The concept of connected vehicles is with great potentials for enhancing the road transportation systems in the future. To support the functions and applications under the connected vehicles frame, the estimation of d...The concept of connected vehicles is with great potentials for enhancing the road transportation systems in the future. To support the functions and applications under the connected vehicles frame, the estimation of dynamic states of the vehicles under the cooperative environments is a fundamental issue. By integrating multiple sensors, localization modules in OBUs(on-board units) require effective estimation solutions to cope with various operation conditions. Based on the filtering estimation framework for sensor fusion, an ensemble Kalman filter(En KF) is introduced to estimate the vehicle's state with observations from navigation satellites and neighborhood vehicles, and the original En KF solution is improved by using the cubature transformation to fulfill the requirements of the nonlinearity approximation capability, where the conventional ensemble analysis operation in En KF is modified to enhance the estimation performance without increasing the computational burden significantly. Simulation results from a nonlinear case and the cooperative vehicle localization scenario illustrate the capability of the proposed filter, which is crucial to realize the active safety of connected vehicles in future intelligent transportation.展开更多
Strong mechanical vibration and acoustical signals of grinding process contain useful information related to load parameters in ball mills. It is a challenge to extract latent features and construct soft sensor model ...Strong mechanical vibration and acoustical signals of grinding process contain useful information related to load parameters in ball mills. It is a challenge to extract latent features and construct soft sensor model with high dimensional frequency spectra of these signals. This paper aims to develop a selective ensemble modeling approach based on nonlinear latent frequency spectral feature extraction for accurate measurement of material to ball volume ratio. Latent features are first extracted from different vibrations and acoustic spectral segments by kernel partial least squares. Algorithms of bootstrap and least squares support vector machines are employed to produce candidate sub-models using these latent features as inputs. Ensemble sub-models are selected based on genetic algorithm optimization toolbox. Partial least squares regression is used to combine these sub-models to eliminate collinearity among their prediction outputs. Results indicate that the proposed modeling approach has better prediction performance than previous ones.展开更多
Urban living in large modern cities exerts considerable adverse effectson health and thus increases the risk of contracting several chronic kidney diseases (CKD). The prediction of CKDs has become a major task in urba...Urban living in large modern cities exerts considerable adverse effectson health and thus increases the risk of contracting several chronic kidney diseases (CKD). The prediction of CKDs has become a major task in urbanizedcountries. The primary objective of this work is to introduce and develop predictive analytics for predicting CKDs. However, prediction of huge samples isbecoming increasingly difficult. Meanwhile, MapReduce provides a feasible framework for programming predictive algorithms with map and reduce functions.The relatively simple programming interface helps solve problems in the scalability and efficiency of predictive learning algorithms. In the proposed work, theiterative weighted map reduce framework is introduced for the effective management of large dataset samples. A binary classification problem is formulated usingensemble nonlinear support vector machines and random forests. Thus, instead ofusing the normal linear combination of kernel activations, the proposed work creates nonlinear combinations of kernel activations in prototype examples. Furthermore, different descriptors are combined in an ensemble of deep support vectormachines, where the product rule is used to combine probability estimates ofdifferent classifiers. Performance is evaluated in terms of the prediction accuracyand interpretability of the model and the results.展开更多
An extremely large（＂freak＂） wave is a typical though rare phenomenon observed in the sea. Special theories（for example, the modulation instability theory） were developed to explain mechanics and appearance of fr...An extremely large（＂freak＂） wave is a typical though rare phenomenon observed in the sea. Special theories（for example, the modulation instability theory） were developed to explain mechanics and appearance of freak waves as a result of nonlinear wave-wave interactions. In this paper, it is demonstrated that the freak wave appearance can be also explained by superposition of linear modes with the realistic spectrum. The integral probability of trough-to-crest waves is calculated by two methods： the first one is based on the results of the numerical simulation of a wave field evolution performed with one-dimensional and two-dimensional nonlinear models.The second method is based on calculation of the same probability over the ensembles of wave fields constructed as a superposition of linear waves with random phases and the spectrum similar to that used in the nonlinear simulations. It is shown that the integral probabilities for nonlinear and linear cases are of the same order of values展开更多
基金Project(4144081)supported by Beijing Natural Science Foundation,ChinaProjects(61403021,U1334211,61490705)supported by the National Natural Science Foundation of China+1 种基金Project(2015RC015)supported by the Fundamental Research Funds for Central Universities,ChinaProject supported by the Foundation of Beijing Key Laboratory for Cooperative Vehicle Infrastructure Systems and Safety Control,China
文摘The concept of connected vehicles is with great potentials for enhancing the road transportation systems in the future. To support the functions and applications under the connected vehicles frame, the estimation of dynamic states of the vehicles under the cooperative environments is a fundamental issue. By integrating multiple sensors, localization modules in OBUs(on-board units) require effective estimation solutions to cope with various operation conditions. Based on the filtering estimation framework for sensor fusion, an ensemble Kalman filter(En KF) is introduced to estimate the vehicle's state with observations from navigation satellites and neighborhood vehicles, and the original En KF solution is improved by using the cubature transformation to fulfill the requirements of the nonlinearity approximation capability, where the conventional ensemble analysis operation in En KF is modified to enhance the estimation performance without increasing the computational burden significantly. Simulation results from a nonlinear case and the cooperative vehicle localization scenario illustrate the capability of the proposed filter, which is crucial to realize the active safety of connected vehicles in future intelligent transportation.
基金Supported partially by the Post Doctoral Natural Science Foundation of China(2013M532118,2015T81082)the National Natural Science Foundation of China(61573364,61273177,61503066)+2 种基金the State Key Laboratory of Synthetical Automation for Process Industriesthe National High Technology Research and Development Program of China(2015AA043802)the Scientific Research Fund of Liaoning Provincial Education Department(L2013272)
文摘Strong mechanical vibration and acoustical signals of grinding process contain useful information related to load parameters in ball mills. It is a challenge to extract latent features and construct soft sensor model with high dimensional frequency spectra of these signals. This paper aims to develop a selective ensemble modeling approach based on nonlinear latent frequency spectral feature extraction for accurate measurement of material to ball volume ratio. Latent features are first extracted from different vibrations and acoustic spectral segments by kernel partial least squares. Algorithms of bootstrap and least squares support vector machines are employed to produce candidate sub-models using these latent features as inputs. Ensemble sub-models are selected based on genetic algorithm optimization toolbox. Partial least squares regression is used to combine these sub-models to eliminate collinearity among their prediction outputs. Results indicate that the proposed modeling approach has better prediction performance than previous ones.
文摘Urban living in large modern cities exerts considerable adverse effectson health and thus increases the risk of contracting several chronic kidney diseases (CKD). The prediction of CKDs has become a major task in urbanizedcountries. The primary objective of this work is to introduce and develop predictive analytics for predicting CKDs. However, prediction of huge samples isbecoming increasingly difficult. Meanwhile, MapReduce provides a feasible framework for programming predictive algorithms with map and reduce functions.The relatively simple programming interface helps solve problems in the scalability and efficiency of predictive learning algorithms. In the proposed work, theiterative weighted map reduce framework is introduced for the effective management of large dataset samples. A binary classification problem is formulated usingensemble nonlinear support vector machines and random forests. Thus, instead ofusing the normal linear combination of kernel activations, the proposed work creates nonlinear combinations of kernel activations in prototype examples. Furthermore, different descriptors are combined in an ensemble of deep support vectormachines, where the product rule is used to combine probability estimates ofdifferent classifiers. Performance is evaluated in terms of the prediction accuracyand interpretability of the model and the results.
基金The Rissian Fund for Basic Research under contract No.#14-05-00422Australian Research Council,Discovery under contract Nos DP1093349 and DP130100227
文摘An extremely large（＂freak＂） wave is a typical though rare phenomenon observed in the sea. Special theories（for example, the modulation instability theory） were developed to explain mechanics and appearance of freak waves as a result of nonlinear wave-wave interactions. In this paper, it is demonstrated that the freak wave appearance can be also explained by superposition of linear modes with the realistic spectrum. The integral probability of trough-to-crest waves is calculated by two methods： the first one is based on the results of the numerical simulation of a wave field evolution performed with one-dimensional and two-dimensional nonlinear models.The second method is based on calculation of the same probability over the ensembles of wave fields constructed as a superposition of linear waves with random phases and the spectrum similar to that used in the nonlinear simulations. It is shown that the integral probabilities for nonlinear and linear cases are of the same order of values
