difference value g t Can also be expressed as
Figure GDA0003789262900000079
Figure GDA00037892629000000710
Can be understood as the optimized difference value g t 。
And step 60, adding the regression prediction curve and the initial logistic fitted curve to obtain an optimized logistic fitted curve, wherein the independent variable of the optimized logistic fitted curve is time, and the dependent variable is the optimized predicted value of the power consumption.
The regression prediction curve obtained in step 50 can be used to compensate for the deviation between the actual saturated power consumption development and the initial logistic fitting curve in step 20, thereby improving the prediction accuracy.
Specifically, the initial logistic fitted curve determined in step 20 is added to the regression prediction curve obtained in step 50 to obtain the final optimized logistic fitted curve
Figure GDA00037892629000000711
Wherein the content of the first and second substances,
Figure GDA00037892629000000712
represents the optimized logistic fitted curve,
Figure GDA0003789262900000081
representing an initial logistic fit curve,
Figure GDA0003789262900000082
representing a regression prediction curve.
Referring to fig. 2, fig. 2 shows the prediction results of the total social power consumption in 2001-2019 of beijing city by using different prediction methods. As can be seen from fig. 2, there is a larger difference between the initial logistic fitted curve and the real power consumption curve, and the optimized logistic fitted curve is almost overlapped with the real power consumption curve, which illustrates that the saturated power consumption can be accurately predicted by using the method for predicting saturated power consumption based on the logistic model provided in the embodiment of the present application.
Continuing to refer to FIG. 3, FIG. 3 shows the absolute percentage error of the prediction of total social power usage in 2001-2019 of Beijing, using different prediction methods. As can be seen from FIG. 3, the gray prediction method, the logistic prediction method before optimization, and the optimized logis provided by the present applicationCompared with a grey prediction method and a logistic prediction method before optimization, the optimized logistic prediction method provided by the application has absolute advantages. Wherein the mean absolute percentage error
Figure GDA0003789262900000083
Wherein t represents time, y t The actual amount of electricity used is indicated,
Figure GDA0003789262900000084
indicating the predicted amount of power usage.
Furthermore, a more accurate logistic fitting curve can be obtained by adopting the optimized logistic prediction method provided by the application, so that the saturated power consumption of the target time period of the target area can be further predicted based on the optimized logistic fitting curve.
In conclusion, the application introduces the difference value g t The prediction value optimizes the initial prediction result of the traditional logistic, so that the deviation condition of the actual urban power consumption and the initial logistic fitting curve caused by the influence of various factors can be compensated, the defect of inaccurate parameter estimation of the logistic model and the difference value g can be compensated t The prediction value is obtained by predicting through a support vector machine regression based on the initial logistic fitting curve, the method improves the condition that the prediction effect of the original single logistic prediction method is poor, the accuracy of the prediction of the saturated power consumption is effectively improved, and the method has good generalization capability, can ensure the balance of power supply and demand, provides important guarantee for improving the living standard of people, and provides powerful support for the healthy development of economy.
Embodiments of the apparatus corresponding to the above-described embodiments of the method are described below.
Referring to fig. 4, the present application provides a saturated power consumption prediction system based on a logistic model, comprising:
the obtaining module 110 is configured to obtain historical power consumption data of a target area;
an initial logistic fit curve determining module 120, configured to determine an initial logistic fit curve of the target area power consumption according to the historical power consumption data of the target area, where an independent variable of the initial logistic fit curve is time, and a dependent variable is an initial predicted value of the target area power consumption;
an initial difference curve calculation module 130, configured to calculate an initial difference curve according to the initial logistic fit curve and the historical power consumption data of the target area, where an independent variable of the initial difference curve is time, and a dependent variable is an initial difference, where the initial difference is a difference between the historical power consumption of the target area and an initial predicted value of the power consumption;
a normalization processing module 140, configured to perform normalization processing on the initial logistic fitted curve and the initial difference curve, respectively, to obtain a training sample data set, where the training sample data set includes sample data of an initial predicted value of power consumption in the target area and sample data of the initial difference;
the regression prediction curve calculation module 150 is configured to perform training learning on the training sample data set through a support vector machine to obtain a regression prediction curve, where an independent variable of the regression prediction curve is time, a dependent variable of the regression prediction curve is a prediction value of a difference value, and the prediction value of the difference value is a difference value between a historical power consumption of the target area and an optimized prediction value of the power consumption;
and an adding module 160, configured to add the regression prediction curve and the initial logistic fitted curve to obtain an optimized logistic fitted curve, where an independent variable of the optimized logistic fitted curve is time, and a dependent variable is an optimized predicted value of power consumption.
Optionally, the system further includes a prediction module, configured to predict, according to the optimized logistic fitting curve, a saturated power consumption in a target time period of the target area.
Optionally, the normalization processing module includes a first normalization processing sub-module and a second normalization processing sub-module;
the first normalization processing sub-module is configured to perform normalization processing on the initial logistic fitted curve according to a second relational expression, where the second relational expression is:
Figure GDA0003789262900000091
wherein, the first and the second end of the pipe are connected with each other,
Figure GDA0003789262900000092
sample data representing an initial predicted value of electricity usage by the target area,
Figure GDA0003789262900000093
an initial predicted value representing a used amount of electricity of the target area,
Figure GDA0003789262900000094
and m represents the initial predicted value of the electricity consumption with the change rate within 1% compared with the historical electricity consumption of the target area.
The second normalization processing sub-module is configured to perform normalization processing on the difference curve according to a third relation, where the third relation is:
Figure GDA0003789262900000095
wherein, the first and the second end of the pipe are connected with each other,
Figure GDA0003789262900000096
sample data representing said initial difference, g t Represents the initial difference, min (g) t ) Represents the minimum value, max (g), of said initial difference t ) Represents the maximum of the initial differences.
Optionally, a gaussian radial basis function is selected as a kernel function for training and learning the training sample data set by using a support vector machine, where a formula of the kernel function is:
Figure GDA0003789262900000097
wherein, K (x) i ,x j ) Representing the kernel function, x i And x j And the parameter gamma is set to be 2.8 according to the sample data of the initial predicted value of the electricity consumption of the target area in different time periods.
In a specific implementation, an embodiment of the present application further provides a computer storage medium, where the computer storage medium may store a program, where the program includes instructions, and when executed, the program may include some or all of the steps of the method for predicting saturated power consumption based on a logistic model provided in the present application. The computer storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), a Random Access Memory (RAM), or the like.
In the above embodiments, all or part may be implemented by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described herein to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, from one website site, computer, server, or data center to another website site, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a Solid State Disk (SSD)), among others.
The saturated power consumption prediction system based on the logistic model and the computer storage medium are used for executing part or all of the steps of the saturated power consumption prediction method based on the logistic model provided by any one of the embodiments, and accordingly have the beneficial effects of the saturated power consumption prediction method based on the logistic model, and are not described herein again.
It should be understood that, in the embodiments of the present application, the execution sequence of each step should be determined by its function and inherent logic, and the size of the sequence number of each step does not mean the execution sequence, and does not limit the implementation process of the embodiments.
In addition, in the description of the present application, "a plurality" means two or more than two unless otherwise specified. In addition, in order to facilitate clear description of technical solutions of the embodiments of the present application, in the embodiments of the present application, terms such as "first" and "second" are used to distinguish the same items or similar items having substantially the same functions and actions. Those skilled in the art will appreciate that the terms "first," "second," etc. do not denote any order or quantity, nor do the terms "first," "second," etc. denote any order or importance.
All parts of this specification are described in a progressive manner, and like parts of the various embodiments can be referred to one another, with emphasis on each embodiment being placed on differences from other embodiments. In particular, for the embodiments of the saturated power consumption prediction method based on the logistic model and the saturated power consumption prediction system based on the logistic model, since the embodiments are basically similar to the method embodiments, the description is simple, and the relevant points can be referred to the description in the method embodiments.
While the preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all alterations and modifications as fall within the scope of the application.
The above-described embodiments of the present application do not limit the scope of the present application.

Claims (6)

1. A saturated power consumption prediction method based on a logistic model is characterized by comprising the following steps:
acquiring historical electricity consumption data of a target area;
determining an initial logistic fit curve of the power consumption of the target area according to the historical power consumption data of the target area, wherein the independent variable of the initial logistic fit curve is time, and the dependent variable is an initial predicted value of the power consumption of the target area;
calculating to obtain an initial difference value curve according to the initial logistic fitting curve and the historical power consumption data of the target area, wherein the independent variable of the initial difference value curve is time, and the dependent variable is an initial difference value, and the initial difference value is the difference value between the historical power consumption of the target area and the initial predicted value of the power consumption;
respectively carrying out normalization processing on the initial logistic fitted curve and the initial difference value curve to obtain a training sample data set, wherein the training sample data set comprises sample data of an initial predicted value of the power consumption of the target area and the sample data of the initial difference value;
training and learning the training sample data set through a support vector machine to obtain a regression prediction curve; selecting a Gaussian radial basis kernel function as a kernel function for training and learning the training sample data set by a support vector machine, wherein the kernel function is as follows:
Figure FDA0003789262890000011
wherein, K (x) i ,x j ) Representing the kernel function, x i And x j Sample data of an initial predicted value of the power consumption of the target area in different time periods are represented, and a parameter gamma is set to be 2.8; the regression prediction curve is
Figure FDA0003789262890000012
Wherein the independent variable x is time, the dependent variable y is the predicted value of the difference, a i And are and
Figure FDA0003789262890000013
the difference value is a Lagrange multiplier, and the predicted value of the difference value is the difference value between the historical electricity consumption of the target area and the optimized predicted value of the electricity consumption;
adding the regression prediction curve and the initial logistic fitted curve to obtain an optimized logistic fitted curve, wherein the independent variable of the optimized logistic fitted curve is time, and the dependent variable is the optimized predicted value of the power consumption;
and predicting the saturated power consumption of the target time period of the target area according to the optimized logistic fitting curve.
2. The method of claim 1, wherein determining an initial logistic fit curve for the target area power usage based on the target area historical power usage data comprises:
estimating parameters a, b and c in a logistic fitting curve by using a Yule algorithm, wherein the logistic fitting curve satisfies the following first relational expression:
Figure FDA0003789262890000014
wherein t represents time, y t Representing the initial predicted value of the electricity consumption of the target area, and a, b and c represent parameters of a logistic fitting curve;
and determining an initial logistic fitted curve of the target area electricity consumption according to the estimated parameters a, b and c.
3. The method of claim 1, wherein the initial logistic fitted curve is normalized according to a second relation, the second relation being:
Figure FDA0003789262890000021
wherein, the first and the second end of the pipe are connected with each other,
Figure FDA0003789262890000022
sample data representing an initial predicted value of electricity usage by the target area,
Figure FDA0003789262890000023
an initial predicted value representing a used amount of electricity of the target area,
Figure FDA0003789262890000024
and m represents the initial predicted value of the electricity consumption with the change rate within 1% compared with the historical electricity consumption of the target area.
4. The method of claim 1, wherein the initial difference curve is normalized according to a third relationship, the third relationship being:
Figure FDA0003789262890000025
wherein the content of the first and second substances,
Figure FDA0003789262890000026
sample data representing said initial difference, g t Represents the initial difference, min (g) t ) Represents the minimum value, max (g), of said initial difference t ) Representing the maximum of said initial differences。
5. A saturated power consumption prediction system based on a logistic model is characterized by comprising the following components:
the acquisition module is used for acquiring historical power consumption data of a target area;
the initial logistic fitted curve determining module is used for determining an initial logistic fitted curve of the power consumption of the target area according to the historical power consumption data of the target area, wherein an independent variable of the initial logistic fitted curve is time, and a dependent variable is an initial predicted value of the power consumption of the target area;
an initial difference curve calculation module, configured to calculate an initial difference curve according to the initial logistic fit curve and the historical power consumption data of the target area, where an independent variable of the initial difference curve is time, and a dependent variable is an initial difference, where the initial difference is a difference between the historical power consumption of the target area and an initial predicted value of the power consumption;
the normalization processing module is used for respectively carrying out normalization processing on the initial logistic fitting curve and the initial difference value curve to obtain a training sample data set, wherein the training sample data set comprises sample data of an initial predicted value of the power consumption of the target area and the sample data of the initial difference value;
the regression prediction curve calculation module is used for training and learning the training sample data set through a support vector machine to obtain a regression prediction curve; selecting a Gaussian radial basis kernel function as a kernel function for training and learning the training sample data set by a support vector machine, wherein the kernel function is as follows:
Figure FDA0003789262890000027
wherein, K (x) i ,x j ) Representing the kernel function, x i And x j Sample data representing initial predicted values of power consumption of the target area in different time periods, wherein a parameter gamma is set to be 2.8; the returnIs returned to the predicted curve
Figure FDA0003789262890000028
Wherein the independent variable x is time, the dependent variable y is the predicted value of the difference, a i And are and
Figure FDA0003789262890000031
the difference value is a Lagrange multiplier, and the predicted value of the difference value is the difference value between the historical electricity consumption of the target area and the optimized predicted value of the electricity consumption;
the addition module is used for adding the regression prediction curve and the initial logistic fitted curve to obtain an optimized logistic fitted curve, wherein the independent variable of the optimized logistic fitted curve is time, and the dependent variable is the optimized predicted value of the power consumption;
and the prediction module is used for predicting the saturated power consumption of the target time period of the target area according to the optimized logistic fitting curve.
6. The system of claim 5, wherein the normalization processing module comprises a first normalization processing sub-module and a second normalization processing sub-module;
the first normalization processing sub-module is configured to perform normalization processing on the initial logistic fitted curve according to a second relational expression, where the second relational expression is:
Figure FDA0003789262890000032
wherein the content of the first and second substances,
Figure FDA0003789262890000033
sample data representing an initial predicted value of electricity usage by the target area,
Figure FDA0003789262890000034
indicating the initial amount of electricity used in the target areaThe value of the initial prediction is obtained,
Figure FDA0003789262890000035
the minimum value of the initial predicted values of the power consumption of the target area is represented, and m represents the initial predicted value of the power consumption of which the change rate is within 1% compared with the historical power consumption of the target area;
the second normalization processing sub-module is configured to perform normalization processing on the initial difference curve according to a third relation, where the third relation is:
Figure FDA0003789262890000036
wherein the content of the first and second substances,
Figure FDA0003789262890000037
sample data representing said initial difference, g t Represents the initial difference, min (g) t ) Represents the minimum value of said initial difference, max (g) t ) Represents the maximum of the initial differences.
CN202110749465.7A 2021-07-02 2021-07-02 Method and system for predicting saturated power consumption based on logistic model Active CN113435653B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110749465.7A CN113435653B (en) 2021-07-02 2021-07-02 Method and system for predicting saturated power consumption based on logistic model

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110749465.7A CN113435653B (en) 2021-07-02 2021-07-02 Method and system for predicting saturated power consumption based on logistic model

Publications (2)

Publication Number Publication Date
CN113435653A CN113435653A (en) 2021-09-24
CN113435653B true CN113435653B (en) 2022-11-04

Family

ID=77758686

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110749465.7A Active CN113435653B (en) 2021-07-02 2021-07-02 Method and system for predicting saturated power consumption based on logistic model

Country Status (1)

Country Link
CN (1) CN113435653B (en)

Citations (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104199870A (en) * 2014-08-19 2014-12-10 桂林电子科技大学 Method for building LS-SVM prediction model based on chaotic search
CN104573854A (en) * 2014-12-23 2015-04-29 国家电网公司 Iron steel electricity consumption forecasting method and device
CN105225006A (en) * 2015-09-30 2016-01-06 杨明 A kind of short-term wind-electricity power nonparametric probability forecasting method
CN105404935A (en) * 2015-11-11 2016-03-16 国网浙江省电力公司经济技术研究院 Electric power system monthly load prediction method considering business expansion increment
AU2015282362A1 (en) * 2015-01-06 2016-07-21 Accenture Global Services Limited Power distribution transformer load prediction analysis system
JP2017049674A (en) * 2015-08-31 2017-03-09 富士通株式会社 Prediction performance curve estimation program, prediction performance curve estimation device and prediction performance curve estimation method
CN106548210A (en) * 2016-10-31 2017-03-29 腾讯科技（深圳）有限公司 Machine learning model training method and device
CN108022008A (en) * 2017-11-30 2018-05-11 国网福建省电力有限公司经济技术研究院 The training method of electricity demand forecasting model
CN108108837A (en) * 2017-12-15 2018-06-01 国网新疆电力有限公司经济技术研究院 A kind of area new energy power supply structure optimization Forecasting Methodology and system
CN109447332A (en) * 2018-10-18 2019-03-08 国网天津市电力公司电力科学研究院 A kind of Middle-long Electric Power Load Forecast method suitable for S type load curve
CN109492818A (en) * 2018-11-16 2019-03-19 上海交通大学 Based on energy development and the entitled electricity demand forecasting method of Shapley value
CN110516831A (en) * 2019-06-18 2019-11-29 国网（北京）节能设计研究院有限公司 A kind of short-term load forecasting method based on MWOA algorithm optimization SVM
CN111259340A (en) * 2020-01-16 2020-06-09 天津大学 Saturated load prediction method based on logistic regression
CN111523674A (en) * 2019-02-01 2020-08-11 阿里巴巴集团控股有限公司 Model training method, device and system
CN112257958A (en) * 2020-11-12 2021-01-22 国网河北省电力有限公司经济技术研究院 Power saturation load prediction method and device
JP2021022051A (en) * 2019-07-25 2021-02-18 富士通株式会社 Machine learning program, machine learning method, and machine learning apparatus
CN112668807A (en) * 2021-01-17 2021-04-16 国网江苏省电力有限公司宿迁供电分公司 Wind speed prediction method for wind power plant power prediction

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11601374B2 (en) * 2014-10-30 2023-03-07 Pearson Education, Inc Systems and methods for data packet metadata stabilization
US20200034665A1 (en) * 2018-07-30 2020-01-30 DataRobot, Inc. Determining validity of machine learning algorithms for datasets

Patent Citations (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104199870A (en) * 2014-08-19 2014-12-10 桂林电子科技大学 Method for building LS-SVM prediction model based on chaotic search
CN104573854A (en) * 2014-12-23 2015-04-29 国家电网公司 Iron steel electricity consumption forecasting method and device
AU2015282362A1 (en) * 2015-01-06 2016-07-21 Accenture Global Services Limited Power distribution transformer load prediction analysis system
JP2017049674A (en) * 2015-08-31 2017-03-09 富士通株式会社 Prediction performance curve estimation program, prediction performance curve estimation device and prediction performance curve estimation method
CN105225006A (en) * 2015-09-30 2016-01-06 杨明 A kind of short-term wind-electricity power nonparametric probability forecasting method
CN105404935A (en) * 2015-11-11 2016-03-16 国网浙江省电力公司经济技术研究院 Electric power system monthly load prediction method considering business expansion increment
CN106548210A (en) * 2016-10-31 2017-03-29 腾讯科技（深圳）有限公司 Machine learning model training method and device
CN108022008A (en) * 2017-11-30 2018-05-11 国网福建省电力有限公司经济技术研究院 The training method of electricity demand forecasting model
CN108108837A (en) * 2017-12-15 2018-06-01 国网新疆电力有限公司经济技术研究院 A kind of area new energy power supply structure optimization Forecasting Methodology and system
CN109447332A (en) * 2018-10-18 2019-03-08 国网天津市电力公司电力科学研究院 A kind of Middle-long Electric Power Load Forecast method suitable for S type load curve
CN109492818A (en) * 2018-11-16 2019-03-19 上海交通大学 Based on energy development and the entitled electricity demand forecasting method of Shapley value
CN111523674A (en) * 2019-02-01 2020-08-11 阿里巴巴集团控股有限公司 Model training method, device and system
CN110516831A (en) * 2019-06-18 2019-11-29 国网（北京）节能设计研究院有限公司 A kind of short-term load forecasting method based on MWOA algorithm optimization SVM
JP2021022051A (en) * 2019-07-25 2021-02-18 富士通株式会社 Machine learning program, machine learning method, and machine learning apparatus
CN111259340A (en) * 2020-01-16 2020-06-09 天津大学 Saturated load prediction method based on logistic regression
CN112257958A (en) * 2020-11-12 2021-01-22 国网河北省电力有限公司经济技术研究院 Power saturation load prediction method and device
CN112668807A (en) * 2021-01-17 2021-04-16 国网江苏省电力有限公司宿迁供电分公司 Wind speed prediction method for wind power plant power prediction

Non-Patent Citations (7)

* Cited by examiner, † Cited by third party
Title
Improved Parametric Estimation of Logistic Model for Saturated Load Forecast;Yudong Jia 等;《2012 Asia-Pacific Power and Energy Engineering Conference》;20120920;第1-4页 *
Short-term electric load forecasting for buildings using logistic mixture vector autoregressive model with curve registration;DongyeonJeong 等;《Applied Energy》;20210115;第282卷;第1-14页 *
基于Logistic回归的数据分类问题研究;邹晓辉;《智能计算机与应用》;20161228;第6卷(第06期);第139-140、143页 *
基于多维度的城市饱和负荷预测方法研究;何淼 等;《沈阳工程学院学报(自然科学版)》;20170415;第13卷(第02期);第164-168、178页 *
基于密度估计的逻辑回归模型;毛毅等;《自动化学报》;20140115;第40卷(第01期);第62-72页 *
情景分析在饱和负荷预测中的应用;陈屹东 等;《电力系统及其自动化学报》;20160615;第28卷(第06期);第7-13页 *
饱和负荷预测中的改进Logistic算;鲍正杰 等;《电力系统及其自动化学报》;20141115;第26卷(第11期);第7-11页 *

Also Published As

Publication number Publication date
CN113435653A (en) 2021-09-24

Similar Documents

Publication Publication Date Title
Li et al. A novel grey forecasting model and its application in forecasting the energy consumption in Shanghai
CN116307215A (en) Load prediction method, device, equipment and storage medium of power system
CN110458601B (en) Method and device for processing resource data, computer equipment and storage medium
CN112668822A (en) Scientific and technological achievement transformation platform sharing system, method, storage medium and mobile phone APP
Chang et al. A mega-trend-diffusion grey forecasting model for short-term manufacturing demand
CN112215398A (en) Power consumer load prediction model establishing method, device, equipment and storage medium
CN113705929B (en) Spring festival holiday load prediction method based on load characteristic curve and typical characteristic value fusion
Tang et al. Forecasting annual foreign tourist arrivals to China by incorporating firefly algorithm into fractional non-homogenous discrete Grey model
CN112257958A (en) Power saturation load prediction method and device
CN113435653B (en) Method and system for predicting saturated power consumption based on logistic model
Su Nonlinear ARIMA models with feedback SVR in financial market forecasting
Cheng A grey CES production function model and its application in Calculating the contribution rate of economic growth factors
Hashorva et al. Some mathematical aspects of price optimisation
Yu et al. Post-averaging inference for optimal model averaging estimator in generalized linear models
Xu et al. Research on Multistage Dynamic Trading Model Based on Gray Model and Auto‐Regressive Integrated Moving Average Model
Mao et al. The impact of R&D on GDP study based on grey delay Lotka-Volterra model
CN112950390A (en) Equipment system investment portfolio selection method, electronic equipment and storage medium
Qin A study on real estate price by using probability statistics theory and grey theory
Zha et al. Container throughput time series forecasting using a hybrid approach
Guo et al. A Fuzzy‐Decomposition Grey Modeling Procedure for Management Decision Analysis
KR101830090B1 (en) System and method for optimizing portfolio, and executable digital media storing computational programs implementing the same
Ma et al. Applying Grey System Theory to Predict the Three Major Industries Structure Evolution in China
Tkachenko et al. Forecasting analytic model of construction energy cluster development in Ukraine
Lan et al. Digital Investment Risk Evaluation Model of Power Grid Enterprises Based on FAHP-AOA-LSSVM
Garcia et al. Filter method-based feature selection process for unattributed-identity multi-target regression problem

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
GR01 Patent grant
GR01 Patent grant