International Professional University of Technology in Nagoya Repository
Not a member yet
15131 research outputs found
Sort by
On Weight-Prioritized Multitask Control of Humanoid Robots
No full textInternational audienceWe propose a formal analysis with some theoretical properties of weight-prioritized multi-task inverse-dynamics-like control of humanoid robots, being a case of redundant " ma-nipulators " with a non-actuated free-floating base and multiple unilateral frictional contacts with the environment. The controller builds on a weighted sum scalarization of a multiobjective optimization problem under equality and inequality constraints, which appears as a straightforward solution to account for state and control input viability constraints characteristic of humanoid robots that were usually absent from early existing pseudo-inverse and null-space projection-based prioritized multi-task approaches. We argue that our formulation is indeed well founded and justified from a theoretical standpoint, and we propose an analysis of some stability properties of the approach: Lyapunov stability is demonstrated for the closed-loop dynamical system that we analytically derive in the unconstrained multiob-jective optimization case. Stability in terms of solution existence, uniqueness, continuity, and robustness to perturbations, is then formally demonstrated for the constrained quadratic program
About detecting steam condensation by means of polymer racetrack micro-resonators: highlighting the dynamics of such a soft-matter process
No full textInternational audienceWe have investigated the effect of brutal steam condensation processes and the behavior of its condensed water prior evaporation, with an integrated resonant photonic structure and dynamic tracking of its transduced signal. The aim of this analysis is to develop a steam condensation lab-on-chip sensor, with the possibility of data treatment with an embedded system. We have designed and fabricated integrated photonic micro-resonators (MRs) devices in polymer UV210 by means of Deep-UV photolithography. The UV210 polymer is a Chemically Amplified (CA) positive resin containing a Photo Acid Generator (PAG), featuring an absorption band in the Deep-UV domain. Such a property allows us to take advantage of a smaller insolation wavelength than that of a traditional i-line photolithography (here, λ insolation =248nm), so as to make smaller and more precise structures. Due to exposure, the PAG entails a cascade of chemical reactions that leads to a change of polarity in the resin, going from a lipophilic state to a hydrophilic one. This changing of state makes possible the development by means of the basic solvent tetra-methyl ammonium hydroxide (TMAH). Thanks to this technique, we have achieved racetrack shaped micro-resonators coupled to an access waveguide. We have made such MRs with different geometrical characteristics while changing respectively; the coupling length (L C), the radius of curvature (R) and the width (w) of the guides. The chosen values for the set of parameters L C-R-w (in µm) are: 5-5-3, 5-5-4, 10-10-3 and 10-10-4. The laser source used with the injection bench is a Gaussian broadband laser (λ central =790nm, FWHM=40nm) allowing us to visualize several resonances at the same time in order to multiplex the relevant measurements. The transduced spectrum is then acquired with an Optical Spectrum Analyzer (OSA) linked to a computer on which Labview and Matlab software record and process the data in real time. The relevant characteristics to be tracked are the Free Spectral Range (FSR), the Full Width at Half Maximum (FWHM), the extinction ratio, the transmitted energy and the shift of the different resonances. These quantities can be linked to the physical characteristics of the structure such as effective refractive index, coupling coefficient and absorption coefficient. The experimental setup also includes various movies, a top-view imaging camera of the chip (MRs) during such a soft matter process, so as to correlate the changes in the transduced spectrum and the behavior of the condensed steam mechanisms (condensation, coalescence and evaporation). Then, the chip is linked to a temperature controller, so as to carry out measurements at different temperatures: 22°C, 25°C, 28°C and 31°C
Alternating Optimisation and Quadrature for Robust Control
No full textInternational audienceBayesian optimisation has been successfully applied to a variety of reinforcement learning problems. However, the traditional approach for learning optimal policies in simulators does not utilise the opportunity to improve learning by adjusting certain environment variables - state features that are randomly determined by the environment in a physical setting but are controllable in a simulator. This paper considers the problem of finding an optimal policy while taking into account the impact of environment variables. We present alternating optimisation and quadrature (ALOQ), which uses Bayesian optimisation and Bayesian quadrature to address such settings. ALOQ is robust to the presence of significant rare events, which may not be observable under random sampling, but have a considerable impact on determining the optimal policy. We provide experimental results demonstrating our approach learning more efficiently than existing methods
Multi-patch and multi-group epidemic models: A new framework
No full textInternational audienceWe develop a multi-patch and multi-group model that captures the dynamics of an infectious disease when the host is structured into an arbitrary number of groups and interacts into an arbitrary number of patches where the infection takes place. In this framework, we model host mobility that depends on its epidemiological status, by a Lagrangian approach. This framework is applied to a general SEIRS model and the basic reproduction number R0 is derived. The effects of heterogeneity in groups, patches and mobility patterns on R0 and disease prevalence are explored. Our results show that for a fixed number of groups, the basic reproduction number increases with respect to the number of patches and the host mobility patterns. Moreover, when the mobility matrix of susceptible individuals is of rank one, the basic reproduction number is explicitly determined and was found to be independent of the latter if the matrix is also stochastic. The cases where mobility matrices are of rank one capture important modeling scenarios. Additionally, we study the global analysis of equilibria for some special cases. Numerical simulations are carried out to showcase the ramifications of mobility pattern matrices on disease prevalence and basic reproduction number
Deep neural network based multichannel audio source separation
No full textInternational audienceThis chapter presents a multichannel audio source separation framework where deep neural networks (DNNs) are used to model the source spectra and combined with the classical multichannel Gaussian model to exploit the spatial information. The parameters are estimated in an iterative expectation-maximization (EM) fashion and used to derive a multichannel Wiener filter. Different design choices and their impact on the performance are discussed. They include the cost functions for DNN training, the number of parameter updates, the use of multiple DNNs, and the use of weighted parameter updates. Finally, we present its application to a speech enhancement task and a music separation task. The experimental results show the benefit of the multichannel DNN-based approach over a single-channel DNN-based approach and the multichannel nonnegative matrix factorization based iterative EM framework
Uncertainty theory as a basis for belief reliability
No full textInternational audienceBelief reliability is a newly developed, model-based reliability metric which considers both what we know (expressed as reliability models) and what we don't know (expressed as epistemic uncertainty in the reliability models) about the reliability. In this paper, we show that due to the explicit representation of epistemic uncertainty, belief reliability should not be regarded as a probability measure; rather, it should be treated as an uncertain measure in uncertainty theory. A minimal cut set-based method is developed to calculate the belief reliability of coherent systems. A numerical algorithm is, then, presented for belief reliability analysis based on fault tree models. The results of application show that the developed methods require less computations than the structure function-based method of classical reliability theory
On the Statistical Resolution Limit (SRL) for Time-Reversal based MIMO radar
No full textInternational audienceIn the single-input multiple-output radar, the system transmits scaled (coherent) versions of a single waveform. The multiple-input multiple-output (MIMO) radar uses multiple antennas to simultaneously transmit several non-coherent waveforms and exploits multiple antennas to receive the reflected signals (echoes). This diversity in term of waveform coding allows to transmit orthogonal waveforms which enables the MIMO radar superiority in several fundamental aspects, including: improved parameter identifiability and estimation and much enhanced flexibility for transmit beam-pattern design. The context of this work is the co-located MIMO radar where the transmit and the receive arrays are close in space. In this paper, we provide a theoretical performance analysis to compare two configurations of MIMO radar: Conventional configuration and Time Reversal (TR) configuration in term of Statistical Resolution Limit (SRL). This study provides new insights on the performance gain of the TR scheme which is discussed and illustrated by appropriate simulation results depending on the receive noise level
Chapter 35: Free Simulation Software and Library
No full textInternational audienceWith the advent of powerful computation technologies and efficient algorithms , simulators became an important tool in most engineering areas. The field of humanoid robotics is no exception; there have been numerous simulation tools developed over the last two decades to foster research and development activities. With this in mind, this chapter is written to introduce and discuss the current-day open source simulators that are actively used in the field. Using a developer-based feedback, we provide an outline regarding the specific features and capabilities of the open-source simulators, with a special emphasis on how they correspond to recent research trends in humanoid robotics. The discussion is centered around the contemporary requirements in humanoid simulation technologies with regards to future of the field
A novel non-intrusive method using design of experiments and smooth approximation to speed up multi-period load-flows in distribution network planning
No full textInternational audienceAlternative solutions to network reinforcement are now being investigated in distribution network planning studies to reduce the costs and periods for integrating renewable energy sources. However, a thorough techno-economic analysis of these solutions requires a large number of multi-period load-flow calculations, which makes it hard to implement in planning tools. A non-intrusive approximation method is therefore proposed to obtain fast and accurate multi-period load-flows. This method builds a surrogate model of the load-flow solver using polynomial regression and kriging, combined with Latin hypercube sampling. Case studies based on real distribution networks show that the proposed method is more efficient for distribution network planning in presence of renewable energy sources than time subsampling and, in some cases, voltage linearization. In particular, accurate 10-minute profiles of voltages, currents, and network power losses are obtained in a satisfactory computation time
On the reversibility of ECAs with fully asynchronous updating: the recurrence point of view
No full textInternational audienceThe reversibility of classical cellular automata is now a well-studied topic but what is reversibility when the evolution of the system is stochastic? In this context, we study a particular form of reversibility: the possibility of returning infinitely often to the initial condition after a random number of time steps. This corresponds to the recurrence property of the system. We analyse this property for the 256 elementary cellular automata with a finite size and a fully asynchronous updating, that is, we update only one cell, randomly chosen, at each time step. We show that there are 46 recurrent rules which almost surely come back to their initial condition. We analyse the structure of the communication graph of the system and find that the number of the communication classes may have different scaling laws, depending on the active transitions of the rules (those for which the state of the cell is modified when an update occurs)