CN113554213A - Natural gas demand prediction method, system, storage medium and equipment - Google Patents
Natural gas demand prediction method, system, storage medium and equipment Download PDFInfo
- Publication number
- CN113554213A CN113554213A CN202110656778.8A CN202110656778A CN113554213A CN 113554213 A CN113554213 A CN 113554213A CN 202110656778 A CN202110656778 A CN 202110656778A CN 113554213 A CN113554213 A CN 113554213A
- Authority
- CN
- China
- Prior art keywords
- neural network
- natural gas
- gas demand
- prediction
- value
- 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
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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06315—Needs-based resource requirements planning or analysis
-
- 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Abstract
The invention provides a natural gas demand forecasting method, a natural gas demand forecasting system, a storage medium and equipment, wherein a natural gas demand historical data sample is obtained, and key factors influencing natural gas demand are screened by using different analysis methods; respectively constructing a neural network model based on the screened key factors, and training and testing the neural network model until the neural network model meets the requirements; determining key factor values based on the belonging area planning or the situation hypothesis by combining the characteristic factor screening result, constructing a prediction sample, and inputting the prediction sample into a neural network to predict the natural gas demand to obtain a prediction result; and comparing the prediction results of the neural network models constructed by different screening methods, and combining and outputting the different prediction results when the deviation among various results is less than a preset value. The method avoids the influence on the neural network caused by the factors which have small influence on the natural gas demand or reflect information overlapping, and improves the accuracy and precision of the neural network prediction.
Description
Technical Field
The invention belongs to the technical field of energy demand utilization, and particularly relates to a natural gas demand prediction method, a natural gas demand prediction system, a natural gas demand prediction storage medium and natural gas demand prediction equipment.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
Natural gas is a low-carbon, green, clean and efficient energy source, and occupies an important position in energy production and utilization in China. In recent years, natural gas energy production and consumption in China are in a growing state, and in order to guarantee stable, safe and continuous supply of natural gas, accurate prediction on natural gas demand is needed.
The traditional natural gas demand forecasting method mainly comprises a regression analysis forecasting method, a time sequence forecasting method, a grey model forecasting method and various model combination forecasting methods, wherein the models only consider the influence of time factors on the natural gas demand, and when the natural gas demand is changed seriously due to an emergency in a certain time period, the stability and the accuracy of the models are difficult to guarantee.
In recent years, artificial neural networks based on big data analysis show good effects in the aspect of natural gas demand prediction, and have received research attention.
However, the inventor knows that in the actual operation process, because the artificial neural network comprehensively considers the influence of various factors on the natural gas demand, the model is established by learning and training according to the historical data of the influencing factors, but the natural gas demand is influenced by a plurality of factors, so that the neural network has more input parameter quantity, the structure of the neural network to be established is complex, the training process and the calculation process are complex, and the calculation quantity is large; on the other hand, most factors selected by the neural network model at present are only based on qualitative analysis, and the prediction accuracy of the neural network is influenced.
Disclosure of Invention
The invention provides a natural gas demand prediction method, a natural gas demand prediction system, a natural gas demand prediction storage medium and natural gas demand prediction equipment, which can reduce the complexity and the calculation amount of a neural network, avoid the influence of factors which have small influence on natural gas demand or reflect information overlapping on the neural network, and improve the precision and the accuracy of the neural network prediction.
According to some embodiments, the invention adopts the following technical scheme:
a natural gas demand forecasting method comprises the following steps:
(1) acquiring a natural gas demand historical data sample, and screening key factors influencing the natural gas demand by using different analysis methods;
(2) respectively constructing a neural network model based on the screened key factors, and training and testing the neural network model until the neural network model meets the requirements;
(3) determining key factor values based on the belonging area planning or the situation hypothesis by combining the characteristic factor screening result, constructing a prediction sample, and inputting the prediction sample into a neural network to predict the natural gas demand to obtain a prediction result;
(4) comparing the prediction results of the neural network models constructed by different screening methods, and when the deviation among various results is smaller than a preset value, combining and outputting the different prediction results; otherwise, returning to the step (2).
As an alternative embodiment, in the step (1), the different analysis methods include a gray correlation analysis method, an average influence value analysis method, and a principal component analysis method.
As a further limitation, in the step (1), a specific process of analysis by a gray correlation analysis method includes:
1-1) taking the natural gas demand as a reference series, taking the rest of influence factors as comparison series, and respectively carrying out non-dimensionalization processing on the reference series and the comparison series;
1-2) solving and comparing the grey correlation coefficient of the number series and the reference number series, and calculating the correlation of the influence factors;
1-3) sorting the association degrees of the influence factors from large to small, and screening the characteristic factors with the association degrees larger than a preset value.
As a further limitation, in the step (1), the specific process of the analysis of the average influence value analysis method includes:
2-1) establishing a neural network model by using the obtained original data sample, performing learning training and outputting the neural network model;
2-2) taking out a certain influence factor of an original sample, respectively adding and subtracting a certain proportion to the original value, keeping the other parameters unchanged, obtaining two new samples, outputting the constructed samples to a neural network to obtain two output results, calculating the difference value of the output results, solving an average value according to the number of the samples, and obtaining the average influence value of the taken out factor;
2-3) repeatedly executing the step 2-2) to obtain the average influence value of other factors, sorting the average influence values from large to small, and screening the characteristic factors of which the average influence value is larger than a set value.
As a further limitation, in the step (1), the specific process of principal component analysis includes:
3-1) carrying out standardization processing on the original influence factor data, and calculating a correlation matrix;
3-2) calculating the eigenvalue of the correlation matrix and the corresponding eigenvector thereof;
3-3) sorting according to the characteristic values from large to small, and calculating the proportion of the characteristic values in the sum of the characteristic values, namely the variance contribution rate corresponding to the characteristic values;
3-4) screening principal components from front to back until the cumulative variance contribution rate exceeds a preset value, and calculating the determined principal component score.
As an alternative embodiment, in the step (2), the neural network model is a BP neural network model.
As an alternative embodiment, in the step (2), the specific process of performing training and testing includes: and (2) selecting a part of samples from the historical data samples in the step (1) as training samples, training the neural network model, and using the rest samples as test samples to test the neural network model.
As an alternative embodiment, in the step (2), the requirement that the neural network model meets the requirement means that the relative error between the test value and the real value is smaller than a set value, and the set value is set according to the quantity of the historical samples.
As an alternative embodiment, in the step (3), determining the value of the key factor based on the plan of the belonging area or based on the scenario assumption includes: determining the value of a key factor according to the planned natural gas demand influence factor level of the region to which the natural gas demand influence factor belongs, or assuming that the factor meets the value of the key factor under the expected change.
As an alternative embodiment, in the step (4), different prediction results are combined and output, and the weights of the prediction results obtained by the analysis methods are the same.
A natural gas demand forecasting system comprising:
the key factor analysis module is configured to obtain a natural gas demand historical data sample and screen key factors influencing the natural gas demand by using different analysis methods;
the prediction model building module is configured to build a neural network model respectively based on the screened key factors, and train and test the neural network model until the neural network model meets the requirements;
the prediction module is configured to determine key factor values based on the area planning or the situation hypothesis by combining the feature factor screening results, construct a prediction sample, and input the prediction sample into a neural network to predict the natural gas demand to obtain a prediction result;
and the result fusion module is configured to compare the prediction results of the neural network models constructed by different screening methods, and when the deviation among various results is smaller than a preset value, combine and output different prediction results.
An electronic device comprising a memory and a processor and computer instructions stored on the memory and executed on the processor, the computer instructions, when executed by the processor, performing the steps of a method for predicting a demand for natural gas as set forth above.
A computer readable storage medium storing computer instructions which, when executed by a processor, perform the steps of a method for natural gas demand forecasting as described above.
Compared with the prior art, the invention has the beneficial effects that:
the invention provides a natural gas prediction method based on characteristic factor screening BP neural network for the first time, and the characteristic factor screening is carried out by simultaneously utilizing grey correlation degree analysis, average influence value analysis and principal component analysis, so that the input parameter quantity of the neural network can be effectively reduced, the complexity and the calculated quantity of the neural network are reduced, the influence of factors which have small influence on natural gas requirements or reflect information overlapping on the neural network is avoided, and the precision and the accuracy of neural network prediction are improved. In addition, the three screening methods can also play a mutual verification role, and the result is output only when the deviation ranges of the three prediction results meet the requirements, so that the accuracy and the stability of the neural network model are improved.
In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the invention and together with the description serve to explain the invention and not to limit the invention.
FIG. 1 is a schematic flow diagram of the present invention.
The specific implementation mode is as follows:
the invention is further described with reference to the following figures and examples.
It is to be understood that the following detailed description is exemplary and is intended to provide further explanation of the invention as claimed. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
A method for predicting natural gas demand, as shown in fig. 1, includes the following steps:
determining natural gas demand influence factors and acquiring related historical data samples;
secondly, screening key factors (characteristic factors) influencing the demand of the natural gas by respectively adopting a grey correlation analysis method, an average influence value analysis method and a principal component analysis method based on the acquired data;
thirdly, constructing a BP neural network based on the characteristic factor screening result, randomly selecting a part of samples (70-95% in the embodiment and adjustable in the range) of the original sample capacity from the original data sample in the first step as training samples, carrying out BP neural network training, taking the rest samples as test samples to carry out BP neural network test, and calculating the relative error between the test value and the true value;
step four, when the error between the test value and the true value is smaller than a certain range, outputting the neural network, otherwise, re-executing the step three;
determining the value of input parameters (characteristic factors) of the neural network based on national or regional planning or scenario hypothesis by combining the characteristic factor screening result, constructing a prediction sample, and inputting the prediction sample into the neural network to predict the natural gas demand;
step six, comparing the BP neural network prediction results constructed by the three characteristic factor screening methods, and when the deviation between the three results is smaller than a threshold (in the embodiment, 10% is taken, in other embodiments, the threshold can be adjusted), combining and outputting the three prediction results; otherwise, re-executing the third step.
The historical data of the first step can be derived from national statistics yearbook, regional annual reports, media public reports, autonomous research and the like.
Further, the grey correlation analysis method in the second step includes the following steps:
1) taking the natural gas demand as a reference series, and carrying out non-dimensionalization processing on the reference series and the comparison series respectively, wherein the rest of influence factors are comparison series;
2) solving and comparing the grey correlation coefficient of the number series and the reference number series, and calculating the correlation of the influence factors;
3) and sorting the relevance degrees of the influence factors from large to small, and screening the characteristic factors with the relevance degrees larger than a certain value.
The average influence value analysis method in the second step comprises the following steps:
1) establishing a BP neural network by using the obtained original data sample, performing learning training and outputting the neural network;
2) taking out a certain influence factor of an original sample, respectively adding and subtracting 10% of the original value, keeping the other parameters unchanged, obtaining two new samples, outputting the constructed samples to a neural network to obtain two output results, calculating the difference value of the output results, solving an average value according to the number of the samples, and obtaining an average influence value of the taken out factor;
3) and (3) repeatedly executing the step 2) to obtain the average influence values of other factors, sequencing the average influence values from large to small, and screening the characteristic factors with the average influence value larger than a certain value.
The principal component analysis method in the second step includes the steps of:
1) carrying out standardization processing on the original influence factor data, and calculating a correlation matrix;
2) calculating the eigenvalue of the correlation matrix and the corresponding eigenvector thereof;
3) sorting according to the characteristic values from large to small, and calculating the proportion of the characteristic values in the sum of the characteristic values, namely the variance contribution rate corresponding to the characteristic values;
4) the principal components are screened from front to back until the cumulative variance contribution exceeds a certain value, and the determined principal component score is calculated.
Preferably, a certain range of step four may be determined according to the number of the original samples obtained, for example, the original sample size is larger (more than 30), the range may be within 3%, the original data sample size is smaller (less than 20), and the range may be within 5%.
Preferably, the country or region planning in the fifth step refers to the natural gas demand influence factor level planned by the country, for example, when the GDP is screened as the characteristic factor, and the country has an 8% increase rate of the annual planning of GDP in the next 3 years, the GDP can be calculated by adopting the 8% increase rate; the scenario assumption of step five is to assume that the factors meet certain changes, such as the GDP maintains the annual growth rate in the next few years.
The above parameters are exemplary, and in other embodiments, adjustments may be made.
Preferably, the combined output is an average of the three predictions.
As a typical application example, neural network modeling is carried out on the natural gas demand and related influence factor data in 2018 years obtained from China statistics yearbook, and the natural gas demand in 2019 years 2025 years is predicted by using the method provided by the invention.
A natural gas demand prediction method based on characteristic factor screening BP neural network is disclosed, as shown in figure 1, and the specific implementation steps are as follows:
step one, acquiring natural gas consumption and relevant influence factor data samples of 2019 in 1995 + of China from China's annual book of statistics, wherein each group of data comprises 14 parameters in total, namely year, population, town population, total nutrient ratio, GDP, second industry output value, third industry output value, resident consumption index, commodity consumption price index, total energy production, natural gas production, total energy consumption, coal consumption and natural gas consumption, wherein the natural gas consumption is the natural gas demand, and the rest parameters are influence factors of the natural gas demand;
secondly, based on the obtained data, performing characteristic screening on the natural gas demand factors by respectively adopting a grey correlation degree analysis method, an average influence value analysis method and a principal component analysis method; the grey correlation degree analysis comprises the following specific steps:
1) selecting the natural gas demand as a reference series, and the other parameters (influence factors) as comparison series, and respectively carrying out non-dimensionalization processing on the reference series and the comparison series;
2) solving the grey correlation coefficient of the comparison array and the reference array, and calculating the correlation of the comparison array;
3) the comparison number with the screening relevance degree larger than 0.85 is a key factor influencing the natural gas demand.
The average influence value analysis comprises the following steps:
1) establishing a BP neural network by using the obtained original data sample, performing learning training and outputting the neural network;
2) taking out a certain influence factor of an original sample, respectively adding and subtracting 10% of the original value, keeping the other parameters unchanged, obtaining two new samples, outputting the constructed samples to a neural network to obtain two output results, calculating the difference value of the output results, solving an average value according to the number of the samples, and obtaining an average influence value of the taken out factor;
3) and (3) repeatedly executing the step 2) to obtain the average influence value of other factors, and screening the average influence value larger than the average value of all the factor influence values as the characteristic factor.
The principal component analysis steps are as follows:
1) carrying out standardization processing on the original influence factor data, and calculating a correlation matrix;
2) calculating the eigenvalue of the correlation matrix and the corresponding eigenvector thereof;
3) sorting according to the characteristic values from large to small, and calculating the proportion of the characteristic values in the sum of the characteristic values, namely the variance contribution rate corresponding to the characteristic values;
4) the principal components are screened from front to back until the cumulative variance contribution exceeds 0.9, and the determined principal component score is calculated.
Thirdly, constructing a BP neural network based on a feature screening result, randomly selecting samples with the quantity of 70% -95% of the original sample capacity from the original data samples in the first step as training samples, carrying out BP neural network training, using the rest samples as test samples to carry out BP neural network test, and calculating the relative error between a test value and a true value;
step four, when the error between the test value and the true value is less than 5%, outputting the neural network, otherwise, re-executing the step three;
step five, combining the characteristic factor screening results, calculating all values of input parameters (characteristic factors) of the neural network based on the situation hypothesis, constructing a prediction sample, and inputting the prediction sample into the neural network to predict the natural gas demand;
step six, comparing BP neural network prediction results constructed by the three characteristic factor screening methods, and when the deviation between the three results is less than 10%, combining and outputting the results to be used as a natural gas demand value predicted by the neural network; otherwise, re-executing the third step.
Through the operation of the steps, the test errors of the neural network models established by the three characteristic factor screening methods are less than 4%, the test error of the neural network without characteristic screening is 9%, and the prediction deviation of the three methods to the natural gas demand in 2019-.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Although the embodiments of the present invention have been described with reference to the accompanying drawings, it is not intended to limit the scope of the present invention, and it should be understood by those skilled in the art that various modifications and variations can be made without inventive efforts by those skilled in the art based on the technical solution of the present invention.
Claims (10)
1. A natural gas demand prediction method is characterized by comprising the following steps: the method comprises the following steps:
(1) acquiring a natural gas demand historical data sample, and screening key factors influencing the natural gas demand by using different analysis methods;
(2) respectively constructing a neural network model based on the screened key factors, and training and testing the neural network model until the neural network model meets the requirements;
(3) determining key factor values based on the belonging area planning or the situation hypothesis by combining the characteristic factor screening result, constructing a prediction sample, and inputting the prediction sample into a neural network to predict the natural gas demand to obtain a prediction result;
(4) comparing the prediction results of the neural network models constructed by different screening methods, and when the deviation among various results is smaller than a preset value, combining and outputting the different prediction results; otherwise, returning to the step (2).
2. A natural gas demand forecasting method as claimed in claim 1, characterized in that: in the step (1), the different analysis methods include a gray correlation analysis method, an average influence value analysis method and a principal component analysis method.
3. A natural gas demand forecasting method as claimed in claim 2, characterized in that: in the step (1), the specific process of grey correlation analysis method analysis includes:
1-1) taking the natural gas demand as a reference series, taking the rest of influence factors as comparison series, and respectively carrying out non-dimensionalization processing on the reference series and the comparison series;
1-2) solving and comparing the grey correlation coefficient of the number series and the reference number series, and calculating the correlation of the influence factors;
1-3) sorting the association degrees of the influence factors from large to small, and screening the characteristic factors with the association degrees larger than a preset value.
4. A natural gas demand forecasting method as claimed in claim 2, characterized in that: in the step (1), the specific process of the analysis of the average influence value analysis method comprises the following steps:
2-1) establishing a neural network model by using the obtained original data sample, performing learning training and outputting the neural network model;
2-2) taking out a certain influence factor of an original sample, respectively adding and subtracting a certain proportion to the original value, keeping the other parameters unchanged, obtaining two new samples, outputting the constructed samples to a neural network to obtain two output results, calculating the difference value of the output results, solving an average value according to the number of the samples, and obtaining the average influence value of the taken out factor;
2-3) repeatedly executing the step 2-2) to obtain the average influence value of other factors, sorting the average influence values from large to small, and screening the characteristic factors of which the average influence value is larger than a set value.
5. A natural gas demand forecasting method as claimed in claim 2, characterized in that: in the step (1), the specific process of principal component analysis method analysis comprises:
3-1) carrying out standardization processing on the original influence factor data, and calculating a correlation matrix;
3-2) calculating the eigenvalue of the correlation matrix and the corresponding eigenvector thereof;
3-3) sorting according to the characteristic values from large to small, and calculating the proportion of the characteristic values in the sum of the characteristic values, namely the variance contribution rate corresponding to the characteristic values;
3-4) screening principal components from front to back until the cumulative variance contribution rate exceeds a preset value, and calculating the determined principal component score.
6. A natural gas demand forecasting method as claimed in claim 1, characterized in that: in the step (2), the neural network model is a BP neural network model;
or, in the step (2), the specific process of training and testing includes: selecting a part of samples from the historical data samples in the step (1) as training samples to train the neural network model, and using the rest samples as test samples to test the neural network model;
or, in the step (2), the condition that the neural network model meets the requirement means that the relative error between the test value and the true value is smaller than a set value, and the set value is set according to the quantity of the historical samples.
7. A natural gas demand forecasting method as claimed in claim 1, characterized in that: in the step (4), different prediction results are combined and output, and the weight of the prediction results obtained by each analysis method is the same.
8. A natural gas demand forecasting system, characterized by: the method comprises the following steps:
the key factor analysis module is configured to obtain a natural gas demand historical data sample and screen key factors influencing the natural gas demand by using different analysis methods;
the prediction model building module is configured to build a neural network model respectively based on the screened key factors, and train and test the neural network model until the neural network model meets the requirements;
the prediction module is configured to determine key factor values based on the area planning or the situation hypothesis by combining the feature factor screening results, construct a prediction sample, and input the prediction sample into a neural network to predict the natural gas demand to obtain a prediction result;
and the result fusion module is configured to compare the prediction results of the neural network models constructed by different screening methods, and when the deviation among various results is smaller than a preset value, combine and output different prediction results.
9. An electronic device, characterized in that: comprising a memory and a processor and computer instructions stored on the memory and executed on the processor, which when executed by the processor, perform the steps of a method for natural gas demand forecasting according to any of claims 1 to 7.
10. A computer-readable storage medium characterized by: for storing computer instructions which, when executed by a processor, perform the steps of a method for natural gas demand forecasting according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110656778.8A CN113554213A (en) | 2021-06-11 | 2021-06-11 | Natural gas demand prediction method, system, storage medium and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110656778.8A CN113554213A (en) | 2021-06-11 | 2021-06-11 | Natural gas demand prediction method, system, storage medium and equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113554213A true CN113554213A (en) | 2021-10-26 |
Family
ID=78130506
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110656778.8A Pending CN113554213A (en) | 2021-06-11 | 2021-06-11 | Natural gas demand prediction method, system, storage medium and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113554213A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114184695A (en) * | 2021-11-09 | 2022-03-15 | 国网内蒙古东部电力有限公司电力科学研究院 | Parameter optimization-based method and system for predicting gas concentration in random forest oil |
| CN115049118A (en) * | 2022-06-02 | 2022-09-13 | 太原理工大学 | Method for realizing optimal capacity of natural gas production facility based on improved particle screening algorithm |
| CN115910193A (en) * | 2022-12-01 | 2023-04-04 | 华南农业大学 | Method for predicting synchronous saccharification and fermentation ethanol process of lignocellulose raw material by BP-MIV |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105894113A (en) * | 2016-03-31 | 2016-08-24 | 中国石油天然气股份有限公司规划总院 | Natural gas short-period demand prediction method |
| CN107092798A (en) * | 2017-04-28 | 2017-08-25 | 成都理工大学 | The method for estimating stability and device of predictive model of landslide |
| CN109191846A (en) * | 2018-10-12 | 2019-01-11 | 国网浙江省电力有限公司温州供电公司 | A kind of traffic trip method for predicting |
| CN109492825A (en) * | 2018-11-26 | 2019-03-19 | 中国水利水电科学研究院 | Medium-long Term Prediction method based on mutual information and the principal component analysis screening factor |
| CN111428926A (en) * | 2020-03-23 | 2020-07-17 | 国网江苏省电力有限公司镇江供电分公司 | Regional power load prediction method considering meteorological factors |
| CN112163745A (en) * | 2020-09-17 | 2021-01-01 | 武汉大学 | Regional security risk analysis method and system based on combined model |
-
2021
- 2021-06-11 CN CN202110656778.8A patent/CN113554213A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105894113A (en) * | 2016-03-31 | 2016-08-24 | 中国石油天然气股份有限公司规划总院 | Natural gas short-period demand prediction method |
| CN107092798A (en) * | 2017-04-28 | 2017-08-25 | 成都理工大学 | The method for estimating stability and device of predictive model of landslide |
| CN109191846A (en) * | 2018-10-12 | 2019-01-11 | 国网浙江省电力有限公司温州供电公司 | A kind of traffic trip method for predicting |
| CN109492825A (en) * | 2018-11-26 | 2019-03-19 | 中国水利水电科学研究院 | Medium-long Term Prediction method based on mutual information and the principal component analysis screening factor |
| CN111428926A (en) * | 2020-03-23 | 2020-07-17 | 国网江苏省电力有限公司镇江供电分公司 | Regional power load prediction method considering meteorological factors |
| CN112163745A (en) * | 2020-09-17 | 2021-01-01 | 武汉大学 | Regional security risk analysis method and system based on combined model |
Non-Patent Citations (3)
| Title |
|---|
| MATLAB中文论坛: "《MATLAB神经网络30个案例分析》", 30 April 2010, 北京航空航天大学出版社 * |
| 梁怡: "基于组合模型的西安市天然气需求预测研究", 《西安文理学院学报:自然科学版》 * |
| 袁颖 等: "岩质边坡稳定性预测模型的评价指标处理方法研究", 《自然灾害学报》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114184695A (en) * | 2021-11-09 | 2022-03-15 | 国网内蒙古东部电力有限公司电力科学研究院 | Parameter optimization-based method and system for predicting gas concentration in random forest oil |
| CN115049118A (en) * | 2022-06-02 | 2022-09-13 | 太原理工大学 | Method for realizing optimal capacity of natural gas production facility based on improved particle screening algorithm |
| CN115910193A (en) * | 2022-12-01 | 2023-04-04 | 华南农业大学 | Method for predicting synchronous saccharification and fermentation ethanol process of lignocellulose raw material by BP-MIV |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113554213A (en) | Natural gas demand prediction method, system, storage medium and equipment | |
| CN106485353B (en) | Air pollutant concentration forecasting procedure and system | |
| US10606862B2 (en) | Method and apparatus for data processing in data modeling | |
| CN103745273B (en) | Semiconductor fabrication process multi-performance prediction method | |
| CN107169628B (en) | Power distribution network reliability assessment method based on big data mutual information attribute reduction | |
| CN112990500B (en) | Transformer area line loss analysis method and system based on improved weighted gray correlation analysis | |
| CN111639783A (en) | Line loss prediction method and system based on LSTM neural network | |
| CN104574209A (en) | Modeling method of urban electrical network distribution transform weight overload mid-term forewarning model | |
| WO2017071369A1 (en) | Method and device for predicting user unsubscription | |
| CN109981749A (en) | A kind of cloud workflow task running time prediction method promoted based on limit gradient | |
| CN115470862A (en) | Dynamic self-adaptive load prediction model combination method | |
| CN105976025A (en) | BP neural network gas prediction method based on genetic algorithm optimization | |
| CN115860214A (en) | Early warning method and device for PM2.5 emission concentration | |
| Zhijun | RBF neural networks optimization algorithm and application on tax forecasting | |
| CN116187835A (en) | Data-driven-based method and system for estimating theoretical line loss interval of transformer area | |
| CN111178605A (en) | Distribution network engineering project construction period prediction method based on feature selection | |
| CN111680452A (en) | Power grid engineering accurate investment decision simulation method based on full-factor data mining | |
| CN110837952A (en) | Game theory-based power grid new technology equipment selection method and system | |
| CN104598998A (en) | Energy demand forecasting method based on economic growth indicators | |
| CN113793057A (en) | Building bidding and tendering data generation method based on regression analysis model | |
| CN111126827A (en) | Input-output accounting model construction method based on BP artificial neural network | |
| CN109840536A (en) | A kind of power grid power supply reliability horizontal clustering method and system | |
| CN115660296B (en) | Automatic issuing method of non-compliance project proposal scheme based on machine learning | |
| CN110866694A (en) | Power grid construction project financial evaluation system and method | |
| WO2022222230A1 (en) | Indicator prediction method and apparatus based on machine learning, and device and storage medium |
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 |