State-space models are successfully used in many areas of science, engineering and economics to model time series and dynamical systems. We present a fully Bayesian approach to inference and learning (i.e. state estim...
详细信息
State-space models are successfully used in many areas of science, engineering and economics to model time series and dynamical systems. We present a fully Bayesian approach to inference and learning (i.e. state estimation and system identification) in nonlinear nonparametric state-space models. We place a Gaussian process prior over the state transition dynamics, resulting in a flexible model able to capture complex dynamical phenomena. To enable efficient inference, we marginalize over the transition dynamics function and, instead, infer directly the joint smoothing distribution using specially tailored Particle Markov Chain Monte Carlo samplers. Once a sample from the smoothing distribution is computed, the state transition predictive distribution can be formulated analytically. Our approach preserves the full nonparametric expressivity of the model and can make use of sparse Gaussian processes to greatly reduce computational complexity.
Flexible manufacturing puts high demands on industrial automation in terms of affordable and competitive solutions and is of key importance for short-lot production typically found in small and medium sized enterprise...
详细信息
ISBN:
(纸本)9783902661432
Flexible manufacturing puts high demands on industrial automation in terms of affordable and competitive solutions and is of key importance for short-lot production typically found in small and medium sized enterprises, situations where today manual labour is extensively used despite e.g., environmental issues and harsh working conditions. Key factors in the successful introduction of new robot automation concepts are that they provide both the desired performance and quality (technical capability), but also that they enable fast deployment and extend available task repertoire. The Gantry-Tau manipulator is a new robot concept which in contrast to other high-performance parallel kinematic manipulators (PKM), has a large, open working range. Its high stiffness makes it ideal for a wide range of tasks such as grinding, deburring, and cutting. An additional aspect of such a PKM is the modularity, which in this work has been studied in terms of possibilities for assembly and mechanical reconfiguration at the end-user site, integration of such a kinematically different robot with a standard industrial controller, and new needs for methods/tools to support simple (re)configuration. A range of software tools and methods were found to be useful and necessary for efficient engineering and integration of the concept in typical SME manufacturing scenarios. For experimental evaluation, two full-scale prototype robots were designed and built, the kinematic software was developed and integrated into a couple of different robot control systems, robot CAD software was adapted to the configuration needs, and both simulations and physical experiments were carried out. Our findings make us believe that enhanced software tools should be integrated on a higher symbolic (or meta-) level to better support transformation of data and code generation, but also that the Gantry-Tau type of robot (with adequate software support) will bring a new dimension of flexibility into SME manufacturing.
The goal of this study was to examine the ability of Neural Networks to recognise the levels of anaesthetic state of a patient. Data obtained under different levels of anaesthesia have been modelled for the purpose. I...
详细信息
The goal of this study was to examine the ability of Neural Networks to recognise the levels of anaesthetic state of a patient. Data obtained under different levels of anaesthesia have been modelled for the purpose. It is shown that inferential parameters can be used to recognise the levels of anaesthesia. In addition to demonstrating the ability of neural networks for classification we were interested in understanding the classification strategy discovered by the neural networks. Multivariate data analysis techniques, namely Principal Components Analysis and Canonical Discriminant Variates, were applied to analyse the resultant networks. (C) 1997 Elsevier Science B.V.
Services on the Internet rely on Internet applications, that is software applications distributed between clients and an Internet server relying on IP-based communication. This work demonstrates that they can benefit ...
详细信息
A novel approach to fault diagnosis in nonlinear stochastic systems is proposed. It is based on the particle filtering (PF) algorithm, a Monte Carlo technique based state estimation method, and the multiple model (MM)...
详细信息
This paper introduces the computer communication network system design of the first CIMS (computer Integrated Manufacturing System) application project of China-CAC (Chengdu Aircraft corporation) CIMS. The network sys...
详细信息
This work presents the study and development of a high-gain hybrid DC-DC converter with switched capacitor for photovoltaic energy applications. Qualitative analyzes and quantitative values of the converter are propos...
详细信息
In this paper, a perceptual audio hashing method in compressed domain is proposed for content identification, in which MDCT coefficients as the intermediate decoding result are selected for perceptual feature extracti...
详细信息
This paper describes the application of a Parallel Genetic Algorithm that solves the weekly timetable construction problem for elementary schools. Timetable construction is NPcomplete and highly constrained problem, a...
详细信息
In the Model Predictive control (MPC) framework a controller is computed using a finite horizon optimal control cost. When linear constraints are imposed on the system with a quadratic cost function, then the MPC prob...
详细信息
暂无评论