CN117557100A - Deep foundation pit excavation induced adjacent building risk assessment method based on deep learning - Google Patents
Deep foundation pit excavation induced adjacent building risk assessment method based on deep learning Download PDFInfo
- Publication number
- CN117557100A CN117557100A CN202311629365.6A CN202311629365A CN117557100A CN 117557100 A CN117557100 A CN 117557100A CN 202311629365 A CN202311629365 A CN 202311629365A CN 117557100 A CN117557100 A CN 117557100A
- Authority
- CN
- China
- Prior art keywords
- foundation pit
- pit excavation
- deep
- probability
- deep learning
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000009412 basement excavation Methods 0.000 title claims abstract description 170
- 238000000034 method Methods 0.000 title claims abstract description 65
- 238000013135 deep learning Methods 0.000 title claims abstract description 54
- 238000012502 risk assessment Methods 0.000 title claims abstract description 37
- 238000012544 monitoring process Methods 0.000 claims abstract description 104
- 238000006073 displacement reaction Methods 0.000 claims abstract description 55
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 6
- 238000012216 screening Methods 0.000 claims description 45
- 238000009826 distribution Methods 0.000 claims description 26
- 238000010219 correlation analysis Methods 0.000 claims description 12
- 238000013136 deep learning model Methods 0.000 claims description 12
- 230000003203 everyday effect Effects 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 10
- 238000004088 simulation Methods 0.000 claims description 8
- 238000013528 artificial neural network Methods 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 6
- 238000000342 Monte Carlo simulation Methods 0.000 claims description 5
- 238000005070 sampling Methods 0.000 claims description 4
- 238000010276 construction Methods 0.000 abstract description 26
- 238000005516 engineering process Methods 0.000 abstract description 2
- 230000015654 memory Effects 0.000 description 11
- 230000008569 process Effects 0.000 description 9
- 230000002829 reductive effect Effects 0.000 description 8
- 238000012549 training Methods 0.000 description 6
- 238000004590 computer program Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 238000011835 investigation Methods 0.000 description 3
- 239000000463 material Substances 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 230000001186 cumulative effect Effects 0.000 description 2
- 125000004122 cyclic group Chemical group 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000004836 empirical method Methods 0.000 description 2
- 230000007787 long-term memory Effects 0.000 description 2
- 230000008447 perception Effects 0.000 description 2
- 230000001681 protective effect Effects 0.000 description 2
- 230000006403 short-term memory Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000012854 evaluation process Methods 0.000 description 1
- 230000000670 limiting effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 230000005641 tunneling Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
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
- G06Q10/06—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
- 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
- G06F18/211—Selection of the most significant subset of features
- G06F18/2113—Selection of the most significant subset of features by ranking or filtering the set of features, e.g. using a measure of variance or of feature cross-correlation
-
- 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
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- 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
- G06Q50/08—Construction
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Entrepreneurship & Innovation (AREA)
- Tourism & Hospitality (AREA)
- Evolutionary Biology (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Probability & Statistics with Applications (AREA)
- Biomedical Technology (AREA)
- Primary Health Care (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a deep foundation pit excavation induced adjacent building risk assessment method based on deep learning, which comprises the following steps: collecting an original data set, and determining foundation pit excavation monitoring parameters for deep learning from the original data set; constructing a vertical displacement prediction model of the adjacent building by taking the foundation pit excavation monitoring parameters as input and the vertical displacement of the adjacent building as output and adopting a deep learning algorithm; establishing a probability model of foundation pit excavation monitoring parameters in a specified time period; generating sample data of foundation pit excavation monitoring parameters based on the probability model, and inputting the sample data into a vertical displacement prediction model of the adjacent building to obtain a prediction result; the probability that neighboring buildings are at different risk levels is obtained. The method is used for solving the problems of high execution difficulty, high cost, low efficiency and the like in the prior art, reducing the severe requirements on the professional skills of deep foundation pit excavation construction managers, and achieving the purpose of deep foundation pit excavation induced adjacent building risk assessment with low cost, high efficiency and high precision by using a deep learning technology.
Description
Technical Field
The invention relates to the field of deep foundation pit excavation, in particular to a deep foundation pit excavation induced adjacent building risk assessment method based on deep learning.
Background
As ground traffic pressure continues to increase, a number of complications are brought. The utilization of the underground space is helpful for relieving the above-ground traffic pressure and providing the diversification and sustainable development of cities. Large-scale underground space development is being performed in more and more large and medium-sized cities and extra large cities in the world. Deep foundation pit excavation is a good-quality scheme for solving underground space development in recent years, and is extremely outstanding in subway construction, underground parking lots and market development. Deep foundation pit engineering generally refers to foundation pit earthwork excavation, supporting and dewatering engineering with excavation depth exceeding 5 m. In recent years, large-scale deep foundation pit excavation projects are developed in different cities successively, and a series of safety accidents such as foundation pit collapse, adjacent building damage and the like occur in different degrees due to insufficient project management experience and lack of perception of risks, so that casualties and huge economic losses are caused. Therefore, the method for sensing the deep foundation pit excavation risk comprehensively and accurately is urgent to provide guidance for deep foundation pit excavation project management and protect safety of surrounding buildings under foundation pit excavation.
In the existing research, an empirical method is widely adopted to predict the vertical displacement of a building near a deep foundation pit, but the method is limited by site engineering conditions and empirical formulas, and is also subject to factors such as inaccurate materials, insufficient site investigation and the like, and the result of the displacement estimation of the adjacent building is easy to generate larger deviation, so that the requirement of the existing construction project on high standard and high precision of risk perception cannot be met, and the empirical method has limited application range and can only be used as a reference of a prediction project. Another prediction method is from numerical simulation, a numerical model is established through a large number of assumptions on construction environment and construction engineering, specific construction conditions are quantized into parameters, the construction of the model and the simplification of specific projects are realized, but a large amount of parameter adjustment time and simulation time are required in the process.
Disclosure of Invention
The invention provides a deep foundation pit excavation induced adjacent building risk assessment method based on deep learning, which aims to solve the problems of high execution difficulty, high cost, low efficiency and the like in a deep foundation pit excavation induced adjacent building risk prediction task in the prior art, and reduces the severe requirements on professional skills of deep foundation pit excavation construction managers by using a deep learning technology so as to achieve the purpose of efficiently, accurately and automatically assessing the deep foundation pit excavation induced adjacent building risk.
The invention is realized by the following technical scheme:
deep foundation pit excavation induced adjacent building risk assessment method based on deep learning comprises the following steps:
collecting an original data set representing excavation characteristics of a foundation pit, determining excavation monitoring parameters of the foundation pit for deep learning from the original data set, and preprocessing;
taking the preprocessed foundation pit excavation monitoring parameters as input and the vertical displacement of the adjacent building as output, and constructing a vertical displacement prediction model of the adjacent building by adopting a deep learning algorithm;
establishing a probability model of the preprocessed foundation pit excavation monitoring parameters in a specified time period;
generating sample data of foundation pit excavation monitoring parameters based on the probability model, and inputting the sample data into the adjacent building vertical displacement prediction model to obtain a prediction result;
and obtaining the probability that the adjacent buildings are at different risk levels based on the prediction result.
Aiming at the problems of high execution difficulty, high cost, low efficiency and the like in the prediction of deep foundation pit excavation induced adjacent building risks in the prior art, the invention provides a deep foundation pit excavation induced adjacent building risk assessment method based on deep learning. Then, constructing a vertical displacement prediction model of the adjacent building by taking the foundation pit excavation monitoring parameters as input and the vertical displacement of the adjacent building as output and adopting a deep learning algorithm; meanwhile, performing independent probability modeling on the foundation pit excavation monitoring parameters within a specified time period to obtain a probability model of the foundation pit excavation monitoring parameters; and finally, dividing intervals by preset risk levels to obtain the probability that the adjacent building is at different risk levels during the current deep foundation pit excavation.
According to the method, a probability model of a vertical displacement prediction model of an adjacent building and a probability model of foundation pit excavation monitoring parameters in a specified time period are established, the two models are fully combined, sample data required by the prediction model are obtained by utilizing the probability model, and a required pre-stored result is obtained based on the sample data; the method can overcome the defects of large estimation deviation, high execution difficulty, low estimation accuracy and the like caused by insufficient field investigation, imperfect materials and the like, does not need a large amount of parameter adjustment time and simulation time, remarkably improves the evaluation efficiency, and realizes the effect of evaluating the risk of the adjacent building induced by deep foundation pit excavation with low cost, high efficiency and high precision.
Further, the original data set for representing the excavation characteristics of the foundation pit comprises observation values of a plurality of primary screening characteristic parameters at different moments;
the method for determining foundation pit excavation monitoring parameters for deep learning from the original dataset comprises the following steps: and carrying out correlation analysis on all the primary screening characteristic parameters in the original data set, removing the primary screening characteristic parameters with higher correlation than the set primary screening characteristic parameters, and taking the residual primary screening characteristic parameters as deep-learning foundation pit excavation monitoring parameters.
Because the foundation pit excavation monitoring parameters acquired by the existing foundation pit information platform are hundreds of features, if all the features are used in the subsequent modeling process, the problems of large calculated amount, low evaluation efficiency and the like are necessarily caused. To overcome this problem, the present solution makes the collected raw data be the feature parameters after the primary screening, which can be determined primarily by engineering experience or reference data, and can affect the excavation monitoring parameters of the foundation pit for the neighboring building. Thus, the data in the original dataset is the observations of each of the primary screening feature parameters at different times.
According to the scheme, final required foundation pit excavation monitoring parameters are determined based on the primary screening characteristic parameters, specifically, all the primary screening characteristic parameters are subjected to correlation analysis, the primary screening characteristic parameters with higher correlation than the set primary screening characteristic parameters are removed, and the residual primary screening characteristic parameters are used as foundation pit excavation monitoring parameters for deep learning. The criterion for judging whether the correlation is too high can be set by a person skilled in the art according to the specific working condition adaptability when the method is implemented.
Further, the correlation analysis is performed by the following formula:
wherein: r is (r) k Is a correlation coefficient; n is the number of observations of the analyzed primary screening feature parameters in the original dataset; x, Y represent two preliminary screening characteristic parameters for correlation analysis, respectively; x is X i ，Y i Respectively representing the ith observation value of the primary screening characteristic parameter X, Y in the original data set; x is X j ，Y j Represents the j observation values of the primary screening characteristic parameters X, Y in the original data set respectively;
if |r k The method includes the steps that (a) one primary screening characteristic parameter is removed and the other primary screening characteristic parameter is reserved in X, Y; wherein alpha is a set threshold and alpha < 1.
According to the correlation analysis of the scheme, all the primary screening characteristic parameters are analyzed pairwise, if the correlation of the two primary screening characteristic parameters is higher than the set threshold value alpha, only one primary screening characteristic parameter is reserved, so that the interference of the subsequent evaluation process and result is avoided, the calculated amount of the subsequent modeling process is obviously reduced, and the estimation efficiency is further improved.
Further, the method for constructing the vertical displacement prediction model of the adjacent building comprises the following steps:
respectively taking M different deep learning networks as frames, and establishing M independent deep learning models;
respectively calculating root mean square errors of the M deep learning models, and selecting the deep learning model with the lowest root mean square error as a vertical displacement prediction model of the adjacent building; wherein M is more than or equal to 3.
The scheme can compare the prediction performances of M different deep learning networks in the aspect of deep foundation pit excavation induced vertical displacement of adjacent buildings, and an optimal training model is selected from the M networks, so that the situation that the prediction precision is not high due to the fact that a worker selects the deep learning network at will can be avoided; meanwhile, the optimal training model can be obtained only by comparing the root mean square error, so that the severe skill requirement of deep foundation pit excavation constructors in the deep learning professional direction is remarkably reduced.
Further, the method for establishing the probability model of the preprocessed foundation pit excavation monitoring parameters in the appointed time period comprises the following steps:
dividing the designated time period into a plurality of small segments;
fitting observation values of excavation monitoring parameters of each foundation pit in each small section according to N different probability distributions respectively to obtain N fitting results; respectively calculating root mean square errors of the N fitting results, and selecting the fitting result with the smallest root mean square error as a sub-probability model in the current small section; wherein N is more than or equal to 3;
and combining the sub-probability models in all the small sections to obtain a probability model of the foundation pit excavation monitoring parameters in the specified time period.
According to the scheme, the probability model is subjected to sectional modeling under a time sequence to obtain sub-probability models in each time segment, and finally all the sub-probability models are combined to obtain the probability model of the foundation pit excavation monitoring parameters in the designated time segment, and a sample required by prediction is generated on the basis of the probability model.
In the process of determining the sub-probability model, the observation values of the excavation monitoring parameters of each foundation pit in each small section are fitted by using N different probability distributions, and the optimal probability distribution mode which is most suitable for the excavation monitoring parameters of the current foundation pit and in the small section of the current time is selected from the N different probability distributions, so that the problem that the precision is not high due to random probability distribution is avoided; meanwhile, the optimal probability distribution can be obtained only by comparing the root mean square error, and the professional harsh skill requirement on deep foundation pit excavation constructors in the probability direction is remarkably reduced.
Further, sample data of the foundation pit excavation monitoring parameters are generated from the probability model through Monte Carlo simulation.
Further, after the prediction result is obtained, the method further includes: and calculating the probability of the adjacent building entering the specified risk level for the first time on the specified calculation date based on the prediction result.
The probability of first entering a specified risk level is calculated by the following formula:
wherein: m represents a specified calculation date; p (O) m ) A probability of first entering a specified risk level at date m;representing the probability of entering a specified risk level at date m; />Representing a sum of probabilities of first entering a specified risk level before date m; m is M m The number of samples of the vertical displacement of the adjacent building exceeding the limit value every day in m days; n (N) m The number of samples that the vertical displacement of the adjacent building does not exceed the limit value every day in m days; m is M m-1 The number of samples with vertical displacement exceeding the limit value in m-1 days is the number of samples with vertical displacement exceeding the limit value in each day; n (N) m-1 Is the number of samples of the vertical displacement of the adjacent building which does not exceed the limit value every day within m-1 days。
The scheme creatively provides a calculation method for the probability and interval of the first entering of the adjacent building into the appointed risk level, provides an important reference for deep foundation pit excavation risk assessment, is beneficial to determining the time point and the time length of occurrence of the higher risk level, and further provides more scientific and reasonable guidance for site construction and corresponding protective measures.
Further, after the prediction result is obtained, the method further includes: constructing a foundation pit excavation monitoring parameter contribution degree model based on a back propagation neural network by taking vertical displacement of adjacent buildings as input and foundation pit excavation monitoring parameters as output; and obtaining the contribution degree of each foundation pit excavation monitoring parameter to different risk levels based on the contribution degree model.
The method for obtaining the contribution degree of each foundation pit excavation monitoring parameter to different risk levels based on the contribution degree model comprises the following steps:
under different risk levels, sampling the vertical displacement of adjacent buildings, and inputting the sampling result into the foundation pit excavation monitoring parameter contribution degree model to obtain the simulation result of each foundation pit excavation monitoring parameter;
and calculating probability distribution of each foundation pit excavation monitoring parameter in different intervals under different risk levels based on simulation results of each foundation pit excavation monitoring parameter.
The inventor also finds that the risk state of the adjacent building is closely related to the range of the foundation pit excavation monitoring parameters in the research process, so that the risk interval of the foundation pit excavation monitoring parameters can be determined by calculating the contribution degree of each foundation pit excavation monitoring parameter when the adjacent building is in different risk levels, and the purpose of protecting the adjacent building through the risk assessment result is achieved by pertinently adjusting the construction working condition and the range of the foundation pit excavation monitoring parameters.
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. the deep foundation pit excavation induced adjacent building risk assessment method based on deep learning can overcome the defects of large risk assessment deviation, high execution difficulty, low prediction accuracy and the like caused by insufficient field investigation, imperfect materials and the like, save parameter adjustment and simulation time, remarkably improve calculation efficiency and realize low-cost, high-efficiency and high-accuracy assessment of deep foundation pit excavation induced adjacent building risks.
2. According to the deep foundation pit excavation induced adjacent building risk assessment method based on deep learning, the optimal training model and parameters can be selected in a self-adaptive mode in the assessment process, the problem that prediction accuracy is low due to the fact that a worker selects a deep learning network at will is avoided, and meanwhile the severe skill requirements on deep foundation pit excavation constructors in the deep learning professional direction are remarkably reduced.
3. According to the deep foundation pit excavation induced adjacent building risk assessment method based on deep learning, independent probability modeling is carried out in sections under a time sequence, sub-probability models in each time period are obtained, finally all the sub-probability models are combined, the probability model of foundation pit excavation monitoring parameters in a specified time period is obtained, and accuracy of the probability model is improved.
4. According to the deep foundation pit excavation induced adjacent building risk assessment method based on deep learning, the optimal probability distribution mode which is most suitable for the current foundation pit excavation monitoring parameters and is in the current time period can be selected in a self-adaptive mode, the problem of low precision caused by random probability distribution selection can be avoided, and meanwhile the professional harsh skill requirements on deep foundation pit excavation constructors in the probability direction are remarkably reduced.
5. The deep foundation pit excavation induced adjacent building risk assessment method based on deep learning provides a method for calculating the probability of the adjacent building entering a specified risk level for the first time, provides an important reference for risk assessment of deep foundation pit excavation, is beneficial to determining the time point and the time length of occurrence of higher risk, and further provides more scientific and reasonable guidance for site construction and corresponding protective measures.
6. According to the deep foundation pit excavation induced adjacent building risk assessment method based on deep learning, a method for calculating the contribution degree of each foundation pit excavation monitoring parameter when adjacent buildings are under different risk levels is provided, the dangerous interval of the foundation pit excavation monitoring parameter can be determined, the construction working condition can be adjusted in a targeted mode, the foundation pit excavation monitoring parameter range can be adjusted, and the purpose of protecting the adjacent buildings through the risk assessment results is achieved.
Drawings
The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiments of the invention. In the drawings:
FIG. 1 is a schematic flow chart of an embodiment of the present invention;
FIG. 2 is a graph of correlation coefficients of foundation pit excavation monitoring parameters in accordance with an embodiment of the present invention;
FIG. 3 is a graph of predictive outcome evaluation for three deep learning networks in accordance with an embodiment of the present invention;
FIG. 4 is a graph showing the fit of three probability distributions to the points of foundation pit excavation monitoring parameters in an embodiment of the present invention;
FIG. 5 is a graph of cumulative probability of adjacent buildings being at different risk levels each day in an embodiment of the invention;
FIG. 6 is a probability map of an affected neighboring building first entering a high risk level in an embodiment of the invention;
FIG. 7 is a graph of probability of an affected neighboring building first entering an ultra-high risk level in an embodiment of the invention;
fig. 8 is a graph showing the contribution of foundation pit excavation monitoring parameters to different risk levels of neighboring buildings according to an embodiment of the present invention.
Detailed Description
For the purpose of making apparent the objects, technical solutions and advantages of the present invention, the present invention will be further described in detail with reference to the following examples and the accompanying drawings, wherein the exemplary embodiments of the present invention and the descriptions thereof are for illustrating the present invention only and are not to be construed as limiting the present invention.
Example 1:
the deep foundation pit excavation induced neighboring building risk assessment method based on deep learning as shown in fig. 1 comprises the following steps:
step one, collecting an original data set representing excavation characteristics of a foundation pit, determining excavation monitoring parameters of the foundation pit for deep learning from the original data set, and preprocessing:
collecting all characteristic parameters from a foundation pit information platform, and selecting the following parameters of a deep foundation pit as primary screening characteristic parameters according to engineering experience: inclinometry data, ground surface subsidence data, upright post subsidence data, wall top subsidence data and various pipeline subsidence data.
And collecting observed values of the primary screening characteristic parameters at different moments as an original data set.
And carrying out correlation analysis on all the primary screening characteristic parameters in the original data set, removing the primary screening characteristic parameters with higher correlation than the set primary screening characteristic parameters, and taking the residual primary screening characteristic parameters as deep-learning foundation pit excavation monitoring parameters.
Wherein, the formula of correlation analysis is:
wherein: r is (r) k Is a correlation coefficient; n is the number of observations of the analyzed primary screening feature parameters in the original dataset; x, Y represent two preliminary screening characteristic parameters for correlation analysis, respectively; x is X i ，Y i Respectively representing the ith observation value of the primary screening characteristic parameter X, Y in the original data set; x is X j ，Y j Represents the j observation values of the primary screening characteristic parameters X, Y in the original data set respectively;
if |r k The method includes the steps that (a) one primary screening characteristic parameter is removed and the other primary screening characteristic parameter is reserved in X, Y; wherein alpha is a set threshold and alpha < 1.
Preferably, α is between 0.8 and 0.9.
After the foundation pit excavation monitoring parameters for deep learning are determined through the steps, preprocessing is carried out on the foundation pit excavation monitoring parameters, and the preprocessing method adopts the existing data processing mode, such as standardized processing.
After the screening of the relevance analysis in the embodiment, the foundation pit excavation monitoring parameters finally obtained are as follows: ground surface subsidence, inclinometry, displacement of electric power pipelines, wall top subsidence and displacement of rainwater pipelines. The correlation analysis of the foundation pit excavation monitoring parameters is shown in fig. 2, wherein X1, X2 and X3 represent foundation pit excavation monitoring parameters related to earth surface subsidence, X4 and X5 represent foundation pit excavation monitoring parameters related to inclinometry, X6 and X7 represent foundation pit excavation monitoring parameters related to displacement of a power pipeline, and X8 and X9 represent foundation pit excavation monitoring parameters related to displacement of a rainwater pipeline.
It can be seen from fig. 2 that the correlation coefficient between any two foundation pit excavation monitoring parameters is low, and no high correlation parameter exists.
Step two, constructing a vertical displacement prediction model of the adjacent building by using the preprocessed foundation pit excavation monitoring parameters as input and the vertical displacement of the adjacent building as output and adopting a deep learning algorithm:
respectively taking a long-short-time memory network and a circulating neural network and a gate-control circulating unit network as a framework to establish 3 independent deep learning models;
the preprocessed data set is separated into a training set and a testing set. Based on the training set, the 3 deep learning models are independently trained, and based on the testing set, the training model with optimal prediction performance is selected, and the specific selection method is as follows: and respectively calculating root mean square errors MSE of the 3 deep learning models, selecting the deep learning model with the lowest MSE as a vertical displacement prediction model of the adjacent building, wherein a calculation formula of the MSE adopts:
wherein: y is i Is the actual value of the current,is a predicted value, p is the number of samples.
Fig. 3 shows MSE values between the predicted and actual results of three deep learning models each day in this embodiment, where LSTM represents a long and short term memory network, RNN represents a cyclic neural network, and GRU represents a gated cyclic unit network, it can be seen that the lowest mse=0.089 is obtained in this embodiment based on the deep learning model established by the long and short term memory neural network. Therefore, in this embodiment, the long-short time memory network model with the best performance is selected as the final adjacent building vertical displacement prediction model.
Thirdly, establishing a probability model of the preprocessed foundation pit excavation monitoring parameters in a specified time period:
the preprocessed foundation pit excavation monitoring parameters record different data of a plurality of measuring points at different times, so that the independent probability modeling is carried out on the foundation pit excavation monitoring parameters in the step. The probability modeling can be used for modeling the problem of uncertainty by using probability distribution, a model is built through the probability distribution under the condition that a small amount of monitoring data is known, and the model can supplement enough reliable data through Monte Carlo simulation to develop the next research.
Specific:
the specified time period in this embodiment is a period to be predicted, such as the entire period of deep pit excavation.
In the embodiment, the appointed time period is divided into a plurality of small segments by taking a day as a unit;
and fitting observation values of excavation monitoring parameters of each foundation pit every day by three different probability distributions of Gaussian distribution, lognormal distribution and Weibull distribution to obtain 3 fitting results. Because the emphasis of the application is on sensing the risk state of the adjacent building, the embodiment tries to fit data points of different parameters every day by three distributions, calculates the root mean square error of the fitting result, and selects the fitting result with the smallest root mean square error as the sub-probability model in the current small section so as to obtain the optimal probability distribution.
As shown in fig. 4, fig. 4 shows the selection of the optimal distribution of two types of parameters X1 and X8 at different times. In fig. 4, three small diagrams from left to right (a), (b), and (c) represent: fitting results of parameter X1 on the first day D1, fitting results of parameter X1 on the twentieth day D20, and fitting results of parameter X8 on the first day D1.
Taking the parameter X1 of the first day D1 as an example, the present embodiment fits the monitored data of the first day with three probability distributions to obtain MSE values of 0.23,0.45 and 1.18, respectively. Therefore, the best effect is achieved under the Gaussian distribution fitting, so that the foundation pit excavation monitoring parameter X1 is selected to be fitted by the Gaussian distribution in the first day D1. Fitting of the foundation pit excavation monitoring parameters X1 in other dates and fitting of the other parameters are obtained in the same way, and finally, the fitting results of the different parameters in each day are combined to obtain a complete probability model of all the foundation pit excavation monitoring parameters in a specified time period.
This step provides a prerequisite for the probability fit of the raw data to create reliable sample data, which is generated by monte carlo simulation under optimal distribution.
And step four, based on the probability model, generating sample data of foundation pit excavation monitoring parameters through Monte Carlo simulation, and inputting the sample data into the adjacent building vertical displacement prediction model to obtain a prediction result.
The step provides a guarantee for sensing the risk of the adjacent building through the trained prediction model in the previous step.
Step five, obtaining the probability that the adjacent building is at different risk levels based on the prediction result:
the risk level of the adjacent building under deep foundation pit excavation is divided into five progressive stages based on foundation pit engineering supervision specifications and engineering experience: the detailed risk classification can provide more definite guidance for construction managers.
According to the embodiment, based on a large amount of engineering experience, warning reference values of vertical displacement of adjacent buildings are determined, the vertical displacement is set to be a boundary between a risk-free area and a low risk area, the risk areas are more clearly divided at intervals of 0.1mm, and risk grades are composed of low risk, medium risk, high risk and ultrahigh risk and correspond to grades II, III, IV and V.
Then, the probability and the cumulative probability of the adjacent building being in different risk level intervals every day can be calculated according to the prediction results obtained above. As shown in fig. 5, after the deep foundation pit excavation work is performed, the probability that the neighboring building is at medium risk, high risk and ultra-high risk is continuously increased until the final stability. It is noted that, this excavation ends on day 112, and in view of the hysteresis of the monitoring result and observing the influence on the neighboring building after the construction is stopped, the designated time period in this embodiment is the construction period of deep foundation pit excavation+10 days, that is, 121 days, and it can be seen that the probability of the corresponding risk level of the neighboring building is basically unchanged within 10 days after the construction is stopped.
Step six, calculating the probability of entering the specified risk level for the first time when the adjacent building is on the specified calculation date based on the prediction result:
in this embodiment, the high risk and the ultra-high risk are used as the specified risk levels, because the probability that the adjacent building first enters the dangerous zone can be used as an important reference basis for evaluating the excavation risk of the deep foundation pit; particularly, the probability of entering the high risk interval IV and the ultrahigh risk interval V for the first time has very important engineering significance for risk assessment of adjacent buildings, and the probability is helpful for determining the time point and the time length of risk occurrence.
According to the embodiment, under the prior probability and the conditional probability, the probability that the adjacent building enters the high-risk interval IV and the ultrahigh-risk interval V for the first time under deep foundation pit excavation is calculated according to a Bayesian formula:
wherein: m represents a specified calculation date; p (O) m ) A probability of first entering a specified risk level at date m;representing the probability of entering a specified risk level at date m; />Representing a sum of probabilities of first entering a specified risk level before date m; m is M m The number of samples of the vertical displacement of the adjacent building exceeding the limit value every day in m days; n (N) m The number of samples that the vertical displacement of the adjacent building does not exceed the limit value every day in m days; m is M m-1 The number of samples with vertical displacement exceeding the limit value in m-1 days is the number of samples with vertical displacement exceeding the limit value in each day; n (N) m-1 Is the number of samples that the vertical displacement of the adjacent building does not exceed the limit value every day within m-1 days.
Through the formula, the probability that the adjacent building enters the high risk level and the ultra-high risk level for the first time can be calculated respectively, and specific calculation results are shown in fig. 6 and 7:
in terms of the first entering of the adjacent building into the high risk level, the probability of the first entering of the adjacent building into the high risk level interval is high in the construction initial stage of deep foundation pit excavation; along with the deep foundation pit excavation construction, the probability is reduced along with the deep foundation pit excavation construction, the probability is continuously increased after the probability is reduced for a period of time, and the probability is continuously reduced after the construction is in the middle period.
In the middle construction period of deep foundation pit excavation, the probability that adjacent buildings enter the ultra-high risk zone for the first time is higher; the prediction result of the method is consistent with the actual condition of engineering. Through the probability analysis, construction manager should pay more attention to the construction stage with high probability of entering the high risk level and the ultrahigh risk level for the first time by adjacent building, and related preparation and prevention work is done in advance.
Step seven, constructing a foundation pit excavation monitoring parameter contribution degree model based on a back propagation neural network by taking vertical displacement of adjacent buildings as input and each foundation pit excavation monitoring parameter as output; and obtaining the contribution degree of each foundation pit excavation monitoring parameter to different risk levels based on the contribution degree model:
the risk state of the adjacent building is closely related to the foundation pit excavation monitoring parameter range, the risk interval of the foundation pit excavation monitoring parameter can be reversely determined by calculating the contribution degree of the foundation pit excavation monitoring parameter of the adjacent building under different risk levels, and then construction managers are guided to scientifically and reasonably change the construction working condition and adjust the foundation pit excavation monitoring parameter range so as to achieve the purpose of protecting the adjacent building.
Specifically, in this embodiment, based on a backward propagation neural network of a bayesian process, predicted vertical displacement data of adjacent buildings are sampled under different risk levels, and then the contribution degrees of foundation pit excavation monitoring parameters in different intervals are obtained through a trained foundation pit excavation monitoring parameter contribution degree model.
Taking the foundation pit excavation monitoring parameter X1 as an example for explanation, the prediction result of the foundation pit excavation monitoring parameter X1 in this embodiment is shown in fig. 8, and the ordinate in fig. 8 is the vertical displacement of the neighboring building, and the abscissa is the probability distribution of the foundation pit excavation monitoring parameter X1. Analysis of fig. 8 shows that the foundation pit excavation monitoring parameter X1 is more likely to cause the risk of settlement of the neighboring building when within 8-9mm, and as the risk level increases, X1 hardly causes the neighboring building to enter high risk and ultra-high risk level when below 6 mm.
It can be seen that the reverse analysis of the foundation pit excavation monitoring parameters provides more visual and scientific basis and guidance for protecting the safety of adjacent buildings: to reduce the probability of the adjacent building entering the high risk level or even the ultra-high risk level, proper measures are required to be taken during deep foundation pit excavation construction, so that the corresponding foundation pit excavation monitoring parameters are prevented from entering the high risk range as much as possible.
Example 2:
a deep foundation pit excavation induced proximity building risk assessment system based on deep learning for performing the risk assessment scheme of embodiment 1, the system comprising:
and a data acquisition module: the method comprises the steps of acquiring an original data set for representing excavation characteristics of a foundation pit, determining excavation monitoring parameters of the foundation pit for deep learning from the original data set, and preprocessing the excavation monitoring parameters of the foundation pit;
a first modeling module: the method comprises the steps of using preprocessed foundation pit excavation monitoring parameters as input and using vertical displacement of adjacent buildings as output, and constructing a vertical displacement prediction model of the adjacent buildings by adopting a deep learning algorithm;
a second modeling module: the method comprises the steps of establishing a probability model of the preprocessed foundation pit excavation monitoring parameters in a specified time period;
a sample generation module: sample data for generating foundation pit excavation monitoring parameters based on the probability model;
and a prediction module: the method comprises the steps of inputting sample data to a vertical displacement prediction model of an adjacent building to obtain a prediction result;
risk assessment module: for deriving and outputting probabilities that neighboring buildings are at different risk levels.
Example 3:
deep foundation pit excavation induced proximity building risk assessment device based on deep learning, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, which processor, when executing the computer program, implements the steps of the risk assessment method as in embodiment 1.
The memory may be used to store the computer program and/or the modules, and the processor may implement various functions of the shield tunneling attitude prediction apparatus of the present invention by executing or executing 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, application programs required for at least one function (such as a sound playing function, an image playing function, etc.), 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, memory, plug-in hard disk, smart memory card, secure digital card, flash memory card, at least one magnetic disk storage device, flash memory device, or other volatile solid state storage device.
Example 4:
a computer readable storage medium storing a computer program which, when executed by a processor, implements the steps of the risk assessment method as in embodiment 1.
The processor may be a central processing unit (CPU, central Processing Unit), other general purpose processors, digital signal processors (digital signal processor), application specific integrated circuits (Application Specific Integrated Circuit), off-the-shelf programmable gate arrays (Field programmable gate array) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. The general purpose processor may be a microprocessor or any conventional processor.
The computer readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash disk, a removable hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory, a random access memory, a point carrier signal, a telecommunication signal, a software distribution medium, and the like. It should be noted that the content of the computer readable medium can be appropriately increased or decreased according to the requirements of the legislation and the patent practice in the jurisdiction.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.
It should be noted that in this document, terms such as "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Claims (10)
1. Deep foundation pit excavation induced adjacent building risk assessment method based on deep learning is characterized by comprising the following steps:
collecting an original data set representing excavation characteristics of a foundation pit, determining excavation monitoring parameters of the foundation pit for deep learning from the original data set, and preprocessing;
taking the preprocessed foundation pit excavation monitoring parameters as input and the vertical displacement of the adjacent building as output, and constructing a vertical displacement prediction model of the adjacent building by adopting a deep learning algorithm;
establishing a probability model of the preprocessed foundation pit excavation monitoring parameters in a specified time period;
generating sample data of foundation pit excavation monitoring parameters based on the probability model, and inputting the sample data into the adjacent building vertical displacement prediction model to obtain a prediction result;
and obtaining the probability that the adjacent buildings are at different risk levels based on the prediction result.
2. The deep foundation pit excavation induced proximity building risk assessment method based on deep learning of claim 1, wherein the raw dataset characterizing foundation pit excavation characteristics comprises observations of several preliminary screening characteristic parameters at different moments;
the method for determining foundation pit excavation monitoring parameters for deep learning from the original dataset comprises the following steps: and carrying out correlation analysis on all the primary screening characteristic parameters in the original data set, removing the primary screening characteristic parameters with higher correlation than the set primary screening characteristic parameters, and taking the residual primary screening characteristic parameters as deep-learning foundation pit excavation monitoring parameters.
3. Deep-learning-based deep foundation pit excavation induced proximity building risk assessment method according to claim 2, characterized in that the correlation analysis is performed by the following formula:
wherein: r is (r) k Is a correlation coefficient; n is the number of observations of the analyzed primary screening feature parameters in the original dataset; x, Y represent two preliminary screening characteristic parameters for correlation analysis, respectively; x is X i ，Y i Respectively representing the ith observation value of the primary screening characteristic parameter X, Y in the original data set; x is X j ，Y j Represents the j observation values of the primary screening characteristic parameters X, Y in the original data set respectively;
if |r k The method includes the steps that (a) one primary screening characteristic parameter is removed and the other primary screening characteristic parameter is reserved in X, Y; wherein alpha is a set threshold and alpha < 1.
4. The deep foundation pit excavation induced proximity building risk assessment method based on deep learning of claim 1, wherein the method of constructing the proximity building vertical displacement prediction model comprises:
respectively taking M different deep learning networks as frames, and establishing M independent deep learning models;
respectively calculating root mean square errors of the M deep learning models, and selecting the deep learning model with the lowest root mean square error as a vertical displacement prediction model of the adjacent building; wherein M is more than or equal to 3.
5. The deep-learning-based deep foundation pit excavation induced proximity building risk assessment method of claim 1, wherein the method of establishing a probabilistic model of the preprocessed foundation pit excavation monitoring parameters over a specified period of time comprises:
dividing the designated time period into a plurality of small segments;
fitting observation values of excavation monitoring parameters of each foundation pit in each small section according to N different probability distributions respectively to obtain N fitting results; respectively calculating root mean square errors of the N fitting results, and selecting the fitting result with the smallest root mean square error as a sub-probability model in the current small section; wherein N is more than or equal to 3;
and combining the sub-probability models in all the small sections to obtain a probability model of the foundation pit excavation monitoring parameters in the specified time period.
6. Deep-learning-based deep foundation pit excavation induced proximity building risk assessment method according to claim 1, characterized in that the sample data of foundation pit excavation monitoring parameters are generated from the probabilistic model by monte carlo simulation.
7. The deep foundation pit excavation induced proximity building risk assessment method based on deep learning according to any one of claims 1 to 6, further comprising, after obtaining the prediction result: and calculating the probability of the adjacent building entering the specified risk level for the first time on the specified calculation date based on the prediction result.
8. The deep foundation pit excavation induced proximity building risk assessment method based on deep learning of claim 7, wherein the probability of first entering a specified risk level is calculated by the following formula:
wherein: m represents a specified calculation date; p (O) m ) A probability of first entering a specified risk level at date m;representing the probability of entering a specified risk level at date m; />Representing a sum of probabilities of first entering a specified risk level before date m; m is M m The number of samples of the vertical displacement of the adjacent building exceeding the limit value every day in m days; n (N) m The number of samples that the vertical displacement of the adjacent building does not exceed the limit value every day in m days; m is M m-1 The number of samples with vertical displacement exceeding the limit value in m-1 days is the number of samples with vertical displacement exceeding the limit value in each day; n (N) m-1 Is the number of samples that the vertical displacement of the adjacent building does not exceed the limit value every day within m-1 days.
9. The deep foundation pit excavation induced proximity building risk assessment method based on deep learning of claim 1, further comprising, after obtaining the prediction result: constructing a contribution degree model of foundation pit excavation monitoring parameters based on a back propagation neural network by taking vertical displacement of adjacent buildings as input and taking excavation monitoring parameters of each foundation pit as output; and obtaining the contribution degree of each foundation pit excavation monitoring parameter to different risk levels based on the contribution degree model.
10. The deep foundation pit excavation induced proximity building risk assessment method based on deep learning of claim 9, wherein the method for obtaining the contribution degree of each foundation pit excavation monitoring parameter to different risk levels based on the contribution degree model comprises:
under different risk levels, sampling the vertical displacement of adjacent buildings, and inputting the sampling result into the contribution degree model to obtain the simulation result of each foundation pit excavation monitoring parameter;
and calculating probability distribution of each foundation pit excavation monitoring parameter in different intervals under different risk levels based on simulation results of each foundation pit excavation monitoring parameter.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311629365.6A CN117557100A (en) | 2023-11-30 | 2023-11-30 | Deep foundation pit excavation induced adjacent building risk assessment method based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311629365.6A CN117557100A (en) | 2023-11-30 | 2023-11-30 | Deep foundation pit excavation induced adjacent building risk assessment method based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117557100A true CN117557100A (en) | 2024-02-13 |
Family
ID=89812524
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311629365.6A Pending CN117557100A (en) | 2023-11-30 | 2023-11-30 | Deep foundation pit excavation induced adjacent building risk assessment method based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117557100A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117933947A (en) * | 2024-03-21 | 2024-04-26 | 辽宁隆祥昌建筑工程服务有限公司 | Building engineering project progress management system and optimization method thereof |
| CN118297487A (en) * | 2024-06-06 | 2024-07-05 | 中国水利水电第十工程局有限公司 | Abnormality detection and root cause analysis method and system for foundation pit monitoring |
-
2023
- 2023-11-30 CN CN202311629365.6A patent/CN117557100A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117933947A (en) * | 2024-03-21 | 2024-04-26 | 辽宁隆祥昌建筑工程服务有限公司 | Building engineering project progress management system and optimization method thereof |
| CN118297487A (en) * | 2024-06-06 | 2024-07-05 | 中国水利水电第十工程局有限公司 | Abnormality detection and root cause analysis method and system for foundation pit monitoring |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117557100A (en) | Deep foundation pit excavation induced adjacent building risk assessment method based on deep learning | |
| CN110008301B (en) | Regional geological disaster susceptibility prediction method and device based on machine learning | |
| CN108595845B (en) | Bridge earthquake risk probability analysis method based on performance design | |
| CN110991046B (en) | Drainage system waterlogging risk rapid early warning method based on response surface function | |
| Yum et al. | Development of vulnerability curves of buildings to windstorms using insurance data: An empirical study in South Korea | |
| CN111144636B (en) | Slope deformation prediction method | |
| Li | Exploring the potential of utilizing unsupervised machine learning for urban drainage sensor placement under future rainfall uncertainty | |
| Arnoux et al. | Coastal flooding event definition based on damages: Case study of Biarritz Grande Plage on the French Basque coast | |
| CN117272232A (en) | Tunnel monitoring method and device for data fusion, computer equipment and storage medium | |
| CN110852415B (en) | Vegetation index prediction method, system and equipment based on neural network algorithm | |
| CN110968929A (en) | Wind power plant wind speed prediction method and device and electronic equipment | |
| AU2021104794A4 (en) | Stage-wise stochastic inversion identification method of aquifer structure based on deep learning | |
| CN118410326A (en) | Deep foundation pit earth surface subsidence space-time prediction method based on graph convolution gate control neural network | |
| CN116307270B (en) | Method and system for evaluating casualties influenced by landslide chain type disasters induced by storm | |
| CN116106856B (en) | Identification model establishment method and identification method for thunderstorm strong wind and computing equipment | |
| CN117609218B (en) | Drainage pipe network GIS data quality control method and system based on dynamic knowledge base | |
| CN117853077B (en) | Piling process management method, device, equipment and storage medium | |
| CN118171381B (en) | Terrain construction method and system based on sponge city design | |
| CN116776636B (en) | Digital twinning-based emergency simulation method and system for flood and drought disasters | |
| CN118349810A (en) | Method, equipment and storage medium for predicting maximum subsidence of earth surface caused by construction | |
| CN117421549A (en) | Shale gas horizontal well fracture pressure prediction method and system | |
| CN117725843A (en) | House structure deformation prediction method based on deep learning | |
| CN117669868A (en) | Genetic intelligent moon exploration site selection method and device considering exploration value index | |
| CN118297280A (en) | Road soil slope stability evaluation method and system | |
| Chen et al. | Eliminating thermal responses of stay cables by partially missing sensors toward more reliable and robust SHM |
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 |