Traditional track dynamic geometric state(TDGS)simulation incurs substantial computational burdens,posing challenges for developing reliability assessment approach that accounts for TDGS.To overcome these,firstly,a si...Traditional track dynamic geometric state(TDGS)simulation incurs substantial computational burdens,posing challenges for developing reliability assessment approach that accounts for TDGS.To overcome these,firstly,a simulation-based TDGS model is established,and a surrogate-based model,grid search algorithm-particle swarm optimization-genetic algorithm-multi-output least squares support vector regression,is established.Among them,hyperparameter optimization algorithm’s effectiveness is confirmed through test functions.Subsequently,an adaptive surrogate-based probability density evolution method(PDEM)considering random track geometry irregularity(TGI)is developed.Finally,taking curved train-steel spring floating slab track-U beam as case study,the surrogate-based model trained on simulation datasets not only shows accuracy in both time and frequency domains,but also surpasses existing models.Additionally,the adaptive surrogate-based PDEM shows high accuracy and efficiency,outperforming Monte Carlo simulation and simulation-based PDEM.The reliability assessment shows that the TDGS part peak management indexes,left/right vertical dynamic irregularity,right alignment dynamic irregularity,and track twist,have reliability values of 0.9648,0.9918,0.9978,and 0.9901,respectively.The TDGS mean management index,i.e.,track quality index,has reliability value of 0.9950.These findings show that the proposed framework can accurately and efficiently assess the reliability of curved low-stiffness track-viaducts,providing a theoretical basis for the TGI maintenance.展开更多
Underground excavation can lead to stress redistribution and result in an excavation damaged zone(EDZ),which is an important factor affecting the excavation stability and support design.Accurately estimating the thick...Underground excavation can lead to stress redistribution and result in an excavation damaged zone(EDZ),which is an important factor affecting the excavation stability and support design.Accurately estimating the thickness of EDZ is essential to ensure the safety of the underground excavation.In this study,four novel hybrid ensemble learning models were developed by optimizing the extreme gradient boosting(XGBoost)and random forest(RF)algorithms through simulated annealing(SA)and Bayesian optimization(BO)approaches,namely SA-XGBoost,SA-RF,BO XGBoost and BO-RF models.A total of 210 cases were collected from Xiangxi Gold Mine in Hunan Province and Fankou Lead-zinc Mine in Guangdong Province,China,including seven input indicators:embedding depth,drift span,uniaxial compressive strength of rock,rock mass rating,unit weight of rock,lateral pressure coefficient of roadway and unit consumption of blasting explosive.The performance of the proposed models was evaluated by the coefficient of determination,root mean squared error,mean absolute error and variance accounted for.The results indicated that the SA-XGBoost model performed best.The Shapley additive explanations method revealed that the embedding depth was the most important indicator.Moreover,the convergence curves suggested that the SA-XGBoost model can reduce the generalization error and avoid overfitting.展开更多
Support vector regression （SVR） method is a novel type of learning machine algorithms, which is seldom applied to the development of urban atmospheric quality models under multiple socio-economic factors. This study...Support vector regression （SVR） method is a novel type of learning machine algorithms, which is seldom applied to the development of urban atmospheric quality models under multiple socio-economic factors. This study presents four SVR models by selecting linear, radial basis, spline, and polynomial functions as kernels, respectively for the prediction of urban dust fall levels. The inputs of the models are identified as industrial coal consumption, population density, traffic flow coefficient, and shopping density coefficient. The training and testing results show that the SVR model with radial basis kernel performs better than the other three both in the training and testing processes. In addition, a number of scenario analyses reveal that the most suitable parameters （insensitive loss function e, the parameter to reduce the influence of error C, and discrete level or average distribution of parameters σ） are 0.001, 0.5, and 2 000, respectively.展开更多
A new parallel architecture for quantified boolean formula(QBF)solving was proposed,and the prediction model based on machine learning technology was proposed for how sharing knowledge affects the solving performance ...A new parallel architecture for quantified boolean formula(QBF)solving was proposed,and the prediction model based on machine learning technology was proposed for how sharing knowledge affects the solving performance in QBF parallel solving system,and the experimental evaluation scheme was also designed.It shows that the characterization factor of clause and cube influence the solving performance markedly in our experiment.At the same time,the heuristic machine learning algorithm was applied,support vector machine was chosen to predict the performance of QBF parallel solving system based on clause sharing and cube sharing.The relative error of accuracy for prediction can be controlled in a reasonable range of 20%30%.The results show the important and complex role that knowledge sharing plays in any modern parallel solver.It shows that the parallel solver with machine learning reduces the quantity of knowledge sharing about 30%and saving computational resource but does not reduce the performance of solving system.展开更多
When detecting deletions in complex human genomes,split-read approaches using short reads generated with next-generation sequencing still face the challenge that either false discovery rate is high,or sensitivity is l...When detecting deletions in complex human genomes,split-read approaches using short reads generated with next-generation sequencing still face the challenge that either false discovery rate is high,or sensitivity is low.To address the problem,an integrated strategy is proposed.It organically combines the fundamental theories of the three mainstream methods(read-pair approaches,split-read technologies and read-depth analysis) with modern machine learning algorithms,using the recipe of feature extraction as a bridge.Compared with the state-of-art split-read methods for deletion detection in both low and high sequence coverage,the machine-learning-aided strategy shows great ability in intelligently balancing sensitivity and false discovery rate and getting a both more sensitive and more precise call set at single-base-pair resolution.Thus,users do not need to rely on former experience to make an unnecessary trade-off beforehand and adjust parameters over and over again any more.It should be noted that modern machine learning models can play an important role in the field of structural variation prediction.展开更多
基金Project(52072412)supported by the National Natural Science Foundation of China。
文摘Traditional track dynamic geometric state(TDGS)simulation incurs substantial computational burdens,posing challenges for developing reliability assessment approach that accounts for TDGS.To overcome these,firstly,a simulation-based TDGS model is established,and a surrogate-based model,grid search algorithm-particle swarm optimization-genetic algorithm-multi-output least squares support vector regression,is established.Among them,hyperparameter optimization algorithm’s effectiveness is confirmed through test functions.Subsequently,an adaptive surrogate-based probability density evolution method(PDEM)considering random track geometry irregularity(TGI)is developed.Finally,taking curved train-steel spring floating slab track-U beam as case study,the surrogate-based model trained on simulation datasets not only shows accuracy in both time and frequency domains,but also surpasses existing models.Additionally,the adaptive surrogate-based PDEM shows high accuracy and efficiency,outperforming Monte Carlo simulation and simulation-based PDEM.The reliability assessment shows that the TDGS part peak management indexes,left/right vertical dynamic irregularity,right alignment dynamic irregularity,and track twist,have reliability values of 0.9648,0.9918,0.9978,and 0.9901,respectively.The TDGS mean management index,i.e.,track quality index,has reliability value of 0.9950.These findings show that the proposed framework can accurately and efficiently assess the reliability of curved low-stiffness track-viaducts,providing a theoretical basis for the TGI maintenance.
基金Project(52204117)supported by the National Natural Science Foundation of ChinaProject(2022JJ40601)supported by the Natural Science Foundation of Hunan Province,China。
文摘Underground excavation can lead to stress redistribution and result in an excavation damaged zone(EDZ),which is an important factor affecting the excavation stability and support design.Accurately estimating the thickness of EDZ is essential to ensure the safety of the underground excavation.In this study,four novel hybrid ensemble learning models were developed by optimizing the extreme gradient boosting(XGBoost)and random forest(RF)algorithms through simulated annealing(SA)and Bayesian optimization(BO)approaches,namely SA-XGBoost,SA-RF,BO XGBoost and BO-RF models.A total of 210 cases were collected from Xiangxi Gold Mine in Hunan Province and Fankou Lead-zinc Mine in Guangdong Province,China,including seven input indicators:embedding depth,drift span,uniaxial compressive strength of rock,rock mass rating,unit weight of rock,lateral pressure coefficient of roadway and unit consumption of blasting explosive.The performance of the proposed models was evaluated by the coefficient of determination,root mean squared error,mean absolute error and variance accounted for.The results indicated that the SA-XGBoost model performed best.The Shapley additive explanations method revealed that the embedding depth was the most important indicator.Moreover,the convergence curves suggested that the SA-XGBoost model can reduce the generalization error and avoid overfitting.
基金Projects(2007JT3018, 2008JT1013, 2009FJ4056) supported by the Key Project in Hunan Science and Technology Program, ChinaProject(20090161120014) supported by the New Teachers Sustentation Fund in Doctoral Program, Ministry of Education, China
文摘Support vector regression （SVR） method is a novel type of learning machine algorithms, which is seldom applied to the development of urban atmospheric quality models under multiple socio-economic factors. This study presents four SVR models by selecting linear, radial basis, spline, and polynomial functions as kernels, respectively for the prediction of urban dust fall levels. The inputs of the models are identified as industrial coal consumption, population density, traffic flow coefficient, and shopping density coefficient. The training and testing results show that the SVR model with radial basis kernel performs better than the other three both in the training and testing processes. In addition, a number of scenario analyses reveal that the most suitable parameters （insensitive loss function e, the parameter to reduce the influence of error C, and discrete level or average distribution of parameters σ） are 0.001, 0.5, and 2 000, respectively.
基金Project(61171141)supported by the National Natural Science Foundation of China
文摘A new parallel architecture for quantified boolean formula(QBF)solving was proposed,and the prediction model based on machine learning technology was proposed for how sharing knowledge affects the solving performance in QBF parallel solving system,and the experimental evaluation scheme was also designed.It shows that the characterization factor of clause and cube influence the solving performance markedly in our experiment.At the same time,the heuristic machine learning algorithm was applied,support vector machine was chosen to predict the performance of QBF parallel solving system based on clause sharing and cube sharing.The relative error of accuracy for prediction can be controlled in a reasonable range of 20%30%.The results show the important and complex role that knowledge sharing plays in any modern parallel solver.It shows that the parallel solver with machine learning reduces the quantity of knowledge sharing about 30%and saving computational resource but does not reduce the performance of solving system.
基金Project(61472026)supported by the National Natural Science Foundation of ChinaProject(2014J410081)supported by Guangzhou Scientific Research Program,China
文摘When detecting deletions in complex human genomes,split-read approaches using short reads generated with next-generation sequencing still face the challenge that either false discovery rate is high,or sensitivity is low.To address the problem,an integrated strategy is proposed.It organically combines the fundamental theories of the three mainstream methods(read-pair approaches,split-read technologies and read-depth analysis) with modern machine learning algorithms,using the recipe of feature extraction as a bridge.Compared with the state-of-art split-read methods for deletion detection in both low and high sequence coverage,the machine-learning-aided strategy shows great ability in intelligently balancing sensitivity and false discovery rate and getting a both more sensitive and more precise call set at single-base-pair resolution.Thus,users do not need to rely on former experience to make an unnecessary trade-off beforehand and adjust parameters over and over again any more.It should be noted that modern machine learning models can play an important role in the field of structural variation prediction.
