CN111047109A - Quantitative prediction method for regional air temperature change - Google Patents
Quantitative prediction method for regional air temperature change Download PDFInfo
- Publication number
- CN111047109A CN111047109A CN201911347400.9A CN201911347400A CN111047109A CN 111047109 A CN111047109 A CN 111047109A CN 201911347400 A CN201911347400 A CN 201911347400A CN 111047109 A CN111047109 A CN 111047109A
- Authority
- CN
- China
- Prior art keywords
- component
- data
- time
- air temperature
- model
- 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
- 238000000034 method Methods 0.000 title claims abstract description 51
- 230000008859 change Effects 0.000 title claims abstract description 24
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 56
- 238000004458 analytical method Methods 0.000 claims abstract description 14
- 238000005065 mining Methods 0.000 claims abstract description 7
- 230000009467 reduction Effects 0.000 claims abstract description 6
- 238000005311 autocorrelation function Methods 0.000 claims description 12
- 238000001228 spectrum Methods 0.000 claims description 10
- 238000007689 inspection Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 abstract description 9
- 230000000694 effects Effects 0.000 description 4
- 238000007619 statistical method Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000013277 forecasting method Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000010792 warming Methods 0.000 description 2
- 241001584785 Anavitrinella pampinaria Species 0.000 description 1
- 238000012271 agricultural production Methods 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000003190 augmentative effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000008713 feedback mechanism Effects 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- YHXISWVBGDMDLQ-UHFFFAOYSA-N moclobemide Chemical compound C1=CC(Cl)=CC=C1C(=O)NCCN1CCOCC1 YHXISWVBGDMDLQ-UHFFFAOYSA-N 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000001556 precipitation Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000001932 seasonal effect Effects 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 238000001744 unit root test Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
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
- 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
- G06Q10/067—Enterprise or organisation modelling
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Educational Administration (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the field of meteorology, in particular to a quantitative prediction method for regional air temperature change, which comprises the following steps: acquiring historical data of temperature in a preset time period aiming at an area to be predicted; performing data noise reduction and feature mining on the historical data by adopting variational modal decomposition to obtain components of the historical data; aiming at each component, establishing a differential autoregressive moving average model for prediction to obtain a prediction result of each component; and carrying out variation modal decomposition reconstruction on the prediction result of each component to obtain the prediction result of the region to be predicted. The invention can reduce the data requirement on the basis of simplifying the analysis process and can meet the prediction precision.
Description
Technical Field
The invention relates to the field of meteorology, in particular to a quantitative prediction method for regional air temperature change.
Background
Climate warming is a global problem facing the 21 st century. The fifth IPCC evaluation report indicated that the global earth surface mean temperature rose by 0.85 ℃ from 1880 to 2012. Global warming has significant impact on water supply, species distribution, glacier and marine ecosystems, as well as agricultural production, human health, and the like. The climate change is actually a process of multi-factor combined action, multi-scale overlapping, human activities and natural factors mixing, besides the direct action of the natural factors, the human activities indirectly change the climate environment by influencing the factors, and meanwhile, the climate change influences all aspects of the human society, and a feedback mechanism between the two is very complex. The study of the temperature change is an important component in the study of the climate change, and the regional temperature change is influenced by the global climate change and has the characteristics and regularity of the regional temperature change. The historical change rule of the regional temperature is analyzed, the future temperature change trend is reasonably and quantitatively predicted, and the method has important practical significance for reflecting the macroscopic evolution process of the regional temperature and serving social production and life.
The theoretical method related to temperature prediction can be mainly divided into a numerical forecasting method and a statistical method. The numerical forecasting is the most used method of official institutions in various countries at present, and utilizes a numerical calculation method to solve a partial differential equation set of atmospheric motion, substitutes a large amount of meteorological observation data into the equation, and solves an approximate value of atmospheric state (air temperature, pressure intensity, wind power and the like) at a preset moment by integrating the initial value of a weather element forwards. There are many mature numerical Prediction modes, such as european global spectrum mode TL511L60, japanese global spectrum mode and far east region spectrum mode, american NCEP mode (National Center for Environmental Prediction mode), cesm (community earth system mode), chinese National weather bureau T213 and T639 mode, chinese academy atmospheric circulation mode, marine coupling climate mode, and so on. The numerical forecasting method can accurately simulate the atmospheric environment of each period, but a large amount of meteorological observation data and a professional meteorological analysis model are needed for establishing the corresponding mode.
The statistical method is based on the principle that a large amount of on-site observation data are mined by a mathematical statistical method to perform fitting and prediction, and the method is simple, needs less data, and is not high in prediction accuracy. With the improvement of the requirement on the prediction accuracy, the traditional statistical analysis method cannot meet the requirements on fitting of a climate system and prediction of meteorological factors, and researchers begin to introduce new theories and models. Paulo AA and the like (2005) combine a heterogeneous Markov chain and a logarithmic linear model, drought monitoring and early warning are achieved according to precipitation and air temperature data, the long-term prediction effect in the model is good, but the Markov transition probability matrix has a strong diagonal trend, short-term prediction is often repeated, and the effect is poor. The Ortiz-Garcia EG et al (2012) used a support vector machine regression algorithm in combination with the Hess-Brezowsky classification to perform local very short term (6 hour) air temperature prediction, verifying the feasibility of the method. Leaf toming et al (2013) propose a temperature prediction model based on deterministic and random time sequences, a polynomial function and a Fourier method are combined with a seasonal autoregressive moving average modeling method to predict global absolute temperature per month, and the model is complex in calculation and suitable for short-term prediction.
Because the factors influencing the temperature change of the area are numerous and the interaction relation is complex, the system correlation is difficult to establish. However, if the historical observation data is directly used, large fluctuation exists, and data noise weakens information reflected in the temperature sequence, so that the prediction accuracy is low. Therefore, a new model and a new method need to be introduced, which can reduce the data requirement, simplify the system analysis process, and meet better prediction accuracy.
Disclosure of Invention
The quantitative prediction method for regional air temperature change provided by the invention can reduce the data requirement on the basis of simplifying the analysis process and can meet the prediction precision.
The invention provides a quantitative prediction method of regional air temperature change, which comprises the following steps:
acquiring historical data of temperature in a preset time period aiming at an area to be predicted;
performing data noise reduction and feature mining on the historical data by adopting variational modal decomposition to obtain components of the historical data;
aiming at each component, establishing a differential autoregressive moving average model for prediction to obtain a prediction result of each component;
and carrying out variation modal decomposition reconstruction on the prediction result of each component to obtain the prediction result of the region to be predicted.
Further, the performing data noise reduction and feature mining on the historical data by using variational modal decomposition to obtain components of the historical data specifically includes:
converting the time-frequency signal corresponding to the historical data into an analysis signal, calculating the instantaneous frequency mean value of the analysis signal, and determining the optimal decomposition layer number;
and performing variable mode decomposition on the time-frequency signals according to the optimal decomposition layer number to obtain a trend component and a detail component of the historical data.
Further, the establishing a differential autoregressive moving average model for each component for fitting and predicting further comprises:
performing unit root inspection on time sequence data corresponding to the components of the historical data, and judging the stationarity of each component;
if the time-series data of a certain component is determined to be non-stationary data, the difference is smoothed for the component.
Further, the establishing a differential autoregressive moving average model for each component to predict specifically includes:
calculating the autocorrelation function and the partial autocorrelation function of each component;
determining an autoregressive hysteresis order and a moving average hysteresis order of each component according to the trailing and truncation conditions of the autocorrelation function and the partial autocorrelation function of each component;
selecting a preliminary model of each component according to the autoregressive hysteresis order and the moving average hysteresis order of each component;
determining the optimal model structure of each component from the preliminary model of each component according to the Chichi information criterion;
determining parameters of an optimal model of each component according to a least square method;
and establishing a differential autoregressive moving average model of each component according to the optimal model structure and the corresponding parameters of each component.
Still further, the performing variation modal decomposition reconstruction on the prediction results of the components specifically includes:
taking the prediction result of each component as new component data;
and carrying out variation modal decomposition reconstruction on the new component data.
Still further, the performing a variational modal decomposition on the time-frequency signal according to the optimal number of decomposition layers specifically includes:
taking the modal function of each time-frequency signal as a finite bandwidth around a center frequency;
decomposing the mode function of each time-frequency signal into a component set with specific bandwidth sparsity;
and determining the center frequency and the broadband of each component set by iteratively searching the optimal solution of the variation model.
In the above technical solution, the determining the center frequency and the bandwidth of each component set by iteratively searching for the optimal solution of the variational model specifically includes:
solving the marginal frequency spectrum of the analytic function corresponding to each time-frequency signal through Hilbert transform;
estimating the center frequency of each time-frequency signal through index mixing modulation, and transferring the corresponding marginal spectrum to a baseband;
estimating the signal bandwidth of each time-frequency signal according to the Gaussian smoothness and the gradient square to obtain the constraint variation problem of each time-frequency signal;
introducing a Lagrange multiplier and a secondary penalty term, and converting the constraint variation problem of each time-frequency signal into a corresponding non-constraint variation problem;
and (4) carrying out iterative solution on each time-frequency signal on the basis of the optimal decomposition layer number until the precision meets the preset requirement.
In the above technical solution, the preliminary model is represented as:
in the formula: epsilontIn order to smooth the white noise,is an autoregressive coefficient, θjP is the autoregressive lag order, q is the moving average lag order, and μ is the mean of the time series.
In the above technical solution, each component is represented as:
in the formula: lambda [ alpha ]n(t) is a Lagrangian multiplier,is composed ofThe fourier transform of (a) the signal,for the component sequence obtained for the m-th iteration,is λn(t) Fourier transform, α is a quadratic penalty term, m represents the number of iterations, ω is the frequency, f is the original data, and k is the number of best decomposition levels.
In the above technical solution, the precision is expressed as:
in the formula:is composed ofThe fourier transform of (a) the signal,is the m-th iterationAnd (4) generating a component sequence, wherein m represents the iteration number, and k is the optimal decomposition layer number.
In the invention, each component of a region to be measured is obtained by utilizing Variational Modal Decomposition (VMD), and then a differential autoregressive moving average model of each component is established for prediction to obtain a prediction result of each component; and then carrying out variation modal decomposition reconstruction on the component prediction result to finally obtain the prediction result of the region to be predicted. The VMD can perform multi-scale decomposition on the non-linear, non-stationary data signal to weaken its volatility while extracting trend features and detail features in the data. For time series modeling, VMD decomposition can provide more data information, and effectively solves the problems of data dependence, complex mechanism and the like. Therefore, the method and the device can reduce the data requirement on the basis of simplifying the analysis process and can meet the prediction precision.
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, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart of a method of an embodiment of the present invention;
FIG. 2 is a data view of historical data in an embodiment of the present invention;
FIG. 3 is a data view of a trend component RES in an embodiment of the present invention;
FIG. 4 is a data perspective diagram of detail component IMF1 in an embodiment of the present invention;
FIG. 5 is a data perspective diagram of detail component IMF2 in an embodiment of the present invention;
FIG. 6 is a comparison graph of the predicted value and the actual value of the trend component RES in the embodiment of the present invention;
FIG. 7 is a comparison graph of predicted values and actual values of detail component IMF1 according to an embodiment of the present invention;
FIG. 8 is a comparison graph of predicted values and actual values of detail component IMF2 according to an embodiment of the present invention;
FIG. 9 is a comparison graph of the predicted result and the actual value of the region to be measured according to the embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, 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 invention.
As shown in fig. 1, the method for quantitatively predicting a change in regional air temperature according to the present invention includes:
101. acquiring historical data of temperature in a preset time period aiming at an area to be predicted; specifically, the method comprises the following steps:
assuming that the area to be tested is wuhan city, the historical data is 1956-.
102. Performing data noise reduction and feature mining on the historical data by adopting variational modal decomposition to obtain components of the historical data; the tool comprises:
1021. converting the time-frequency signal corresponding to the historical data into an analysis signal, calculating the instantaneous frequency mean value of the analysis signal, and determining the optimal decomposition layer number;
in contrast to Empirical Mode Decomposition (EMD), VMD models need to set the number of decomposition levels during decomposition, i.e., there is an optimal decomposition scale. Most of the current methods use the VMD model, and the method for quantitatively determining the decomposition layer number is lacked, wherein the number of the layers is usually obtained by referring to the EMD model for self-adaptation or by repeatedly testing to obtain the VMD decomposition layer number.
In this embodiment, the mean curve of the instantaneous frequency is used as a standard for determining the optimal decomposition level, and the historical data is used as a series of time-frequency signals. Firstly, Hilbert function is used for Hilbert transformation of historical data; changing a real signal into an analytic signal, then calculating instantaneous frequency and the mean value of the analytic signal by using an instfreq function and a mean function, calculating the mean value of the instantaneous frequency at different decomposition layers through a for loop, and drawing a line graph.
And determining the optimal decomposition layer number according to the trend of the line graph. If the number of decomposition layers is lower, the information contained in the original data in different scales cannot be well reflected; if the number of decomposition layers is too large, the components are too discrete, and are particularly noticeable in a high frequency region. This results in a lower average instantaneous frequency even at high frequencies, with a significant drop in the instantaneous frequency profile.
The optimal number of decomposition layers in this example was determined by calculation to be 3 layers.
1022. Performing variable mode decomposition on the time-frequency signal according to the optimal decomposition layer number to obtain a trend component and a detail component of historical data; specifically, the method comprises the following steps:
10221. taking the modal function of each time-frequency signal as a finite bandwidth around a center frequency;
10222. decomposing the mode function of each time-frequency signal into a component set with specific bandwidth sparsity;
10223. determining the center frequency and the broadband of each component set by iteratively searching the optimal solution of the variation model; specifically, the method comprises the following steps:
102231, solving the marginal frequency spectrum of the analytic function corresponding to each time-frequency signal through Hilbert transform; the marginal spectrum is:
in the formula: δ (t) represents a pulse function,representing the phase, xnk(t) is a component of the history data.
102232, estimating the center frequency of each time-frequency signal through index mixing modulation, and transferring the corresponding marginal spectrum to a baseband;
102233, estimating the signal bandwidth of each time-frequency signal according to Gaussian smoothness and gradient square to obtain the constraint variation problem of each time-frequency signal;
in the formula: x is the number ofnkIs a sequence of components, wnkIn order to realize the purpose,is the sum of all components, f is the original data,which means partial derivative of t, which means the number of analysis signals,an exponential harmonic term representing the analytical signal resulting from the hilbert transform.
102234, introducing Lagrange multipliers and secondary punishment items, and converting the constraint variation problem of each time-frequency signal into a corresponding non-constraint variation problem;
the saddle point of the augmented Lagrange formula is represented by formula (3) as the solution of the constraint variation problem:
in the formula: f (t) is the raw data, λn(t) is a Lagrange multiplier, and k is the optimal number of decomposition layers.
102235, iteratively solving each time-frequency signal on the basis of the optimal decomposition layer number until the precision meets the preset requirement;
the components are represented as:
in the formula:is composed ofThe fourier transform of (a) the signal,for the component sequence obtained for the m-th iteration,is λn(t) fourier transform, α being a secondary penalty term, m representing the number of iterations, ω being the frequency;
the precision is expressed as:
10224. calculating to obtain a trend component and a detail component;
as shown in fig. 2 to 5, the present embodiment obtains one trend component RES, two detail components IMF1 and IMF 2.
103. Performing stationarity verification on each component, specifically:
1031. performing unit root inspection on time sequence data corresponding to the components of the historical data, and judging the stationarity of each component;
1032. if the time-series data of a certain component is determined to be non-stationary data, the difference is smoothed for the component.
In the embodiment, in actual operation, the components obtained in 102 are regarded as a set of time series data, and the stationarity of the time series data is judged by a unit root test:
if the calculated ADF (amplified dictionary-Fuller test) value is smaller than the t statistic under the confidence level, the sequence is a stable sequence;
conversely, the sequence is not smooth, and needs to be d-order differentiated (d is 1,2, …) until smooth.
The historical data in the embodiment is checked, and the result shows that the detail components (IMF1 and IMF2) are smooth sequences, the trend component (RES) is a non-smooth sequence, and the sequence becomes smooth after 1-order difference is carried out on the sequence.
104. Aiming at each component, establishing a differential autoregressive moving average model for prediction to obtain a prediction result of each component;
1041. calculating the autocorrelation function and the partial autocorrelation function of each component;
1042. determining an autoregressive hysteresis order and a moving average hysteresis order of each component according to the trailing and truncation conditions of the autocorrelation function and the partial autocorrelation function of each component;
1043. selecting a preliminary model of each component according to the autoregressive lag order p and the moving average lag order q of each component;
and calculating to obtain a fitting result of the preliminary model and relevant statistics, wherein the relevant statistics comprise goodness of fit, adjusted goodness of fit, AIC (Chi information criterion) values, variances and the like.
1044. Determining the optimal model structure of each component from the preliminary model of each component according to the Chichi information criterion;
1045. determining parameters of an optimal model of each component according to a least square method;
1046. establishing a differential autoregressive moving average model of each component according to the optimal model structure and the corresponding parameters of each component;
determining the optimal model structure from the AIC values, and then estimating the model parameters using the least squares methodAnd thetajI.e. to express the certainty and randomness of the historical data as a linear model consisting of an autoregressive part and a moving average part.
Although the time series set is influenced by various factors in the objective world and has certain randomness, when the same sequence changes along with time, a corresponding dependency relationship or a related relationship exists, so that the time series can be analyzed and predicted according to the correlation. The ARIMA (p, d, q) model is to perform d-order differential smoothing on an original data sequence (usually a non-stationary sequence), and then an Autoregressive (AR) model is used for fitting a deterministic trend, namely a multivariate linear correlation relationship between a sequence value at the time t and p time sequences lagging behind the sequence value; a Moving Average (MA) model fits the random residual, i.e., the correlation of the sequence value at time t with q white noise factors behind it.
The preliminary model is represented as:
in the formula: epsilontIn order to smooth the white noise,is an autoregressive coefficient, θjP is the autoregressive lag order, q is the moving average lag order, and μ is the mean of the time series.
1047. Predicting by using the difference autoregressive moving average model of each component to obtain the prediction result of each component;
as shown in fig. 6 to 8, on the basis of this, static prediction is performed on the training data (1956-. Table 1 shows the component model parameters and partial statistics.
TABLE 1 model parameters and statistics for each component
105. And carrying out variation modal decomposition reconstruction on the prediction result of each component to obtain the prediction result of the region to be predicted.
1051. Taking the prediction result of each component as new component data;
1052. and carrying out variation modal decomposition reconstruction on the new component data.
And regarding each prediction sequence obtained in the step 104 as a new trend component and a new detail component, and reconstructing through a VMD to obtain a prediction result of the original air temperature sequence.
Table 2 shows the comparison between the predicted result of the integrated Model proposed by the present invention and the relative error (MRE), absolute error (MAE), Root Mean Square Error (RMSE) of the common gray Model (Grey Model) and time series Model, and fig. 9 is the comparison between the predicted result and the actual data of the weather station.
TABLE 2 analysis of temperature prediction error for empirical data of different methods
In the invention, each component of a region to be measured is obtained by utilizing Variational Modal Decomposition (VMD), and then a differential autoregressive moving average model of each component is established for prediction to obtain a prediction result of each component; and then carrying out variation modal decomposition reconstruction on the component prediction result to finally obtain the prediction result of the region to be predicted. The VMD can perform multi-scale decomposition on the non-linear, non-stationary data signal to weaken its volatility while extracting trend features and detail features in the data. For time series modeling, VMD decomposition can provide more data information, and effectively solves the problems of data dependence, complex mechanism and the like. Therefore, the method and the device can reduce the data requirement on the basis of simplifying the analysis process and can meet the prediction precision.
In addition, the invention provides an integrated modeling idea, namely, noise reduction and multi-scale information mining are carried out on time series data, components reflecting different characteristics are respectively predicted, and reference value is provided for similar researches in the aspects of weather, hydrology, social economy and the like.
It should be understood that the specific order or hierarchy of steps in the processes disclosed is an example of exemplary approaches. Based upon design preferences, it is understood that the specific order or hierarchy of steps in the processes may be rearranged without departing from the scope of the present disclosure. The accompanying method claims present elements of the various steps in a sample order, and are not intended to be limited to the specific order or hierarchy presented.
In the foregoing detailed description, various features are grouped together in a single embodiment for the purpose of streamlining the disclosure. This method of disclosure is not to be interpreted as reflecting an intention that the claimed embodiments of the subject matter require more features than are expressly recited in each claim. Rather, as the following claims reflect, invention lies in less than all features of a single disclosed embodiment. Thus, the following claims are hereby expressly incorporated into the detailed description, with each claim standing on its own as a separate preferred embodiment of the invention.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. To those skilled in the art; various modifications to these embodiments will be readily apparent, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the disclosure. Thus, the present disclosure is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
What has been described above includes examples of one or more embodiments. It is, of course, not possible to describe every conceivable combination of components or methodologies for purposes of describing the aforementioned embodiments, but one of ordinary skill in the art may recognize that many further combinations and permutations of various embodiments are possible. Accordingly, the embodiments described herein are intended to embrace all such alterations, modifications and variations that fall within the scope of the appended claims. Furthermore, to the extent that the term "includes" is used in either the detailed description or the claims, such term is intended to be inclusive in a manner similar to the term "comprising" as "comprising" is interpreted when employed as a transitional word in a claim. Furthermore, any use of the term "or" in the specification of the claims is intended to mean a "non-exclusive or".
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (10)
1. A method for quantitatively predicting a change in a zone air temperature, comprising:
acquiring historical data of temperature in a preset time period aiming at an area to be predicted;
performing data noise reduction and feature mining on the historical data by adopting variational modal decomposition to obtain components of the historical data;
aiming at each component, establishing a differential autoregressive moving average model for prediction to obtain a prediction result of each component;
and carrying out variation modal decomposition reconstruction on the prediction result of each component to obtain the prediction result of the region to be predicted.
2. The method for quantitatively predicting the change of the regional air temperature according to claim 1, wherein the data denoising and feature mining are performed on the historical data by using the variational modal decomposition to obtain the components of the historical data, and specifically comprises:
converting the time-frequency signal corresponding to the historical data into an analysis signal, calculating the instantaneous frequency mean value of the analysis signal, and determining the optimal decomposition layer number;
and performing variable mode decomposition on the time-frequency signals according to the optimal decomposition layer number to obtain a trend component and a detail component of the historical data.
3. The method of claim 1, wherein the fitting and predicting by building a differential auto-regressive moving average model for each component further comprises:
performing unit root inspection on time sequence data corresponding to the components of the historical data, and judging the stationarity of each component;
if the time-series data of a certain component is determined to be non-stationary data, the difference is smoothed for the component.
4. The method for quantitatively predicting regional air temperature change according to claim 3, wherein the establishing a differential auto-regressive moving average model for each component to predict the regional air temperature change specifically comprises:
calculating the autocorrelation function and the partial autocorrelation function of each component;
determining an autoregressive hysteresis order and a moving average hysteresis order of each component according to the trailing and truncation conditions of the autocorrelation function and the partial autocorrelation function of each component;
selecting a preliminary model of each component according to the autoregressive hysteresis order and the moving average hysteresis order of each component;
determining the optimal model structure of each component from the preliminary model of each component according to the Chichi information criterion;
determining parameters of an optimal model of each component according to a least square method;
and establishing a differential autoregressive moving average model of each component according to the optimal model structure and the corresponding parameters of each component.
5. The method for quantitatively predicting a change in regional air temperature according to claim 1, wherein the performing a metamorphic modal decomposition reconstruction of the prediction results of the components specifically includes:
taking the prediction result of each component as new component data;
and carrying out variation modal decomposition reconstruction on the new component data.
6. The method for quantitatively predicting the change of the regional air temperature according to claim 2, wherein the performing the variational modal decomposition on the time-frequency signal according to the optimal decomposition layer number specifically comprises:
taking the modal function of each time-frequency signal as a finite bandwidth around a center frequency;
decomposing the mode function of each time-frequency signal into a component set with specific bandwidth sparsity;
and determining the center frequency and the broadband of each component set by iteratively searching the optimal solution of the variation model.
7. The method for quantitatively predicting regional air temperature variation according to claim 6, wherein the determining the center frequency and the wide band of each component set by iteratively searching for the optimal solution of the variation model comprises:
solving the marginal frequency spectrum of the analytic function corresponding to each time-frequency signal through Hilbert transform;
estimating the center frequency of each time-frequency signal through index mixing modulation, and transferring the corresponding marginal spectrum to a baseband;
estimating the signal bandwidth of each time-frequency signal according to the Gaussian smoothness and the gradient square to obtain the constraint variation problem of each time-frequency signal;
introducing a Lagrange multiplier and a secondary penalty term, and converting the constraint variation problem of each time-frequency signal into a corresponding non-constraint variation problem;
and (4) carrying out iterative solution on each time-frequency signal on the basis of the optimal decomposition layer number until the precision meets the preset requirement.
8. The method of quantitatively predicting regional air temperature changes according to claim 4, characterized in that the preliminary model is represented as:
9. The method of quantitatively predicting regional air temperature changes according to claim 7, characterized in that each component is represented by:
in the formula: lambda [ alpha ]n(t) is a Lagrangian multiplier,is composed ofThe fourier transform of (a) the signal,is λn(t) a Fourier transform of the (t),for the component sequence obtained by the mth iteration, α is a secondary penalty term, m represents the iteration number, ω is frequency, f is original data, and k is the optimal decomposition layer number.
10. The method of quantitatively predicting regional air temperature changes according to claim 7, wherein the accuracy is expressed as:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911347400.9A CN111047109A (en) | 2019-12-24 | 2019-12-24 | Quantitative prediction method for regional air temperature change |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911347400.9A CN111047109A (en) | 2019-12-24 | 2019-12-24 | Quantitative prediction method for regional air temperature change |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111047109A true CN111047109A (en) | 2020-04-21 |
Family
ID=70238847
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911347400.9A Pending CN111047109A (en) | 2019-12-24 | 2019-12-24 | Quantitative prediction method for regional air temperature change |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111047109A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111784050A (en) * | 2020-07-01 | 2020-10-16 | 青岛洪锦智慧能源技术有限公司 | Short-term outdoor temperature prediction method based on particle filter algorithm |
CN112991088A (en) * | 2021-02-04 | 2021-06-18 | 国网浙江省电力有限公司 | Monthly power demand prediction method based on VMD-ANFIS-ARIMA |
CN113267256A (en) * | 2021-04-14 | 2021-08-17 | 国网山东省电力公司济宁供电公司 | Distribution line contact temperature prediction system and method |
CN113984833A (en) * | 2021-10-29 | 2022-01-28 | 江苏徐工工程机械研究院有限公司 | Environmental temperature equivalent and accelerated test method |
CN114997055A (en) * | 2022-06-06 | 2022-09-02 | 安徽理工大学 | Sea level temperature time-frequency domain change characteristic analysis method |
CN116362428A (en) * | 2023-06-01 | 2023-06-30 | 国网天津市电力公司城南供电分公司 | Short-term load prediction method based on VMD-PCF-ARIMA |
CN117457096A (en) * | 2023-12-26 | 2024-01-26 | 山东省海洋资源与环境研究院(山东省海洋环境监测中心、山东省水产品质量检验中心) | Dynamic monitoring and adjusting system for simulating carbon dioxide dissolution in ocean acidification device |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107392364A (en) * | 2017-07-12 | 2017-11-24 | 河海大学 | The short-term load forecasting method of variation mode decomposition and depth belief network |
CN109118000A (en) * | 2018-08-07 | 2019-01-01 | 广东工业大学 | A kind of short-term wind speed forecasting method based on CEEMD-VMD-GA-ORELM model |
-
2019
- 2019-12-24 CN CN201911347400.9A patent/CN111047109A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107392364A (en) * | 2017-07-12 | 2017-11-24 | 河海大学 | The short-term load forecasting method of variation mode decomposition and depth belief network |
CN109118000A (en) * | 2018-08-07 | 2019-01-01 | 广东工业大学 | A kind of short-term wind speed forecasting method based on CEEMD-VMD-GA-ORELM model |
Non-Patent Citations (2)
Title |
---|
HUAN WANG ET AL.: "An Integrated Variational Mode Decomposition and ARIMA Model to Forecast Air Temperature", 《SUSTAINABILITY》 * |
KONSTANTIN DRAGOMIRETSKIY ET AL.: "Variational Mode Decomposition", 《IEEE TRANSACTIONS ON SIGNAL PROCESSING》 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111784050A (en) * | 2020-07-01 | 2020-10-16 | 青岛洪锦智慧能源技术有限公司 | Short-term outdoor temperature prediction method based on particle filter algorithm |
CN111784050B (en) * | 2020-07-01 | 2024-03-22 | 青岛洪锦智慧能源技术有限公司 | Short-term outdoor temperature prediction method based on particle filter algorithm |
CN112991088A (en) * | 2021-02-04 | 2021-06-18 | 国网浙江省电力有限公司 | Monthly power demand prediction method based on VMD-ANFIS-ARIMA |
CN113267256A (en) * | 2021-04-14 | 2021-08-17 | 国网山东省电力公司济宁供电公司 | Distribution line contact temperature prediction system and method |
CN113984833A (en) * | 2021-10-29 | 2022-01-28 | 江苏徐工工程机械研究院有限公司 | Environmental temperature equivalent and accelerated test method |
CN113984833B (en) * | 2021-10-29 | 2024-03-01 | 江苏徐工工程机械研究院有限公司 | Environment temperature equivalent and acceleration test method |
CN114997055A (en) * | 2022-06-06 | 2022-09-02 | 安徽理工大学 | Sea level temperature time-frequency domain change characteristic analysis method |
CN114997055B (en) * | 2022-06-06 | 2024-04-05 | 安徽理工大学 | Sea level temperature time-frequency domain change characteristic analysis method |
CN116362428A (en) * | 2023-06-01 | 2023-06-30 | 国网天津市电力公司城南供电分公司 | Short-term load prediction method based on VMD-PCF-ARIMA |
CN117457096A (en) * | 2023-12-26 | 2024-01-26 | 山东省海洋资源与环境研究院(山东省海洋环境监测中心、山东省水产品质量检验中心) | Dynamic monitoring and adjusting system for simulating carbon dioxide dissolution in ocean acidification device |
CN117457096B (en) * | 2023-12-26 | 2024-03-22 | 山东省海洋资源与环境研究院(山东省海洋环境监测中心、山东省水产品质量检验中心) | Dynamic monitoring and adjusting system for simulating carbon dioxide dissolution in ocean acidification device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111047109A (en) | Quantitative prediction method for regional air temperature change | |
Leutbecher et al. | Stochastic representations of model uncertainties at ECMWF: State of the art and future vision | |
Zheng et al. | Response of ENSO amplitude to global warming in CESM large ensemble: uncertainty due to internal variability | |
Shamshad et al. | First and second order Markov chain models for synthetic generation of wind speed time series | |
Wang et al. | Evaluating climate field reconstruction techniques using improved emulations of real-world conditions | |
CN111665575B (en) | Medium-and-long-term rainfall grading coupling forecasting method and system based on statistical power | |
Kober et al. | Blending a probabilistic nowcasting method with a high‐resolution numerical weather prediction ensemble for convective precipitation forecasts | |
Bothe et al. | Continental-scale temperature variability in PMIP3 simulations and PAGES 2k regional temperature reconstructions over the past millennium | |
CN116642560B (en) | Ultrasonic gas meter metering correction method and system based on intelligent gas Internet of things | |
CN102855634A (en) | Image detection method and image detection device | |
CN104764868A (en) | Soil organic carbon predication method based on geographically weighted regression | |
CN111639803A (en) | Prediction method applied to future vegetation index of area under climate change scene | |
Cornes et al. | Twentieth-century trends in the annual cycle of temperature across the Northern Hemisphere | |
CN115587666A (en) | Load prediction method and system based on seasonal trend decomposition and hybrid neural network | |
Keune et al. | Multivariate probabilistic analysis and predictability of medium-range ensemble weather forecasts | |
CN105303051A (en) | Air pollutant concentration prediction method | |
Koster et al. | Soil moisture initialization error and subgrid variability of precipitation in seasonal streamflow forecasting | |
CN116719984B (en) | Intelligent fuel gas data management method, internet of things system and storage medium | |
Frogner et al. | Ensemble prediction with different spatial resolutions for the 2014 Sochi Winter Olympic Games: The effects of calibration and multimodel approaches | |
CN116186493B (en) | Meteorological forecasting method, device, equipment and medium based on numerical mode correction | |
Chikhar et al. | Impact of lateral boundary conditions on regional analyses | |
Arnold | Stochastic parametrisation and model uncertainty | |
KR102500534B1 (en) | Recurrent neural network based water resource information generating device and method | |
Zhan et al. | Impulse Weibull distribution for daily precipitation and climate change in China during 1961–2011 | |
Benáček et al. | Satellite bias correction in the regional model ALADIN/CZ: Comparison of different varBC approaches |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200421 |
|
RJ01 | Rejection of invention patent application after publication |