scholarly journals Learning Driver Behaviors Using A Gaussian Process Augmented State-Space Model

2020 ◽  
Author(s):  
Anton Kullberg ◽  
Isaac Skog ◽  
Gustaf Hendeby
Keyword(s):  
Gaussian Process ◽  
State Space ◽  
State Space Model ◽  
Space Model ◽  
Augmented State

Estimation of Forcing Function, Mistuning and Modal Damping in a Bladed Rotor

2019 ◽  
Author(s):  
Arjun Singh Chauhan ◽  
Alok Sinha
Keyword(s):  
Kalman Filter ◽  
State Space ◽  
Unscented Kalman Filter ◽  
State Space Model ◽  
Space Model ◽  
Vibration Data ◽  
Forcing Function ◽  
Modal Damping ◽  
Bladed Rotor ◽  
Augmented State

Abstract This paper deals with the estimation of forcing function, modal damping and mistuned modal stiffnesses in a bladed rotor. Previous research on parameter estimation in a mistuned bladed rotor relies on the knowledge of the forcing function as well as the vibration data. This paper presents two novel approaches. The first approach relies on knowledge of both the forcing and vibration data. The parameters are treated as states of the system and an augmented state space model is created. Unscented Kalman Filter is then used on the steady state data to estimate the parameters. The second approach eliminates the dependence on forcing data. Both the forcing and parameters are now treated as states of the system to construct an augmented state space model. Unscented Kalman Filter is then used on transient vibration data for estimation. Numerical results are presented for a simple model of a mistuned bladed rotor which considers a single mode of vibration per blade.

scholarly journals Financial time series volatility analysis using Gaussian process state-space models

2021 ◽  
Author(s):  
Jianan Han
Keyword(s):  
Time Series ◽  
Gaussian Process ◽  
State Space ◽  
Financial Time Series ◽  
State Space Model ◽  
Likelihood Methods ◽  
Nonparametric Modeling ◽  
Space Model ◽  
Financial Time ◽  
Maximum Likelihood Methods

In this thesis, we propose a novel nonparametric modeling framework for financial time series data analysis, and we apply the framework to the problem of time varying volatility modeling. Existing parametric models have a rigid transition function form and they often have over-fitting problems when model parameters are estimated using maximum likelihood methods. These drawbacks effect the models' forecast performance. To solve this problem, we take Bayesian nonparametric modeling approach. By adding Gaussian process prior to the hidden state transition process, we extend the standard state-space model to a Gaussian process state-space model. We introduce our Gaussian process regression stochastic volatility (GPRSV) model. Instead of using maximum likelihood methods, we use Monte Carlo inference algorithms. Both online particle filter and offline particle Markov chain Monte Carlo methods are studied to learn the proposed model. We demonstrate our model and inference methods with both simulated and empirical financial data.

Estimation of Forcing Function, Mistuning, and Modal Damping in a Bladed Rotor

2020 ◽  
Vol 142 (5) ◽  
Author(s):  
Arjun Singh Chauhan ◽  
Alok Sinha
Keyword(s):  
State Space ◽  
Unscented Kalman Filter ◽  
State Space Model ◽  
Transient Vibration ◽  
Space Model ◽  
Vibration Data ◽  
Forcing Function ◽  
Modal Damping ◽  
Bladed Rotor ◽  
Augmented State

Abstract This paper deals with the estimation of forcing function, modal damping, and mistuned modal stiffnesses in a bladed rotor. Previous research on parameter estimation in a mistuned bladed rotor relies on the knowledge of the forcing function as well as the vibration data. This paper presents two novel approaches. The first approach relies on knowledge of both the forcing and vibration data. The parameters are treated as states of the system and an augmented state space model is created. Unscented Kalman filter (UKF) is then used on the steady-state data to estimate the parameters. The second approach eliminates the dependence on forcing data. Both the forcing and parameters are now treated as states of the system to construct an augmented state space model. UKF is then used on transient vibration data for estimation. Numerical results are presented for a simple model of a mistuned bladed rotor which considers a single mode of vibration per blade.

scholarly journals Financial time series volatility analysis using Gaussian process state-space models

2021 ◽  
Author(s):  
Jianan Han
Keyword(s):  
Time Series ◽  
Gaussian Process ◽  
State Space ◽  
Financial Time Series ◽  
State Space Model ◽  
Likelihood Methods ◽  
Nonparametric Modeling ◽  
Space Model ◽  
Financial Time ◽  
Maximum Likelihood Methods

In this thesis, we propose a novel nonparametric modeling framework for financial time series data analysis, and we apply the framework to the problem of time varying volatility modeling. Existing parametric models have a rigid transition function form and they often have over-fitting problems when model parameters are estimated using maximum likelihood methods. These drawbacks effect the models' forecast performance. To solve this problem, we take Bayesian nonparametric modeling approach. By adding Gaussian process prior to the hidden state transition process, we extend the standard state-space model to a Gaussian process state-space model. We introduce our Gaussian process regression stochastic volatility (GPRSV) model. Instead of using maximum likelihood methods, we use Monte Carlo inference algorithms. Both online particle filter and offline particle Markov chain Monte Carlo methods are studied to learn the proposed model. We demonstrate our model and inference methods with both simulated and empirical financial data.

scholarly journals Gaussian Process-Integrated State Space Model for Continuous Joint Angle Prediction from EMG and Interactive Force in a Human-Exoskeleton System

2019 ◽  
Vol 9 (8) ◽  
pp. 1711 ◽  
Author(s):  
Yan Zeng ◽  
Jiantao Yang ◽  
Yuehong Yin
Keyword(s):  
Gaussian Process ◽  
State Space ◽  
Muscle Activation ◽  
Joint Angle ◽  
State Space Model ◽  
Measurement Model ◽  
The State ◽  
State Model ◽  
Space Model ◽  
Emg Signal

As one of the most direct indicators of the transparency between a human and an exoskeleton, interactive force has rarely been fused with electromyography (EMG) in the control of human-exoskeleton systems, the performances of which are largely determined by the accuracy of the continuous joint angle prediction. To achieve intuitive and naturalistic human intent learning, a state space model (SSM) for continuous angle prediction of knee joint is developed. When the influence of the interactive force is often ignored in the existing models of human-exoskeleton systems, interactive force is applied as the measurement model output of the proposed SSM, and the EMG signal is used as the state model input signal to indicate muscle activation. The forward dynamics of joint motion and the human-machine interaction mechanism, i.e., the biomechanical interpretations of the interactive force generation mechanism, are derived as the bases for the state model and measurement model based on Hill’s muscle model and semiphenomenological (SP) muscular model, respectively. Gaussian process (GP)-based nonlinear autoregressive with the exogenous inputs (NARX) model and back-propagation neural network (BPNN) are applied to provide better adaptivity for the SSM in practical applications. Corresponding experimental results demonstrate the validity and superiority of the method.

scholarly journals Active Learning in Gaussian Process State Space Model

2021 ◽  
pp. 346-361
Author(s):  
Hon Sum Alec Yu ◽  
Dingling Yao ◽  
Christoph Zimmer ◽  
Marc Toussaint ◽  
Duy Nguyen-Tuong
Keyword(s):  
Active Learning ◽  
Gaussian Process ◽  
State Space ◽  
State Space Model ◽  
Space Model
Sign in / Sign up

Export Citation Format

Share Document