CN116050564A - Algorithm flow based on tailing pond dam displacement prediction - Google Patents

Algorithm flow based on tailing pond dam displacement prediction Download PDF

Info

Publication number
CN116050564A
CN116050564A CN202211114732.4A CN202211114732A CN116050564A CN 116050564 A CN116050564 A CN 116050564A CN 202211114732 A CN202211114732 A CN 202211114732A CN 116050564 A CN116050564 A CN 116050564A
Authority
CN
China
Prior art keywords
tailing pond
index
matrix
data
indexes
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
Application number
CN202211114732.4A
Other languages
Chinese (zh)
Inventor
陈爱武
李世航
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.)
Anhui Changjiang Industrial Big Data Technology Co ltd
Original Assignee
Anhui Changjiang Industrial Big Data Technology Co ltd
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 Anhui Changjiang Industrial Big Data Technology Co ltd filed Critical Anhui Changjiang Industrial Big Data Technology Co ltd
Priority to CN202211114732.4A priority Critical patent/CN116050564A/en
Publication of CN116050564A publication Critical patent/CN116050564A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • 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
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0633—Workflow analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0635—Risk analysis of enterprise or organisation activities
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • 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
    • G06Q50/02—Agriculture; Fishing; Forestry; Mining
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/26—Government or public services
    • G06Q50/265—Personal security, identity or safety
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A10/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE at coastal zones; at river basins
    • Y02A10/40—Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Tourism & Hospitality (AREA)
  • Marketing (AREA)
  • Mathematical Physics (AREA)
  • Development Economics (AREA)
  • General Business, Economics & Management (AREA)
  • Data Mining & Analysis (AREA)
  • Software Systems (AREA)
  • Educational Administration (AREA)
  • General Engineering & Computer Science (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Game Theory and Decision Science (AREA)
  • Computing Systems (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Primary Health Care (AREA)
  • Computational Linguistics (AREA)
  • Computational Mathematics (AREA)
  • Molecular Biology (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Mathematical Analysis (AREA)
  • Computer Security & Cryptography (AREA)
  • Databases & Information Systems (AREA)
  • Mining & Mineral Resources (AREA)
  • Animal Husbandry (AREA)

Abstract

The invention belongs to the technical field of algorithm prediction, in particular to an algorithm flow based on tailing pond dam body displacement prediction, which is used for carrying out advanced prediction on ground surface displacement through rainfall factors of the area where a dam body is positioned and the ground surface displacement condition of the dam body history to obtain a ground surface displacement value in a period of time after the current moment; a large number of sample training is carried out through a cluster analysis model, an influence factor analysis model, a correlation analysis model, an index analysis model, a risk prediction analysis model and the like, so that the system can have the prediction capability of calculating and outputting information such as abnormal monitoring data, influence factor influence data, index correlation data, risk trend data and the like according to actual data of each factor. According to the invention, through an algorithm flow, each factor of the index system, the possible tailing pond accidents and the corresponding preventive and correction measures are subjected to correlation analysis, the safety supervision of the tailing pond is enhanced, and a powerful support is provided for timely and effectively preventing and resolving the risk of the tailing pond.

Description

Algorithm flow based on tailing pond dam displacement prediction
Technical Field
The invention relates to the technical field of algorithm prediction, in particular to an algorithm flow based on tailing pond dam displacement prediction.
Background
The safe production of the tailing pond is regarded as the important thing in the national non-coal mine safe production work, and the dam body breaks, so that serious casualties, property loss and bad social influence are caused. The dam displacement prediction algorithm flow can strengthen safety supervision and monitoring early warning on the tailing pond, and has important application value and practical significance on preventing and resolving important safety risks.
Disclosure of Invention
(one) solving the technical problems
Aiming at the defects of the prior art, the invention provides an algorithm flow based on the displacement prediction of the tailing pond dam body, and the method is used for acquiring important risk early warning information of the tailing pond, comprehensively grasping the risk distribution of enterprises of the tailing pond and realizing dynamic monitoring and early warning of the risk of the important areas; by the method, correlation analysis is carried out on each factor of the index system, the possible tailing pond accidents and the corresponding precaution and improvement measures, internal connection among the factors is established, and the operation result of the tailing pond risk trend is formed.
(II) technical scheme
The invention adopts the following technical scheme for realizing the purposes:
and on the basis of an algorithm flow of dam displacement prediction of the tailing pond, the earth surface displacement is predicted in advance through rainfall factors of the area where the dam is located and the earth surface displacement condition of the dam history, so that the earth surface displacement value in a period of time after the current moment is obtained. A large number of sample training is carried out through a cluster analysis model, an influence factor analysis model, a correlation analysis model, an index analysis model, a risk prediction analysis model and the like, so that the system can have the prediction capability of calculating and outputting information such as abnormal monitoring data, influence factor influence data, index correlation data, risk trend data and the like according to actual data of each factor;
the method specifically comprises the following steps:
s1, evaluating data quality:
before data analysis is carried out, the quality of the data is required to be ensured, and a method for preprocessing the data for processing displacement deformation values is designed by combining machine learning, interpolation and filtering aiming at the possible situations of data loss, abnormality and noise of the monitored data, so that the real-time analysis and processing of the monitored data on a dam body are realized, and the quality of the data is improved;
s2, feature engineering:
the purpose of feature engineering is to extract features from raw data to the maximum for use by algorithms and models;
feature selection has two main purposes:
(1) The feature quantity and dimension are reduced, so that the generalization capability of the model is stronger, and the overfitting is reduced;
(2) Enhancing understanding between features and feature values;
after the data preprocessing is completed, meaningful characteristic indexes are required to be selected to be input into a machine learning algorithm and model for training, because the selection of the tailings pond risk evaluation indexes is an important factor affecting the tailings pond risk evaluation prediction credibility. In general, features are selected from two considerations:
(1) Whether the features diverge: if a feature does not diverge, e.g., the variance is close to 0, that is, the sample has substantially no difference in this feature, this feature is not useful for distinguishing samples;
(2) Correlation of features with targets: this is evident in that the features highly relevant to the target should preferably be selected. In addition to the variance method, other methods described herein are considered from the correlation;
the feature selection method can be further classified into three types according to the form of feature selection:
(1) Filter: the filtering method is used for scoring each characteristic according to divergence or correlation, setting a threshold value or the number of threshold values to be selected, and selecting the characteristic;
(2) Wrapper: packaging, selecting a number of features at a time, or excluding a number of features, based on an objective function (typically a predictive effect score);
(3) An Embedded: the embedding method comprises the steps of training by using certain machine learning algorithms and models to obtain weight coefficients of all the features, and selecting the features from large to small according to the coefficients. Similar to the Filter method, but by training to determine the merits of the features;
in the safety monitoring of the tailing pond, the monitoring types of selective arrangement comprise surface displacement monitoring, internal displacement monitoring, infiltration line monitoring, dry beach monitoring, pond water level monitoring, rainfall monitoring, seepage flow monitoring and the like according to different grades of the tailing pond.
According to different monitoring types, the same type of monitoring data are concentrated and processed:
the surface displacement monitoring data contains three indicators: longitudinal displacement, lateral displacement and vertical displacement;
the internal displacement monitoring data contains three indicators: longitudinal displacement, lateral displacement and vertical displacement;
the immersion line monitoring data includes two indicators: the hydraulic pressure height (osmotic pressure) and the empty pipe distance;
the dry beach monitoring data includes an index: dry beach length, dry beach slope, safe superelevation and beach top elevation;
the reservoir level monitoring contains an indicator: a water level;
the rainfall monitoring comprises an index: rainfall;
the seepage monitoring comprises an index: seepage flow;
among various risk indexes, the characteristic indexes of the tailing pond have complex connection attributes besides the positioning characteristics in space; the action relation is reflected in the self-correlation characteristic of the tailing pond characteristic index and the variable area unit problem, scale and edge effect associated with the self-correlation characteristic; after a feature model is constructed through feature indexes, spatial autocorrelation inspection is carried out through a Geary coefficient, so that mutual independence among the space indexes is ensured; the correlation coefficient between the indexes takes a value between [0,2] and is expressed by gamma, wherein gamma <1 represents positive correlation, gamma >1 represents negative correlation, and gamma=1 represents uncorrelation; carrying out correlation calculation among the indexes, recognizing the risk indexes with larger correlation as the same factor, discarding the risk indexes, and finally establishing a multi-level hierarchical structure model of a tailing pond risk assessment index system as shown in the figure; the model hierarchical structure is divided into four stages of P, U, R and K; wherein the P level is a target tailing pond; the U level is the second level, and six main factors affecting the safety of the tailing pond, such as underground displacement, pond water level and the like, are selected; the R groups respectively select factors influencing three different spatial positions of the underground displacement and the underground inclination angle; the K level is a fourth level, and the sensors representing different spatial positions are used as fourth level factors, so that the influence of the spatial information on the tailing pond is increased;
s3, model building and training:
the tailing pond is provided with n safety indexes A1, A2, … and An, the importance degree of the safety indexes can be expressed as An importance vector w= (A1, A2, … and An) T, and if the safety indexes are compared with each other, the ratio of the safety indexes can form An n by n matrix A, namely:
Figure SMS_1
/>
if (a-nI) w=0 is obtained by right-multiplication a transformation of the importance vector, it is known from matrix theory that: w is a feature vector, and n is a feature value; w can be judged by a judging person according to the knowledge of tailings, so that the matrix A is known, and the judging matrix is marked as
Figure SMS_2
Because the objective ratio of each factor deviates from the subjective judgment, the characteristic value and the characteristic vector also existDeviation, therefore, also need to be measured +.>
Figure SMS_3
Is the consistency of (3);
in order to obtain a scientific and quantitative judgment matrix by pairwise comparison between factors, a 1-9 proportion scale method is adopted according to the limit capability of people for distinguishing information grades, a judgment matrix is established for pairwise comparison of tailing pond indexes, the ratio of the influence degree of factors ai and aj relative to membership targets is expressed by aij, the scale aij= {2,4,6,8,1/2,1/4,1/6,1/8} is generally used as the scale of degree comparison by numbers 1-9 and the inverse thereof, and the importance grade is expressed as the scale and meaning of the judgment matrix in the table between aij= {1,3,5,7,9,1/3,1/5,1/7,1/9}, as follows:
Figure SMS_4
the tailing pond and the secondary indexes form a fuzzy judgment matrix, which is as follows:
Figure SMS_5
the fuzzy judgment matrix is composed of the upper part, the middle part and the lower part of the secondary index underground displacement and the tertiary index underground displacement:
Figure SMS_6
the fuzzy judgment matrix formed by the upper part, the middle part and the lower part of the secondary index underground inclination angle and the tertiary index underground inclination angle is as follows:
Figure SMS_7
the fuzzy judgment matrix formed between the four-level space index and the previous-level index is as follows:
Figure SMS_8
the matrix eigenvector W is obtained by applying a square root method to approximate calculation and solving, namely:
Figure SMS_9
w＝[w 1 ，w 2 ，…，w n ] T
wherein aij is the ith row and jth column element of the n-dimensional evaluation matrix A;
when the relative weights of the factors of each layer are obtained, carrying out combination weight calculation; a four-layer laminated model formed by P, U, R, K is arranged; the relative weight of the target tailing pond to the U layer is as follows:
Figure SMS_10
the relative weights of the indexes Ui of the U layer to the n indexes of the R layer are as follows:
Figure SMS_11
then the relative weights of the bottom-most index, i.e. the spatial index, for the target tailings pond are obtained by weight combination, i.e.:
Figure SMS_12
and->
Figure SMS_13
In order to prevent the model from being fitted, regularization is added in an implicit layer, and the model updates a network weight matrix and a bias vector by using a gradient descent algorithm;
firstly, initializing all parameters of a model; then Bi-LSTM extracts the forward and backward information of the time sequence as the input of the full connection layer, calculates the output of the forward hidden layer and the backward hidden layer respectively, and obtains the weight training set output Y through splicing and linear operation ～ Y ～ ；
Secondly, updating model parameters by using a loss function to reversely propagate to obtain an optimal solution;
and finally, testing the trained model by using the test set data, comparing the model prediction result with the real data, and evaluating the prediction accuracy of the model by using the prediction error.
S4, model verification: and (3) carrying out iterative upgrading on the model and the data, based on the established model, establishing a judgment matrix for each sensor representing different spatial information of indexes on the basis of the original established judgment matrix, verifying the consistency of the matrix by using a method root method, ensuring the rigor and scientificity of the judgment matrix by utilizing the mahalanobis distance, evaluating the model layer by layer from bottom to top until the judgment on the current safety situation of the tailing pond is completed, analyzing the hidden danger of dam break of the tailing pond by combining the spatial information, and predicting the future risk in real time.
(III) beneficial effects
Compared with the prior art, the invention provides an algorithm flow based on tailing pond dam body displacement prediction, which has the following beneficial effects:
according to the invention, a tailing pond dam displacement prediction method is realized, correlation analysis is carried out on each factor of an index system, the possible tailing pond accidents and corresponding preventive modification measures through an algorithm flow, internal connection between the factors is established, and an operation result of a tailing pond risk trend is formed; the safety supervision of the tailing pond is enhanced, and a powerful support is provided for timely and effectively preventing and resolving the risk of the tailing pond.
Drawings
FIG. 1 is a schematic diagram of a hierarchical structure model according to the present invention;
FIG. 2 is a schematic diagram of a model training process according to the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Examples
As shown in fig. 1 and 2, one embodiment of the present invention proposes: and on the basis of an algorithm flow of dam displacement prediction of the tailing pond, the earth surface displacement is predicted in advance through rainfall factors of the area where the dam is located and the earth surface displacement condition of the dam history, so that the earth surface displacement value in a period of time after the current moment is obtained. A large number of sample training is carried out through a cluster analysis model, an influence factor analysis model, a correlation analysis model, an index analysis model, a risk prediction analysis model and the like, so that the system can have the prediction capability of calculating and outputting information such as abnormal monitoring data, influence factor influence data, index correlation data, risk trend data and the like according to actual data of each factor;
the method specifically comprises the following steps:
s1, evaluating data quality:
before data analysis is carried out, the quality of the data is required to be ensured, and a method for preprocessing the data for processing displacement deformation values is designed by combining machine learning, interpolation and filtering aiming at the possible situations of data loss, abnormality and noise of the monitored data, so that the real-time analysis and processing of the monitored data on a dam body are realized, and the quality of the data is improved;
s2, feature engineering:
the purpose of feature engineering is to extract features from raw data to the maximum for use by algorithms and models;
feature selection has two main purposes:
(1) The feature quantity and dimension are reduced, so that the generalization capability of the model is stronger, and the overfitting is reduced;
(2) Enhancing understanding between features and feature values;
after the data preprocessing is completed, meaningful characteristic indexes are required to be selected to be input into a machine learning algorithm and model for training, because the selection of the tailings pond risk evaluation indexes is an important factor affecting the tailings pond risk evaluation prediction credibility. In general, features are selected from two considerations:
(1) Whether the features diverge: if a feature does not diverge, e.g., the variance is close to 0, that is, the sample has substantially no difference in this feature, this feature is not useful for distinguishing samples;
(2) Correlation of features with targets: this is evident in that the features highly relevant to the target should preferably be selected. In addition to the variance method, other methods described herein are considered from the correlation;
the feature selection method can be further classified into three types according to the form of feature selection:
(1) Filter: the filtering method is used for scoring each characteristic according to divergence or correlation, setting a threshold value or the number of threshold values to be selected, and selecting the characteristic;
(2) Wrapper: packaging, selecting a number of features at a time, or excluding a number of features, based on an objective function (typically a predictive effect score);
(3) An Embedded: the embedding method comprises the steps of training by using certain machine learning algorithms and models to obtain weight coefficients of all the features, and selecting the features from large to small according to the coefficients. Similar to the Filter method, but by training to determine the merits of the features;
in the safety monitoring of the tailing pond, the monitoring types of selective arrangement comprise surface displacement monitoring, internal displacement monitoring, infiltration line monitoring, dry beach monitoring, pond water level monitoring, rainfall monitoring, seepage flow monitoring and the like according to different grades of the tailing pond.
According to different monitoring types, the same type of monitoring data are concentrated and processed:
the surface displacement monitoring data contains three indicators: longitudinal displacement, lateral displacement and vertical displacement;
the internal displacement monitoring data contains three indicators: longitudinal displacement, lateral displacement and vertical displacement;
the immersion line monitoring data includes two indicators: the hydraulic pressure height (osmotic pressure) and the empty pipe distance;
the dry beach monitoring data includes an index: dry beach length, dry beach slope, safe superelevation and beach top elevation;
the reservoir level monitoring contains an indicator: a water level;
the rainfall monitoring comprises an index: rainfall;
the seepage monitoring comprises an index: seepage flow;
among various risk indexes, the characteristic indexes of the tailing pond have complex connection attributes besides the positioning characteristics in space; the action relation is reflected in the self-correlation characteristic of the tailing pond characteristic index and the variable area unit problem, scale and edge effect associated with the self-correlation characteristic; after a feature model is constructed through feature indexes, spatial autocorrelation inspection is carried out through a Geary coefficient, so that mutual independence among the space indexes is ensured; the correlation coefficient between the indexes takes a value between [0,2] and is expressed by gamma, wherein gamma <1 represents positive correlation, gamma >1 represents negative correlation, and gamma=1 represents uncorrelation; carrying out correlation calculation among the indexes, recognizing the risk indexes with larger correlation as the same factor, discarding the risk indexes, and finally establishing a multi-level hierarchical structure model of a tailing pond risk assessment index system as shown in the figure; the model hierarchical structure is divided into four stages of P, U, R and K; wherein the P level is a target tailing pond; the U level is the second level, and six main factors affecting the safety of the tailing pond, such as underground displacement, pond water level and the like, are selected; the R groups respectively select factors influencing three different spatial positions of the underground displacement and the underground inclination angle; the K level is a fourth level, and the sensors representing different spatial positions are used as fourth level factors, so that the influence of the spatial information on the tailing pond is increased;
s3, model building and training:
the tailing pond is provided with n safety indexes A1, A2, … and An, the importance degree of the safety indexes can be expressed as An importance vector w= (A1, A2, … and An) T, and if the safety indexes are compared with each other, the ratio of the safety indexes can form An n by n matrix A, namely:
Figure SMS_14
if the important vector is multiplied by AThe transformation can be obtained (a-nI) w=0, as known from matrix theory: w is a feature vector, and n is a feature value; w can be judged by a judging person according to the knowledge of tailings, so that the matrix A is known, and the judging matrix is marked as
Figure SMS_15
Because the objective ratio of each factor deviates from the subjective judgment, the characteristic value and the characteristic vector also deviate, and therefore, the +.>
Figure SMS_16
Is the consistency of (3);
in order to obtain a scientific and quantitative judgment matrix by pairwise comparison between factors, a 1-9 proportion scale method is adopted according to the limit capability of people for distinguishing information grades, a judgment matrix is established for pairwise comparison of tailing pond indexes, the ratio of the influence degree of factors ai and aj relative to membership targets is expressed by aij, the scale aij= {2,4,6,8,1/2,1/4,1/6,1/8} is generally used as the scale of degree comparison by numbers 1-9 and the inverse thereof, and the importance grade is expressed as the scale and meaning of the judgment matrix in the table between aij= {1,3,5,7,9,1/3,1/5,1/7,1/9}, as follows:
Figure SMS_17
the tailing pond and the secondary indexes form a fuzzy judgment matrix, which is as follows:
Figure SMS_18
the fuzzy judgment matrix is composed of the upper part, the middle part and the lower part of the secondary index underground displacement and the tertiary index underground displacement:
Figure SMS_19
/>
the fuzzy judgment matrix formed by the upper part, the middle part and the lower part of the secondary index underground inclination angle and the tertiary index underground inclination angle is as follows:
Figure SMS_20
the fuzzy judgment matrix formed between the four-level space index and the previous-level index is as follows:
Figure SMS_21
the matrix eigenvector W is obtained by applying a square root method to approximate calculation and solving, namely:
Figure SMS_22
w＝[w 1 ，w 2 ，…，w n ] T
wherein aij is the ith row and jth column element of the n-dimensional evaluation matrix A;
when the relative weights of the factors of each layer are obtained, carrying out combination weight calculation; a four-layer laminated model formed by P, U, R, K is arranged; the relative weight of the target tailing pond to the U layer is as follows:
Figure SMS_23
the relative weights of the indexes Ui of the U layer to the n indexes of the R layer are as follows:
Figure SMS_24
then the relative weights of the bottom-most index, i.e. the spatial index, for the target tailings pond are obtained by weight combination, i.e.:
Figure SMS_25
and->
Figure SMS_26
In order to prevent the model from being fitted, regularization is added in an implicit layer, and the model updates a network weight matrix and a bias vector by using a gradient descent algorithm;
firstly, initializing all parameters of a model; then Bi-LSTM extracts the forward and backward information of the time sequence as the input of the full connection layer, calculates the output of the forward hidden layer and the backward hidden layer respectively, and obtains the weight training set output Y through splicing and linear operation ～ Y ～ ；
Secondly, updating model parameters by using a loss function to reversely propagate to obtain an optimal solution;
and finally, testing the trained model by using the test set data, comparing the model prediction result with the real data, and evaluating the prediction accuracy of the model by using the prediction error.
S4, model verification:
the model verification and optimization process is an iterative upgrade process for the model and the data, but the factors affecting the on-line effect of the model are required to be clear, from basic data to constructed features, from algorithm selection to experimental strategy, and from the given sequencing result to front-end position display, the influence is possible; in the whole process of applying the algorithm model, each action or each modification can influence the effect expression of the model, and when problems occur, the problems are often positioned and interpreted from multiple aspects;
(1) Consistency test:
according to matrix theory, when making a decision by using a fuzzy evaluation matrix, the transmissibility of the pairwise comparison sequence among elements is maintained, and the acceptability of the overall deviation is maintained, so that the concept of calculating a satisfactory consistency matrix is defined, and a matrix consistency check formula is defined as follows:
Figure SMS_27
wherein: RI is an average random consistency index, the larger the dimension of the judgment matrix is, the worse the consistency of the judgment matrix is, and the random consistency index of the matrix can be obtained through the following table; (AW) i is the i-th element of the vector; a is a judgment matrix; CR is a random consistency ratio; generally, when CR is less than 0.10, the degree of inconsistency of a is considered to be within the allowable range, there is satisfactory consistency, the normalized eigenvector thereof can be used as the weight vector, otherwise, the judgment matrix is to be reconstructed, aij is adjusted, and the matrix random consistency index is as follows:
Figure SMS_28
the difference between all peer evaluation weight vectors in a certain tailing pond is measured by using the mahalanobis distance, so that subjectivity caused by determining a judgment matrix by an expert is reduced, and the most reasonable vector Wk is established as the final evaluation weight for the judgment matrix given by an expert by the obtained feature vectors Wi-Wj; the covariance matrix is denoted as S and the mahalanobis distance between the vectors Wi and Wj is defined as:
Figure SMS_29
Figure SMS_30
each element in the covariance matrix is represented by Cov (X, Y), which is a mathematical expectation, between the vector elements, cov (X, Y) =e { [ X-E (X) ] [ Y-E (Y) ]};
the fuzzy mathematics are used for carrying out overall evaluation on things or objects limited by various factors, so that the fuzzy and difficult-to-quantify problems can be well solved, and the fuzzy and difficult-to-quantify fuzzy evaluation method is suitable for solving various nondeterminacy problems; in order to reduce subjective assessment of human factors on a tailings pond risk assessment index, a fuzzy transformation principle and a maximum membership principle are applied, and a judgment set V= { V1, V2, …, vn } is made on the overall risk condition correlation degree of the tailings pond by taking the factor set U= { U1, U2, …, un } into consideration, and the judgment of a single factor ui (i=1, 2, …, n) is carried out to obtain fuzzy sets (ri 1, ri2, …, rin) on V, so that the fuzzy mapping from U to V is realized;
if any element x in the domain U is provided with a number F (x) E [0,1] corresponding to the element x, the element x is called as a fuzzy set on the domain, F (x) is called as membership degree [1,9] of x to F, and the membership degree of the x to F is represented by a membership function F (x) with a value in a range [0,1], so that the index score of the tailing pond is established by establishing a membership function instead of expert scoring;
the number of the risk evaluation indexes of the tailing pond is large, the fuzzy characteristics of the indexes are different, the evaluation values of indexes such as the pond water level, the surface displacement, the underground inclination angle, the underground water level and the rainfall are increased along with the reduction of the attribute values, and the evaluation value of the dry beach length of the indexes is reduced along with the reduction of the attribute values; the data value set of the tailing pond index in one year contains a large number of numerical values, the sensor sampling frequency is high, and the index characteristic enables the values in the set to be equivalent to continuous, so that a general piecewise linear membership function with simple and efficient calculation is adopted, namely, the formula is that: x (t) is defined as data acquired by a sensor at time t; s (t) is a t moment index evaluation score; a. b represents the minimum and maximum values of the index data quantity a=minx (t) and b=maxx (t) in the tailing pond collecting time range respectively;
(2) And (3) comprehensive evaluation:
in a complex system, as the factors are more and the layers are divided among the factors, in order to better compare the priority order among the transactions in the system, a meaningful judgment result is obtained, and a multi-level comprehensive judgment is needed to be carried out, so that the security risk level is given in real time;
for example: 500 pieces of data acquired by a 5-class risk index sensor of a certain tailing pond are used as test samples, a data preprocessing technology is used for filling in vacant values of a small amount of missing data, smooth noise data is carried out, a judgment matrix is established, the relative importance among all levels of factors of the tailing pond is determined, a U, R, K-level nine-level scale method is obtained, a plurality of experts measure and judge disaster causing factors of each layer and write out all judgment matrix types, the most suitable feature vectors are selected by utilizing the mahalanobis distance to serve as weight vectors, and random consistency ratios CRu, CRRi, CRRj, CRk are respectively obtained according to random consistency indexes of corresponding order matrixes and judgment matrixes corresponding to final weight vectors, so that matrix consistency requirements are verified to be met;
(3) Space-time information of tailing pond:
predicting the development trend of a future tailing pond according to the judgment of the risk situation of the current tailing pond, selecting time, space position and earth surface displacement evaluation scores to establish a tailing pond index evaluation three-dimensional graph, selecting time and earth surface displacement comprehensive evaluation scores to establish an earth surface displacement historical safety condition graph, and observing the safety condition of a single sensor along with the time;
example (1): the safety conditions of the surface displacement indexes reflected by a certain displacement sensor on the tailing pond are larger in fluctuation condition along with time and inconsistent with each other, so that the movement conditions of the tailing pond at different surface positions are different, but if the water level of a pond area is higher at the moment, the burial depth of a dam body infiltration line is possibly raised, so that the safety conditions of the tailing pond cannot be judged according to the rapid movement of a single index or the relative stability of the single index;
example (2): the mine tailing pond can increase the potential energy of the earth surface to enlarge the dam break risk due to large-area snowfall, and when weather forecast reports that the area where the mine tailing pond area is located has large-area snowfall in a certain period of time in the future, the earth surface displacement safety condition can be reduced, and the earth surface displacement comprehensive evaluation score can be reduced;
example (3): the influence of the safety conditions of the tailings pond reflected by the upper, middle and lower spatial positions of the underground inclined angle on the overall safety conditions of the tailings pond at a specific moment is achieved, the safety condition scores of the different positions of the inclined angle are compared with the overall scores of the tailings pond to find out the tailings pond areas deformed due to various unstable factors, so that the tailings pond management personnel can monitor the tailings pond areas in a key way, and dam break accidents are prevented from happening;
based on the established model, on the basis of the original established judgment matrix, the judgment matrix is also established for each sensor representing different spatial information of indexes, the consistency of the matrix is verified by a method, the strictness and scientificity of the judgment matrix are ensured by utilizing the mahalanobis distance, the judgment on the safety status quo of the tailing pond is evaluated layer by layer from bottom to top until the judgment on the safety status quo of the tailing pond is completed, and the hidden danger of the tailing pond in dam break is analyzed by combining the spatial information, and real-time prediction is carried out on future risks.
Finally, it should be noted that: the foregoing description is only a preferred embodiment of the present invention, and the present invention is not limited thereto, but it is to be understood that modifications and equivalents of some of the technical features described in the foregoing embodiments may be made by those skilled in the art, although the present invention has been described in detail with reference to the foregoing embodiments. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (6)

1. The algorithm flow based on the tailing pond dam displacement prediction is characterized by comprising the following steps of:
s1, evaluating data quality: before data analysis, preprocessing is carried out for the possible situations of data loss, abnormality and noise of the monitoring data, so that the monitoring data on the dam body are analyzed and processed in real time, and the quality of the data is improved;
s2, feature engineering: extracting features from the raw data to the maximum for use by algorithms and models;
s3, model building and training: firstly, initializing all parameters of a model; then Bi-LSTM extracts forward and backward information of the time sequence as input of a full connection layer, calculates output of the forward hidden layer and the backward hidden layer respectively, and obtains weight training set output Y-Y through splicing and linear operation; secondly, updating model parameters by using a loss function to reversely propagate to obtain an optimal solution; finally, testing the trained model by using the test set data, comparing the model prediction result with the real data, and evaluating the prediction accuracy of the model by using the prediction error;
s4, model verification: and (3) carrying out iterative upgrading on the model and the data, based on the established model, establishing a judgment matrix for each sensor representing different spatial information of indexes on the basis of the original established judgment matrix, verifying the consistency of the matrix by using a method root method, ensuring the rigor and scientificity of the judgment matrix by utilizing the mahalanobis distance, evaluating the model layer by layer from bottom to top until the judgment on the current safety situation of the tailing pond is completed, analyzing the hidden danger of dam break of the tailing pond by combining the spatial information, and predicting the future risk in real time.
2. The algorithm flow based on tailing pond dam displacement prediction according to claim 1, wherein the algorithm flow is characterized in that: the monitoring data includes:
the surface displacement monitoring data contains three indicators: longitudinal displacement, lateral displacement and vertical displacement;
the internal displacement monitoring data contains three indicators: longitudinal displacement, lateral displacement and vertical displacement;
the immersion line monitoring data includes two indicators: the hydraulic pressure height (osmotic pressure) and the empty pipe distance;
the dry beach monitoring data includes an index: dry beach length, dry beach slope, safe superelevation and beach top elevation;
the reservoir level monitoring contains an indicator: a water level;
the rainfall monitoring comprises an index: rainfall;
the seepage monitoring comprises an index: seepage flow.
3. The algorithm flow based on tailing pond dam displacement prediction according to claim 1, wherein the algorithm flow is characterized in that: the feature selection in the feature engineering comprises the following three modes:
(1) Filter: the filtering method is used for scoring each characteristic according to divergence or correlation, setting a threshold value or the number of threshold values to be selected, and selecting the characteristic;
(2) Wrapper: packaging, selecting a number of features at a time, or excluding a number of features, based on an objective function (typically a predictive effect score);
(3) An Embedded: the embedding method comprises the steps of training by using certain machine learning algorithms and models to obtain weight coefficients of all the features, selecting the features from large to small according to the coefficients, and determining the advantages and disadvantages of the features through training.
4. The algorithm flow based on tailing pond dam displacement prediction according to claim 1, wherein the algorithm flow is characterized in that: the training and practice process of the model is as follows:
the tailing pond is provided with n safety indexes A1, A2, … and An, the importance degree of the safety indexes can be expressed as An importance vector w= (A1, A2, … and An) T, and if the safety indexes are compared with each other, the ratio of the safety indexes can form An n by n matrix A, namely:
Figure FDA0003845036010000021
/>
if (a-nI) w=0 is obtained by right-multiplication a transformation of the importance vector, it is known from matrix theory that: w is a feature vector, n is a feature value, and W can be judged by a judging person according to the knowledge of tailings, so that the matrix A is known, and the judging matrix is marked as
Figure FDA0003845036010000034
Because the objective ratio of each factor deviates from the subjective judgment, the characteristic value and the characteristic vector also deviate, and therefore, the +.>
Figure FDA0003845036010000035
Is the consistency of (3);
in order to obtain a scientific quantitative judgment matrix by pairwise comparison among factors, a 1-9 proportion scale method is adopted according to the limit capability of people for distinguishing information grades, a judgment matrix is established for pairwise comparison of tailing pond indexes, the ratio of the influence degree of factors ai and aj relative to membership targets is expressed by aij, the scale aij= {2,4,6,8,1/2,1/4,1/6,1/8} is generally used as the scale of degree comparison by numbers 1-9 and the inverse thereof, and the importance grade is expressed as the scale and meaning of the judgment matrix in a table between aij= {1,3,5,7,9,1/3,1/5,1/7,1/9 }:
Figure FDA0003845036010000031
the tailing pond and the secondary indexes form a fuzzy judgment matrix, which is as follows:
Figure FDA0003845036010000032
the fuzzy judgment matrix is composed of the upper part, the middle part and the lower part of the secondary index underground displacement and the tertiary index underground displacement:
Figure FDA0003845036010000033
the fuzzy judgment matrix formed by the upper part, the middle part and the lower part of the secondary index underground inclination angle and the tertiary index underground inclination angle is as follows:
Figure FDA0003845036010000041
the fuzzy judgment matrix formed between the four-level space index and the previous-level index is as follows:
Figure FDA0003845036010000042
the matrix eigenvector W is obtained by applying a square root method to approximate calculation and solving, namely:
Figure FDA0003845036010000043
W＝[W 1 ，W 2 ，…，W n ] T
wherein aij is the ith row and jth column element of the n-dimensional evaluation matrix A;
after the relative weights of the factors of each layer are obtained, carrying out combination weight calculation, and providing a four-layer hierarchical model composed of P, U, R, K, wherein the relative weights of the target tailing pond to the U layers are as follows:
Figure FDA0003845036010000044
the relative weights of the indexes Ui of the U layer to the n indexes of the R layer are as follows:
Figure FDA0003845036010000045
then the relative weights of the bottom-most index, i.e. the spatial index, for the target tailings pond are obtained by weight combination, i.e.:
Figure FDA0003845036010000046
5. the algorithm flow based on tailing pond dam displacement prediction according to claim 1, wherein the algorithm flow is characterized in that: the model verification includes consistency verification: according to matrix theory, when making a decision by using a fuzzy evaluation matrix, the transmissibility of the pairwise comparison sequence among elements is maintained, and the acceptability of the overall deviation is maintained, so that the concept of calculating a satisfactory consistency matrix is defined, and a matrix consistency check formula is defined as follows:
Figure FDA0003845036010000051
wherein: RI is an average random consistency index, the larger the dimension of the judgment matrix is, the worse the consistency of the judgment matrix is, and the random consistency index of the matrix can be obtained through the following table; (AW) i is the i-th element of the vector; a is a judgment matrix; CR is a random consistency ratio; generally, when CR is less than 0.10, the degree of inconsistency of a is considered to be within the allowable range, and there is satisfactory consistency, and the normalized eigenvector thereof can be used as the weight vector, otherwise, the judgment matrix is reconfigured to adjust aij.
6. The algorithm flow based on tailing pond dam displacement prediction according to claim 1, wherein the algorithm flow is characterized in that: the difference between all peer evaluation weight vectors in a certain tailing pond is measured by using the mahalanobis distance, so that subjectivity caused by determining a judgment matrix by an expert is reduced, and the most reasonable vector Wk is established as the final evaluation weight for the judgment matrix given by an expert by the obtained feature vectors Wi-Wj;
the covariance matrix is denoted as S and the mahalanobis distance between the vectors Wi and Wj is defined as:
Figure FDA0003845036010000052
Figure FDA0003845036010000053
each element in the covariance matrix is represented by Cov (X, Y), which is a mathematical expectation, between the vector elements, cov (X, Y) =e { [ X-E (X) ] [ Y-E (Y) ]};
the fuzzy transformation principle and the maximum membership principle are applied, and the factor set U= { U1, U2, …, un } is considered to make a judgment set V= { V1, V2, …, vn } on the overall risk condition correlation degree of the tailing pond, and the fuzzy set (ri 1, ri2, …, rin) on V is obtained by judging the single factor ui (i=1, 2, …, n), so that the fuzzy mapping from U to V is realized;
if any element x in the domain U is provided with a number F (x) E [0,1] corresponding to the element x, the element x is called as a fuzzy set on the domain, F (x) is called as membership degree [1,9] of x to F, and the membership degree of the x to F is represented by a membership function F (x) with a value in a range [0,1], so that the index score of the tailing pond is established by establishing a membership function instead of expert scoring;
the number of the tailing pond risk evaluation indexes is large, the fuzzy characteristics of the indexes are different, the evaluation values of indexes such as pond water level, surface displacement, underground inclination angle, underground water level and rainfall are increased along with the reduction of the attribute values, meanwhile, the evaluation value of the length of the dry beach of the indexes is reduced along with the reduction of the attribute values, a data value set of the tailing pond indexes in one year contains a large number of numerical values, the sampling frequency of a sensor is high, and the index characteristics enable the values in the set to be equivalent to continuous, so that a general piecewise linear membership function with simple and efficient calculation is adopted, namely, in the formula: x (t) is defined as data acquired by a sensor at time t; s (t) is a t moment index evaluation score; a. b represents the minimum and maximum values of the index data quantity a=minx (t) and b=maxx (t) in the tailing pond collecting time range respectively.
CN202211114732.4A 2022-09-14 2022-09-14 Algorithm flow based on tailing pond dam displacement prediction Pending CN116050564A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211114732.4A CN116050564A (en) 2022-09-14 2022-09-14 Algorithm flow based on tailing pond dam displacement prediction

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211114732.4A CN116050564A (en) 2022-09-14 2022-09-14 Algorithm flow based on tailing pond dam displacement prediction

Publications (1)

Publication Number Publication Date
CN116050564A true CN116050564A (en) 2023-05-02

Family

ID=86126530

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211114732.4A Pending CN116050564A (en) 2022-09-14 2022-09-14 Algorithm flow based on tailing pond dam displacement prediction

Country Status (1)

Country Link
CN (1) CN116050564A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117009831B (en) * 2023-10-07 2023-12-08 山东世纪阳光科技有限公司 Fine chemical accident risk prediction assessment method

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117009831B (en) * 2023-10-07 2023-12-08 山东世纪阳光科技有限公司 Fine chemical accident risk prediction assessment method

Similar Documents

Publication Publication Date Title
CN107943880B (en) Geological disaster susceptibility improvement and evaluation method based on analytic hierarchy process
CN113807570B (en) XGBoost-based reservoir dam risk level assessment method and system
CN111415492A (en) Slope landslide early warning method and system based on fuzzy comprehensive evaluation algorithm
CN114818076B (en) Machine learning-based fault closed hydrocarbon column height evaluation method
CN115063020B (en) Multi-dimensional safety scheduling device and method for cascade hydropower station based on risk monitoring fusion
CN112257942B (en) Stress corrosion cracking prediction method and system
CN111797364A (en) Landslide multilevel safety evaluation method based on composite cloud model
CN116050564A (en) Algorithm flow based on tailing pond dam displacement prediction
CN111597751A (en) Crude oil film absolute thickness inversion method based on self-expansion depth confidence network
CN116777223A (en) Urban underground pipe network safety comprehensive risk assessment method and system
CN116992522A (en) Deep foundation pit support structure deformation prediction method, device, equipment and storage medium
CN114580940A (en) Grouting effect fuzzy comprehensive evaluation method based on grey correlation degree analysis method
CN112001600B (en) Water leakage risk monitoring method based on SVM and DS theory
CN117574321A (en) Dam safety assessment method based on mass monitoring and inspection information fusion model
CN116403093B (en) Multi-source remote sensing interpretation method and device for potential landslide along railway
CN115660221B (en) Oil and gas reservoir economic recoverable reserve assessment method and system based on hybrid neural network
CN115019476B (en) Landslide space-time information monitoring and early warning method based on multi-source information fusion
CN111582634A (en) Multi-factor safety grading method and system for underground large-space construction
CN115563874A (en) Tunnel health degree analysis method and system based on big data
CN106372352B (en) Landslide area detection device and method
Kanoun et al. An enhanced deep learning approach for tectonic fault and fracture extraction in very high resolution optical images
CN113837674A (en) Dyke dangerous case sensitivity analysis method based on weight model
CN113888384B (en) Landslide hazard spatial distribution characteristic analysis method and system
CN112906793B (en) Monitoring data repairing method and system for bridge health monitoring system
CN117592856A (en) Algorithm flow based on petrochemical enterprise output prediction

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