CN114493023A - Diagnosis and interpolation method in abnormal water regime data based on RF-Adaboost model - Google Patents
Diagnosis and interpolation method in abnormal water regime data based on RF-Adaboost model Download PDFInfo
- Publication number
- CN114493023A CN114493023A CN202210116677.6A CN202210116677A CN114493023A CN 114493023 A CN114493023 A CN 114493023A CN 202210116677 A CN202210116677 A CN 202210116677A CN 114493023 A CN114493023 A CN 114493023A
- Authority
- CN
- China
- Prior art keywords
- data
- abnormal
- adaboost
- value
- water regime
- 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
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 61
- 230000002159 abnormal effect Effects 0.000 title claims abstract description 45
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000003745 diagnosis Methods 0.000 title claims abstract description 18
- 238000012544 monitoring process Methods 0.000 claims abstract description 22
- 238000001914 filtration Methods 0.000 claims abstract description 13
- 238000012549 training Methods 0.000 claims description 33
- 230000009471 action Effects 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000007637 random forest analysis Methods 0.000 abstract description 17
- 238000010586 diagram Methods 0.000 abstract description 8
- 238000012731 temporal analysis Methods 0.000 abstract description 2
- 238000000700 time series analysis Methods 0.000 abstract description 2
- 230000008859 change Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012546 transfer Methods 0.000 description 3
- 230000006978 adaptation Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000011160 research Methods 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
- 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
- G06F18/2148—Generating training patterns; Bootstrap methods, e.g. bagging or boosting characterised by the process organisation or structure, e.g. boosting cascade
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- 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
- G06Q10/20—Administration of product repair or maintenance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Primary Health Care (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- General Health & Medical Sciences (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a diagnosis and interpolation method in abnormal water regime data based on an RF-Adaboost model, which relates to the technical field of water regime monitoring and comprises the following steps: acquiring real-time water regime data, performing time series analysis on the water regime data by adopting median filtering, drawing a 3-Sigma diagram, and performing real-time identification and diagnosis on abnormal data based on the 3-Sigma diagram; the method has the advantages that an Adaboost model improved based on Random Forest (RF) is constructed and trained, the RF-Adaboost model is applied to predict the water regime monitoring data in real time, and the abnormal data are interpolated.
Description
Technical Field
The invention relates to the technical field of regimen monitoring, in particular to a real-time abnormity diagnosis and interpolation method of flow monitoring data.
Background
With the development of intelligent water diversion, a plurality of automatic water condition monitoring devices are arranged along the established open channel water diversion project and are used for monitoring water condition information such as water level, flow and gate opening, water condition data are abnormal due to various device random faults, man-made regular maintenance and other interference factors, when the water condition data are abnormal, a large amount of information is lost due to direct deletion, the integrity, the continuity and the consistency of the water condition data are damaged, the reliability and the precision of a data set are reduced seriously due to simple modification, and great influence on gate hydraulic calculation and dispatching control is possibly generated in serious cases. Therefore, on the premise of ensuring the integrity, continuity and consistency of the regimen data, missing and abnormal data are identified in time and are predicted and interpolated, and the method has important application value and scientific significance for improving the reliability of the data, objectively reflecting the regimen change and effectively guiding engineering scheduling.
The interpolation prediction of abnormal data by using a single certain machine learning method such as a three-Spline (Spline) interpolation method, a Random Forest (RF) interpolation method and the like often has large subjective one-sidedness, the interpolation prediction effect is greatly different from an actually measured value, the real water situation cannot be reflected, and the application to the real-time interpolation of monitoring data is greatly limited. Therefore, the method effectively identifies abnormal data and utilizes reasonable data to interpolate and predict in real time, and is a key problem to be solved for monitoring the water regime of the open channel water transfer engineering.
Therefore, it is desirable to find a method for predicting and diagnosing flow rate to solve the above technical problems.
Disclosure of Invention
The invention aims to provide a diagnosis and interpolation method in abnormal water regime data based on an RF-Adaboost model, so as to solve the problems in the prior art.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
a diagnosis and interpolation method in abnormal water situation data based on an RF-Adaboost model comprises the following steps:
s1, carrying out median filtering analysis on the acquired regimen monitoring data to obtain a trend term and a residual term of a time sequence;
s2, carrying out real-time diagnosis on the abnormal data by adopting a 3-Sigma graph method;
s3, constructing and training an improved Adaboost model based on RF, inputting the model as the time of eliminating the water regime data of a noise value, a missing value and an abnormal value, and outputting the model as the flow of the water regime data;
and S4, performing real-time prediction and interpolation of abnormal values on the water regime data by applying an RF-Adaboost model, and outputting a prediction variable.
Preferably, the predictive variable refers to real-time monitoring data of the water regime, and comprises water level and flow of a check gate.
Preferably, step S1 specifically includes:
s11, defining a long window with a length L of 3 for the acquired regimen data; at a certain moment, the signal sample in the window is x (i-1), x (i), and x (i +1), wherein x (i) is the signal sample value located in the center of the window; after the 3 signal samples are arranged according to the sequence of sample values from small to large, wherein the sample value at the position i is defined as an output value of median filtering, so that an isolated noise point is eliminated, and the isolated noise point is a trend item of the time sequence;
and S12, calculating the difference between the trend term and the measured value of the time series, namely the residual term.
Preferably, step S2 specifically includes:
the real-time diagnosis standard specifically comprises the following steps: data less than μ - β σ or greater than μ + β σ are outlier data, where μ is the mean, σ is the standard deviation, and β is a variable parameter.
Preferably, step S3 specifically includes:
given a training data set: (x)1,y1),......,(xN,yN) Wherein y isiE {1, -1} used to represent training samplesA category label;
s31, initializing sample set weight as
In the formula: d1(i) Representing an initial weight distribution of a training sample set;
w1ieach training sample is initially given the same weight;
n is the number of samples;
s32, performing multiple iterations, where m is 1,2, and N is the first iteration
e) Selecting Random Forest (RF) as basic classifier, and weighting with data distribution DmLearning and calculating weak classifier Gm(x);
The above formula represents the weak classifier Gm(x) Classifying the sample data by a threshold, wherein all data on one side of the threshold is classified into a class-1, and data on the other side is classified into a class + 1;
f) computing the weak classifier Gm(x) Classification error rate on training data set
In the formula: e.g. of the typemIs the error rate;
P(Gm(xi)≠yi) Is Gm(xi)≠yiThe probability of (1), i.e. the probability of a mistake;
I(Gm(xi)≠yi): if is Gm(xi)=yiWhen I is 1, otherwise I is 0;
wmifor the ith training in the mth iterationWeights assigned to training samples;
from this the data weight distribution D can be seenmAnd the weak classifier Gm(x) The relation of the classification error rate of (1);
g) computing the weak classifier Gm(x) Coefficient (c):
αmrepresents the weak classifier Gm(x) Degree of importance in the final classifier whenWhen α ismNot less than 0 and alphamWith emDecrease and increase;
the classifier is now: f. ofm(x)=αmGm(x)
h) Updating the weight distribution of the training data for the next iteration:
Dm+1=(wm+1,1,wm+1,2...wm+1,i...,wm+1,N) (4)
where Z ismIs a normalization factor such that Dm+1Becomes a probability distribution:
s33, according to the weak classifier weight alphamCombining weak classifiers, i.e.
Through the action of sign function sign, a strong classifier is obtained as follows:
preferably, the RF-Adaboost-based model established in step S3 is:
classifying the sample data set of Adaboost by adopting a Random Forest (RF) as a weak classifier;
after the step S1 is finished, removing the water regime data of the isolated noise points after median filtering, and training the constructed RF-Adaboost model based on the actually-measured water regime data of the abnormal values after the step S2 is finished;
the training data is actually measured water regime data within 2 hours of a year with a time span;
in each round of training, a new weak classifier is obtained by classifying through Random Forest (RF), namely, by changing the weight of the sample, especially, the previous error sample is more heavily weighted until the error rate is lower than a specified value or a preset maximum iteration number is reached.
Preferably, the regimen monitoring data acquired in step S1 is updated to 2-hour time-series data, the RF-Adaboost-based model trained in step S3 is input, and the output predictive variable values are corrected values of real-time predictive values and abnormal values.
The invention has the beneficial effects that:
the invention discloses a diagnosis and interpolation method in abnormal water regime data based on an RF-Adaboost model, which comprises the steps of obtaining water regime monitoring data, carrying out median filtering to eliminate obvious noise points, and analyzing a residual error term; drawing a 3-Sigma diagram, and identifying and diagnosing abnormal data in real time based on the 3-Sigma diagram; constructing and training an Adaboost model based on Random Forest (RF) improvement; and (3) predicting the regimen monitoring data in real time by applying an RF-Adaboost model, and interpolating abnormal data. By adopting the method, the real-time diagnosis, prediction and interpolation of the flow monitoring data can be effectively improved, so that the reliability of the data is improved, the water regime change is objectively reflected, and the engineering scheduling is effectively guided.
Drawings
FIG. 1 is a schematic flowchart of a method for diagnosing and interpolating abnormal water regimen data based on an RF-Adaboost model in embodiment 1;
FIG. 2 is the result of the 3-Sigma plot identifying abnormal values of the diagnostic throttle flow data in example 1;
fig. 3 shows the results of prediction and interpolation of throttle valve actual flow data using the RF-Adaboost-based model in example 1.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings. It should be understood that the detailed description and specific examples, while indicating the invention, are intended for purposes of illustration only and are not intended to limit the scope of the invention.
Example 1
The embodiment adopts a specific embodiment, takes the north-south water tune central line as an example, and provides a diagnosis and interpolation method in abnormal water situation data based on an RF-Adaboost model, which specifically comprises the following steps:
step 1, performing median filtering analysis on the acquired water regime monitoring data to obtain a trend term and a residual term of a time sequence.
The method is influenced by various factors such as hydraulic characteristics of a canal, engineering operation period, labor intensity of personnel and the like in the open channel water transfer engineering, and the water regime monitoring frequency of the open channel water transfer engineering is 2 hours, so that the water regime change can be objectively reflected, and engineering scheduling can be effectively guided. The brake-crossing flow data of the Hutuo river regulation gate from 1 month and 1 day in 2018 to 12 month and 31 day in 2018 are selected as research objects.
And extracting a trend item and a residual item of the flow data time sequence by adopting a median filtering method, wherein a frequency distribution diagram of the residual item obtained after the flow data is subjected to median filtering presents an obvious normal distribution characteristic.
And 2, adopting a 3-Sigma graph method to diagnose abnormal data in real time.
Drawing a 3-Sigma graph, depicting the discrete distribution of the data, introducing a parameter beta into a 3-Sigma method to meet the abnormal data diagnosis requirements of different data sets, and taking data smaller than mu-beta Sigma or larger than mu + beta Sigma as judgment standards of abnormal data, wherein mu is a mean value, Sigma is a standard deviation, and beta is 1.62, and regulating the abnormal value result of the brake flow data as shown in a figure 2.
Step 3, constructing and training an RF-Adaboost-based model:
given a training data set: (x)1,y1),......,(xN,yN) Wherein y isiE {1, -1} is used to represent the class label of the training sample.
(1) Initializing sample set weights as
In the formula: d1(i) Training initial weight distribution of a sample set;
w1ieach training sample is initially given the same weight;
n is the number of samples;
(2) a number of iterations are performed, with m being 1,2
Selecting Random Forest (RF) as basic classifier, and weighting DmLearning and calculating weak classifier Gm(x);
The above formula represents the weak classifier Gm(x) Classifying the sample data by a threshold, wherein all data on one side of the threshold is classified into a class-1, and data on the other side is classified into a class + 1;
② calculating Gm(x) Classification error Rate on training data set
In the formula: e.g. of the typemIs an errorRate;
P(Gm(xi)≠yi) Is Gm(xi)≠yiThe probability of (1), i.e. the probability of a mistake;
I(Gm(xi)≠yi): if is Gm(xi)=yiWhen I is 1, otherwise I is 0;
wmithe weight assigned to the ith training sample in the mth iteration;
from this, the data weight distribution D can be seenmAnd a basic classifier Gm(x) The classification error rate of (1).
③ calculating weak classifiers Gm(x) Coefficient (c):
αmrepresents Gm(x) Degree of importance in the final classifier whenWhen a ism30 and αmWith emSo that the smaller the classification error rate, the greater the contribution of the basic classifier in the final classifier.
The classifier is as follows: f. ofm(x)=αmGm(x)
And fourthly, updating the weight distribution of the training data for the next iteration.
Dm+1=(wm+1,1,wm+1,2...wm+1,i...,wm+1,N) (4)
Where Z ismIs a normalization factor such that Dm+1Becomes a probability distribution:
thus, the weak classifier Gm(x) The weight of misclassified samples is expanded while the weight of correctly classified samples is reduced.
(3) According to weak classifier weight alphamCombining weak classifiers, i.e.
Through the action of sign function sign, a strong classifier is obtained as follows:
the model inputs are the time sequence of normal data after eliminating the noise value after median filtering and the abnormal value diagnosed by the 3-Sigma diagram, the flow value of the normal data is output, a new weak classifier is added in each round of training, namely the weight of the sample is changed through a decision tree, particularly the previous wrong sample can obtain larger weight until reaching a low enough error rate or reaching a specified maximum iteration number.
And 4, performing predictive interpolation on the noise value and the abnormal value of the flow monitoring by using an RF-Adaboost model.
The 2-hour time series data of the regimen monitoring data acquired in step 1 is input as an RF-Adaboost model, applied to the RF-Adaboost model trained in step S3, and the output predictive variable value is a real-time predictive value and a correction value of an abnormal value, as shown in fig. 3.
By adopting the technical scheme disclosed by the invention, the following beneficial effects are obtained:
the invention discloses a diagnosis and interpolation method in abnormal water regime data based on an RF-Adaboost model, which relates to the technical field of water regime monitoring and comprises the following steps: acquiring real-time water regime data, performing time series analysis on the water regime data by adopting median filtering, drawing a 3-Sigma diagram, and performing real-time identification and diagnosis on abnormal data based on the 3-Sigma diagram; the method has the advantages that an Adaboost model improved based on Random Forest (RF) is constructed and trained, the RF-Adaboost model is applied to predict the water regime monitoring data in real time, and the abnormal data are interpolated.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, many modifications and adaptations can be made without departing from the principle of the present invention, and such modifications and adaptations should also be considered to be within the scope of the present invention.
Claims (7)
1. A diagnosis and interpolation method in abnormal water situation data based on an RF-Adaboost model is characterized by comprising the following steps:
s1, carrying out median filtering analysis on the acquired regimen monitoring data to obtain a trend term and a residual term of a time sequence;
s2, carrying out real-time diagnosis on the abnormal data by adopting a 3-Sigma graph method;
s3, constructing and training an improved Adaboost model based on RF, inputting the model as the time of eliminating the water regime data of a noise value, a missing value and an abnormal value, and outputting the model as the flow of the water regime data;
and S4, performing real-time prediction and interpolation of abnormal values on the water regimen data by applying an RF-Adaboost model, and outputting a prediction variable.
2. The method for diagnosing and interpolating abnormal water regime data based on the RF-Adaboost model as claimed in claim 1, wherein the predictive variable refers to water regime real-time monitoring data, and comprises a check gate water level and a check gate flow rate.
3. The method for diagnosing and interpolating abnormal water regimen data based on the RF-Adaboost model according to claim 1, wherein the step S1 specifically includes:
s11, defining a long window with the length L being 3 for the acquired water regime data; at a certain moment, the signal sample in the window is x (i-1), x (i), and x (i +1), wherein x (i) is the signal sample value located in the center of the window; after the 3 signal samples are arranged according to the sequence of sample values from small to large, wherein the sample value at the position i is defined as an output value of median filtering, so that an isolated noise point is eliminated, and the isolated noise point is a trend item of the time sequence;
and S12, calculating the difference between the trend term and the measured value of the time series, namely the residual term.
4. The method for diagnosing and interpolating an RF-Adaboost regression model in abnormal water situation data according to claim 1, wherein the step S2 specifically includes:
the real-time diagnosis standard specifically comprises the following steps: data less than μ - β σ or greater than μ + β σ are outlier data, where μ is the mean, σ is the standard deviation, and β is a variable parameter.
5. The method for diagnosing and interpolating abnormal water regimen data based on the RF-Adaboost model according to claim 1, wherein the step S3 specifically includes:
given a training data set: (x)1,y1),......,(xN,yN) Wherein y isiE {1, -1} is used to represent a class label for the training sample;
s31, initializing sample set weight as
In the formula: d1(i) Representing an initial weight distribution of a training sample set;
w1ieach training sample is initially given the same weight;
n is the number of samples;
s32, performing multiple iterations, where m is 1,2, and N is the first iteration
a) Selecting RF as basic classifier, and weighting with data weight distribution DmLearning and calculating weak classifier Gm(x);
The above formula represents the weak classifier Gm(x) Classifying the sample data by a threshold, wherein all data on one side of the threshold is classified into a class-1, and data on the other side is classified into a class + 1;
b) computing the weak classifier Gm(x) Classification error rate on training data set
In the formula: e.g. of the typemIs the error rate;
P(Gm(xi)≠yi) Is Gm(xi)≠yiThe probability of (1), i.e. the probability of a mistake;
I(Gm(xi)≠yi): if is Gm(xi)=yiWhen I is 1, otherwise I is 0;
wmithe weight assigned to the ith training sample in the mth iteration;
from this the data weight distribution D can be seenmAnd the weak classifier Gm(x) The relation of the classification error rate of (1);
c) computing the weak classifier Gm(x) Coefficient (c):
αmrepresents the weak classifier Gm(x) In the final classifierOf importance whenWhen is αmNot less than 0 and alphamWith emDecrease and increase;
the classifier is now: f. ofm(x)=αmGm(x)
d) Updating the weight distribution of the training data for the next iteration:
Dm+1=(wm+1,1,wm+1,2...wm+1,i...,wm+1,N) (4)
where Z ismIs a normalization factor such that Dm+1Becomes a probability distribution:
s33, according to the weak classifier weight alphamCombining weak classifiers, i.e.
Through the action of sign function sign, a strong classifier is obtained as follows:
6. the method for diagnosing and interpolating abnormal water situation data based on the RF-Adaboost model as claimed in claim 1, wherein the RF-Adaboost model established in the step S3 is:
classifying the sample data set of Adaboost by using RF as a basic classifier;
after the step S1 is finished, removing the water regime data of the isolated noise points after median filtering, and training the constructed RF-Adaboost model based on the actually-measured water regime data of the abnormal values after the step S2 is finished;
the training data is actually measured water regime data within 2 hours of a year with a time span;
in each round of training, classification by RF results in a new weak classifier, i.e. by changing the weights of the samples, especially the previous erroneous samples are weighted more, until the error rate is lower than a specified value or a preset maximum number of iterations is reached.
7. The method for diagnosing and interpolating abnormal water regime data based on RF-Adaboost regression model according to claim 1, wherein the water regime monitoring data obtained in step S1 is updated to 2-hour time series data, the improved Adaboost model based on RF trained in step S3 is inputted, and the output predictive variable value is the real-time predicted value and the corrected value of the abnormal value.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210116677.6A CN114493023A (en) | 2022-02-07 | 2022-02-07 | Diagnosis and interpolation method in abnormal water regime data based on RF-Adaboost model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210116677.6A CN114493023A (en) | 2022-02-07 | 2022-02-07 | Diagnosis and interpolation method in abnormal water regime data based on RF-Adaboost model |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114493023A true CN114493023A (en) | 2022-05-13 |
Family
ID=81478520
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210116677.6A Pending CN114493023A (en) | 2022-02-07 | 2022-02-07 | Diagnosis and interpolation method in abnormal water regime data based on RF-Adaboost model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114493023A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113406354A (en) * | 2021-06-21 | 2021-09-17 | 湖南国天电子科技有限公司 | Data optimization method and system for ADCP |
CN114925731A (en) * | 2022-06-06 | 2022-08-19 | 华电金沙江上游水电开发有限公司叶巴滩分公司 | Method for detecting abnormal value of monitoring data of flexible inclinometer |
CN114925731B (en) * | 2022-06-06 | 2024-05-31 | 华电金沙江上游水电开发有限公司叶巴滩分公司 | Method for detecting abnormal value of monitoring data of flexible inclinometer |
-
2022
- 2022-02-07 CN CN202210116677.6A patent/CN114493023A/en active Pending
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113406354A (en) * | 2021-06-21 | 2021-09-17 | 湖南国天电子科技有限公司 | Data optimization method and system for ADCP |
CN114925731A (en) * | 2022-06-06 | 2022-08-19 | 华电金沙江上游水电开发有限公司叶巴滩分公司 | Method for detecting abnormal value of monitoring data of flexible inclinometer |
CN114925731B (en) * | 2022-06-06 | 2024-05-31 | 华电金沙江上游水电开发有限公司叶巴滩分公司 | Method for detecting abnormal value of monitoring data of flexible inclinometer |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108875771B (en) | Fault classification model and method based on sparse Gaussian Bernoulli limited Boltzmann machine and recurrent neural network | |
CN116757534B (en) | Intelligent refrigerator reliability analysis method based on neural training network | |
CN108875772B (en) | Fault classification model and method based on stacked sparse Gaussian Bernoulli limited Boltzmann machine and reinforcement learning | |
CN111160582B (en) | Equipment fault identification method and device and computer readable storage medium | |
Guh et al. | An effective application of decision tree learning for on-line detection of mean shifts in multivariate control charts | |
CN113642754B (en) | Complex industrial process fault prediction method based on RF noise reduction self-coding information reconstruction and time convolution network | |
CN112116198B (en) | Data-driven process industrial state perception network key node screening method | |
CN113204914B (en) | Flight data abnormity interpretation method based on multi-flight data characterization modeling | |
CN111861272A (en) | Multi-source data-based complex electromechanical system abnormal state detection method | |
CN111738520A (en) | System load prediction method fusing isolated forest and long-short term memory network | |
CN112598144B (en) | CNN-LSTM burst fault early warning method based on correlation analysis | |
CN114282443B (en) | Residual service life prediction method based on MLP-LSTM supervised joint model | |
CN109240276B (en) | Multi-block PCA fault monitoring method based on fault sensitive principal component selection | |
CN109298633A (en) | Chemical production process fault monitoring method based on adaptive piecemeal Non-negative Matrix Factorization | |
CN114493023A (en) | Diagnosis and interpolation method in abnormal water regime data based on RF-Adaboost model | |
WO2022188425A1 (en) | Deep learning fault diagnosis method integrating prior knowledge | |
CN115081514A (en) | Industrial equipment fault identification method under data imbalance condition | |
CN114298413A (en) | Hydroelectric generating set runout trend prediction method | |
CN114637620A (en) | Database system abnormity classification prediction method based on SVM algorithm | |
CN110288724B (en) | Batch process monitoring method based on wavelet function principal component analysis | |
CN112561153A (en) | Scenic spot crowd gathering prediction method based on model integration | |
CN116756619B (en) | Equipment intelligent diagnosis method and system based on big data | |
CN116643550A (en) | Industrial fault traceability analysis method based on time sequence causal mining network | |
CN114513328B (en) | Network traffic intrusion detection method based on concept drift and deep learning | |
US11687439B2 (en) | Automatic window generation for process trace |
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 |