The traditional least squares support vector regression(LS-SVR)model,using cross validation to determine the regularization parameter and kernel parameter,is time-consuming.We propose a Bayesian evidence framework t...The traditional least squares support vector regression(LS-SVR)model,using cross validation to determine the regularization parameter and kernel parameter,is time-consuming.We propose a Bayesian evidence framework to infer the LS-SVR model parameters.Three levels Bayesian inferences are used to determine the model parameters,regularization hyper-parameters and tune the nuclear parameters by model comparison.On this basis,we established Bayesian LS-SVR time-series gas forecasting models and provide steps for the algorithm.The gas outburst data of a Hebi 10th mine working face is used to validate the model.The optimal embedding dimension and delay time of the time series were obtained by the smallest differential entropy method.Finally,within a MATLAB7.1 environment,we used actual coal gas data to compare the traditional LS-SVR and the Bayesian LS-SVR with LS-SVMlab1.5 Toolbox simulation.The results show that the Bayesian framework of an LS-SVR significantly improves the speed and accuracy of the forecast.展开更多
Reliability and remaining useful life(RUL)estimation for a satellite rechargeable lithium battery(RLB)are significant for prognostic and health management(PHM).A novel Bayesian framework is proposed to do reliability ...Reliability and remaining useful life(RUL)estimation for a satellite rechargeable lithium battery(RLB)are significant for prognostic and health management(PHM).A novel Bayesian framework is proposed to do reliability analysis by synthesizing multisource data,including bivariate degradation data and lifetime data.Bivariate degradation means that there are two degraded performance characteristics leading to the failure of the system.First,linear Wiener process and Frank Copula function are used to model the dependent degradation processes of the RLB's temperature and discharge voltage.Next,the Bayesian method,in combination with Markov Chain Monte Carlo(MCMC)simulations,is provided to integrate limited bivariate degradation data with other congeneric RLBs'lifetime data.Then reliability evaluation and RUL prediction are carried out for PHM.A simulation study demonstrates that due to the data fusion,parameter estimations and predicted RUL obtained from our model are more precise than models only using degradation data or ignoring the dependency of different degradation processes.Finally,a practical case study of a satellite RLB verifies the usability of the model.展开更多
A precise spacecraft attitude dynamics model is essential for accurately predicting a satellite’s orientation in orbit,with such predictions being centerpiece to mission safety,operational control,and long-term risk ...A precise spacecraft attitude dynamics model is essential for accurately predicting a satellite’s orientation in orbit,with such predictions being centerpiece to mission safety,operational control,and long-term risk management.However,the highly nonlinear nature of spacecraft dynamics,compounded by uncertain and varying space perturbations such as atmospheric drag,solar radiation pressure,and magnetic torques,poses a substantial challenge to model fidelity.The problem is further exacerbated by the limited availability of in-orbit attitude measurements,which constrains direct calibration efforts.This work proposes a Bayesian stochastic model updating framework to systematically calibrate complex attitude dynamics models under epistemic and aleatory uncertainties.The methodology leverages approximate Bayesian computation with Euclidean and Bhattacharyya distance-based likelihoods,integrated within a transitional Markov chain Monte Carlo sampling scheme.A pseudo-online updating process is developed to incorporate sparse,sequential attitude measurements,enabling continual refinement of uncertain model parameters and improved characterization of stochastic dynamics.A numerical case study involving a rigid-body satellite subject to hybrid perturbations is presented to demonstrate the effectiveness of the proposed approach.The results show successful convergence of posterior distributions around true values,a significant reduction in epistemic uncertainty,and an improved predictive capability for attitude propagation in data-sparse scenarios.This framework offers a promising direction for enhancing attitude modeling reliability in the context of increasingly congested and observation-limited space environments.展开更多
In the paper, an iterative method is presented to the optimal control of batch processes. Generally it is very difficult to acquire an accurate mechanistic model for a batch process. Because support vector machine is ...In the paper, an iterative method is presented to the optimal control of batch processes. Generally it is very difficult to acquire an accurate mechanistic model for a batch process. Because support vector machine is powerful for the problems characterized by small samples, nonlinearity, high dimension and local minima, support vector regression models are developed for the optimal control of batch processes where end-point properties are required. The model parameters are selected within the Bayesian evidence framework. Based on the model, an iterative method is used to exploit the repetitive nature of batch processes to determine the optimal operating policy. Numerical simulation shows that the iterative optimal control can improve the process performance through iterations.展开更多
Fluid identification and anisotropic parameters characterization are crucial for shale reservoir exploration and development.However,the anisotropic reflection coefficient equation,based on the transverse isotropy wit...Fluid identification and anisotropic parameters characterization are crucial for shale reservoir exploration and development.However,the anisotropic reflection coefficient equation,based on the transverse isotropy with a vertical axis of symmetry(VTI)medium assumption,involves numerous parameters to be inverted.This complexity reduces its stability and impacts the accuracy of seismic amplitude variation with offset(AVO)inversion results.In this study,a novel anisotropic equation that includes the fluid term and Thomsen anisotropic parameters is rewritten,which reduces the equation's dimensionality and increases its stability.Additionally,the traditional Markov Chain Monte Carlo(MCMC)inversion algorithm exhibits a high rejection rate for random samples and relies on known parameter distributions such as the Gaussian distribution,limiting the algorithm's convergence and sample randomness.To address these limitations and evaluate the uncertainty of AVO inversion,the IADR-Gibbs algorithm is proposed,which incorporates the Independent Adaptive Delayed Rejection(IADR)algorithm with the Gibbs sampling algorithm.Grounded in Bayesian theory,the new algorithm introduces support points to construct a proposal distribution of non-parametric distribution and reselects the rejected samples according to the Delayed Rejection(DR)strategy.Rejected samples are then added to the support points to update the proposal distribution function adaptively.The equation rewriting method and the IADR-Gibbs algorithm improve the accuracy and robustness of AVO inversion.The effectiveness and applicability of the proposed method are validated through synthetic gather tests and practical data applications.展开更多
The command and control(C2) is a decision-making process based on human cognition,which contains operational,physical,and human characteristics,so it takes on uncertainty and complexity.As a decision support approac...The command and control(C2) is a decision-making process based on human cognition,which contains operational,physical,and human characteristics,so it takes on uncertainty and complexity.As a decision support approach,Bayesian networks(BNs) provide a framework in which a decision is made by combining the experts' knowledge and the specific data.In addition,an expert system represented by human cognitive framework is adopted to express the real-time decision-making process of the decision maker.The combination of the Bayesian decision support and human cognitive framework in the C2 of a specific application field is modeled and executed by colored Petri nets(CPNs),and the consequences of execution manifest such combination can perfectly present the decision-making process in C2.展开更多
Seismic inversion is a highly ill-posed problem,due to many factors such as the limited seismic frequency bandwidth and inappropriate forward modeling.To obtain a unique solution,some smoothing constraints,e.g.,the Ti...Seismic inversion is a highly ill-posed problem,due to many factors such as the limited seismic frequency bandwidth and inappropriate forward modeling.To obtain a unique solution,some smoothing constraints,e.g.,the Tikhonov regularization are usually applied.The Tikhonov method can maintain a global smooth solution,but cause a fuzzy structure edge.In this paper we use Huber-Markov random-field edge protection method in the procedure of inverting three parameters,P-velocity,S-velocity and density.The method can avoid blurring the structure edge and resist noise.For the parameter to be inverted,the Huber-Markov random-field constructs a neighborhood system,which further acts as the vertical and lateral constraints.We use a quadratic Huber edge penalty function within the layer to suppress noise and a linear one on the edges to avoid a fuzzy result.The effectiveness of our method is proved by inverting the synthetic data without and with noises.The relationship between the adopted constraints and the inversion results is analyzed as well.展开更多
基金Financial support for this work,provided by the National Natural Science Foundation of China(No.60974126)the Natural Science Foundation of Jiangsu Province(No.BK2009094)
摘要The traditional least squares support vector regression(LS-SVR)model,using cross validation to determine the regularization parameter and kernel parameter,is time-consuming.We propose a Bayesian evidence framework to infer the LS-SVR model parameters.Three levels Bayesian inferences are used to determine the model parameters,regularization hyper-parameters and tune the nuclear parameters by model comparison.On this basis,we established Bayesian LS-SVR time-series gas forecasting models and provide steps for the algorithm.The gas outburst data of a Hebi 10th mine working face is used to validate the model.The optimal embedding dimension and delay time of the time series were obtained by the smallest differential entropy method.Finally,within a MATLAB7.1 environment,we used actual coal gas data to compare the traditional LS-SVR and the Bayesian LS-SVR with LS-SVMlab1.5 Toolbox simulation.The results show that the Bayesian framework of an LS-SVR significantly improves the speed and accuracy of the forecast.
基金Project(71371182) supported by the National Natural Science Foundation of China
摘要Reliability and remaining useful life(RUL)estimation for a satellite rechargeable lithium battery(RLB)are significant for prognostic and health management(PHM).A novel Bayesian framework is proposed to do reliability analysis by synthesizing multisource data,including bivariate degradation data and lifetime data.Bivariate degradation means that there are two degraded performance characteristics leading to the failure of the system.First,linear Wiener process and Frank Copula function are used to model the dependent degradation processes of the RLB's temperature and discharge voltage.Next,the Bayesian method,in combination with Markov Chain Monte Carlo(MCMC)simulations,is provided to integrate limited bivariate degradation data with other congeneric RLBs'lifetime data.Then reliability evaluation and RUL prediction are carried out for PHM.A simulation study demonstrates that due to the data fusion,parameter estimations and predicted RUL obtained from our model are more precise than models only using degradation data or ignoring the dependency of different degradation processes.Finally,a practical case study of a satellite RLB verifies the usability of the model.
摘要A precise spacecraft attitude dynamics model is essential for accurately predicting a satellite’s orientation in orbit,with such predictions being centerpiece to mission safety,operational control,and long-term risk management.However,the highly nonlinear nature of spacecraft dynamics,compounded by uncertain and varying space perturbations such as atmospheric drag,solar radiation pressure,and magnetic torques,poses a substantial challenge to model fidelity.The problem is further exacerbated by the limited availability of in-orbit attitude measurements,which constrains direct calibration efforts.This work proposes a Bayesian stochastic model updating framework to systematically calibrate complex attitude dynamics models under epistemic and aleatory uncertainties.The methodology leverages approximate Bayesian computation with Euclidean and Bhattacharyya distance-based likelihoods,integrated within a transitional Markov chain Monte Carlo sampling scheme.A pseudo-online updating process is developed to incorporate sparse,sequential attitude measurements,enabling continual refinement of uncertain model parameters and improved characterization of stochastic dynamics.A numerical case study involving a rigid-body satellite subject to hybrid perturbations is presented to demonstrate the effectiveness of the proposed approach.The results show successful convergence of posterior distributions around true values,a significant reduction in epistemic uncertainty,and an improved predictive capability for attitude propagation in data-sparse scenarios.This framework offers a promising direction for enhancing attitude modeling reliability in the context of increasingly congested and observation-limited space environments.
基金Project supported by the National Natural Science Foundation of China(Grant No.60504033)
摘要In the paper, an iterative method is presented to the optimal control of batch processes. Generally it is very difficult to acquire an accurate mechanistic model for a batch process. Because support vector machine is powerful for the problems characterized by small samples, nonlinearity, high dimension and local minima, support vector regression models are developed for the optimal control of batch processes where end-point properties are required. The model parameters are selected within the Bayesian evidence framework. Based on the model, an iterative method is used to exploit the repetitive nature of batch processes to determine the optimal operating policy. Numerical simulation shows that the iterative optimal control can improve the process performance through iterations.
基金the sponsorship of the Key Technology for Geophysical Prediction of Ultra-Deep Carbonate Reservoirs(P24240)the National Natural Science Foundation of China(U24B2020)the National Science and Technology Major Project of China for New Oil and Gas Exploration and Development(Grant No.2024ZD1400102)。
摘要Fluid identification and anisotropic parameters characterization are crucial for shale reservoir exploration and development.However,the anisotropic reflection coefficient equation,based on the transverse isotropy with a vertical axis of symmetry(VTI)medium assumption,involves numerous parameters to be inverted.This complexity reduces its stability and impacts the accuracy of seismic amplitude variation with offset(AVO)inversion results.In this study,a novel anisotropic equation that includes the fluid term and Thomsen anisotropic parameters is rewritten,which reduces the equation's dimensionality and increases its stability.Additionally,the traditional Markov Chain Monte Carlo(MCMC)inversion algorithm exhibits a high rejection rate for random samples and relies on known parameter distributions such as the Gaussian distribution,limiting the algorithm's convergence and sample randomness.To address these limitations and evaluate the uncertainty of AVO inversion,the IADR-Gibbs algorithm is proposed,which incorporates the Independent Adaptive Delayed Rejection(IADR)algorithm with the Gibbs sampling algorithm.Grounded in Bayesian theory,the new algorithm introduces support points to construct a proposal distribution of non-parametric distribution and reselects the rejected samples according to the Delayed Rejection(DR)strategy.Rejected samples are then added to the support points to update the proposal distribution function adaptively.The equation rewriting method and the IADR-Gibbs algorithm improve the accuracy and robustness of AVO inversion.The effectiveness and applicability of the proposed method are validated through synthetic gather tests and practical data applications.
基金supported by the National Natural Science Foundation of China (60874068)
摘要The command and control(C2) is a decision-making process based on human cognition,which contains operational,physical,and human characteristics,so it takes on uncertainty and complexity.As a decision support approach,Bayesian networks(BNs) provide a framework in which a decision is made by combining the experts' knowledge and the specific data.In addition,an expert system represented by human cognitive framework is adopted to express the real-time decision-making process of the decision maker.The combination of the Bayesian decision support and human cognitive framework in the C2 of a specific application field is modeled and executed by colored Petri nets(CPNs),and the consequences of execution manifest such combination can perfectly present the decision-making process in C2.
基金supported by the National Basic Research Program of China(973 Program)(No.2013CB228603)National Science and Technology major projects(No.2011ZX05024 and 2011ZX05010)the National Natural Science Foundation of China(No.41174119)
摘要Seismic inversion is a highly ill-posed problem,due to many factors such as the limited seismic frequency bandwidth and inappropriate forward modeling.To obtain a unique solution,some smoothing constraints,e.g.,the Tikhonov regularization are usually applied.The Tikhonov method can maintain a global smooth solution,but cause a fuzzy structure edge.In this paper we use Huber-Markov random-field edge protection method in the procedure of inverting three parameters,P-velocity,S-velocity and density.The method can avoid blurring the structure edge and resist noise.For the parameter to be inverted,the Huber-Markov random-field constructs a neighborhood system,which further acts as the vertical and lateral constraints.We use a quadratic Huber edge penalty function within the layer to suppress noise and a linear one on the edges to avoid a fuzzy result.The effectiveness of our method is proved by inverting the synthetic data without and with noises.The relationship between the adopted constraints and the inversion results is analyzed as well.