CN112163316A - Tunneling parameter prediction method of hard rock tunnel boring machine based on deep learning - Google Patents

Tunneling parameter prediction method of hard rock tunnel boring machine based on deep learning Download PDF

Info

Publication number
CN112163316A
CN112163316A CN202010893005.7A CN202010893005A CN112163316A CN 112163316 A CN112163316 A CN 112163316A CN 202010893005 A CN202010893005 A CN 202010893005A CN 112163316 A CN112163316 A CN 112163316A
Authority
CN
China
Prior art keywords
tunneling
section
deep learning
layer
sample set
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202010893005.7A
Other languages
Chinese (zh)
Inventor
赵生捷
刘诗洋
张�林
张荣庆
吕征南
马慧生
李凡
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tongji University
Original Assignee
Tongji University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tongji University filed Critical Tongji University
Priority to CN202010893005.7A priority Critical patent/CN112163316A/en
Publication of CN112163316A publication Critical patent/CN112163316A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • E—FIXED CONSTRUCTIONS
    • E21—EARTH DRILLING; MINING
    • E21D—SHAFTS; TUNNELS; GALLERIES; LARGE UNDERGROUND CHAMBERS
    • E21D9/00—Tunnels or galleries, with or without linings; Methods or apparatus for making thereof; Layout of tunnels or galleries
    • E21D9/10—Making by using boring or cutting machines
    • E21D9/11—Making by using boring or cutting machines with a rotary drilling-head cutting simultaneously the whole cross-section, i.e. full-face machines
    • 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

Abstract

The invention relates to a hard rock tunnel boring machine tunneling parameter prediction method based on deep learning, which specifically comprises the following steps: s1, acquiring tunneling data continuously recorded by a sensor, dividing an original tunneling section for optimization, acquiring a target tunneling section group, and dividing the target tunneling section group into a training sample set and a test sample set; s2, constructing a convolutional neural network model, inputting a training sample set into the convolutional neural network model for deep learning, and obtaining a preliminary prediction model; s3, obtaining a test sample set, inputting a preliminary prediction model for testing, and adjusting the hyper-parameters of the preliminary prediction model according to a test result; and S4, repeating the step S3, recording the prediction error of the model after each adjustment of the hyper-parameters, and outputting the current model as a final prediction model when the prediction error is smaller than a set threshold value. Compared with the prior art, the method has the advantages of improving the prediction precision of the rotating speed and the propelling speed of the cutter head of the tunneling machine, avoiding economic loss, time delay of construction period and the like.

Description

Tunneling parameter prediction method of hard rock tunnel boring machine based on deep learning
Technical Field
The invention relates to the field of intelligent control of hard rock tunnel boring machines, in particular to a method for predicting boring parameters of a hard rock tunnel boring machine based on deep learning.
Background
A hard rock tunnelling Machine (TBM) is a large Machine tool that is used specifically for tunnelling. In the traditional tunneling process, the tunneling parameters of the TBM are usually controlled by means of manual experience, the tunneling parameters need to be adjusted in real time according to sensor data of the TBM, and due to the fact that an operator cannot analyze rock-machine action information and pre-judge geological conditions, engineering accidents such as blocking and the like occur sometimes, economic loss and construction period delay are caused. Therefore, finding an intelligent prediction algorithm for a set of TBM tunneling parameters is very important for tunnel engineering.
Disclosure of Invention
The invention aims to overcome the defects of economic loss and construction period delay caused by analysis of rock-machine action information and insufficient prediction of geological conditions in the prior art, and provides a hard rock tunnel boring machine tunneling parameter prediction method based on deep learning.
The purpose of the invention can be realized by the following technical scheme:
a method for predicting tunneling parameters of a hard rock tunnel boring machine based on deep learning specifically comprises the following steps:
s1, acquiring tunneling data continuously recorded by a sensor of a tunneling machine, dividing an original tunneling section according to the tunneling data, optimizing the original tunneling section to obtain a target tunneling section group, and dividing the target tunneling section group into a training sample set and a test sample set;
s2, constructing a convolutional neural network model, wherein the convolutional neural network model comprises a convolutional layer, a pooling layer, a linear rectifying layer and a full-link layer, and inputting the training sample set into the convolutional neural network model for deep learning to obtain a preliminary prediction model;
s3, obtaining the test sample set, inputting the preliminary prediction model for testing, and adjusting the hyper-parameters of the preliminary prediction model according to a test result;
and S4, repeating the step S3, recording the prediction error of the model after each adjustment of the hyper-parameters, and outputting the current model as a final prediction model when the prediction error is smaller than a set threshold value.
The step S1 specifically includes the following steps:
s101, removing a shutdown section according to basic parameters of the heading machine, dividing an original heading section, and removing a heading section with too short heading time and poor geological shutdown in the original heading section to obtain a target heading section group;
s102, according to the rising trend of the propelling speed of the heading machine, further dividing each heading section in the target heading section group into an idle pushing section, a rising section and a stable section;
s103, arranging the sensor data of the target time period before the ascending section into an ascending section number matrix together, and establishing a stable section tunneling parameter matrix according to the stable section;
and S104, dividing the ascending section number matrix and the stable section tunneling parameter matrix of all the tunneling sections in the target tunneling section group into a training sample set and a testing sample set according to a preset proportion.
The storage form of the training sample set and the test sample set is a data file, so that the model can be read quickly.
Further, the basic parameters of the heading machine in the step S101 include thrust, torque, thrust speed and cutterhead rotation speed of the heading machine.
Further, the steady-state tunneling parameter matrix in step S103 specifically includes an average cutter head rotation speed and an average propulsion speed of the steady-state segment.
Further, the depth of the rise segment number matrix has the same value as the duration of the target period.
Further, the step S2 specifically includes the following steps:
s201, constructing a convolutional neural network model comprising a convolutional layer, a pooling layer, a linear rectifying layer and a full-connection layer;
s202, the convolutional neural network model obtains a training sample set, and according to a preset loss function and an optimization algorithm, an ascending section numerical matrix in the training sample set is used as input, and a stable section tunneling parameter matrix is used as input to conduct supervised learning of the convolutional neural network model.
Further, the structure of the convolutional neural network model specifically includes 1 input layer, 1 convolutional layer, a combination of 1 pooling layer and 2 convolutional layers, a combination of 1 pooling layer and 3 convolutional layers, a combination of 1 pooling layer and 1 linear rectifying layer, 1 full-connection layer and 1 full-connection layer.
Further, the loss function is specifically a mean square error function, and the formula is specifically as follows:
Figure BDA0002657479710000021
wherein, yiFor the (i) th predicted value,
Figure BDA0002657479710000022
n is the actual value and the number of samples.
Further, the optimization algorithm is specifically an Adam algorithm, and the flow is shown in the following formula:
mt＝β1mt-1+(1-β1)gt
Figure BDA0002657479710000031
Figure BDA0002657479710000032
Figure BDA0002657479710000033
Figure BDA0002657479710000034
where eta is the initial learning rate, beta1Exponential decay Rate, beta, estimated for the first moment2Exponential decay Rate estimated for the second moment, e is a parameter that prevents division by zero, mtMomentum at time t, vtFor small batch random gradients at time t, an exponentially weighted moving average of the squares of the elements, gtFor a small batch of random gradients at time t,
Figure BDA0002657479710000035
for the moment of momentum after the correction t,
Figure BDA0002657479710000036
the modified small batch random gradient at time t is an exponentially weighted moving average of the squares of the elements, thetatAre neural network model parameters.
Further, the hyper-parameters of the model comprise batch size, the number of neurons of the full connection layer and parameters of an optimization algorithm, and the method for adjusting the hyper-parameters is a grid search method.
Compared with the prior art, the invention has the following beneficial effects:
according to the method, the tunneling data continuously recorded by the sensor of the tunneling machine is obtained, the convolutional neural network model comprising the convolutional layer, the pooling layer, the linear rectifying layer and the full-connection layer is constructed to predict the rotating speed and the propelling speed of the cutter head of the stable section, the error between the predicted value and the true value of the tunneling effect of the tunneling machine is reduced, the prediction precision of the rotating speed and the propelling speed of the cutter head of the tunneling machine is improved, an operator can conveniently adjust the advancing route and the working mode of the tunneling machine according to the prediction result, and economic loss and construction period delay are avoided.
Drawings
FIG. 1 is a schematic flow diagram of the present invention;
FIG. 2 is a schematic diagram of the present invention dividing the idle push section, the rise section and the stable section;
FIG. 3 is a schematic diagram of the structure of a convolutional neural network model according to the present invention;
fig. 4 is a diagram illustrating the influence of MSE and MAE initial learning rates on prediction error according to the present invention.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments. The present embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a specific operation process are given, but the scope of the present invention is not limited to the following embodiments.
Example one
As shown in fig. 1, a method for predicting tunneling parameters of a hard rock tunnel boring machine based on deep learning specifically includes the following steps:
s1, acquiring tunneling data continuously recorded by a sensor of a tunneling machine, dividing an original tunneling section according to the tunneling data, optimizing the original tunneling section to obtain a target tunneling section group, and dividing the target tunneling section group into a training sample set and a test sample set;
s2, constructing a convolutional neural network model, wherein the convolutional neural network model comprises a convolutional layer, a pooling layer, a linear rectifying layer and a full-link layer, and inputting a training sample set into the convolutional neural network model for deep learning to obtain a preliminary prediction model;
s3, obtaining a test sample set, inputting a preliminary prediction model for testing, and adjusting the hyper-parameters of the preliminary prediction model according to a test result;
and S4, repeating the step S3, recording the prediction error of the model after each adjustment of the hyper-parameters, and outputting the current model as a final prediction model when the prediction error is smaller than a set threshold value.
In the embodiment, the tunneling data is data recorded for 802 days continuously, the data recorded every day comprises 1-10 tunneling sections, each tunneling section comprises 198 characteristics, and the system related to the characteristics comprises a propulsion system, a main driving system and a steering system.
In this embodiment, the number of heading segments in the target heading segment group is 13099.
Step S1 specifically includes the following steps:
s101, removing a shutdown section according to basic parameters of the heading machine, dividing an original heading section, and removing a heading section with too short heading time and poor geological shutdown in the original heading section to obtain a target heading section group;
s102, as shown in the figure 2, according to the rising trend of the propelling speed of the heading machine, each heading section in the target heading section group is further divided into an idle pushing section, a rising section and a stable section;
s103, removing two constant items from 198 characteristic data of the previous 30 seconds of the ascending section, arranging the constant items into a 14x14x30 ascending section number matrix M1 of the ascending section, and establishing a stable section tunneling parameter matrix according to the stable section;
and S104, dividing the ascending section number matrix and the stable section tunneling parameter matrix of all the tunneling sections in the target tunneling section group into a training sample set and a testing sample set according to the ratio of 7: 3.
The training sample set and the testing sample set are stored in the form of npy data files, so that the model can be read quickly.
The basic parameters of the heading machine in the step S101 comprise the thrust, the torque, the propulsion speed and the cutter head rotating speed of the heading machine.
In step S103, the steady-state tunneling parameter matrix specifically includes the average cutter head rotation speed and the average propulsion speed of the steady-state segment.
Step S2 specifically includes the following steps:
s201, constructing a convolutional neural network model based on a VGG-16 model and comprising a convolutional layer, a pooling layer, a linear rectifying layer and a full-connection layer;
s202, acquiring a training sample set by the convolutional neural network model, and performing supervised learning of the convolutional neural network model by taking an ascending section numerical matrix in the training sample set as input and taking a stable section tunneling parameter matrix as input according to a preset loss function and an optimization algorithm.
As shown in fig. 3, the structure of the convolutional neural network model specifically includes 1 input layer, 1 convolutional layer, a combination of 1 pooling layer and 2 convolutional layers, a combination of 1 pooling layer and 3 convolutional layers, a combination of 1 pooling layer and 1 linear rectifying layer, 1 fully-connected layer, and 1 fully-connected layer.
The loss function is specifically a mean square error function (MSE), and the formula is specifically as follows:
Figure BDA0002657479710000051
wherein, yiFor the (i) th predicted value,
Figure BDA0002657479710000052
n is the actual value and the number of samples.
The optimization algorithm is specifically Adam algorithm, and the flow is shown as the following formula:
mt＝β1mt-1+(1-β1)gt
Figure BDA0002657479710000053
Figure BDA0002657479710000054
Figure BDA0002657479710000055
Figure BDA0002657479710000056
where eta is the initial learning rate, beta1Exponential decay Rate, beta, estimated for the first moment2Exponential decay Rate estimated for the second moment, e is a parameter that prevents division by zero, mtMomentum at time t, vtFor small batch random gradients at time t, an exponentially weighted moving average of the squares of the elements, gtRandom gradient for small batch at time t，
Figure BDA0002657479710000057
For the moment of momentum after the correction t,
Figure BDA0002657479710000058
the modified small batch random gradient at time t is an exponentially weighted moving average of the squares of the elements, thetatAre neural network model parameters.
The hyper-parameters of the model comprise batch size, the number of neurons of the full connection layer and parameters of an optimization algorithm, and the method for adjusting the hyper-parameters is a grid search method.
As shown in fig. 4, for the initial learning rate of the mean square error function (MSE) and the Mean Absolute Error (MAE), when other parameters are not changed, the overall error is the smallest when the initial learning rate is set to 0.0001, and therefore the initial learning rate of the Adam algorithm in the present invention is 0.0001.
In addition, it should be noted that the specific implementation examples described in this specification may have different names, and the above contents described in this specification are only illustrations of the structures of the present invention. All equivalent or simple changes in the structure, characteristics and principles of the invention are included in the protection scope of the invention. Various modifications or additions may be made to the described embodiments or methods may be similarly employed by those skilled in the art without departing from the scope of the invention as defined in the appending claims.

Claims (10)

1. A method for predicting tunneling parameters of a hard rock tunnel boring machine based on deep learning is characterized by comprising the following steps:
s1, acquiring tunneling data continuously recorded by a sensor of a tunneling machine, dividing an original tunneling section according to the tunneling data, optimizing the original tunneling section to obtain a target tunneling section group, and dividing the target tunneling section group into a training sample set and a test sample set;
s2, constructing a convolutional neural network model, wherein the convolutional neural network model comprises a convolutional layer, a pooling layer, a linear rectifying layer and a full-link layer, and inputting the training sample set into the convolutional neural network model for deep learning to obtain a preliminary prediction model;
s3, obtaining the test sample set, inputting the preliminary prediction model for testing, and adjusting the hyper-parameters of the preliminary prediction model according to a test result;
and S4, repeating the step S3, recording the prediction error of the model after each adjustment of the hyper-parameters, and outputting the current model as a final prediction model when the prediction error is smaller than a set threshold value.
2. The hard rock tunnel boring machine tunneling parameter prediction method based on deep learning according to claim 1, wherein the step S1 specifically comprises the following steps:
s101, removing a shutdown section according to basic parameters of the heading machine, dividing an original heading section, and removing a heading section with too short heading time and poor geological shutdown in the original heading section to obtain a target heading section group;
s102, according to the rising trend of the propelling speed of the heading machine, further dividing each heading section in the target heading section group into an idle pushing section, a rising section and a stable section;
s103, arranging the sensor data of the target time period before the ascending section into an ascending section number matrix together, and establishing a stable section tunneling parameter matrix according to the stable section;
and S104, dividing the ascending section number matrix and the stable section tunneling parameter matrix of all the tunneling sections in the target tunneling section group into a training sample set and a testing sample set according to a preset proportion.
3. The hard rock tunneling machine tunneling parameter prediction method based on deep learning of claim 2, wherein the basic parameters of the tunneling machine in the step S101 include thrust, torque, propulsion speed and cutterhead rotation speed of the tunneling machine.
4. The hard rock tunnel boring machine tunneling parameter prediction method based on deep learning of claim 3, wherein the steady-state tunneling parameter matrix in step S103 specifically includes an average cutterhead rotation speed and an average propulsion speed of the steady state.
5. The hard rock tunnel boring machine tunneling parameter prediction method based on deep learning of claim 2, characterized in that the numerical value of the depth of the ascent stage numerical matrix is the same as the numerical value of the duration of the target time period.
6. The hard rock tunnel boring machine tunneling parameter prediction method based on deep learning according to claim 2, wherein the step S2 specifically comprises the following steps:
s201, constructing a convolutional neural network model comprising a convolutional layer, a pooling layer, a linear rectifying layer and a full-connection layer;
s202, the convolutional neural network model obtains a training sample set, and according to a preset loss function and an optimization algorithm, an ascending section numerical matrix in the training sample set is used as input, and a stable section tunneling parameter matrix is used as input to conduct supervised learning of the convolutional neural network model.
7. The hard rock tunneling machine tunneling parameter prediction method based on deep learning of claim 6, characterized in that the convolutional neural network model has a structure specifically including 1 input layer, 1 convolutional layer, a combination of 1 pooling layer and 2 convolutional layers, a combination of 1 pooling layer and 3 convolutional layers, a combination of 1 pooling layer and 1 linear rectifying layer, 1 full-connection layer and 1 full-connection layer.
8. The hard rock tunnel boring machine tunneling parameter prediction method based on deep learning of claim 6, characterized in that the loss function is specifically a mean square error function, and the formula is specifically as follows:
Figure FDA0002657479700000021
wherein, yiFor the (i) th predicted value,
Figure FDA0002657479700000022
n is the actual value and the number of samples.
9. The hard rock tunnel boring machine tunneling parameter prediction method based on deep learning of claim 6 is characterized in that the optimization algorithm is specifically Adam algorithm, and the flow is shown as the following formula:
mt＝β1mt-1+(1-β1)gt
Figure FDA0002657479700000023
Figure FDA0002657479700000024
Figure FDA0002657479700000025
Figure FDA0002657479700000026
where eta is the initial learning rate, beta1Exponential decay Rate, beta, estimated for the first moment2Exponential decay Rate estimated for the second moment, e is a parameter that prevents division by zero, mtMomentum at time t, vtFor small batch random gradients at time t, an exponentially weighted moving average of the squares of the elements, gtFor a small batch of random gradients at time t,
Figure FDA0002657479700000027
for the moment of momentum after the correction t,
Figure FDA0002657479700000031
the modified small batch random gradient at time t is an exponentially weighted moving average of the squares of the elements, thetatAre neural network model parameters.
10. The hard rock tunnel boring machine tunneling parameter prediction method based on deep learning of claim 6, wherein the hyper-parameters of the model include batch size, the number of neurons in a fully connected layer, and parameters of an optimization algorithm, and the method for adjusting the hyper-parameters is a grid search method.
CN202010893005.7A 2020-08-31 2020-08-31 Tunneling parameter prediction method of hard rock tunnel boring machine based on deep learning Pending CN112163316A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010893005.7A CN112163316A (en) 2020-08-31 2020-08-31 Tunneling parameter prediction method of hard rock tunnel boring machine based on deep learning

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010893005.7A CN112163316A (en) 2020-08-31 2020-08-31 Tunneling parameter prediction method of hard rock tunnel boring machine based on deep learning

Publications (1)

Publication Number Publication Date
CN112163316A true CN112163316A (en) 2021-01-01

Family

ID=73859618

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010893005.7A Pending CN112163316A (en) 2020-08-31 2020-08-31 Tunneling parameter prediction method of hard rock tunnel boring machine based on deep learning

Country Status (1)

Country Link
CN (1) CN112163316A (en)

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113222273A (en) * 2021-05-26 2021-08-06 中铁十八局集团有限公司 TBM tunneling rate prediction method based on GWO-FW-MKL-SVR algorithm
CN113377017A (en) * 2021-07-19 2021-09-10 中国铁建重工集团股份有限公司 Earth pressure balance shield machine and propelling speed prediction method, device and medium thereof
CN113420506A (en) * 2021-06-30 2021-09-21 北京交通大学 Method for establishing prediction model of tunneling speed, prediction method and device
CN113642082A (en) * 2021-08-24 2021-11-12 上海交通大学 A-CNN method and system for predicting TBM utilization rate
CN113685188A (en) * 2021-08-16 2021-11-23 中铁十八局集团有限公司 TBM tunneling optimization method based on physical characteristics of rock slag
CN114722697A (en) * 2022-03-09 2022-07-08 山东拓新电气有限公司 Method and device for determining control parameters of heading machine based on machine learning
CN115268272A (en) * 2022-08-11 2022-11-01 北京交通大学 TBM control parameter decision method and device based on tunneling load prediction
CN115618222A (en) * 2022-06-21 2023-01-17 北京交通大学 Prediction method of tunneling response parameters
CN116936119A (en) * 2023-09-15 2023-10-24 山东优杰生物科技有限公司 Blood bank intelligent scheduling management system and method

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110852423A (en) * 2019-11-12 2020-02-28 中铁工程装备集团有限公司 Tunnel boring machine excavation performance and control parameter prediction method based on transfer learning
CN111144635A (en) * 2019-12-20 2020-05-12 山东大学 TBM operation parameter decision method and system based on deep learning
CN111582610A (en) * 2020-07-13 2020-08-25 清华四川能源互联网研究院 Prediction method for family energy decomposition based on convolutional neural network

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110852423A (en) * 2019-11-12 2020-02-28 中铁工程装备集团有限公司 Tunnel boring machine excavation performance and control parameter prediction method based on transfer learning
CN111144635A (en) * 2019-12-20 2020-05-12 山东大学 TBM operation parameter decision method and system based on deep learning
CN111582610A (en) * 2020-07-13 2020-08-25 清华四川能源互联网研究院 Prediction method for family energy decomposition based on convolutional neural network

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113222273A (en) * 2021-05-26 2021-08-06 中铁十八局集团有限公司 TBM tunneling rate prediction method based on GWO-FW-MKL-SVR algorithm
CN113420506A (en) * 2021-06-30 2021-09-21 北京交通大学 Method for establishing prediction model of tunneling speed, prediction method and device
CN113377017A (en) * 2021-07-19 2021-09-10 中国铁建重工集团股份有限公司 Earth pressure balance shield machine and propelling speed prediction method, device and medium thereof
CN113685188A (en) * 2021-08-16 2021-11-23 中铁十八局集团有限公司 TBM tunneling optimization method based on physical characteristics of rock slag
CN113642082A (en) * 2021-08-24 2021-11-12 上海交通大学 A-CNN method and system for predicting TBM utilization rate
CN113642082B (en) * 2021-08-24 2024-04-09 上海交通大学 A-CNN method and system for predicting TBM utilization rate
CN114722697A (en) * 2022-03-09 2022-07-08 山东拓新电气有限公司 Method and device for determining control parameters of heading machine based on machine learning
CN115618222A (en) * 2022-06-21 2023-01-17 北京交通大学 Prediction method of tunneling response parameters
CN115268272A (en) * 2022-08-11 2022-11-01 北京交通大学 TBM control parameter decision method and device based on tunneling load prediction
CN116936119A (en) * 2023-09-15 2023-10-24 山东优杰生物科技有限公司 Blood bank intelligent scheduling management system and method

Similar Documents

Publication Publication Date Title
CN112163316A (en) Tunneling parameter prediction method of hard rock tunnel boring machine based on deep learning
CN110852423B (en) Tunnel boring machine excavation performance and control parameter prediction method based on transfer learning
CN110083125B (en) Machine tool thermal error modeling method based on deep learning
CN110096827B (en) Shield tunneling machine parameter optimization method based on deep neural network
CN112347580A (en) Shield tunneling machine cutter head torque real-time prediction method and system
CN110147875A (en) A kind of shield machine auxiliary cruise method based on LSTM neural network
CN114418469B (en) LGBM-NSGA-III-based shield proximity construction parameter multi-objective optimization method and device
CN107248026A (en) The quantitative approach of shield driving parameter is predicted using equivalent rock mass basic quality's index
Wallace et al. A system for real-time drilling performance optimization and automation based on statistical learning methods
CN110895730A (en) TBM tunneling parameter prediction method based on LSTM algorithm
Sun et al. Optimized throughput improvement of assembly flow line with digital twin online analytics
CN112650053A (en) Genetic algorithm optimization-based motor PID self-tuning method for BP neural network
CN115481565A (en) Earth pressure balance shield tunneling parameter prediction method based on LSTM and ant colony algorithm
CN114662793A (en) Business process remaining time prediction method and system based on interpretable hierarchical model
CN115329853A (en) Equipment parameter prediction and knowledge transfer method based on multi-source domain migration
CN113361824B (en) Earth pressure balance shield machine and propelling speed prediction method, device and storage medium thereof
CN117252086A (en) Method, system and equipment for health evaluation and degradation prediction of shield tunneling machine cutterhead
CN116522777A (en) Drilling process drilling rate online prediction method and system based on multi-source information fusion
CN113435055B (en) Self-adaptive migration prediction method and system in shield cutter head torque field
CN116432855A (en) Tunnel collapse condition prediction method based on tunneling data
CN111946258B (en) GRU-based sliding orientation intelligent control method
CN114818500A (en) Method for predicting soil bin pressure based on LSTM algorithm
CN113742913A (en) Python-based ADAMS post-processing file K & C parameter extraction method and system
Salloum et al. Agent-Based Simulation Model for the Real-Time Evaluation of Tunnel Boring Machines Using Deep Learning
CN109138969B (en) Prediction method and device for drilling state variable and storage device

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