Showing papers by "John B. Moore published in 1995"
TL;DR: In this paper, the problem of risk-sensitive filtering and smoothing for discrete-time Hidden Markov Models (HMM) with finite-discrete states is addressed, where the objective is to minimize the expectation of the exponential of the squared estimation error weighted by a risk sensitive parameter.
55 citations
21 Jun 1995
TL;DR: In this article, the authors address the risk-sensitive filtering and smoothing problem for discrete-time nonlinear and linear Gauss-Markov state-space models and describe the connection between L/sub 2/ filtering and risk sensitive filtering via the limiting results when the risk sensitive parameter tends to zero.
Abstract: In this paper, we address the risk-sensitive filtering and smoothing problem for discrete-time nonlinear and linear Gauss-Markov state-space models. Also, connection between L/sub 2/ filtering (termed here risk-neutral filtering) and risk-sensitive filtering is described via the limiting results when the risk-sensitive parameter tends to zero. The technique used in this paper is the so-called reference probability method which defines a new probability measure where the observations are independent. The optimisation problem is in the new measure and the results are interpreted as solutions in the original measure.
31 citations
TL;DR: In this paper, the risk-sensitive nonlinear stochastic filtering problem is addressed in both continuous and discrete-time for quite general finite-dimensional signal models, including also discrete state hidden Markov models (HMMs).
27 citations
TL;DR: Extended Kalman filtering (EKF) and hidden Markov model (HMM) signal processing techniques are coupled in order to demodulate frequency modulated signals in noisy fading channels to help cope with coloured noise.
21 citations
21 May 1995
TL;DR: The authors view the task of grasping force optimization as a linearly constrained semidefinite programming problem for which there are known globally exponentially convergent solutions via gradient flows.
Abstract: For dextrous robotic hand grasping a key goal is to balance external and internal grasping forces to assure a stable grasp and low grasping energy. In this paper, the authors view the task of grasping force optimization as a linearly constrained semidefinite programming problem for which there are known globally exponentially convergent solutions via gradient flows. A key observation is that the nonlinear friction constraint at each contact point is equivalent to the positive definiteness of a suitable matrix containing the contact wrench intensities. A recursive version is presented, and numerical examples demonstrate the simplicity, the good numerical properties, and optimality of the approach.
15 citations
13 Dec 1995
TL;DR: It is shown that the effect of initial conditions on these filters dies out geomet rically fast under very reasonable observability assump tions.
Abstract: Hidden Markov models have proved suitable for many in teresting applications which can be modelled using some unobservable finite state Markov process, influencing mea sured signals. This can be used to describe bursty telecom munications traffic, or the faults in a complicated systems, for modelling the activity in neurons, for modelling speech patterns, etc. In all these applications, one has to estimate the unobservable underlying state of the Markov process, using the observed signals. Optimal recursive filters are well known for this estimation problem. Recently risk sensitive filters for the same problem have also been ob tained. An important question in studying the quality of such filters is the rate at which arbitrarily assigned initial conditions are forgotten. In this paper we show that the effect of initial conditions on these filters dies out geomet rically fast under very reasonable observability assump tions. The proof is given in the simplest case of finite state space and of a finite, quantised, observations spacc. How ever the method can be extended to more general models by continuity arguments.
11 citations
05 Aug 1995
TL;DR: The goal is to develop a fast and robust algorithm for pose estimation using range data using algebraic techniques in a two stage optimization procedure involving least squares estimation, or better the method of instrumental variables, and 3/spl times/3 matrix diagonalizations.
Abstract: A key problem in robotics is the estimation of the location and orientation of objects from surface measurement data. This is termed pose estimation. A fundamental task is the pose estimation of known quadratic surfaces from, possibly noisy, data. A solution for this task facilitates pose estimation for more complex objects. Current algorithms frequently converge to local minima of the performance index and/or pay a high computing cost and/or are sensitive to noise, that are unsuited for online applications because of the intensive computer effort required. The goal is to develop a fast and robust algorithm for pose estimation using range data. Here, pose estimation is carried out using algebraic techniques in a two stage optimization procedure involving least squares estimation, or better the method of instrumental variables, and 3/spl times/3 matrix diagonalizations. The procedure leads to zero pose estimation error in the noise free finite data case, and in the case of infinite data with additive white noise.
11 citations
TL;DR: In this paper, the authors derived risk sensitive optimality results for finite-dimensional controllers for nonlinear stochastic systems and performance indices for which the controllers are optimal for the nonlinear plants.
11 citations
TL;DR: In this paper, a discrete-time partially observed stochastic control problem with exponential running cost is considered, where the dynamics are linear and the running cost quadratic in the state variable, but the control may enter nonlinearly.
9 citations
13 Dec 1995
TL;DR: In this article, interior-point techniques for solving quadratic programming problems in a Hilbert space were developed and applied to the linear-quadratic control problem with linear inequality constraints.
Abstract: We develop an interior-point technique for solving quadratic programming problem in a Hilbert space. As an example we consider an application of these results to the linear-quadratic control problem with linear inequality constraints. It is shown that the Newton step in this situation is basically reduced to solving the standard linear-quadratic control problem.
5 citations
13 Dec 1995
TL;DR: In this article, a dynamic programming equation solution is given to an optimal risk-sensitive dual control problem penalizing outputs, rather than the states, for a reasonably general class of nonlinear signal models.
Abstract: In this paper, we develop new results concerning the risk-sensitive dual control problem for output feedback nonlinear systems, with unknown time-varying parameters. A dynamic programming equation solution is given to an optimal risk-sensitive dual control problem penalizing outputs, rather than the states, for a reasonably general class of nonlinear signal models. This equation, in contrast to earlier formulations in the literature, clearly shows the dual aspects of the risk-sensitive controller regarding control and estimation. The extensive computational burden for solving this equation motivates our study of risk-sensitive versions for one-step horizon cost indices and suboptimal risk-sensitive dual control. The idea of a more generalized optimal risk-sensitive dual controller is briefly introduced.
TL;DR: A recursive prediction error algorithm is presented which addresses the problem of computational complexity for on-line identification of hidden Markov models (HMMs) and results in a sub-optimal reduced order identification scheme.
01 Jan 1995
TL;DR: In this article, the pose estimation problem is converted to a non-linear optimization problem that minimizes an error objective function between the measured surface data and one of CAD model, and the projected gradient flow of the objective function onto the manifold SO(3)xR3 is de- rived and converge to an equilibrium point as usual steeplest decent methods.
Abstract: A key problem in robotics is the es- timation of the location and orientation of objects from surface measurement data. This is termed pose estimation. Our pose estimation problem is converted to a non-linear optimization problem that minimize an error objective function between the measured surface data and one of CAD model. The authors study gradient flows on the Lie groups to- ward a solution of the pose estimation problem of quadratic surfaces. In this paper, the projected gradient flow of the objective function onto the manifold SO(3)xR3 is de- rived and converge to an equilibrium point as usual steeplest decent methods. Discretizations of flow lead to recursive numerical methods for pose esti- mation.
01 Jan 1995
06 Nov 1995
TL;DR: In this paper, a nonlinear optimization problem that minimizes an error objective function between the measured surface data and one of a CAD model is formulated and solved using gradient flows on the Lie groups.
Abstract: A key problem in robotics is the estimation of the location and orientation of objects from surface measurement data. This is termed pose estimation. The authors' pose estimation problem is converted to a nonlinear optimization problem that minimizes an error objective function between the measured surface data and one of a CAD model. The authors study gradient flows on the Lie groups toward a solution of the pose estimation problem of quadratic surfaces. In this paper, the projected gradient flow of the objective function onto the manifold SO(3)/spl times/R/sup 3/ is derived and converge to an equilibrium point as is usual in steepest descent methods. Discretizations of flow lead to recursive numerical methods for pose estimation.