CN112085926A - River water pollution early warning method and system - Google Patents

River water pollution early warning method and system Download PDF

Info

Publication number
CN112085926A
CN112085926A CN202010765837.0A CN202010765837A CN112085926A CN 112085926 A CN112085926 A CN 112085926A CN 202010765837 A CN202010765837 A CN 202010765837A CN 112085926 A CN112085926 A CN 112085926A
Authority
CN
China
Prior art keywords
pollution
physical quantity
model
sensor
quantity data
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
Application number
CN202010765837.0A
Other languages
Chinese (zh)
Other versions
CN112085926B (en
Inventor
姜春涛
谢佳璇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Foshan University
Original Assignee
Foshan University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Foshan University filed Critical Foshan University
Priority to CN202010765837.0A priority Critical patent/CN112085926B/en
Publication of CN112085926A publication Critical patent/CN112085926A/en
Application granted granted Critical
Publication of CN112085926B publication Critical patent/CN112085926B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
    • G08B21/02—Alarms for ensuring the safety of persons
    • G08B21/12—Alarms for ensuring the safety of persons responsive to undesired emission of substances, e.g. pollution alarms
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
    • G01N33/18—Water
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
    • 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
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B25/00—Alarm systems in which the location of the alarm condition is signalled to a central station, e.g. fire or police telegraphic systems
    • G08B25/01—Alarm systems in which the location of the alarm condition is signalled to a central station, e.g. fire or police telegraphic systems characterised by the transmission medium
    • G08B25/10—Alarm systems in which the location of the alarm condition is signalled to a central station, e.g. fire or police telegraphic systems characterised by the transmission medium using wireless transmission systems

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Health & Medical Sciences (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Chemical & Material Sciences (AREA)
  • Mathematical Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Computational Mathematics (AREA)
  • Emergency Management (AREA)
  • Business, Economics & Management (AREA)
  • Probability & Statistics with Applications (AREA)
  • Algebra (AREA)
  • Immunology (AREA)
  • Biochemistry (AREA)
  • Analytical Chemistry (AREA)
  • Operations Research (AREA)
  • Pathology (AREA)
  • Food Science & Technology (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Medicinal Chemistry (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • Environmental & Geological Engineering (AREA)
  • Toxicology (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a river water pollution early warning method and a system, wherein sensor nodes are arranged in a river at equal intervals; physical quantity data are collected through sensor nodes; preprocessing physical quantity data to extract key factors; constructing a grey prediction model to calculate the predicted value of the key factor; predicting the pollution degree of the water area where the sensor node is located through the SVR model and the predicted value; the method has the advantages that the main characteristics are determined by using an information gain method, the requirement on the quality of the factors is further reduced by combining a Lasso characteristic selection algorithm, more effective key factors are determined, the prediction precision is greatly improved, a large amount of calculation is reduced, the requirement on the quality of the factors by the traditional support vector machine prediction method is reduced, a better effect is achieved, the position of the pollution source in a water function area can be intelligently judged in real time, the pollution point can be timely positioned, and an alarm is given.

Description

River water pollution early warning method and system
Technical Field
The disclosure relates to the field of water environment pollution early warning, in particular to a river water pollution early warning method and system.
Background
In the work of river regulation, most pollution monitoring stations are only arranged at the important river basin access boundary line, the distance between each pollution monitoring station is generally hundreds of kilometers, in the interval of hundreds of kilometers, a large number of drain outlets exist, the water pollution condition in the river is difficult to analyze and early warn, the monitoring work is difficult to reach real-time performance, when serious pollution occurs, the monitoring work can take several hours, even several days, the pollution monitoring stations arranged by river monitoring departments can monitor the pollution condition, the river pollution condition can not be quickly positioned and monitored, the accurate positioning monitoring is difficult to realize, the accuracy of pollution enterprises can not be realized, and the pollution cause can be quickly analyzed.
Disclosure of Invention
The invention aims to provide a river water pollution early warning method and a river water pollution early warning system, which are used for solving one or more technical problems in the prior art and at least provide a beneficial choice or creation condition.
In order to achieve the above objects, according to an aspect of the present disclosure, there is provided a river water pollution early warning method, including the steps of:
step 1, arranging sensor nodes in a river channel at equal intervals;
step 2, physical quantity data are collected through sensor nodes;
step 3, preprocessing the physical quantity data to extract key factors;
step 4, constructing a grey prediction model to calculate the predicted value of the key factor;
step 5, constructing an SVR model and predicting the pollution degree of the water area where the sensor node is located through the SVR model and the predicted value;
and 6, when the pollution degree of the water area where the sensor node is located is larger than a pollution threshold value, pushing pollution early warning information to mobile equipment of a manager.
Further, in step 1, the sensor nodes are composed of an active temperature sensor, an active conductivity sensor, an active COD sensor, an active pH sensor and an active residual chlorine sensor, and the sensor nodes supply power and transmit physical quantity data to a base station through a communication line; the base station is used for receiving the sensor number of the sensor node and physical quantity data acquired by the sensor node; the equidistant sensor nodes are arranged in the river channel, and the equidistant sensor nodes are arranged in the water of the river channel at intervals of 500 meters.
Further, the parameters of the temperature sensor are: precision: ± 0.1 ℃, range: -5-50 ℃ (23-122 ° F), resolution: 0.01 ℃; the parameters of the conductivity sensor are: precision: 0-10000 muS/cm is plus or minus 0.5% +1 muS/cm of the reading; precision: 100000-: 0-350,000. mu.S/cm, resolution: 0.1 mu S/cm; the parameters of the COD sensor are: precision: . + -. 0.1mg/L, range: 0-20 mg/L; resolution ratio: 0.01 mg/L; the parameters of the pH sensor were: precision: pH ± 0.1 or better, range: pH 0-14, resolution: 0.01 pH; the parameters of the residual chlorine sensor are as follows: the measuring range is 0.00-4.00 mg/L, the residual chlorine value is 0-30mg/L, the temperature is-10.0- +60.0 ℃, and the output is 4-20mA/0-20 mA.
Further, in step 2, the physical quantity data includes temperature, conductivity, chemical oxygen demand, pH value, residual chlorine amount.
Further, in step 3, the method for preprocessing the physical quantity data to extract the key factors includes:
step 3.1, respectively recording physical quantity data: the temperature, the conductivity, the chemical oxygen demand, the pH value and the residual chlorine amount are x1，x2，x3，x4，x5Recording the pollution degree of the water area where the sensor node is located as y, wherein the calculation method of the pollution degree y comprises the following steps: y is x1，x2，x3，x4，x5The number of any physical quantity data greater than a preset value;
wherein x is1The preset value of (A) is 35 ℃, and the preset value of x2 is 1500-2000uS/cm, x3The preset value of (a) is 300-350 mg/L, x4The preset value of (A) is 4.0-9.5, x5The preset value of the pressure sensor is 9.17kgCL/h, and the preset values can be manually adjusted;
step 3.2, preprocessing the data by using an information gain method and a Lasso characteristic selection algorithm; the information gain is based on the concept of entropy in information theory, which is a measure of uncertainty on attributes of events; the larger the entropy of an attribute is, the larger uncertain information it contains is; therefore, by adopting the greedy algorithm ID3, the attribute with the highest information gain is always selected as the test attribute of the current node, so that the main factor influencing the pollution degree is judged; the information gain method and the Lasso characteristic selection algorithm in the step are mainly used for selecting key characteristics influencing the pollution degree, and the specific processing process is as follows:
step 3.2.1, influence factors x of the pollution degree y and the pollution degree1，x2，x3，x4，x5The data are subjected to standardization preprocessing: a transformation function of
Figure BDA0002614554430000021
Wherein a isijDenotes xi(i ═ 1,2, 3.., 5.) corresponding to physical quantity data of the jth previous sensor node arranged at equal distance in the river where the current sensor node is located, bijThe physical quantity data after conversion (namely, the data of different attributes acquired by the sensor at different nodes are standardized, so that the subsequent calculation is convenient, and the subsequent calculation directly uses 0 or 1); m is the number of sensor types, n is the serial number of the physical quantity, wherein m is 5, and n is 5; the standardization method is convenient for calculating the factors influencing the pollution degree, and the factors influencing the pollution degree are divided by the pollution degree to ensure that the entropy of each factor is further solved, so that the information gain is further solved, and the main factors influencing the pollution degree are judged;
wherein step 3.2.1 corresponds to a normalization for the subsequent calculation of SiReference may be made to the article "regression analysis based housing price models and predictions" (queen matches), noting, for example, that the temperature is α1(the different attributes are denoted as αj) To temperature alpha1Data α at 5 nodes11、α21、α31、α41、α51Performing standardization (comparing each data with the average value of the 5 data, wherein the average value is greater than 1 and less than 0), and so on, performing the same treatment on each attribute and the data of the pollution degree, then counting the number of samples of 0 and 1 in the pollution degree, and recording the pollution degree as 0 as s1The degree of contamination is 1 and is denoted as s2。
Step 3.2.2, obtaining expected information required by sample classification of the physical quantity data of the current sensor node;
let S be a set of samples of S current sensor node physical quantity data, i.e. physical quantity data collected by a current sensor node in a given period of time, (e.g. the last 30 minutes), assuming that the class label attribute has m different values, m different classes C are definedi(i is 1,2,3 … m), let siIs of the class CiThe number of samples in (1), the desired information required for a given sample classification is given by:
Figure BDA0002614554430000031
wherein p isiIs that any sample belongs to CiProbability of, by siEstimating by the/s;
step 3.2.3, determining the attribute (i.e. the influence factor of the contamination degree) x1Dividing the entropy of the subsets; let SijIs the subset SjClass CiThe number of samples of (a); according to x1Entropy of partitioning subsets into
Figure BDA0002614554430000032
Wherein
Figure BDA0002614554430000033
Act as weights for the jth subset, and are equal to the subset (i.e., x)1A value of ajAnd j is a value range [1, v ]]V is the number of samples) divided by the total number of samples in s,
Figure BDA0002614554430000034
wherein
Figure BDA0002614554430000035
Is SjThe middle sample belongs to class CiProbability (set attribute x)1There are v different values a1,a2,…av}. Can use the attribute x1Divide S into v values S1,S2,…SvIn which S isjIncluding samples of S at x1Has a value ofaj)，v＝5；
Step 3.2.4, the corresponding information gain can be obtained according to the expected information and the entropy: for in x1The Gain of information to be obtained by the upper branch is Gain (x) obtained by the following formula1)＝I(s1,s2…sm)-E(x1)；
Step 3.2.5, similarly through step 3.2.3 to step 3.2.4, gets the respective attribute (X)2-X5) After the comparison from large to small, more than half of the attributes are selected as the predicted attributes, which are marked as c1,c2,…,cn；
Step 3.2.6, selecting key factors influencing the pollution degree by combining the selected attributes (namely the main factors influencing the pollution degree) with a Lasso characteristic selection algorithm; lasso realizes the purpose of compressing the coefficients of the features and changing some regression coefficients into 0 by constructing a penalty function so as to achieve the purpose of feature selection; it is a compression estimation method taking a reduced variable set as an idea; model selection is essentially a process that seeks for sparse representation of the model, and this process can be accomplished by optimizing a "loss" + "penalty" function problem; the Lasso method further screens the pollution degree influence factors screened by the information gain method, so that the quality of the factors is improved; the Lasso parameter estimate is defined as:
Figure BDA0002614554430000041
wherein beta is a regression coefficient vector, and lambda is a nonnegative regular parameter, and the complexity of the model is controlled; the larger the lambda is, the greater the penalty degree of the linear model with more characteristics is, so as to finally obtain a model with less characteristics,
Figure BDA0002614554430000042
referred to as a penalty term, ρ is the number of attributes, that is, ρ is 5; the parameter lambda can be determined by adopting a cross validation method, and a lambda value with the minimum cross validation error is selected; finally, fitting the model again according to the obtained lambda value,
Figure BDA0002614554430000043
is a key factor.
Further, in step 4, the method for constructing the gray prediction model to calculate the predicted value of the key factor includes:
grey prediction is a method for predicting a system containing uncertain factors and is based on a grey model; selecting a common GM (1,1) model;
step 4.1, assume the current sensor node xiIs defined as x(0)＝{x(0)(i)}，xiCorresponding data is x(0)(i) I ═ 1,2, …,5, the gray prediction model was built as follows:
step 4.2, for x(0)Performing a first accumulation to obtain a first accumulation sequence x(1)＝{x(1)(k),k＝0,1,2,…,m}；
Step 4.3, for x(0)Establishing a first order linear differential equation of the formula
Figure BDA0002614554430000044
Namely the GM (1,1) model, where ∈ is the coefficient of development, μ is the amount of gray effect, and the value of μ is set as the value of the key factor; m is the number of sensor types of the sensor nodes and is set to be 5; n value range [1,5 ]]；
Step 4.4, solving a differential equation to obtain a prediction model with the formula of
Figure BDA0002614554430000045
x(1)And x*(1)Respectively representing the accumulation sequence and a prediction model, wherein the numbers in the upper right corner represent parameters used in the prediction model, and the parameters are consistent with the definition;
step 4.5, obtaining data x of GM (1,1) model*(1)(k +1) is reduced to x by cumulative subtraction*(0)The gray prediction model of (k +1), i.e., x (0), is formulated as
Figure BDA0002614554430000046
Finally, the predicted value of each key factor for predicting the influence on the pollution degree is recorded as
Figure BDA0002614554430000047
Wherein, 0<m≤n。
Further, in step 5, the method for constructing the SVR model and predicting the pollution degree of the water area where the sensor node is located through the SVR model and the predicted value is that the SVR adopts the thought of a support vector machine to carry out regression analysis on data when fitting, and then the SVR carries out regression analysis on the data
Figure BDA0002614554430000051
As a training set, among others,
Figure BDA0002614554430000052
training the prediction model, and updating a training sample in real time before predicting the next moment, namely adding the actual pollution degree and the selected principal component data of the previous moment and removing the most original data; for the sample
Figure BDA0002614554430000053
Output f (Out) according to the modeli) And true value
Figure BDA0002614554430000054
The difference between them to calculate the loss, if and only if
Figure BDA0002614554430000055
Time, loss is zero, f (Out)i) The key factor prediction value is output by the SVR model; allow f (Out)i) And
Figure BDA0002614554430000056
the most deviation there between; only when
Figure BDA0002614554430000057
Calculating the loss only when the loss is not calculated; when in use
Figure BDA0002614554430000058
Then, the pollution degree of the water area where the sensor node is located at the next moment is recorded as yn。
Further, the pollution threshold is an arithmetic average value of current pollution degrees of all sensor nodes in the river channel.
Further, the method for pushing the pollution early warning information to the mobile device of the manager comprises the following steps: and the base station pushes the alarm information to the mobile equipment of the manager through the Internet.
The invention also provides a river water pollution early warning system, and the device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the computer program to operate in the units of:
the node setting unit is used for setting sensor nodes in a river channel at equal distance;
the sensor acquisition unit is used for acquiring physical quantity data through the sensor nodes;
the key factor extraction unit is used for preprocessing the physical quantity data to extract key factors;
the predicted value calculating unit is used for constructing a gray prediction model to calculate the predicted value of the key factor;
the pollution degree prediction unit is used for constructing an SVR model and predicting the pollution degree of the water area where the sensor node is located through the SVR model and the predicted value;
and the alarm pushing unit is used for pushing the pollution early warning information to the mobile equipment of the manager when the pollution degree of the water area where the sensor node is located is greater than the pollution threshold value.
The beneficial effect of this disclosure does: the invention provides a river water pollution early warning method and a river water pollution early warning system, which are characterized in that an information gain method is used for determining main characteristics, and then a Lasso characteristic selection algorithm is combined, so that the requirement on the quality of factors is further reduced, more effective key factors are determined, the prediction precision is greatly improved, a large amount of calculation is reduced, the requirement on the quality of the factors by the traditional support vector machine prediction method is reduced, a better effect is achieved, the pollution source position of a water functional area can be intelligently judged in real time, a pollution point can be timely positioned and an alarm can be issued, early warning investigation can be rapidly given when an abnormality occurs, a large amount of time cost is saved, the environment early warning efficiency is improved, and a pollution degree prediction value with high precision can be obtained through a prediction model only by collecting pollution data sets of influence factors in a small period of time.
Drawings
The foregoing and other features of the present disclosure will become more apparent from the detailed description of the embodiments shown in conjunction with the drawings in which like reference characters designate the same or similar elements throughout the several views, and it is apparent that the drawings in the following description are merely some examples of the present disclosure and that other drawings may be derived therefrom by those skilled in the art without the benefit of any inventive faculty, and in which:
FIG. 1 is a flow chart of a method for early warning of river water pollution;
fig. 2 is a structural diagram of a river water pollution early warning system.
Detailed Description
The conception, specific structure and technical effects of the present disclosure will be clearly and completely described below in conjunction with the embodiments and the accompanying drawings to fully understand the objects, aspects and effects of the present disclosure. It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.
Fig. 1 is a flowchart illustrating a method for early warning of river water pollution according to the present disclosure, and a method for early warning of river water pollution according to an embodiment of the present disclosure is described below with reference to fig. 1.
The utility model provides a river course water pollution early warning method, specifically includes the following steps:
step 1, arranging sensor nodes in a river channel at equal intervals;
step 2, physical quantity data are collected through sensor nodes;
step 3, preprocessing the physical quantity data to extract key factors;
step 4, constructing a grey prediction model to calculate the predicted value of the key factor;
step 5, constructing an SVR model and predicting the pollution degree of the water area where the sensor node is located through the SVR model and the predicted value;
and 6, when the pollution degree of the water area where the sensor node is located is larger than a pollution threshold value, pushing pollution early warning information to mobile equipment of a manager.
Further, in step 1, the sensor nodes are composed of an active temperature sensor, an active conductivity sensor, an active COD sensor, an active pH sensor and an active residual chlorine sensor, and the sensor nodes supply power and transmit physical quantity data to a base station through a communication line; the base station is used for receiving the sensor number of the sensor node and physical quantity data acquired by the sensor node; the equidistant sensor nodes are arranged in the river channel, and the equidistant sensor nodes are arranged in the water of the river channel at intervals of 500 meters.
Further, the parameters of the temperature sensor are: precision: ± 0.1 ℃, range: -5-50 ℃ (23-122 ° F), resolution: 0.01 ℃; the parameters of the conductivity sensor are: precision: 0-10000 muS/cm is plus or minus 0.5% +1 muS/cm of the reading; precision: 100000-: 0-350,000. mu.S/cm, resolution: 0.1 mu S/cm; the parameters of the COD sensor are: precision: . + -. 0.1mg/L, range: 0-20 mg/L; resolution ratio: 0.01 mg/L; the parameters of the pH sensor were: precision: pH ± 0.1 or better, range: pH 0-14, resolution: 0.01 pH; the parameters of the residual chlorine sensor are as follows: the measuring range is 0.00-4.00 mg/L, the residual chlorine value is 0-30mg/L, the temperature is-10.0- +60.0 ℃, and the output is 4-20mA/0-20 mA.
Further, in step 2, the physical quantity data includes temperature, conductivity, chemical oxygen demand, pH value, residual chlorine amount.
Further, in step 3, the method for preprocessing the physical quantity data to extract the key factors includes:
step 3.1, respectively recording physical quantity data: the temperature, the conductivity, the chemical oxygen demand, the pH value and the residual chlorine amount are x1，x2，x3，x4，x5Recording the pollution degree of the water area where the sensor node is located as y, wherein the calculation method of the pollution degree y comprises the following steps: y is x1，x2，x3，x4，x5Number of physical quantity data in which any one physical quantity data is larger than a predetermined valueAn amount;
wherein x is1The preset value of (A) is 35 ℃, and the preset value of x2 is 1500-2000uS/cm, x3The preset value of (a) is 300-350 mg/L, x4The preset value of (A) is 4.0-9.5, x5The preset value of the pressure sensor is 9.17kgCL/h, and the preset values can be manually adjusted;
step 3.2, preprocessing the data by using an information gain method and a Lasso characteristic selection algorithm; the information gain is based on the concept of entropy in information theory, which is a measure of uncertainty on attributes of events; the larger the entropy of an attribute is, the larger uncertain information it contains is; therefore, ID3 always selects the attribute with the highest information gain as the test attribute of the current node, thereby determining the main factor affecting the degree of pollution; the information gain method and the Lasso characteristic selection algorithm in the step are mainly used for selecting key characteristics influencing the pollution degree, and the specific processing process is as follows:
step 3.2.1, influence factors x of the pollution degree y and the pollution degree1，x2，x3，x4，x5The data are subjected to standardization preprocessing: a transformation function of
Figure BDA0002614554430000071
Wherein a isijDenotes xi(i ═ 1,2, 3.., 5.) corresponding to physical quantity data of the jth previous sensor node arranged at equal distance in the river where the current sensor node is located, bijConverted physical quantity data; m is the number of sensor types, n is the serial number of physical quantity, m is 5, n is 5; the standardization method is convenient for calculating the factors influencing the pollution degree, and the factors influencing the pollution degree are divided by the pollution degree to ensure that the entropy of each factor is further solved, so that the information gain is further solved, and the main factors influencing the pollution degree are judged;
step 3.2.2, obtaining expected information required by sample classification of the physical quantity data of the current sensor node;
let S be a set of samples of S current sensor node physical quantity data, i.e. physical quantity data collected by a current sensor node in a given period of time, (e.g. the last 30 minutes), assuming that the class label attribute has m different values, m different classes C are definedi(i＝1,2,3… m), set siIs of the class CiThe number of samples in (1), the desired information required for a given sample classification is given by:
Figure BDA0002614554430000081
wherein p isiIs that any sample belongs to CiProbability of, by siEstimating by the/s;
step 3.2.3, determining the attribute (i.e. the influence factor of the contamination degree) x1Dividing the entropy of the subsets; let SijIs the subset SjClass CiThe number of samples of (a); according to x1Entropy of partitioning subsets into
Figure BDA0002614554430000082
Wherein
Figure BDA0002614554430000083
Act as weights for the jth subset, and are equal to the subset (i.e., x)1A value of aj) Divided by the total number of samples in s,
Figure BDA0002614554430000084
wherein
Figure BDA0002614554430000085
Is SjThe middle sample belongs to class CiProbability (set attribute x)1There are v different values a1,a2,…av}. Can use the attribute x1Divide S into v values S1,S2,…SvIn which S isjIncluding samples of S at x1Has a value of aj) V is 5, and v is the number of samples;
step 3.2.4, the corresponding information gain can be obtained according to the expected information and the entropy: for in x1The Gain of information to be obtained by the upper branch is Gain (x) obtained by the following formula1)＝I(s1,s2…sm)-E(x1)；
Step 3.2.5, the same asThe process goes through step 3.2.3 to step 3.2.4 to obtain the respective attributes (X)2-X5) After the comparison from large to small, more than half of the attributes are selected as the predicted attributes, which are marked as c1,c2,…,cn；
Step 3.2.6, selecting key factors influencing the pollution degree by combining the selected attributes (namely the main factors influencing the pollution degree) with a Lasso characteristic selection algorithm; lasso realizes the purpose of compressing the coefficients of the features and changing some regression coefficients into 0 by constructing a penalty function so as to achieve the purpose of feature selection; it is a compression estimation method taking a reduced variable set as an idea; model selection is essentially a process that seeks for sparse representation of the model, and this process can be accomplished by optimizing a "loss" + "penalty" function problem; the Lasso method further screens the pollution degree influence factors screened by the information gain method, so that the quality of the factors is improved; the Lasso parameter estimate is defined as:
Figure BDA0002614554430000086
wherein beta is a regression coefficient vector, and lambda is a nonnegative regular parameter, and the complexity of the model is controlled; the larger the lambda is, the greater the penalty degree of the linear model with more characteristics is, so as to finally obtain a model with less characteristics,
Figure BDA0002614554430000091
referred to as a penalty term, ρ is the number of attributes, that is, ρ is 5; the parameter lambda can be determined by adopting a cross validation method, and a lambda value with the minimum cross validation error is selected; finally, fitting the model again according to the obtained lambda value,
Figure BDA0002614554430000092
is a key factor.
Further, in step 4, the method for constructing the gray prediction model to calculate the predicted value of the key factor includes:
grey prediction is a method for predicting a system containing uncertain factors and is based on a grey model; selecting a common GM (1,1) model;
step 41, assume current sensor node xiIs defined as x(0)＝{x(0)(i)}，xiCorresponding data is x(0)(i) I ═ 1,2, …,5, the gray prediction model was built as follows:
step 4.2, for x(0)Performing a first accumulation to obtain a first accumulation sequence x(1)＝{x(1)(k) K is 0,1,2, …, m }; m is the number of sensor types of the sensor nodes and is set to be 5; n value range [1,5 ]]；
Step 4.3, for x(0)Establishing a first order linear differential equation of the formula
Figure BDA0002614554430000093
Namely the GM (1,1) model, where ∈ is the coefficient of development, μ is the amount of gray effect, and the value of μ is set as the value of the key factor; m is the number of sensor types of the sensor nodes and is set to be 5; n value range [1,5 ]]；
Step 4.4, solving a differential equation to obtain a prediction model with the formula of
Figure BDA0002614554430000094
x(1)And x*(1)Respectively representing the accumulation sequence and a prediction model, wherein the numbers in the upper right corner represent parameters used in the prediction model, and the parameters are consistent with the definition;
step 4.5, because the GM (1,1) model obtains the first accumulation amount, the data x obtained by the GM (1,1) model is used*(1)(k +1) is reduced to x by cumulative subtraction*(0)(k +1), i.e. x(0)The gray prediction model formula is
Figure BDA0002614554430000095
Figure BDA0002614554430000096
Finally, the predicted value of each key factor for predicting the influence on the pollution degree is recorded as
Figure BDA0002614554430000097
Figure BDA0002614554430000098
Further, in step 5, the method for constructing the SVR model and predicting the pollution degree of the water area where the sensor node is located through the SVR model and the predicted value is that the SVR adopts the thought of a support vector machine to carry out regression analysis on data when fitting, and then the SVR carries out regression analysis on the data
Figure BDA0002614554430000099
As a training set, among others,
Figure BDA00026145544300000910
training the prediction model, and updating a training sample in real time before predicting the next moment, namely adding the actual pollution degree and the selected principal component data of the previous moment and removing the most original data; for the sample
Figure BDA0002614554430000101
Output f (c) according to the modeli) And true value
Figure BDA0002614554430000102
The difference between them to calculate the loss, if and only if
Figure BDA0002614554430000103
The loss is zero. Allowing f (c)i) And
Figure BDA0002614554430000104
with the most deviation therebetween. Only when
Figure BDA0002614554430000105
The loss is calculated. When in use
Figure BDA0002614554430000106
The prediction is considered accurate. The pollution degree of the water area where the sensor node is located at the next moment is recorded as yn。
Further, in step 5The method for constructing the SVR model and predicting the pollution degree of the water area where the sensor node is located through the SVR model and the predicted value comprises the steps of carrying out regression analysis on data by adopting the thought of a support vector machine when fitting is carried out on the SVR, and carrying out regression analysis on the data
Figure BDA0002614554430000107
As a training set, among others,
Figure BDA0002614554430000108
training the prediction model, and updating a training sample in real time before predicting the next moment, namely adding the actual pollution degree and the selected principal component data of the previous moment and removing the most original data; for the sample
Figure BDA0002614554430000109
Output f (Out) according to the modeli) And true value
Figure BDA00026145544300001010
The difference between them to calculate the loss, if and only if
Figure BDA00026145544300001011
Time, loss is zero, f (Out)i) The key factor prediction value is output by the SVR model; allow f (Out)i) And
Figure BDA00026145544300001012
the most deviation there between; only when
Figure BDA00026145544300001013
Calculating the loss only when the loss is not calculated; when in use
Figure BDA00026145544300001014
Then, the pollution degree of the water area where the sensor node is located at the next moment is recorded as yn。
Further, the pollution threshold is an arithmetic average value of current pollution degrees of all sensor nodes in the river channel.
Further, the method for pushing the pollution early warning information to the mobile device of the manager comprises the following steps: and the base station pushes the alarm information to the mobile equipment of the manager through the Internet.
The embodiment of the present disclosure provides a river water pollution early warning system, is shown as fig. 2 as a river water pollution early warning system structure chart of the present disclosure, and a river water pollution early warning system of the present embodiment includes: the early warning system comprises a processor, a memory and a computer program which is stored in the memory and can run on the processor, wherein the processor executes the computer program to realize the steps in the embodiment of the early warning system for river water pollution.
The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the computer program to operate in the units of:
the node setting unit is used for setting sensor nodes in a river channel at equal distance;
the sensor acquisition unit is used for acquiring physical quantity data through the sensor nodes;
the key factor extraction unit is used for preprocessing the physical quantity data to extract key factors;
the predicted value calculating unit is used for constructing a gray prediction model to calculate the predicted value of the key factor;
the pollution degree prediction unit is used for constructing an SVR model and predicting the pollution degree of the water area where the sensor node is located through the SVR model and the predicted value;
and the alarm pushing unit is used for pushing the pollution early warning information to the mobile equipment of the manager when the pollution degree of the water area where the sensor node is located is greater than the pollution threshold value.
The river water pollution early warning system can be operated in computing equipment such as desktop computers, notebooks, palm computers and cloud servers. The river water pollution early warning system can be operated by devices including, but not limited to, a processor and a memory. It will be understood by those skilled in the art that the example is merely an example of a river water pollution early warning system, and does not constitute a limitation of a river water pollution early warning system, and may include more or less components than the river water pollution early warning system, or combine some components, or different components, for example, the river water pollution early warning system may further include an input and output device, a network access device, a bus, and the like.
The Processor may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. The general processor can be a microprocessor or the processor can also be any conventional processor and the like, the processor is a control center of the operating device of the river water pollution early warning system, and various interfaces and lines are utilized to connect all parts of the operable device of the whole river water pollution early warning system.
The memory can be used for storing the computer program and/or the module, and the processor realizes various functions of the river water pollution early warning system by running or executing the computer program and/or the module stored in the memory and calling data stored in the memory. The memory may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required by at least one function (such as a sound playing function, an image playing function, etc.), and the like; the storage data area may store data (such as audio data, a phonebook, etc.) created according to the use of the cellular phone, and the like. In addition, the memory may include high speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), at least one magnetic disk storage device, a Flash memory device, or other volatile solid state storage device.
While the present disclosure has been described in considerable detail and with particular reference to a few illustrative embodiments thereof, it is not intended to be limited to any such details or embodiments or any particular embodiments, but it is to be construed as effectively covering the intended scope of the disclosure by providing a broad, potential interpretation of such claims in view of the prior art with reference to the appended claims. Furthermore, the foregoing describes the disclosure in terms of embodiments foreseen by the inventor for which an enabling description was available, notwithstanding that insubstantial modifications of the disclosure, not presently foreseen, may nonetheless represent equivalent modifications thereto.

Claims (8)

1. A river water pollution early warning method is characterized by comprising the following steps:
step 1, arranging sensor nodes in a river channel at equal intervals;
step 2, physical quantity data are collected through sensor nodes;
step 3, preprocessing the physical quantity data to extract key factors;
step 4, constructing a grey prediction model to calculate the predicted value of the key factor;
step 5, constructing an SVR model and predicting the pollution degree of the water area where the sensor node is located through the SVR model and the predicted value;
and 6, when the pollution degree of the water area where the sensor node is located is larger than a pollution threshold value, pushing pollution early warning information to mobile equipment of a manager.
2. The river water pollution early warning method according to claim 1, wherein in the step 1, the sensor nodes are composed of an active temperature sensor, an active conductivity sensor, an active COD sensor, an active pH sensor and an active residual chlorine sensor, and the sensor nodes supply power and transmit physical quantity data to the base station through a communication line; the base station is used for receiving the sensor number of the sensor node and physical quantity data acquired by the sensor node; the equidistant sensor nodes are arranged in the river channel, and the equidistant sensor nodes are arranged in the water of the river channel at intervals of 500 meters.
3. The method as claimed in claim 1, wherein in step 2, the physical quantity data includes temperature, conductivity, chemical oxygen demand, pH value and residual chlorine amount.
4. The river water pollution early warning method according to claim 1, wherein in the step 3, the method for preprocessing the physical quantity data and extracting the key factors comprises the following steps:
step 3.1, respectively recording physical quantity data: the temperature, the conductivity, the chemical oxygen demand, the pH value and the residual chlorine amount are x1，x2，x3，x4，x5Recording the pollution degree of the water area where the sensor node is located as y, wherein the calculation method of the pollution degree y comprises the following steps: y is x1，x2，x3，x4，x5The number of any physical quantity data greater than a preset value;
step 3.2, preprocessing the data by using an information gain method and a Lasso characteristic selection algorithm, wherein the specific processing process is as follows:
step 3.2.1, influence factors x of the pollution degree y and the pollution degree1，x2，x3，x4，x5The data are subjected to standardization preprocessing: a transformation function of
Figure FDA0002614554420000011
Wherein a isijDenotes xiCorresponding physical quantity data of the jth previous sensor node arranged at equal distance in the river where the current sensor node is located, m is the number of sensor types, n is the serial number of the physical quantity, bijConverted physical quantity data;
step 3.2.2, obtaining expected information required by sample classification of the physical quantity data of the current sensor node;
let S be a set of samples of physical quantity data of the current sensor node, i.e. physical quantity data collected by the current sensor node in a given period of time, assume that the class label attribute has m different values, and define m different classes CiLet siIs of the class CiThe number of samples in (a) is,the desired information needed for a given sample classification is given by:
Figure FDA0002614554420000021
wherein p isiIs that any sample belongs to CiProbability of, by siEstimating by the/s;
step 3.2.3, solve Attribute x1Dividing the entropy of the subsets; let SijIs the subset SjClass CiThe number of samples of (a); according to x1Entropy of partitioning subsets into
Figure FDA0002614554420000022
Wherein
Figure FDA0002614554420000023
Act as a weight for the jth subset, and equal to the subset, x1Divided by the total number of samples in s,
Figure FDA0002614554420000024
wherein
Figure FDA0002614554420000025
Is SjThe middle sample belongs to class CiProbability of (2), setting attribute x1There are v different values a1,a2,…av}; by attribute x1Divide S into v values S1,S2,…SvIn which S isjIncluding samples of S at x1Has a value of ajV is 5, and v is the number of samples;
step 3.2.4, obtaining corresponding information gain according to the expected information and the entropy: for in x1The Gain of information to be obtained by the upper branch is Gain (x) obtained by the following formula1)＝I(s1,s2…sm)-E(x1)；
Step 3.2.5, similarly through step 3.2.3 to step 3.2.4, gets the respective attribute (X)2-X5) After the comparison from large to small, more than half of the attributes are selected as the predicted attributes, which are marked as c1,c2,…,cn；
Step 3.2.6, defining Lasso parameter estimation as:
Figure FDA0002614554420000026
wherein beta is a regression coefficient vector, and lambda is a nonnegative regular parameter, and the complexity of the model is controlled; the larger the lambda is, the greater the penalty degree of the linear model with more characteristics is, so as to finally obtain a model with less characteristics,
Figure FDA0002614554420000027
referred to as a penalty term, ρ is the number of attributes, that is, ρ is 5; the parameter lambda can be determined by adopting a cross validation method, and a lambda value with the minimum cross validation error is selected; finally, fitting the model again according to the obtained lambda value,
Figure FDA0002614554420000031
is a key factor.
5. The river water pollution early warning method according to claim 1, wherein in step 4, the method for constructing a gray prediction model to calculate the predicted value of the key factor comprises the following steps:
step 4.1, assume the current sensor node xiIs defined as x(0)＝{x(0)(i)}，xiCorresponding data is x(0)(i) I ═ 1,2, …,5, the gray prediction model was built as follows:
step 4.2, for x(0)Performing a first accumulation to obtain a first accumulation sequence x(1)＝{x(1)(k)，k＝0，1，2，…，m}；
Step 4.3, for x(0)Establishing a first order linear differential equation of the formula
Figure FDA0002614554420000032
Namely the GM (1,1) model, where ∈ is the coefficient of development, μ is the amount of gray effect, and the value of μ is set as the value of the key factor;
step 4.4, solving a differential equation to obtain a prediction model with the formula of
Figure FDA0002614554420000033
x(1)And x*(1)Respectively representing the accumulation sequence and a prediction model, wherein the numbers in the upper right corner represent parameters used in the prediction model, and the parameters are consistent with the definition;
step 4.5, obtaining data x of GM (1,1) model*(1)(k +1) is reduced to x by cumulative subtraction*(0)The gray prediction model of (k +1), i.e., x (0), is formulated as
Figure FDA0002614554420000034
Finally, the predicted value of each key factor for predicting the influence on the pollution degree is recorded as
Figure FDA0002614554420000035
6. The river water pollution early warning method according to claim 1, wherein in the step 5, the method for constructing the SVR model and predicting the pollution degree of the water area where the sensor node is located through the SVR model and the predicted value comprises the following steps:
handle
Figure FDA0002614554420000036
As a training set, among others,
Figure FDA0002614554420000037
training the prediction model, and updating a training sample in real time before predicting the next moment, namely adding the actual pollution degree and the selected principal component data of the previous moment and removing the most original data; for the sample
Figure FDA0002614554420000038
Output f (Out) according to the modeli) And true value
Figure FDA0002614554420000039
The difference between them to calculate the loss, if and only if
Figure FDA00026145544200000310
Time, loss is zero, f (Out)i) The key factor prediction value is output by the SVR model; allow f (Out)i) And
Figure FDA00026145544200000311
the most deviation there between; only when
Figure FDA00026145544200000312
Calculating the loss only when the loss is not calculated; when in use
Figure FDA00026145544200000313
Then, the pollution degree of the water area where the sensor node is located at the next moment is recorded as yn。
7. The river water pollution early warning method according to claim 1, wherein the pollution threshold is an arithmetic mean of current pollution degrees of all sensor nodes in the river.
8. The utility model provides a river course water pollution early warning system which characterized in that, the device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the computer program to operate in the units of:
the node setting unit is used for setting sensor nodes in a river channel at equal distance;
the sensor acquisition unit is used for acquiring physical quantity data through the sensor nodes;
the key factor extraction unit is used for preprocessing the physical quantity data to extract key factors;
the predicted value calculating unit is used for constructing a gray prediction model to calculate the predicted value of the key factor;
the pollution degree prediction unit is used for constructing an SVR model and predicting the pollution degree of the water area where the sensor node is located through the SVR model and the predicted value;
and the alarm pushing unit is used for pushing the pollution early warning information to the mobile equipment of the manager when the pollution degree of the water area where the sensor node is located is greater than the pollution threshold value.
CN202010765837.0A 2020-08-03 2020-08-03 River water pollution early warning method and system Active CN112085926B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010765837.0A CN112085926B (en) 2020-08-03 2020-08-03 River water pollution early warning method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010765837.0A CN112085926B (en) 2020-08-03 2020-08-03 River water pollution early warning method and system

Publications (2)

Publication Number Publication Date
CN112085926A true CN112085926A (en) 2020-12-15
CN112085926B CN112085926B (en) 2022-09-13

Family

ID=73736036

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010765837.0A Active CN112085926B (en) 2020-08-03 2020-08-03 River water pollution early warning method and system

Country Status (1)

Country Link
CN (1) CN112085926B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116819029A (en) * 2023-08-09 2023-09-29 水利部珠江水利委员会水文局 River water pollution monitoring method and system
CN118094453A (en) * 2024-04-28 2024-05-28 山东维平信息安全测评技术有限公司 Environment-friendly information datamation management monitoring system

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6021664A (en) * 1998-01-29 2000-02-08 The United States Of America As Represented By The Secretary Of The Interior Automated groundwater monitoring system and method
CN101424679A (en) * 2008-11-25 2009-05-06 烟台迪特商贸有限公司 Synthesis monitoring system for water quality mutation and monitoring method thereof
US20090150088A1 (en) * 2007-12-06 2009-06-11 Seo Il-Won Method of analyzing behavior of pollutants through prediction of transverse dispersion coefficient using basic hydraulic data in stream
CN105260585A (en) * 2015-07-31 2016-01-20 河海大学 Two-dimensional water quality influence prediction method for sewage draining space with large water yield
CN205879929U (en) * 2015-12-09 2017-01-11 安徽海聚信息科技有限责任公司 Water quality automatic checkout device
CN109270232A (en) * 2018-08-08 2019-01-25 佛山科学技术学院 A kind of monitoring of water pollution big data and method for early warning and device
CN110956310A (en) * 2019-11-14 2020-04-03 佛山科学技术学院 Fish feed feeding amount prediction method and system based on feature selection and support vector
CN111103414A (en) * 2019-12-18 2020-05-05 上一环境科技（金华）有限公司 Intelligent river channel supervision system

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6021664A (en) * 1998-01-29 2000-02-08 The United States Of America As Represented By The Secretary Of The Interior Automated groundwater monitoring system and method
US20090150088A1 (en) * 2007-12-06 2009-06-11 Seo Il-Won Method of analyzing behavior of pollutants through prediction of transverse dispersion coefficient using basic hydraulic data in stream
CN101424679A (en) * 2008-11-25 2009-05-06 烟台迪特商贸有限公司 Synthesis monitoring system for water quality mutation and monitoring method thereof
CN105260585A (en) * 2015-07-31 2016-01-20 河海大学 Two-dimensional water quality influence prediction method for sewage draining space with large water yield
CN205879929U (en) * 2015-12-09 2017-01-11 安徽海聚信息科技有限责任公司 Water quality automatic checkout device
CN109270232A (en) * 2018-08-08 2019-01-25 佛山科学技术学院 A kind of monitoring of water pollution big data and method for early warning and device
CN110956310A (en) * 2019-11-14 2020-04-03 佛山科学技术学院 Fish feed feeding amount prediction method and system based on feature selection and support vector
CN111103414A (en) * 2019-12-18 2020-05-05 上一环境科技（金华）有限公司 Intelligent river channel supervision system

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
包炳钦，陈杰明，陈俊雯，黎晓霞: "《前海湾"两河一涌"水质改善及环境修复对策研究》", 《广东化工》 *
熊和金: "《智能信息处理 第2版》", 31 August 2012, 国防工业出版社 *
赵宇明: "《模式识别》", 31 October 2013, 上海交通大学出版社 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116819029A (en) * 2023-08-09 2023-09-29 水利部珠江水利委员会水文局 River water pollution monitoring method and system
CN116819029B (en) * 2023-08-09 2024-02-09 水利部珠江水利委员会水文局 River water pollution monitoring method and system
CN118094453A (en) * 2024-04-28 2024-05-28 山东维平信息安全测评技术有限公司 Environment-friendly information datamation management monitoring system

Also Published As

Publication number Publication date
CN112085926B (en) 2022-09-13

Similar Documents

Publication Publication Date Title
Engelke et al. Estimation of hüsler–reiss distributions and brown–resnick processes
CN110427654B (en) Landslide prediction model construction method and system based on sensitive state
CN112085926B (en) River water pollution early warning method and system
CN113554175B (en) Knowledge graph construction method and device, readable storage medium and terminal equipment
CN111008726A (en) Class image conversion method in power load prediction
CN110956310A (en) Fish feed feeding amount prediction method and system based on feature selection and support vector
CN114912720A (en) Memory network-based power load prediction method, device, terminal and storage medium
CN114565196B (en) Multi-event trend prejudging method, device, equipment and medium based on government affair hotline
Sundararajan et al. Nonparametric change point detection in multivariate piecewise stationary time series
CN110232130B (en) Metadata management pedigree generation method, apparatus, computer device and storage medium
CN116860562B (en) Method and system for monitoring data quality of data center
Weiß Fully observed INAR (1) processes
CN116306030A (en) New energy prediction dynamic scene generation method considering prediction error and fluctuation distribution
CN113538239B (en) Interpolation method based on space-time autoregressive neural network model
CN114821140A (en) Image clustering method based on Manhattan distance, terminal device and storage medium
CN113139673A (en) Method, device, terminal and storage medium for predicting air quality
CN110728615B (en) Steganalysis method based on sequential hypothesis testing, terminal device and storage medium
CN116936103B (en) User health prediction management method and system based on homodromous network
CN116541252B (en) Computer room fault log data processing method and device
CN117095216B (en) Model training method, system, equipment and medium based on countermeasure generation network
CN117932280B (en) Long sequence data prediction method, long sequence data prediction device, computer equipment, medium and long sequence data prediction product
CN118228069A (en) Method, apparatus, device, medium and program product for predicting electric load
CN109766255B (en) Equipment state analysis method and device, storage medium and electronic equipment
CN116565851A (en) Wind farm power prediction method, device, equipment and medium based on clustering algorithm
Pierrot Forecasting offshore wind energy: Online learning, bounds and missing data

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant