Analysis of a micro–macro acceleration method with minimum relative entropy moment matching

We analyse convergence of a micro–macro acceleration method for the simulation of stochastic differential equations with time-scale separation. The method alternates short bursts of path simulations with the extrapolation of macroscopic state variables forward in time. After extrapolation, a new microscopic state is constructed, consistent with the extrapolated macroscopic state, that minimises the perturbation caused by the extrapolation in a relative entropy sense. We study local errors and numerical stability of the method to prove its convergence to the full microscopic dynamics when the extrapolation time step tends to zero and the number of macroscopic state variables tends to infinity.     

Shape optimisation problem for stability of Navier–Stokes flow field

ABSTRACT

This paper presents the numerical results of a shape optimisation problem with regard to delaying the transition of a Navier–Stokes flow field from laminar to turbulent by using the theory developed by Nakazawa and Azegami. The theory was reviewed within the framework of functional analysis and updated with another expression of the shape derivative with respect to the objective function. A computer program was developed with the FreeFEM++. Numerical analyses were performed for two types of problems: a two-dimensional Poiseuille flow field with a sudden expansion and a two-dimensional uniform flow field around an isolated body. From the first example, two local minimum points of symmetric and asymmetric flow fields were determined, and the asymmetric flow field was found to be more stable. With regard to the second example, we reached the local minimum point of an elliptical shape, and infrequently determined a solution converging to an elliptical shape with the bluff in the leeward direction. By comparison, the superiority of the elliptical shape was obvious.     

Dynamics of a Mobile Crane and Optimisation of the Slewing Motion of Its Upper Structure

In the paper the 3-D model of a telescopic mobile crane is presented. In the model flexibilities and damping have been taken into account and also the number of degrees of freedom can be chosen according to the complexity of an approach. The algorithm of optimisation of drive functions for the slewing upper structure is given. The main goal of optimisation was to ensure load positioning at the end point of a work cycle in the case when oscillations are minimal. In order to achieve appropriate numerical effectiveness, the optimisation problem was solved for the simple, fully rigid model of a mobile crane. The method of compensating neglected flexibilities of the base structure of the crane and inner disturbances connected with the feedback control system via digital PID controller is presented.     

Une technique de lissage de données basée sur la théorie de la régularisationA filter technique based on regularization theory

A simple filter technique based on the regularization theory is presented. We consider the problem as an optimization one. The regularization theory gives us a suitable theoretical framework to define a functional to minimize. We make a numerical comparation between this method and a classical Fourier technique. To cite this article: J.-M. Fullana, C. R. Mecanique 330 (2002) 647–652.     

New Interfaces and Approaches to Machine Learning When Classifying Gestures within Music

Interactive music uses wearable sensors (i.e., gestural interfaces—GIs) and biometric datasets to reinvent traditional human–computer interaction and enhance music composition. In recent years, machine learning (ML) has been important for the artform. This is because ML helps process complex biometric datasets from GIs when predicting musical actions (termed performance gestures). ML allows musicians to create novel interactions with digital media. Wekinator is a popular ML software amongst artists, allowing users to train models through demonstration. It is built on the Waikato Environment for Knowledge Analysis (WEKA) framework, which is used to build supervised predictive models. Previous research has used biometric data from GIs to train specific ML models. However, previous research does not inform optimum ML model choice, within music, or compare model performance. Wekinator offers several ML models. Thus, we used Wekinator and the Myo armband GI and study three performance gestures for piano practice to solve this problem. Using these, we trained all models in Wekinator and investigated their accuracy, how gesture representation affects model accuracy and if optimisation can arise. Results show that neural networks are the strongest continuous classifiers, mapping behaviour differs amongst continuous models, optimisation can occur and gesture representation disparately affects model mapping behaviour; impacting music practice.     

POD enriched boundary models and their optimal stabilisation

Modes obtained using the Proper Orthogonal Decomposition are used as boundary enrichment functions within a variational multiscale method for the stochastically forced Burgers equation. Initially, large increases in accuracy are obtained using the enrichment functions without stabilisation terms. Then, optimal coefficients for the stabilisation parameter τ of the unresolved scale model are calibrated using a goal‐oriented model‐constrained optimisation technique, resulting in further improvements. As both the determination of the enrichment functions and the optimisation of the coefficients requires high‐accuracy reference data, a scaling procedure is introduced to allow their use over range of conditions. Numerical experiments confirm that the scaling procedure is effective. Copyright © 2014 John Wiley & Sons, Ltd.     

Structural evolution and properties of small-size thiol-protected gold nanoclusters

Ligand-protected gold clusters are widely used in biosensors and catalysis. Understanding the structural evolution of these kinds of nanoclusters is important for experimental synthesis. Herein, based on the particle swarm optimisation algorithm and density functional theory method, we use [Au₁(SH)₂]_n, [Au₂(SH)₃]_n, [Au₃(SH)₄]_n (n?=?1–3) as basic units to research the structural evolution relationships from building blocks to the final whole structures. Results show that there is a ‘line-ring-core’ structural evolution pattern in the growth process of the nanoclusters. The core structures of the ligand-protected gold clusters consist of Au₃, Au₄, Au₆ and Au₇ atoms. The electronics and optics analysis reflects that stability and optical properties gradually enhance with increase in size. These results can be used to understand the initial growth stage and design new ligand-protected nanoclusters.     

Using radial basis functions to construct local volatility surfaces

This article presents a new method for constructing a volatility surface for use in local volatility option pricing models. It builds on previous work focussing on non-parametric regression approaches using a set of radial basis functions, specifically thin plate splines. Optimal parameters are found using a trust region optimisation approach. While there is still much work to be done, the results are encouraging and show that the method is relatively tractable, stable and accurate.