Issue Of Computational Cost Research Articles

Model-free learning on robot kinematic chains using a nested multi-agent topology

This paper proposes a model-free learning scheme for the developmental acquisition of robot kinematic control and dexterous manipulation skills. The approach is based on a nested-hierarchical multi-agent architecture that intuitively encapsulates the topology of robot kinematic chains, where the activity of each independent degree-of-freedom (DOF) is finally mapped onto a distinct agent. Each one of those agents progressively evolves a local kinematic control strategy in a game-theoretic sense, that is, based on a partial (local) view of the whole system topology, which is incrementally updated through a recursive communication process according to the nested-hierarchical topology. Learning is thus approached not through demonstration and training but through an autonomous self-exploration process. A fuzzy reinforcement learning scheme is employed within each agent to enable efficient exploration in a continuous state–action domain. This paper constitutes in fact a proof of concept, demonstrating that global dexterous manipulation skills can indeed evolve through such a distributed iterative learning of local agent sensorimotor mappings. The main motivation behind the development of such an incremental multi-agent topology is to enhance system modularity, to facilitate extensibility to more complex problem domains and to improve robustness with respect to structural variations including unpredictable internal failures. These attributes of the proposed system are assessed in this paper through numerical experiments in different robot manipulation task scenarios, involving both single and multi-robot kinematic chains. The generalisation capacity of the learning scheme is experimentally assessed and robustness properties of the multi-agent system are also evaluated with respect to unpredictable variations in the kinematic topology. Furthermore, these numerical experiments demonstrate the scalability properties of the proposed nested-hierarchical architecture, where new agents can be recursively added in the hierarchy to encapsulate individual active DOFs. The results presented in this paper demonstrate the feasibility of such a distributed multi-agent control framework, showing that the solutions which emerge are plausible and near-optimal. Numerical efficiency and computational cost issues are also discussed.

Using an equivalent source with positivity for low-latitude reduction to the pole without striation

We present a reformulation of reduction to the pole (RTP) of magnetic data at low latitudes and the equator using equivalent sources. The proposed method addresses both the theoretical difficulty of low-latitude instability and the practical issue of computational cost. We prove that a positive equivalent source exists when the magnetic data are produced by normal induced magnetization, and we show that the positivity is sufficient to overcome the low-latitude instability in the space domain. We further apply a regularization term directly to the recovered RTP field to improve the solution. The use of equivalent source also naturally enables the processing of data acquired on uneven surface. The result is a practical algorithm that is effective at the equatorial region and can process large-scale data sets with uneven observation heights.

Programming strategy for efficient modeling of dynamics in a population of heterogeneous cells

Heterogeneity is a ubiquitous property of biological systems. Even in a genetically identical population of a single cell type, cell-to-cell differences are observed. Although the functional behavior of a given population is generally robust, the consequences of heterogeneity are fairly unpredictable. In heterogeneous populations, synchronization of events becomes a cardinal problem-particularly for phase coherence in oscillating systems. The present article presents a novel strategy for construction of large-scale simulation programs of heterogeneous biological entities. The strategy is designed to be tractable, to handle heterogeneity and to handle computational cost issues simultaneously, primarily by writing a generator of the 'model to be simulated'. We apply the strategy to model glycolytic oscillations among thousands of yeast cells coupled through the extracellular medium. The usefulness is illustrated through (i) benchmarking, showing an almost linear relationship between model size and run time, and (ii) analysis of the resulting simulations, showing that contrary to the experimental situation, synchronous oscillations are surprisingly hard to achieve, underpinning the need for tools to study heterogeneity. Thus, we present an efficient strategy to model the biological heterogeneity, neglected by ordinary mean-field models. This tool is well posed to facilitate the elucidation of the physiologically vital problem of synchronization. The complete python code is available as Supplementary Information. bjornhald@gmail.com or pgs@kiku.dk Supplementary data are available at Bioinformatics online.

High dimensional real parameter optimization with teaching learning based optimization

Article history: Received 12 April 2012 Received in revised format 12 June 2012 Accepted June 22 2012 Available online 27 June 2012 In this paper, a new optimization technique known as Teaching–Learning-Based Optimization (TLBO) is implemented for solving high dimensional function optimization problems. Even though there are several other approaches to address this issue but the cost of computations are more in handling high dimensional problems. In this work we simulate TLBO for high dimensional benchmark function optimizations and compare its results with very widely used alternate techniques like Differential Evolution (DE) and Particle Swarm Optimization (PSO). Results clearly reveal that TLBO is able to address the computational cost issue for all simulated functions to a large dimensions compared to other two techniques. © 2012 Growing Science Ltd. All rights reserved

On the Dual Basis for Solving Electromagnetic Surface Integral Equations

A powerful technique for solving electromagnetic (EM) surface integral equations (SIEs) for inhomogenous objects by the method of moments (MoM) involves the well-known Rao-Wilton-Glisson (RWG) basis function to represent the electric current and, for field orthogonality and numerical stability reasons, a variation of the RWG basis known as the ntilde X RWG basis (where ntilde is a unit normal vector at the object surface) to represent the magnetic current. Though this combination provides a numerically efficient and effective solution that has been demonstrated on a variety of structures, one cannot feel entirely comfortable because of the presence of fictitious magnetic current associated with the modified basis. Chen and Wilton proposed a different, smoother basis in 1990 that avoids the fictitious line charges, but because of computational cost issues it has not been used beyond Chen's dissertation. Recently, this basis was rediscovered and has received considerable attention. Our work reexamines the dual basis, exploring issues not addressed by Chen and Wilton and showing accurate solutions for a variety of EM scattering structures.

Ensemble Monte Carlo and Full-Wave Electrodynamic Models Implemented Self-Consistently on a Parallel Processor Using Perfectly Matched Layer Boundary Conditions

We have been using a self-consistent formulation of full-wave electromagnetic solvers and ensemble Monte Carlo techniques to model ultrafast photoconductivity. Our simulations are running on a MasPar machine. This paper will address aspects ofthis simulation which may interest workers who are simulating not only photoconductive systems but other systems as well which involve electrodynamics, waves and wave phenomena and ensemble Monte Carlo transport models. In particular, we will report on the inclusion of perfectly matched layer approaches to absorbing boundary conditions for electromagnetic waves. These have in the past several years become widely used in computational electromagnetics codes because they reduce error due to spurious numerical wave reflection off of an absorbing boundary by several orders of magnitude. We will also address the issue of computational cost and show that a full-wave electromagnetic approach is more competitive with a Poisson's equation approach than one might believe. Lastly, our system has the feature that the active portion where the electrons and holes lie is in fact a small fraction of the total experimental system's volume. Unless care is exerted one either has a very significant load imbalance problem or high communications overhead. We compare two different tradeoffs between load imbalance and communications overhead.

On robust control-oriented identification of discrete and continuous-time systems

Robust control-oriented system identification is considered for discrete and continuous time systems. It is shown here how largely to eliminate the small noise-to- signal ratio and high computational cost issues which have been claimed to limit the usefulness of the new theory. This is achieved through a linear data preprocessing and compression step to obtain small nonlinear constrained optimization problems which are easy to solve even in real-time. In particular, a method based on sample correlation quantities and the Chebyshev criterion is considered. Furthermore, worst-case L1 identification is studied for BIBO stable linear continuous time systems. The existence of robustly convergent identification algorithms is established in the Banach space L1 in a constructive way. The topic of L1 identification of continuous time systems is nicely motivated by the recently developed robust sampled-data control theory of continuous-time systems.

A new method of magnetic resonance image reconstruction with short acquisition time and truncation artifact reduction

A method for the reconstruction of magnetic resonance images that allows for a substantial reduction of the quantity of measured data and, therefore, of the acquisition time is described. The truncation artifact is also reduced, improving the image quality. The method is based on techniques for getting superresolution in spectral analysis such as autoregressive modeling and Prony's method. Moreover, some new ideas about the autoregressive order selection are introduced. The method is compared to the standard one for reconstructing simulated, phantom, and medical magnetic resonance images. The numerical stability and the computational cost issues of the resulting algorithm are also addressed.