CN110807475A - Flood classification, identification and forecast method based on certainty coefficient - Google Patents
Flood classification, identification and forecast method based on certainty coefficient Download PDFInfo
- Publication number
- CN110807475A CN110807475A CN201910982313.4A CN201910982313A CN110807475A CN 110807475 A CN110807475 A CN 110807475A CN 201910982313 A CN201910982313 A CN 201910982313A CN 110807475 A CN110807475 A CN 110807475A
- Authority
- CN
- China
- Prior art keywords
- flood
- model
- forecast
- event
- classification
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 55
- 230000008569 process Effects 0.000 claims abstract description 25
- 238000011161 development Methods 0.000 claims abstract description 9
- 230000007246 mechanism Effects 0.000 claims abstract description 7
- 238000013145 classification model Methods 0.000 claims description 16
- 238000004088 simulation Methods 0.000 claims description 16
- 238000012549 training Methods 0.000 claims description 16
- 239000011159 matrix material Substances 0.000 claims description 13
- 230000000694 effects Effects 0.000 claims description 7
- 238000001914 filtration Methods 0.000 claims description 7
- 238000006424 Flood reaction Methods 0.000 claims description 5
- 230000008878 coupling Effects 0.000 claims description 4
- 238000010168 coupling process Methods 0.000 claims description 4
- 238000005859 coupling reaction Methods 0.000 claims description 4
- 238000013277 forecasting method Methods 0.000 claims description 4
- 230000031877 prophase Effects 0.000 claims description 4
- 238000011156 evaluation Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 abstract description 3
- 230000007547 defect Effects 0.000 abstract description 2
- 235000008733 Citrus aurantifolia Nutrition 0.000 description 7
- 235000011941 Tilia x europaea Nutrition 0.000 description 7
- 239000004571 lime Substances 0.000 description 7
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 4
- 238000011160 research Methods 0.000 description 3
- 101000767534 Arabidopsis thaliana Chorismate mutase 2 Proteins 0.000 description 2
- 101000986989 Naja kaouthia Acidic phospholipase A2 CM-II Proteins 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000002950 deficient Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000007621 cluster analysis Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 150000002678 macrocyclic compounds Chemical class 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000002844 melting Methods 0.000 description 1
- 230000008018 melting Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A10/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE at coastal zones; at river basins
- Y02A10/40—Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Business, Economics & Management (AREA)
- General Engineering & Computer Science (AREA)
- Marketing (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- General Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Primary Health Care (AREA)
- Health & Medical Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A flood classification, identification and forecast method based on certainty coefficients belongs to the technical field of hydrological flood forecast. The method provides a framework with simpler classification and identification and less requirement on other data. The framework is based on the flood generation and development process, and combines the certainty coefficient to classify the flood and construct a corresponding flood forecast sub-model; in the real-time forecasting process, a forgetting mechanism is introduced to identify the sub-model from a certainty coefficient so as to realize forecasting. The method has the advantages that the classification principle is clear, the classified sub-models can reflect the nonlinearity of various types of flood, and the defect that a single non-classified model is biased to be optimal integrally in the parameter identification process is overcome; in the process of model identification, the influence of the information which is more and more distant on the current forecast result is considered to be smaller, and a forgetting mechanism is introduced to identify the forecast submodel in real time. The method has the advantages of clear principle, simplicity, less indexes, low data requirement, convenience in calculation and better popularization.
Description
Technical Field
The invention belongs to the technical field of hydrological flood forecasting, and relates to a flood classification identification forecasting method based on a certainty coefficient.
Background
The flood process is influenced by many factors such as natural geography, hydrology, weather and human activities, and has larger uncertainty, so that the difficulty of flood simulation forecast is large. How to improve the flood forecasting precision is always the most important problem in the hydrology field.
Classification is gradually introduced into flood forecasting as a way to make irregular events more orderly. At present, the idea of flood classification and identification has been widely applied and accepted (liu yubang and wainscot (2011). "rainstorm flood grading research based on weather cause and principal component projection analysis." water conservancy project (01):98-104 "). For example, Zhanping Nu et al (Zhanping Ping and Zhouhui Cheng (2012) 'flood classification research based on rainstorm flood similarity analysis.' hydroelectric energy science (09):50-54.) use the similarity coefficient of the rainstorm flood evolution process to classify the rainstorm flood, and provide a foundation for dynamic control of flood limiting water level of the reservoir; li Yue Jade and the like (Li Yue Jade and Zhou Jian Yi and the like 2016). "basin flood real-time classification correction based on K-means cluster analysis". "Chinese rural water conservancy and hydropower (12): 160-; the method comprises the following steps of (1) classifying royal macrocycles and the like (2017) 'classification application of entropy weight-based fuzzy clustering models in mountain snow melting flood runoff types.' China rural water conservancy and hydropower (05): 114-; xu wei et al (xu wei and Lianghua et al (2013) 'research on a real-time flood classification forecasting model based on second-order clustering and rough set' hydropower science forecasting (02):60-67.), and a flood classification forecasting model based on a second-order clustering method is established by considering a plurality of influence factors such as rainfall intensity, underlying surface characteristics, human activities and a weather system.
However, these studies are based on flood influence factors (climate factors, etc.) that lead to the occurrence and development of flood, and carry out flood classification and identification, ignoring the occurrence and development of flood itself. In addition, for the data-deficient drainage basin, the difficulty of extracting and identifying factors influencing flood variation factors is high, and the interaction between classification factors is likely to cause the difficulty of flood classification and identification, so a flood forecasting method which is simpler and has less requirements on other data is needed to be sought.
The invention provides a flood classification, identification and forecast method based on a certainty coefficient. The method is based on a flood forecasting framework coupling DBM and Kalman filtering, starts from a flood generation and development process, classifies flood by taking a certainty coefficient as a judgment index, and constructs a corresponding flood forecasting sub-model. In the real-time forecasting process, a forgetting mechanism is combined, and the model is identified in real time by the early forecasting certainty coefficient, so that forecasting is realized. And taking a lime kiln basin as an example, the method is applied to flood forecasting, and the reasonability of the method is verified.
Disclosure of Invention
At present, the method for identifying domestic flood mainly takes flood influence factors (climate factors and the like) which cause the flood generation and development process as starting points to carry out flood classification identification, and ignores the flood self generation and development process. In addition, for a data-deficient watershed, the difficulty in extracting and identifying factors influencing flood variation factors is high, and the flood classification identification is difficult due to interaction among classification factors. Aiming at the problems in the prior art, the invention provides a framework which is simpler in classification and identification and has less requirements on other data, and provides a flood classification, identification and forecast method based on a deterministic coefficient.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
the method comprises the following steps: establishing a flood forecasting model, and adopting a DBM flood forecasting model of coupling Kalman filtering;
step two: flood forecasting classification
Firstly, each flood event is substituted into a DBM flood forecasting model for training, flood model parameter identification is carried out on each flood event to obtain the optimal model parameter of each flood event, a model corresponding to each flood is established according to the optimal model parameter of each flood event, and L models are established on the assumption that L fields of training flood events exist. Secondly, forecasting the flood events of the other (L-1) fields by adopting a model established by any flood event, and if the certainty coefficients of forecasting values obtained by mutually forecasting the flood events of the two fields are both larger than a certain threshold value RRthrAnd if the flood is smaller than the preset threshold, the two floods are determined not to belong to the same type of flood. And finally, repeating the steps for L times until the flood events of the other fields are completely forecasted by the L-field model, and obtaining the final classification result of the flood events. The threshold value RRthrGenerally 0.7-1.0, as the case may be.
The flood event classification process comprises the following sub-steps:
(1) substituting any flood event in the L flood events into a DBM flood forecasting model of the coupled Kalman filtering to train to obtain model parameters of the flood of the field, and identifying the flood model parameters to obtain optimal model parameters;
(2) simulating the rest (L-1) flood events by adopting the optimal model parameters of any one flood event, and calculating the certainty coefficient of the forecast value under the maximum forecast period;
(3) repeating the steps (1) and (2) to obtain a deterministic coefficient matrix for simulating the flood of the rest (L-1) field by any model, wherein the deterministic coefficient matrix is as follows:
wherein, RRljA certainty factor representing a prediction value obtained by predicting the jth flood event by using a model established by the ith flood event; l is the total field number of flood events. (4) Determining a deterministic coefficient RRljWhether a certain threshold value RR is exceededthrIf the certainty factor exceeds a certain threshold value RRthrI.e. RRlj≥RRthrThen it is considered feasible for the ith model to forecast the jth flood event, making VVlj1. If RRlj<RRthrThen it is not feasible for the ith model to forecast the jth flood event, VVlj0. The matrix RR is transformed into matrix VV, as shown in (2).
(5) And classifying each flood event according to the matrix VV. When RR islj≥RRthrAnd RRjl≥RRthrWhen is VVlj·VVjlAnd (1), considering that the first flood event and the jth flood event can accurately forecast each other, and the occurrence and development processes of the first flood event and the jth flood event have greater similarity, and determining that the first flood event and the jth flood event belong to a class of flood events. For a plurality of flood events, whether the flood events (k (1), k (2), …, k (p)) belong to the same type of flood event is judged, and the final classification result can be judged according to the formula (3).
Wherein p represents the number of events, t1 represents the model established by the t1 th flood event, and t2 represents the t2 th flood event.
Step three: and on the basis of the flood classification obtained in the step two, respectively training each type of flood by adopting the DBM flood forecasting model established in the step one, identifying model parameters of various types of flood events, and further obtaining a model corresponding to each type of flood.
Step four:
(1) and D, forecasting the current flood event in real time by using the plurality of models obtained in the step three respectively to obtain an early-stage forecasting result, and calculating a certainty coefficient of each model.
(2) According to the certainty coefficient of each model, a forgetting mechanism is introduced into the evaluation of the early-stage forecast result, and the method specifically comprises the following steps:
first, the whole early stage is divided into a high-impact stage with relatively high impact on the current simulation result and a low-impact stage with relatively low impact on the current simulation result, as shown in fig. 3. Then, the two stages of certainty coefficients are calculated using the formulas (4) and (5), respectively. Then, according to the two-stage weight ω1And omega2And (5) evaluating the simulation effect of the submodel h by using the formula (6).
RCM-h＝ω1·Rlow+ω2·Rhigh(6)
Wherein, H is 1,2, …, H represents the number of classification groups, i.e. the number of submodels; rlowAnd RhighThe deterministic coefficients are a low impact stage and a high impact stage at the early stage; omega1And omega2Is a two-stage weight, satisfies omega1+ω21. s represents the number of high-influence prophase sequences; r represents the number of low-influence prophase sequences; qh(t | t-delta) represents the forecast flow value of the h-th sub-classification model with forecast period delta days at the (t-delta) th time, m3/s；
Represents the mean of all actual flow values of Q0(i) between times (i-s-1) and (i-s-r), m3S; q0(t) represents the corresponding actual flow rate value on day t, m3S; since the k and s values are integers, the invention accordingly takes the possible values of k and sAnd (5) screening k and s values with optimal simulation effect in a simulation recognition training model to serve as the final number of the early-stage influence stages.
(3) Selection of RCM-hAnd (H-1, 2, …, H) the maximum model is used as the model for forecasting the current time, and the forecast value of the model is used as the final forecast value. And (3) repeating the steps (1) and (2) along with the time, so that the flood forecast value can be obtained in real time.
Compared with the prior art, the invention has the following advantages and effects:
the invention divides the existing flood through the certainty coefficient, the divided sub-models can reflect the nonlinearity of each flood, and in the real-time training and forecasting, the forgetting mechanism is introduced to identify the sub-models through comparing the current flood early-stage actual value with the early-stage forecast value of the classification sub-models, thereby realizing the forecasting, increasing the result precision and overcoming the defect that the single non-classification model only deviates to the integral optimum in the parameter identification process. In addition, the method has the advantages of clear principle, simple method, less indexes, low data requirement, convenience in calculation and better popularization.
Drawings
Fig. 1 is a flow chart of a flood classification recognition forecasting method.
Fig. 2 is a schematic diagram of flood event classification.
Fig. 3 is a view of a sub-model identification window.
Fig. 4 is a process diagram of flood identification in 1970 of a lime kiln basin.
Detailed Description
The present invention is further illustrated by the following specific examples.
The invention discloses a flood classification, identification and forecast method based on certainty coefficients, which takes a lime kiln sub-basin at the upstream of a Nieji reservoir as an example to carry out specific implementation introduction and comprises the following steps:
the method comprises the following steps: establishing a flood forecasting model, wherein the DBM flood forecasting model of a coupling Kalman filtering method is adopted, as shown in an attached figure 1;
step two: firstly, each flood event is brought into a DBM flood forecast model for training, flood model parameter identification is respectively carried out on each flood event to obtain the optimal model parameter of each flood event, and a model corresponding to each flood is established according to the optimal model parameter of each flood event. Secondly, forecasting the rest flood events by adopting a model established by each flood event, if the certainty factors of forecasting values obtained by the mutual forecasting of the two flood events are both larger than a certain threshold value, the two flood events are determined to belong to the same type of flood, if the certainty factors are smaller than the certain threshold value, the two flood events are determined not to belong to the same type of flood, and the steps are repeated, so that the final classification result of the flood events can be obtained. Example detailed steps for classification were carried out as follows:
(1) the lime kiln basin has 19 flood events in total, 15 flood events are defined as training flood events, 4 flood events are simulated flood events, each flood event in the 15 training flood events is substituted into a DBM (database management system) and a flood forecasting model of a coupled Kalman filtering method for training, model parameters of the flood are obtained, and 15 groups of model parameters are calculated in total, namely 15 models.
(2) And (5) simulating the rest 14 flood events by using the model of the ith flood, and calculating a certainty coefficient with the forecast period of 2 days.
(3) Repeating (1) and (2) to obtain the deterministic coefficient matrix RR shown in formula (1). Here, when the certainty factor is set to exceed 0.85, namely RRthr0.85, it is considered feasible that the model generated by the ith event forecasts the jth event.
(4) From the matrix RR, a matrix VV is determined. Coefficient of certainty RRljGreater than or equal to 0.85, then VVljOtherwise, VVlj＝0。
(5) And classifying the flood events according to the matrixes RR and VV. When RR islj≥RRthrAnd RRjl≥RRthrWhen is VVlj·VVjlAnd (1), considering that the first flood event and the jth flood event can accurately forecast each other, and the occurrence and development processes of the first flood event and the jth flood event have greater similarity, and determining that the first flood event and the jth flood event belong to a class of flood events. For multiple events, if judging whether the events (k (1), k (2), …, k (h)) belong to the same flood event, the final classification result of the flood event can be obtained by judging according to the formula (3)As shown in table 1, 15-field floods can be classified into 4 types.
TABLE 1 Classification of sub-basin sub-models of lime kiln
Step three: and on the basis of the flood classification obtained in the step two, respectively training each type of flood by adopting the DBM flood forecasting model established in the step one, identifying model parameters of various types of flood events, and further obtaining models CM-1, CM-2, CM-3 and CM-4 corresponding to each type of flood.
Step four:
(1) and (4) forecasting the current flood event in real time by respectively using the 4 models obtained in the step three to obtain an early-stage forecasting result, and calculating a certainty coefficient of each model.
(2) And according to the certainty coefficient of each model, introducing a forgetting mechanism into the evaluation of the early-stage forecast result, and realizing real-time identification by using a formula (4-6). To further illustrate the identification process of the flood real-time identification model (formula (4-6)) from the example, the invention takes the flood event in 1970 as an example, and the flood forecast identification process is specifically described as shown in fig. 4. As can be seen from the figure, in the dashed box, when rainfall reaches the effective input, the process is certain, i.e. the nonlinear process is the same. The submodels are different from each other in the linear simulation process of the DBM, the result of the CM-4 is obviously more similar to the actual value, and other submodels (CM-1, CM-2, CM-3 and NCM) are obviously larger (wherein the NCM represents a model which is not classified), so that the flood forecasting is mainly carried out by relying on the CM-4 in the next flood forecasting, and the future forecasting precision is improved.
The method provided by the invention is adopted to obtain the forecasting process of the secondary flood simulation training under the 2-day forecasting period of the lime kiln basin, and as can be seen from the table 2, part of the simulation forecasting flow of the non-classification model has larger difference with the actual process, the result of the classification model training is closer to the actual process than the simulation forecasting result of the non-classification model, and the peak value and the peak present time are better than the result of the non-classification model. Meanwhile, as can be seen from the comparison table of the certainty coefficients of the prediction results of the classification model and the classification model in different prediction periods, the result of the classification model is greatly improved compared with that of the classification model, in the simulation year, the certainty coefficient mean values of the prediction results of the classification model in the prediction periods of 1 day and 2 days are respectively 0.93 and 0.87, and are respectively 0.03 and 0.05 higher than those of the classification models of 0.90 and 0.82. In the forecast year, the forecast result certainty coefficients of the classification model in the forecast period of 1 day and 2 days are 0.93 and 0.86, which are higher than 0.06 and 0.16 of the classification model in the forecast period of 0.87 and 0.70. When the forecast period is 2 days, the determinacy coefficient of 18 fields of flood is larger than 0.80, and is obviously improved compared with 14 fields of non-classification models. The above results verify the rationality and effectiveness of the present invention. In addition, the method has the advantages of clear principle, simple method, less indexes, low data requirement, convenience in calculation and strong popularization.
TABLE 2 comparison table of lime kiln flood prediction certainty coefficients under different forecast periods
The above-mentioned embodiments only express the embodiments of the present invention, but not should be understood as the limitation of the scope of the invention patent, it should be noted that, for those skilled in the art, many variations and modifications can be made without departing from the concept of the present invention, and these all fall into the protection scope of the present invention.
Claims (3)
1. A flood classification, identification and forecast method based on certainty coefficients is characterized by comprising the following steps:
the method comprises the following steps: establishing a flood forecasting model, and adopting a DBM flood forecasting model of coupling Kalman filtering;
step two: firstly, each flood event is substituted into a DBM flood forecast model for training, flood model parameter identification is carried out on each flood event to obtain the optimal model parameter of each flood event, a model corresponding to each flood is established according to the optimal model parameter of each flood event, and if L fields of training flood events exist, the model is establishedEstablishing L models; secondly, forecasting the flood events of the other (L-1) fields by adopting a model established by any flood event, and if the certainty coefficients of forecasting values obtained by mutually forecasting the flood events of the two fields are both larger than a certain threshold value RRthrDetermining that the two floods belong to the same type of flood, and if the two floods are smaller than the same type of flood, determining that the two floods do not belong to the same type of flood; finally, repeating the steps for L times until the flood events of the other fields are completely forecasted by the L field model, and obtaining the final classification result of the flood events;
step three: on the basis of the flood classification obtained in the step two, respectively training each type of flood by adopting the DBM flood forecasting model established in the step one, and identifying model parameters of various types of flood events to obtain a model corresponding to each type of flood;
step four:
(1) forecasting the current flood event in real time by using the plurality of models obtained in the step three respectively to obtain an early-stage forecasting result, and calculating a certainty coefficient of each model;
(2) according to the certainty coefficient of each model, a forgetting mechanism is introduced into the evaluation of the early-stage forecast result, and the method specifically comprises the following steps:
firstly, dividing the whole earlier stage into a high-influence stage with relatively high influence on the current simulation result and a low-influence stage with relatively low influence on the current simulation result; secondly, the deterministic coefficients of the two stages are calculated by respectively using formulas (4) and (5); then, according to the two-stage weight ω1And omega2Evaluating the simulation effect of the submodel h by using a formula (6);
RCM-h＝ω1·Rlow+ω2·Rhigh(6)
wherein H is 1,2, …, H, H represents the number of classification groups, i.e. subgroupsThe number of models; rlowAnd RhighThe deterministic coefficients are a low impact stage and a high impact stage at the early stage; omega1And omega2Is a two-stage weight, satisfies omega1+ω21 is ═ 1; s represents the number of high-influence prophase sequences; r represents the number of low-influence prophase sequences; qh(t | t-delta) represents the forecast flow value of the h-th sub-classification model with forecast period delta days at the (t-delta) th time, m3/s；
Represents the mean of all actual flow values of Q0(i) between times (i-s-1) and (i-s-r), m3S; q0(t) represents the corresponding actual flow rate value on day t, m3S; because the k and s values are integers, the possible k and s values are introduced into a simulation recognition training model according to the method, and the k and s values with the optimal simulation effect are screened out and used as the final number of the early-stage influence stages;
(3) selection of RCM-h(H ═ 1,2, …, H) the largest model is taken as the model forecasted at the current moment, and the forecast value of the model is taken as the final forecast value; and (3) repeating the steps (1) and (2) along with the time, so that the flood forecast value can be obtained in real time.
2. The flood classification, identification and forecast method based on certainty factor as claimed in claim 1, wherein the flood event classification process in the second step comprises the following sub-steps:
(1) substituting any flood event in the L flood events into a DBM flood forecasting model of the coupled Kalman filtering to train to obtain model parameters of the flood of the field, and identifying the flood model parameters to obtain optimal model parameters; (2) simulating the flood events of the rest (L-1) fields by adopting the optimal model parameters, and calculating a certainty coefficient of a forecast value under the maximum forecast period;
(3) repeating the steps (1) and (2) to obtain a deterministic coefficient matrix for simulating the flood of the rest (L-1) field by any model, wherein the deterministic coefficient matrix is as follows:
wherein, RRljA certainty factor representing a prediction value obtained by predicting the jth flood event by using a model established by the ith flood event; l is the total field times of flood events; (4) determining a deterministic coefficient RRljWhether a certain threshold value RR is exceededthrIf the certainty factor exceeds a certain threshold value RRthrI.e. RRlj≥RRthrThen it is considered feasible for the ith model to forecast the jth flood event, making VVlj1 is ═ 1; if RRlj<RRthrThen it is not feasible for the ith model to forecast the jth flood event, VVlj0; the matrix RR is transformed into a matrix VV as shown in (2);
(5) classifying each flood event according to the matrix VV; when RR islj≥RRthrAnd RRjl≥RRthrWhen is VVlj·VVjl1, considering that the first flood event and the jth flood event can accurately forecast each other, and the occurrence and development processes of the first flood event and the jth flood event have greater similarity, and then determining that the first flood event and the jth flood event belong to a class of flood events; judging whether the flood events (k (1), k (2), …, k (p)) belong to the same type of flood event or not for a plurality of flood events, and judging a final classification result according to the formula (3);
wherein p represents the number of events, t1 represents the model established by the t1 th flood event, and t2 represents the t2 th flood event.
3. The flood classification and forecasting method based on certainty coefficient as claimed in claim 1, wherein the threshold value RR isthrIs determined according to specific conditions and is 0.7-1.0.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910982313.4A CN110807475B (en) | 2019-10-16 | 2019-10-16 | Flood classification, identification and forecast method based on certainty coefficient |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910982313.4A CN110807475B (en) | 2019-10-16 | 2019-10-16 | Flood classification, identification and forecast method based on certainty coefficient |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110807475A true CN110807475A (en) | 2020-02-18 |
| CN110807475B CN110807475B (en) | 2022-11-18 |
Family
ID=69488719
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910982313.4A Active CN110807475B (en) | 2019-10-16 | 2019-10-16 | Flood classification, identification and forecast method based on certainty coefficient |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110807475B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112183607A (en) * | 2020-09-23 | 2021-01-05 | 浙江水利水电学院 | Southeast coastal region flood classification method based on fuzzy theory |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104090974A (en) * | 2014-07-18 | 2014-10-08 | 河海大学 | Dynamic data mining method and system of extension reservoir subsequent floods |
| US20170168195A1 (en) * | 2015-12-15 | 2017-06-15 | Wuhan University | Method for forecasting floods for multiple lead times |
| CN107992961A (en) * | 2017-11-21 | 2018-05-04 | 中国水利水电科学研究院 | A kind of adaptive basin Medium-and Long-Term Runoff Forecasting model framework method |
| CN109993372A (en) * | 2019-04-12 | 2019-07-09 | 淮河水利委员会水文局(信息中心) | One kind being based on the probabilistic flood probability forecasting procedure of multi-source |
-
2019
- 2019-10-16 CN CN201910982313.4A patent/CN110807475B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104090974A (en) * | 2014-07-18 | 2014-10-08 | 河海大学 | Dynamic data mining method and system of extension reservoir subsequent floods |
| US20170168195A1 (en) * | 2015-12-15 | 2017-06-15 | Wuhan University | Method for forecasting floods for multiple lead times |
| CN107992961A (en) * | 2017-11-21 | 2018-05-04 | 中国水利水电科学研究院 | A kind of adaptive basin Medium-and Long-Term Runoff Forecasting model framework method |
| CN109993372A (en) * | 2019-04-12 | 2019-07-09 | 淮河水利委员会水文局(信息中心) | One kind being based on the probabilistic flood probability forecasting procedure of multi-source |
Non-Patent Citations (1)
| Title |
|---|
| GUOZHEN WEI等: "Nierji reservoir flood forecasting based on a Data-Based Mechanistic", 《JOURNAL OF HYDROLOGY》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112183607A (en) * | 2020-09-23 | 2021-01-05 | 浙江水利水电学院 | Southeast coastal region flood classification method based on fuzzy theory |
| CN112183607B (en) * | 2020-09-23 | 2023-11-07 | 浙江水利水电学院 | Flood classification method for southeast coastal areas based on fuzzy theory |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110807475B (en) | 2022-11-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113379109B (en) | Runoff forecasting method based on prediction model self-adaption | |
| Liu et al. | Deriving optimal refill rules for multi-purpose reservoir operation | |
| CN106650767B (en) | Flood forecasting method based on cluster analysis and real-time correction | |
| Lin et al. | Storm surge return levels induced by mid-to-late-twenty-first-century extratropical cyclones in the Northeastern United States | |
| Yu et al. | Switching ARIMA model based forecasting for traffic flow | |
| CN104715292A (en) | City short-term water consumption prediction method based on least square support vector machine model | |
| CN104933483A (en) | Wind power forecasting method dividing based on weather process | |
| CN107886160B (en) | BP neural network interval water demand prediction method | |
| CN106055846A (en) | Method for identifying photovoltaic inverter model based on BSNN (B-Spline Neural Network)-ARX (Auto Regressive Exogenous system) | |
| CN113435630B (en) | Basin hydrological forecasting method and system with self-adaptive runoff yield mode | |
| CN105005825A (en) | Method and system for predicting photovoltaic power based on dynamic neural network | |
| CN113554466A (en) | Short-term power consumption prediction model construction method, prediction method and device | |
| CN107527117A (en) | Reservoir adaptability dispatching method based on D S evidence theories | |
| CN111079086A (en) | Multi-element joint distribution-based multiple risk assessment method for water resource system | |
| CN113807545A (en) | River and lake ecological flow forecasting and early warning method based on deep learning and physical model | |
| Shen et al. | Real-time correction of water stage forecast using combination of forecasted errors by time series models and Kalman filter method | |
| CN110807475B (en) | Flood classification, identification and forecast method based on certainty coefficient | |
| CN108694475B (en) | Short-time-scale photovoltaic cell power generation capacity prediction method based on hybrid model | |
| CN107330538A (en) | A kind of method of climate lower storage reservoir adaptability scheduling rule establishment | |
| Gastaldi et al. | Short-term forecasting of municipal load through a Kalman filtering based approach | |
| Mawada et al. | Climate change impacts on the extreme rainfall for selected sites in North Western England | |
| Gao et al. | The study of GRNN for wind speed forecasting based on Markov Chain | |
| KR101859109B1 (en) | Control methods of low impact development facilities using rainfall forecasting | |
| CN115330085A (en) | Wind speed prediction method based on deep neural network and without future information leakage | |
| CN115034159A (en) | Power prediction method, device, storage medium and system for offshore wind farm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |