CN114611833A - Dam body deep learning model construction method based on dual-drive combination - Google Patents
Dam body deep learning model construction method based on dual-drive combination Download PDFInfo
- Publication number
- CN114611833A CN114611833A CN202210324685.XA CN202210324685A CN114611833A CN 114611833 A CN114611833 A CN 114611833A CN 202210324685 A CN202210324685 A CN 202210324685A CN 114611833 A CN114611833 A CN 114611833A
- Authority
- CN
- China
- Prior art keywords
- material strength
- main body
- safety monitoring
- structural
- monitoring index
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000013136 deep learning model Methods 0.000 title claims abstract description 72
- 238000010276 construction Methods 0.000 title claims abstract description 15
- 239000000463 material Substances 0.000 claims abstract description 187
- 238000012544 monitoring process Methods 0.000 claims abstract description 173
- 238000013507 mapping Methods 0.000 claims abstract description 40
- 230000009977 dual effect Effects 0.000 claims abstract description 21
- 230000007246 mechanism Effects 0.000 claims abstract description 12
- 238000012549 training Methods 0.000 claims description 42
- 238000000034 method Methods 0.000 claims description 28
- 238000013461 design Methods 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000005284 excitation Effects 0.000 description 17
- 230000004044 response Effects 0.000 description 13
- 238000013135 deep learning Methods 0.000 description 8
- 230000008569 process Effects 0.000 description 8
- 238000011156 evaluation Methods 0.000 description 7
- 238000004088 simulation Methods 0.000 description 7
- 238000012937 correction Methods 0.000 description 6
- 238000010248 power generation Methods 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 4
- 230000002457 bidirectional effect Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 4
- 238000011002 quantification Methods 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 229910052799 carbon Inorganic materials 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 239000000047 product Substances 0.000 description 2
- 231100000279 safety data Toxicity 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000006386 neutralization reaction Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000005316 response function Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06N3/045—Combinations of networks
-
- 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
-
- 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
- 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
-
- 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
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Biophysics (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Quality & Reliability (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- Life Sciences & Earth Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Educational Administration (AREA)
- Primary Health Care (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention relates to a dam body deep learning model construction method based on dual drive combination, which comprises the steps of constructing a deep learning model of a building structure, wherein the deep learning model takes material strength as input and takes a structure main body safety monitoring index as output; constructing an inverse mapping model, wherein the inverse mapping model is used for determining the material strength according to the safety index of the structural main body, verifying the risk degree of the material strength of the current structural main body according to the calculated material strength, determining the safety monitoring index of the building structure according to the material strength of the current structural main body, and determining the risk degree of the building structure; and determining an alarm mechanism according to the risk degree of the material and the risk degree of the structural body so as to alarm the safety degree of the structural body. By adjusting the data parameters of the constructed model, the model after adjustment is more accurate, the building structure is more accurately evaluated, and the accuracy and reliability of prediction are greatly improved.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a dam body deep learning model construction method based on dual drive combination.
Background
China is the first major country of global power production, and the power generation amount accounts for about one fourth of the total amount of the global power generation amount. The safety, the high efficiency and the green of the power production are the key of the healthy development of the related countries and the enterprises. Under the guidance of the medium and long-term development targets of 'carbon peak reaching and carbon neutralization', the requirements of energy structure optimization are embodied as full absorption of new energy power generation and gap supplement of thermal power generation. In new energy, hydroelectric power generation occupies half of the share, and has an extremely important position, so that the dam body safety is ensured to have important significance on the safety production of hydroelectric power generation enterprises, and meanwhile, the dam body safety has extremely important significance on environment, social economy and personal safety.
The dam safety monitoring aims to master the operation state of the dam by arranging monitoring equipment and classifying, arranging, calculating and analyzing collected time sequence data, so as to guarantee the safe operation of the dam. Due to the fact that the dam safety monitoring is various in types, the equipment is distributed, the monitoring data processing difficulty is high, and the monitoring data processing difficulty is not easy to perceive in time varying, various data are redundant and information is dispersed in the management process, and the engineering safety management difficulty is increased. Meanwhile, the change of the dam structure and the material strength is analyzed from the monitoring quantity, and then the safety risk of the dam is analyzed, so that the dam safety monitoring system has great technical challenge.
The principle of the current dam safety risk quantification early warning technology is that according to a dam body structure design scheme, a structural strength parameter is used as an independent variable, detection indexes such as structural deformation, stress, cracks and seepage are used as dependent variables, a mapping model from the independent variable to the dependent variables is established, and dam body safety is evaluated by monitoring whether structural safety indexes reach a threshold value or not. The evaluation of the whole safety of the dam is obtained by carrying out layered weighting on various dam safety detection indexes, and the risk of the dam is difficult to quantitatively and accurately analyze depending on the experience of experts, so that the method has certain limitation.
Disclosure of Invention
Therefore, the dam body deep learning model construction method based on dual drive combination can solve the technical problem that the dam body deep learning model is high in limitation due to the fact that artificial experience judgment is excessively relied on.
In order to achieve the aim, the invention provides a dam body deep learning model construction method based on dual drive combination, which comprises the following steps:
constructing a deep learning model of a building structure, wherein the deep learning model is input according to material strength and output according to a safety monitoring index of a structural main body, and when any material strength value is input and passes through the deep learning model, the safety monitoring index of the structural main body is uniquely determined;
constructing an inverse mapping model, wherein the network structure type and the layer parameter of the inverse mapping model are the same as those of the deep learning model, the inverse mapping model is used for determining the material strength according to the safety index of a structural main body, verifying the risk degree of the material strength of the current structural main body according to the calculated material strength, determining the safety monitoring index of a building structure according to the material strength of the current structural main body, and determining the risk degree of the building structure;
determining an alarm mechanism according to the risk degree of the material and the risk degree of the structural body so as to alarm the safety degree of the structural body;
when a deep learning model of a building structure is constructed, acquiring a structural design model of the building structure and material strength theoretical data based on the structural design model;
acquiring historical monitoring data of the building structure, wherein the historical monitoring data comprises a structure main body safety monitoring index set, the structure main body safety monitoring index set is formed on the basis of a plurality of monitoring parameters, and the number of the monitoring parameters is set to be n;
and when an alarm mechanism is determined, adjusting the number of monitoring parameters in the historical monitoring data according to the actual values of the risk degree of the material and the risk degree of the structural body.
Further, a material parameter threshold range is set, when the risk degree of the material is evaluated, the material strength S0 calculated according to the inverse mapping model is compared with an actual material strength value S, if the actual material strength value S belongs to the material parameter threshold range, but the calculated material strength S0 is equal to the actual material strength value S, it indicates that the risk degree of the material is low, and the value is set as D1, and a second time period T2 is adopted as an early warning period;
if the actual material strength value S belongs to the material parameter threshold range, but the calculated material strength S0 is not equal to the actual material strength value S, the risk of the material is moderate, D2 is set, and a first time period T1 is used as an early warning period;
if the actual material strength value S does not belong to the material parameter threshold range, but the calculated material strength S0 is not equal to the actual material strength value S, the risk degree of the material is high, the value is set to D3, and instant early warning is adopted.
Further, when a first time interval T1 is adopted as an early warning period, adjusting the length of the early warning period according to the relation between the calculated material strength S0 and the actual material strength value S;
if the calculated material strength S0 is greater than the actual material strength value S, correcting the first time period T1 by adopting a first coefficient k 1;
if the calculated material strength S0< the actual material strength value S, the first time period T1 is corrected using the second coefficient k 2.
Further, when the first time interval T1 is corrected by using the first coefficient k1, an adjustment parameter of the first coefficient k1 is determined according to the relationship between the calculated actual difference value between the material strength S0 and the actual material strength value S and the preset standard difference value Δ S10;
if S0-S is less than or equal to delta S10, the first coefficient k1 is adjusted to be 0.5 xk 1;
if S0-S > Δ S10, the first coefficient k1 is adjusted to 0.2 × k 1.
Further, when an instant alarm is carried out, the early warning volume of the early warning period is adjusted according to the relation between the calculated material strength S0 and the actual material strength value S;
if the calculated material strength S0 is greater than the actual material strength value S, alarming by adopting a first volume V1;
and if the calculated material strength S0 is less than the actual material strength value S, alarming by adopting a second volume V2, and enabling the first volume V1 to be less than a second volume V2.
Further, a structural main body safety monitoring index set is preset, a plurality of discontinuous structural main body safety monitoring values are arranged in the structural main body safety monitoring index set, whether the structural main body safety monitoring index belongs to the structural main body safety monitoring index set or not is judged according to the structural main body safety monitoring index calculated by the deep learning model, and if the structural main body safety monitoring index belongs to the structural main body safety monitoring index set, the structural main body safety under the current material strength is represented;
if not, the risk degree of the structural body under the current material strength is high.
Further, when the structural subject safety monitoring index does not belong to the structural subject safety monitoring index set, determining a maximum value M10 and a minimum value M20 in the structural subject safety monitoring index set;
if the maximum value M10 is greater than the structure main body safety monitoring index is greater than the minimum value M20, adding the corresponding strength and the structure main body safety monitoring index into a historical data set to serve as historical data of a deep learning model, and increasing training times;
and if the structural main body safety monitoring index is larger than the maximum value M10, determining the material strength corresponding to the current structural main body safety monitoring index as the unqualified strength.
Further, when the training times are increased, selecting the amplitude of the increase of the training times according to the relationship between the safety monitoring index of the structural main body and the safety monitoring values of two adjacent structural main bodies;
if the safety monitoring index of the structural main body is the average value of the safety monitoring values of two adjacent structural main bodies, increasing the training times by adopting a first increment I1;
and if the safety monitoring index of the structural main body is close to any one of the two adjacent safety monitoring values of the structural main body, increasing the training times by adopting a second increment I2.
Further, when the training times are increased by adopting a second increment I2, setting the safety monitoring values of two adjacent structural bodies as A1 and A2 respectively, wherein A1 is greater than A2, and if the absolute value of the difference between the safety monitoring index of the structural body and A1 is smaller than the absolute value of the difference between the safety monitoring index of the structural body and A2, adjusting the second increment I2 by adopting a first parameter alpha;
and if the absolute value of the difference between the safety monitoring index of the structural main body and A1 is larger than the absolute value of the difference between the safety monitoring index of the structural main body and A2, adjusting the second increment I2 by using a second parameter beta.
Further, when a second increment I2 is adjusted by using the first parameter α, the adjusted second increment is I21' = I2 × (1 + α), and the training times are increased by using the corrected second increment;
when a second increment I2 is adjusted by using a second parameter beta, the adjusted second increment is I22' = I2 x (1 + beta), and the training times are increased by using the corrected second increment, wherein the first parameter alpha and the second parameter beta are both decimals larger than 0 and smaller than 1;
the building structure is a dam, and the structure main body is a dam body.
Compared with the prior art, the method has the advantages that the safety of the current building structure is determined according to the actual values of the risk degree of the material and the risk degree of the structure main body, but the risk degree evaluation of the material of the building structure and the risk evaluation of the structure main body are calculated based on the deep learning model and the inverse mapping model, so that the model after adjustment is more accurate by adjusting the data parameters of the constructed model, the evaluation of the building structure is more accurate, and the accuracy and the reliability of prediction are greatly improved.
Particularly, the material strength S0 calculated according to the inverse mapping model is compared with the actual material strength S, different risk degrees are set according to the comparison result, and different early warning periods are set under the different risk degrees, so that the learning model provided by the embodiment of the invention can find potential safety hazards of the structural body in time in the construction and use processes, early warning is carried out in time, and the safety prediction of the structural body is improved.
In particular, the first time period is corrected by adopting different correction coefficients according to the relationship between the calculated material strength S0 and the actual material strength value S, so that the early warning of the structural subject danger is more efficient and intelligent, the use efficiency of the inverse mapping model is greatly improved, the dynamic adjustment mechanism of the early warning is improved, and the accuracy of the early warning is improved.
Particularly, the preset standard difference value delta S10 is set, the actual difference value is compared with the preset standard difference value delta S10, the first coefficient k1 is finely adjusted according to the comparison result, the first coefficient directly acts on the first time interval, therefore, the early warning period is more accurately selected, the early warning efficiency is improved, different warning periods are set according to the deviation of the actual material strength, the actual state of the structural body is effectively monitored and early warned, and the early warning efficiency is improved.
Especially, different volume levels are set according to the volume when the real-time alarm is given, so that different volumes correspond to different urgency levels, the actual value of the actual material strength is determined, the dynamic monitoring in the real-time early warning process is greatly improved, the effective determination of the risk level of the building structure is greatly improved, and the fine adjustment of the early warning is improved.
Particularly, a plurality of discontinuous structural main body safety monitoring values are set to form a structural main body safety monitoring index set, the structural main body safety monitoring index set is used for judging the output structural main body safety monitoring index in the process of carrying out input-output on the deep learning model to determine the safety of the structural main body, whether the structural main body is safe or not can be judged by directly carrying out data comparison through determining the safety data set of the structural main body, and the sensitivity of an early warning mechanism of the structural main body is improved.
Particularly, by analyzing the structural subject safety monitoring indexes under various possible conditions one by one, the relation between the actual value of the structural subject safety monitoring index and the maximum value M10 and the minimum value M20 in the structural subject safety monitoring index set is determined whether the strength of the material is set to be qualified or not, whether the strength of the material can be used as historical data or not and the like, a data set formed based on the learning model is selected, the accuracy of the learning model is further improved, and the safety of the structural subject can be judged more accurately.
Especially, the deep learning model is adjusted by increasing the training times, so that the training data are more sufficient, the deep learning model is corrected, the input and the output of the deep learning model are more matched, the structure main body safety monitoring indexes are more accurately judged, the judgment accuracy is improved, the safety judgment of the structure main body is more accurate, and the judgment efficiency is improved.
Especially, the second increment is adjusted to different degrees through the first parameter and the second parameter, so that the determination on the training times is more accurate, the training data of the deep learning model is more sufficient, the deep learning model is corrected, the input and the output of the deep learning model are more matched, the structure main body safety monitoring index is more accurately judged, the judgment accuracy is improved, the safety judgment of the structure main body is more accurate, and the judgment efficiency is improved.
Particularly, the product of the original increment and the first parameter or the second parameter is added on the basis of the original increment to serve as the training times after correction, the calculation complexity is low, the operation speed is increased, the judgment on the structure main body is quicker, and the result output efficiency of the deep learning model is improved.
Particularly, a bidirectional mapping model from dam structure strength parameters to dam monitoring index data is established by means of risk quantification early warning aiming at dam structure safety and a deep learning method integrating knowledge driving and data driving dual driving, and prediction of the monitoring index data and calibration and prediction of the dam structure strength parameters are achieved.
Drawings
Fig. 1 is a schematic flow chart of a dam depth learning model construction method based on dual drive combination according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of another solid manner of the dam depth learning model construction method based on dual drive combination according to the embodiment of the present invention;
fig. 3 is an application flowchart of the dam depth learning model building method based on dual drive combination according to the embodiment of the present invention.
Detailed Description
In order that the objects and advantages of the invention will be more clearly understood, the invention is further described in conjunction with the following examples; it should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only for explaining the technical principle of the present invention, and do not limit the scope of the present invention.
It should be noted that in the description of the present invention, the terms of direction or positional relationship indicated by the terms "upper", "lower", "left", "right", "inner", "outer", etc. are based on the directions or positional relationships shown in the drawings, which are only for convenience of description, and do not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation, and thus, should not be construed as limiting the present invention.
Furthermore, it should be noted that, in the description of the present invention, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood by those skilled in the art according to specific situations.
Referring to fig. 1, a dam depth learning model construction method based on dual drive combination according to an embodiment of the present invention includes:
step S100: constructing a deep learning model of a building structure, wherein the deep learning model is input according to material strength and output according to a safety monitoring index of a structural main body, and when any material strength value is input and passes through the deep learning model, the safety monitoring index of the structural main body is uniquely determined;
step S200: constructing an inverse mapping model, wherein the network structure type and the layer parameter of the inverse mapping model are the same as those of the deep learning model, the inverse mapping model is used for determining the material strength according to the safety index of a structural main body, verifying the risk degree of the material strength of the current structural main body according to the calculated material strength, determining the safety monitoring index of a building structure according to the material strength of the current structural main body, and determining the risk degree of the building structure;
step S300: determining an alarm mechanism according to the risk degree of the material and the risk degree of the structural body so as to alarm the safety degree of the structural body;
specifically, in step S100, in constructing a deep learning model of an architectural structure, a structural design model of the architectural structure and material strength theoretical data based on the structural design model are acquired;
acquiring historical monitoring data of the building structure, wherein the historical monitoring data comprises a structure main body safety monitoring index set, the structure main body safety monitoring index set is formed on the basis of a plurality of monitoring parameters, and the number of the monitoring parameters is set to be n;
in step S300, when determining the alarm mechanism, the number of monitoring parameters in the historical monitoring data is adjusted according to the actual values of the risk of the material and the risk of the structural subject.
Specifically, the safety of the current building structure is determined according to the actual values of the risk degree of the material and the risk degree of the structure main body, but the risk degree evaluation of the material of the building structure and the risk evaluation of the structure main body are calculated based on the deep learning model and the inverse mapping model, so that the model after adjustment is more accurate by adjusting the data parameters of the constructed model, the evaluation of the building structure is more accurate, and the accuracy and the reliability of prediction are greatly improved.
Specifically, a material parameter threshold range is preset, when the risk degree of the material is evaluated, the material strength S0 calculated according to the inverse mapping model is compared with an actual material strength value S, if the actual material strength value S belongs to the material parameter threshold range, but the calculated material strength S0 is equal to the actual material strength value S, it indicates that the risk degree of the material is low, D1 is set, and a second time period T2 is adopted as an early warning period;
if the actual material strength value S belongs to the material parameter threshold range, but the calculated material strength S0 is not equal to the actual material strength value S, the risk of the material is moderate, D2 is set, and a first time period T1 is used as an early warning period;
if the actual material strength value S does not belong to the material parameter threshold range, but the calculated material strength S0 is not equal to the actual material strength value S, the risk degree of the material is high, the value is set to D3, and instant early warning is adopted.
Specifically, the material strength S0 calculated according to the inverse mapping model is compared with the actual material strength S, different risk degrees are set according to the comparison result, and different early warning periods are set under the different risk degrees, so that the learning model provided in the embodiment of the invention can find potential safety hazards of the structural body in time during the construction and use processes, early warning is performed in time, and the safety prediction of the structural body is improved.
Specifically, when a first time interval T1 is adopted as an early warning period, the length of the early warning period is adjusted according to the relation between the calculated material strength S0 and the actual material strength value S;
if the calculated material strength S0> the actual material strength value S, the first time period T1 is corrected by using the first coefficient k 1;
if the calculated material strength S0< the actual material strength value S, the first time period T1 is corrected using the second coefficient k 2.
Specifically, according to the embodiment of the invention, the first time period is corrected by adopting different correction coefficients through the relationship between the calculated material strength S0 and the actual material strength value S, so that the early warning of the risk of the structural main body is more efficient and intelligent, the use efficiency of an inverse mapping model is greatly improved, the dynamic adjustment mechanism of the early warning is improved, and the accuracy of the early warning is improved.
Specifically, when the first time interval T1 is corrected by using the first coefficient k1, an adjustment parameter of the first coefficient k1 is determined according to the relationship between the calculated actual difference between the material strength S0 and the actual material strength value S and the preset standard difference Δ S10;
if S0-S is less than or equal to delta S10, the first coefficient k1 is adjusted to be 0.5 xk 1;
if S0-S > Δ S10, the first coefficient k1 is adjusted to 0.2 × k 1.
Specifically, according to the embodiment of the invention, the preset standard difference value Δ S10 is set, the actual difference value is compared with the preset standard difference value Δ S10, the first coefficient k1 is finely adjusted according to the comparison result, and the first coefficient directly acts on the first time period, so that the selection of the early warning period is more accurate, the early warning efficiency is improved, different alarm periods are set according to the deviation of the actual material strength, the actual state of the structural body is effectively monitored and early warned, and the early warning efficiency is improved.
Specifically, when an instant alarm is given, the early warning volume of the early warning period is adjusted according to the relation between the calculated material strength S0 and the actual material strength value S;
if the calculated material strength S0 is greater than the actual material strength value S, alarming by adopting a first volume V1;
and if the calculated material strength S0 is less than the actual material strength value S, alarming by adopting a second volume V2, and enabling the first volume V1 to be less than a second volume V2.
Specifically, according to the embodiment of the invention, different volume levels are set for the volume when the instant alarm is given, so that different volumes correspond to different urgency levels, the actual value of the actual material strength is determined, the dynamic monitoring in the instant alarm process is greatly improved, the effective determination of the risk level of the building structure is greatly improved, and the fine adjustment of the early alarm is improved.
Specifically, a structural main body safety monitoring index set is preset, a plurality of discontinuous structural main body safety monitoring values are arranged in the structural main body safety monitoring index set, whether the structural main body safety monitoring index belongs to the structural main body safety monitoring index set or not is judged according to the structural main body safety monitoring index calculated by the deep learning model, and if the structural main body safety monitoring index belongs to the structural main body safety monitoring index set, the structural main body safety under the current material strength is represented;
if not, the risk degree of the structural body under the current material strength is high.
Specifically, the embodiment of the invention sets a plurality of discontinuous structural subject safety monitoring values to form a structural subject safety monitoring index set, is used for judging the output structural subject safety monitoring index in the process of carrying out input-output on the deep learning model and determining the safety of the structural subject, can judge whether the structural subject is safe or not by determining the safety data set of the structural subject and directly carrying out data comparison, and improves the sensitivity of an early warning mechanism of the structural subject.
Specifically, when the structural subject safety monitoring index does not belong to the structural subject safety monitoring index set, determining a maximum value M10 and a minimum value M20 in the structural subject safety monitoring index set;
if the maximum value M10 is greater than the structure main body safety monitoring index is greater than the minimum value M20, adding the corresponding strength and the structure main body safety monitoring index into a historical data set to serve as historical data of a deep learning model, and increasing training times;
and if the structural main body safety monitoring index is larger than the maximum value M10, determining the material strength corresponding to the current structural main body safety monitoring index as the unqualified strength.
And if the structural main body safety monitoring index is less than the minimum value M20, determining the material strength corresponding to the current structural main body safety monitoring index as the unqualified strength.
Specifically, the embodiment of the present invention performs one-to-one analysis on multiple possible situations of the safety monitoring indexes of the structural subject, so as to determine whether the strength of the material should be set to be qualified or not and whether the strength of the material can be used as historical data or not with respect to the relationship between the actual value of the safety monitoring index of the structural subject and the maximum value M10 and the minimum value M20 in the safety monitoring index set of the structural subject, and select a data set formed based on the learning model, thereby further improving the accuracy of the learning model and facilitating more accurate judgment on the safety of the structural subject.
Specifically, when the training times are increased, the amplitude of the increase of the training times is selected according to the relationship between the safety monitoring index of the structural main body and the safety monitoring values of two adjacent structural main bodies;
if the safety monitoring index of the structural main body is the average value of the safety monitoring values of two adjacent structural main bodies, increasing the training times by adopting a first increment I1;
and if the safety monitoring index of the structural main body is close to any one of the two adjacent safety monitoring values of the structural main body, increasing the training times by adopting a second increment I2.
Specifically, the deep learning model is adjusted by increasing the training times, so that the training data are more sufficient, the deep learning model is corrected, the input and the output of the deep learning model are more matched, the safety monitoring indexes of the structural subject are more accurately judged, the judgment accuracy is improved, the safety judgment of the structural subject is more accurate, and the judgment efficiency is improved.
Specifically, when the training times are increased by adopting a second increment I2, setting the safety monitoring values of two adjacent structural bodies as A1 and A2 respectively, wherein A1 is greater than A2, and if the absolute value of the difference between the safety monitoring index of the structural body and A1 is smaller than the absolute value of the difference between the safety monitoring index of the structural body and A2, adjusting the second increment I2 by adopting a first parameter alpha;
and if the absolute value of the difference between the safety monitoring index of the structural main body and A1 is larger than the absolute value of the difference between the safety monitoring index of the structural main body and A2, adjusting the second increment I2 by using a second parameter beta.
Specifically, the second increment is adjusted to different degrees through the first parameter and the second parameter, so that the determination of the training times is more accurate, the training data of the deep learning model is more sufficient, the deep learning model is corrected, the input and the output of the deep learning model are more matched, the safety monitoring indexes of the structural main body are more accurately judged, the judgment accuracy is improved, the safety judgment of the structural main body is more accurate, and the judgment efficiency is improved.
Specifically, when the first parameter α is used to adjust the second increment I2, the adjusted second increment is I21' = I2 × (1 + α), and the number of training times is increased by using the corrected second increment;
when the second increment I2 is adjusted by using the second parameter β, the adjusted second increment is I22' = I2 × (1 + β), and the number of training times is increased by using the corrected second increment, where the first parameter α and the second parameter β are both decimals greater than 0 and smaller than 1.
Specifically, the embodiment of the invention increases the product of the original increment and the first parameter or the second parameter as the training times after correction on the basis of the original increment, has low calculation complexity, improves the operation speed, further more rapidly judges the structural subject and improves the result output efficiency of the deep learning model.
Specifically, the building structure is a dam, and the structure body is a dam body.
Specifically, in the embodiment of the present invention, the building structure and the structural main body are limited, and in practical applications, the building structure may include a dam but is not limited to a dam, and may also include other building structures, such as a house, a bridge, and the like, which are not listed one by one, and the corresponding structural main body is in a corresponding relationship with the building structure.
In another embodiment of the present invention, as shown in fig. 2, a dam body deep learning model building method based on dual drive combination is further provided, where the method includes:
step S1100: acquiring a structural design model of a building structure and material strength theoretical data based on the structural design model;
step S1200: acquiring historical monitoring data of the building structure, wherein the historical monitoring data comprises a structure main body safety monitoring index set, and the structure main body safety monitoring index set is formed on the basis of a plurality of monitoring parameters;
step S1300: constructing a first constitutive model of a building structure, wherein the first constitutive model is selected and determined according to the structure design model and material strength theoretical data, and the first constitutive model is corrected by taking a plurality of monitoring parameters as a first excitation set and a structure main body safety monitoring index set as a first response set;
step S1400: establishing a deep learning model and an inverse mapping model, determining a network structure type and layer parameters of the deep learning model according to a first constitutive model, obtaining a first response according to a first excitation input randomly, wherein the first excitation belongs to a first excitation set, the first response belongs to a first response set, performing primary pre-training on the deep learning model based on the first excitation and the first response to form a sample data set, performing secondary pre-training on the deep learning model by using the sample data set to complete the correction of the deep learning model, wherein the inverse mapping model adopts a multilayer perceptron model, the layer parameters of the multilayer perceptron model are the same as those of the deep learning model, training the inverse mapping model by using the sample data set to obtain a data set of the inverse mapping model with material strength as a response, verifying the actually measured material strength by using the material strength output by the inverse mapping model to obtain the risk degree of the material, and obtaining the risk degree of the structural body by using a structural body safety monitoring index set output by the deep learning model;
step S1500: and determining an alarm mechanism according to the risk degree of the material and the risk degree of the structural body so as to alarm the safety degree of the structural body.
Specifically, the embodiment of the invention provides a deep learning method integrating knowledge drive and data drive by aiming at risk quantification early warning of dam structure safety, establishes a bidirectional mapping model from dam structure strength parameters to dam monitoring index data, and realizes prediction of the monitoring index data and calibration and prediction of the dam structure strength parameters.
Specifically, the embodiment of the invention provides a dam body deep learning model construction method based on dual drive combination, which is shown in fig. 3 and comprises the following 6 steps:
the method comprises the following steps: data collection
A dam structure design model: a dam structure design scheme and a construction scheme are adopted, and a specific dam structure model is determined;
collecting dam structural strength data: design strength indexes of various materials of the dam and material strength actually measured in engineering completion acceptance;
historical data of dam safety monitoring: monitoring an index set and historical data of associated indexes of dam body safety in the dam operation process;
step two: finite element model and finite difference model modeling of dam structure safety analysis
Selecting a constitutive model according to a dam structure design model and material strength parameters, and primarily establishing a finite element model or a finite difference model;
and (3) correcting the constitutive model by taking the relevant elements of the dam safety monitoring indexes as external excitation and taking the dam safety monitoring index set as response, and establishing a three-dimensional structure model which accords with the reality.
Step three: deep learning modeling for dam structure safety monitoring index prediction
Data simulation based on finite element model or finite difference model: randomly loading input excitation to obtain the response of various monitoring indexes to form a sample data set of the deep learning model pre-training;
deep learning modeling: determining the node number and the attribute of an input layer and an output layer of deep learning according to a finite element model and a constitutive model of a finite difference model; designing a network structure type and hidden layer parameters of a deep learning model;
deep learning model training of dam safety monitoring index set prediction: pre-training the deep learning model by using a simulation data set, and finely adjusting the deep learning model by using actual measurement data;
step four: dam excitation-response inverse mapping model design
Designing an inverse mapping model of dam safety monitoring indexes: adopting a multilayer sensor model, wherein the parameters of an input layer and an output layer of the multilayer sensor model are the same as those of a deep learning model for predicting dam safety monitoring indexes;
training an inverse mapping model: taking external excitation and material strength parameters of the deep learning model as input data and output predicted values as sample marking values, and training a multilayer perceptron inverse mapping model;
calculating the strength parameters of the dam material: calculating an inverse mapping model from the safety monitoring index to the material strength parameter by adopting a correlation layer-by-layer propagation algorithm (LRP algorithm);
step five: correction of constitutive model of dam finite element analysis
And (3) verifying the strength parameters of the dam material: verifying the material strength estimation data of the inverse mapping model by using the actually measured dam material strength detection data;
calibrating a constitutive model: calibrating the inverse mapping model by using the material strength data detected by the dam structure at a safe and regular period and corresponding monitoring data to obtain the latest data of the material strength, and updating the constitutive model;
step six: quantitative early warning modeling for dam safety risk
Risk quantification early warning of dam safety monitoring indexes: predicting the numerical value of a dam safety monitoring index set by using a deep learning model based on the monitored external related indexes, and calculating confidence coefficient to obtain the risk degree of the index set;
risk quantitative prediction of dam material strength parameters: and calculating the numerical value of the dam material strength parameter by using an inverse mapping model based on the data of the safety monitoring index, predicting the change of the numerical value, and obtaining the risk degree of the material parameter.
The method for constructing the deep learning model of the excitation-response function simulation of the dam structure safety monitoring index set comprises the following steps: determining a deep learning network structure according to a finite element model and a finite difference model, using external excitation factors and dam material strength parameters as input data and dam safety monitoring indexes as output data, using a simulation data set generated by the finite element analysis model and the finite difference model to pre-train a deep learning model, using actual monitoring data to finely tune the deep learning model, realizing a deep learning modeling method for more accurately fitting an actual excitation-response relation, accurately predicting the change of a dam safety index set under various external excitations, and realizing the quantitative early warning of dam safety according to the dam safety index set;
the deep learning model construction method of the excitation-response bidirectional mapping modeling of dam material strength prediction comprises the following steps: on the basis of deep learning of excitation-response mapping relation simulation, a deep learning model of reverse mapping function simulation from response to excitation is designed, a bidirectional mapping simulation method between external excitation and structural strength parameters on an index set of dam safety monitoring is established, and whether the dam strength parameters change or not is determined according to a monitoring data set, so that quantitative early warning of dam safety is realized.
So far, the technical solutions of the present invention have been described in connection with the preferred embodiments shown in the drawings, but it is easily understood by those skilled in the art that the scope of the present invention is obviously not limited to these specific embodiments. Equivalent changes or substitutions of related technical features can be made by those skilled in the art without departing from the principle of the invention, and the technical scheme after the changes or substitutions can fall into the protection scope of the invention.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention; various modifications and alterations to this invention will become apparent to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A dam body deep learning model building method based on dual drive combination is characterized by comprising the following steps:
constructing a deep learning model of a building structure, wherein the deep learning model is input according to material strength and output according to a safety monitoring index of a structural main body, and when any material strength value is input and passes through the deep learning model, the safety monitoring index of the structural main body is uniquely determined;
constructing an inverse mapping model, wherein the network structure type and the layer parameter of the inverse mapping model are the same as those of the deep learning model, the inverse mapping model is used for determining the material strength according to the safety index of a structural main body, verifying the risk degree of the material strength of the current structural main body according to the calculated material strength, determining the safety monitoring index of the building structure according to the material strength of the current structural main body, and determining the risk degree of the building structure;
determining an alarm mechanism according to the risk degree of the material and the risk degree of the structural body so as to alarm the safety degree of the structural body;
when a deep learning model of a building structure is constructed, acquiring a structural design model of the building structure and material strength theoretical data based on the structural design model;
acquiring historical monitoring data of the building structure, wherein the historical monitoring data comprises a structure main body safety monitoring index set, the structure main body safety monitoring index set is formed on the basis of a plurality of monitoring parameters, and the number of the monitoring parameters is set to be n;
and when an alarm mechanism is determined, adjusting the number of monitoring parameters in the historical monitoring data according to the actual values of the risk degree of the material and the risk degree of the structural body.
2. The dam body depth learning model building method based on dual drive combination according to claim 1,
firstly, setting a material parameter threshold range, comparing a material strength S0 obtained by calculation according to an inverse mapping model with an actual material strength value S when evaluating the risk degree of the material, if the actual material strength value S belongs to the material parameter threshold range, but the calculated material strength S0 is equal to the actual material strength value S, indicating that the risk degree of the material is low, setting the material as D1, and adopting a second time period T2 as an early warning period;
if the actual material strength value S belongs to the material parameter threshold range, but the calculated material strength S0 is not equal to the actual material strength value S, the risk of the material is moderate, D2 is set, and a first time period T1 is used as an early warning period;
if the actual material strength value S does not belong to the material parameter threshold range, but the calculated material strength S0 is not equal to the actual material strength value S, the risk degree of the material is high, the value is set to D3, and instant early warning is adopted.
3. The dam body depth learning model building method based on dual drive combination according to claim 2,
when a first time interval T1 is adopted as an early warning period, adjusting the length of the early warning period according to the relation between the calculated material strength S0 and the actual material strength value S;
if the calculated material strength S0 is greater than the actual material strength value S, correcting the first time period T1 by adopting a first coefficient k 1;
if the calculated material strength S0< the actual material strength value S, the first time period T1 is corrected using the second coefficient k 2.
4. The dam body depth learning model building method based on dual drive combination according to claim 3,
when the first time interval T1 is corrected by adopting the first coefficient k1, determining an adjusting parameter of the first coefficient k1 according to the relation between the calculated actual difference value of the material strength S0 and the actual material strength value S and the preset standard difference value delta S10;
if S0-S is less than or equal to delta S10, the first coefficient k1 is adjusted to be 0.5 xk 1;
if S0-S > Δ S10, the first coefficient k1 is adjusted to 0.2 × k 1.
5. The dam body depth learning model building method based on dual drive combination according to claim 4,
when an instant alarm is carried out, the early warning volume of the early warning period is adjusted according to the relation between the calculated material strength S0 and the actual material strength value S;
if the calculated material strength S0 is greater than the actual material strength value S, alarming by adopting a first volume V1;
and if the calculated material strength S0 is less than the actual material strength value S, alarming by adopting a second volume V2, and enabling the first volume V1 to be less than a second volume V2.
6. The dam body depth learning model building method based on dual drive combination according to claim 5,
a structural main body safety monitoring index set is preset, a plurality of discontinuous structural main body safety monitoring values are arranged in the structural main body safety monitoring index set, whether the structural main body safety monitoring index belongs to the structural main body safety monitoring index set or not is judged according to the structural main body safety monitoring index calculated by the deep learning model, and if the structural main body safety monitoring index belongs to the structural main body safety monitoring index set, the structural main body safety under the current material strength is represented;
if not, the risk degree of the structural body under the current material strength is high.
7. The dam body depth learning model building method based on dual drive combination according to claim 6,
when the structural subject safety monitoring index does not belong to the structural subject safety monitoring index set, determining a maximum value M10 and a minimum value M20 in the structural subject safety monitoring index set;
if the maximum value M10 is greater than the structure main body safety monitoring index is greater than the minimum value M20, adding the corresponding strength and the structure main body safety monitoring index into a historical data set to serve as historical data of a deep learning model, and increasing training times;
and if the structural main body safety monitoring index is larger than the maximum value M10, determining the material strength corresponding to the current structural main body safety monitoring index as the unqualified strength.
8. The dam body depth learning model building method based on dual drive combination according to claim 7,
when the training times are increased, selecting the amplitude of the increase of the training times according to the relationship between the safety monitoring index of the structural main body and the safety monitoring values of two adjacent structural main bodies;
if the safety monitoring index of the structural main body is the average value of the safety monitoring values of two adjacent structural main bodies, increasing the training times by adopting a first increment I1;
and if the safety monitoring index of the structural main body is close to any one of the two adjacent safety monitoring values of the structural main body, increasing the training times by adopting a second increment I2.
9. The dam body deep learning model construction method based on dual drive combination according to claim 8,
when the training times are increased by adopting a second increment I2, setting the safety monitoring values of two adjacent structure bodies as A1 and A2 respectively, wherein A1 is greater than A2, and if the absolute value of the difference between the safety monitoring index of the structure body and A1 is smaller than the absolute value of the difference between the safety monitoring index of the structure body and A2, adjusting the second increment I2 by adopting a first parameter alpha;
and if the absolute value of the difference between the safety monitoring index of the structural main body and A1 is larger than the absolute value of the difference between the safety monitoring index of the structural main body and A2, adjusting the second increment I2 by using a second parameter beta.
10. The dam depth learning model building method based on dual drive combination according to claim 9, wherein when a first parameter α is used to adjust a second increment I2, the adjusted second increment is I21' = I2 x (1 + α), and the training times are increased by using the modified second increment;
when a second increment I2 is adjusted by using a second parameter beta, the adjusted second increment is I22' = I2 x (1 + beta), and the training times are increased by using the corrected second increment, wherein the first parameter alpha and the second parameter beta are both decimals larger than 0 and smaller than 1;
the building structure is a dam, and the structure main body is a dam body.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210324685.XA CN114611833B (en) | 2022-03-30 | 2022-03-30 | Dam body deep learning model construction method based on dual-drive combination |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210324685.XA CN114611833B (en) | 2022-03-30 | 2022-03-30 | Dam body deep learning model construction method based on dual-drive combination |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114611833A true CN114611833A (en) | 2022-06-10 |
| CN114611833B CN114611833B (en) | 2022-08-30 |
Family
ID=81867168
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210324685.XA Active CN114611833B (en) | 2022-03-30 | 2022-03-30 | Dam body deep learning model construction method based on dual-drive combination |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114611833B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103984856A (en) * | 2014-06-19 | 2014-08-13 | 江苏新瑞峰信息科技有限公司 | System and method for predicting longitudinal crack width of wall body based on neural network |
| CN107480341A (en) * | 2017-07-21 | 2017-12-15 | 河海大学 | A kind of dam safety comprehensive method based on deep learning |
| US20190017911A1 (en) * | 2017-07-11 | 2019-01-17 | Qatar University | Real-time structural damage detection by convolutional neural networks |
| CN113191055A (en) * | 2021-05-06 | 2021-07-30 | 河海大学 | Dam material performance parameter inversion method based on deep reinforcement network |
| CN113569908A (en) * | 2021-06-24 | 2021-10-29 | 河南大学 | Bridge technical condition and disease prediction back-stepping calculation method based on deep learning |
-
2022
- 2022-03-30 CN CN202210324685.XA patent/CN114611833B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103984856A (en) * | 2014-06-19 | 2014-08-13 | 江苏新瑞峰信息科技有限公司 | System and method for predicting longitudinal crack width of wall body based on neural network |
| US20190017911A1 (en) * | 2017-07-11 | 2019-01-17 | Qatar University | Real-time structural damage detection by convolutional neural networks |
| CN107480341A (en) * | 2017-07-21 | 2017-12-15 | 河海大学 | A kind of dam safety comprehensive method based on deep learning |
| CN113191055A (en) * | 2021-05-06 | 2021-07-30 | 河海大学 | Dam material performance parameter inversion method based on deep reinforcement network |
| CN113569908A (en) * | 2021-06-24 | 2021-10-29 | 河南大学 | Bridge technical condition and disease prediction back-stepping calculation method based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 郭张军等: "基于深度学习的大坝变形预测模型", 《水电能源科学》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114611833B (en) | 2022-08-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110414117B (en) | Method for predicting sealing reliability of soft package lithium ion battery | |
| CN107797067A (en) | Lithium ion battery life migration prediction method based on deep learning | |
| CN108803486A (en) | Numerical control machining tool heat error prediction based on deep learning network in parallel and compensation method | |
| CN111579993B (en) | Lithium battery capacity online estimation method based on convolutional neural network | |
| CN115238394B (en) | Multi-source uncertainty hybrid reliability digital twin modeling method for composite material structure | |
| CN113917334A (en) | Battery health state estimation method based on evolution LSTM self-encoder | |
| CN114280490B (en) | Lithium ion battery state of charge estimation method and system | |
| CN118095813B (en) | Visual monitoring method and system for foundation settlement based on BIM technology | |
| CN114611833B (en) | Dam body deep learning model construction method based on dual-drive combination | |
| CN117874655A (en) | Dynamic dam safety monitoring index planning method considering multiple influence factors | |
| CN112985692B (en) | Atmospheric pressure sensor error calibration method integrating polynomial and learning model | |
| CN112966399B (en) | Pulse tube refrigerator working condition prediction method and system based on machine learning | |
| CN112327169B (en) | Lithium battery residual life prediction method | |
| CN110990916B (en) | Integration method for evaluating and predicting long-term operation safety of dam by considering hysteresis effect | |
| CN105301985A (en) | Method and system for measuring length of calcium carbide furnace electrode | |
| CN117408108A (en) | Health monitoring method based on finite element and real-time monitoring | |
| CN114800036B (en) | Equipment health state assessment method | |
| CN114417650B (en) | Dam body safety monitoring method and system | |
| CN110057742B (en) | On-line detection method for porosity of lithium ion battery pole piece and application thereof | |
| CN110750884B (en) | Instant on-line instrument checksum diagnosis method through self-adaptive support vector machine algorithm | |
| CN118144571B (en) | Multi-model fusion power battery thermal runaway early warning system and method | |
| CN113821914B (en) | Low-cost prediction method for cycle life of lithium ion battery | |
| CN116028757B (en) | Optimal soft measurement model generation method and system based on multi-source information fusion | |
| CN118184105B (en) | Intelligent smelting temperature monitoring method and system for optical glass | |
| CN117057027B (en) | Wall heat preservation and insulation performance evaluation method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |