Ebook: Simulation and Modeling Related to Computational Science and Robotics Technology

This volume includes a selection of papers presented at the International Workshop on Simulation and Modeling related to Computational Science and Robotics Technology (SiMCRT2011) was organized by Kobe University during November 1–3, 2011 at the Takikawa Memorial Hall, Kobe University Japan. This workshop was also co-sponsored by AFOSR/AOARD under Grant No. FA23861111041. Historically, this workshop was the third international meeting for providing a forum for discussing recent developments in the growing field of engineering science and mathematical sciences. The first workshop took place in 1997 in Osaka Institute of Technology, Osaka, Japan. That was one of the satellite meetings of IEEE Conference on Decision and Control that was held in Kobe Japan. The second meeting was held in Kobe University in 2007 as the International Symposium on Mathematical Modeling and Computational Methods in Science and Engineering (MMCOM2007). At the course of continuing efforts on the international collaborations, those meetings have been extended to the academic research fields. In that respect, the objectives of this workshop are to i) bring together a diverse group of researchers in these areas in order to share and compare the different approaches to simulation and modeling for broad range of applications in computational science and robotics technology, ii) provide an evaluation of the state of the art in related computational methods and iii) present specific applications where there is a need to develop new computational methods in order to address important design and control problems. The workshop was also aimed at establishing collaborative links between engineering researchers of information and robotics technology (IRT) and applied mathematicians working in modeling and computational methods for design and control. In addition to the formal presentations, time will be allotted for researchers to discuss and evaluate possible applications of new design strategies.

F. Kojima, F. Kobayashi, H. Nakamoto

Editors

Continuous time Markov chains are often used in the literature to model the dynamics of a system with low species count and uncertainty in transitions. In this paper, we investigate three particular algorithms that can be used to numerically simulate continuous time Markov chain models (a stochastic simulation algorithm, explicit and implicit tau-leaping algorithms). To compare these methods, we used them to analyze two stochastic infection models with different level of complexity. One of these models describes the dynamics of Vancomycin-Resistant Enterococcus (VRE) infection in a hospital, and the other is for the early infection of Human Immunodeficiency Virus (HIV) within a host. The relative efficiency of each algorithm is determined based on computational time and degree of precision required. The numerical results suggest that all three algorithms have similar computational efficiency for the VRE model due to the low number of species and small number of transitions. However, we found that with the larger and more complex HIV model, implementation and modification of tau-Leaping methods are preferred.

We present a nonsmooth optimization method for solving the elastic contact problem. The Signorini contact problem is a variational problem that minimizes the elastic deformation energy subject to the contact inequality, i.e., the normal displacement at a given point of the boundary bounded above by an obstacle function. The Coulomb friction problem is a minimization of the deformable energy with a L¹ friction term at the boundary. We develop an effective numerical optimization method using the semi-smooth Newton method for the both variational problems. The method is of the form of Primal-Dual active set methods for Lagrange multiplier methods. We approximate these variational formulations with a multi-moment scheme based on Adini's elements which involves the use of the function values as well as the gradient values at nodes. The Primal-Dual active set method are then applied to these approximations. Finally we combine the solutions to the Signorini and Coulomb friction problems to solve the full contact problem.

The use of surrogate species is a tool commonly used to predict the effects of toxicants on endangered/threatened or economically important species. While use of surrogate species has been critized as being overly simplistic, a quantitative measure linking life history traits and population predictions has been sorely missing. We derive here a closed-form expression aimed at determining conditions under which sublethal effects of a toxicant on surrogate species population outcomes will reliably predict responses of species of concern. We derive a simple inequality that allows us to compare critical thresholds in fecundity reduction across species and thereby pinpoint the level below which surrogate species outcomes indicate a positive population growth, while the listed species actually is driven to extinction. We thus establish a means of determining conditions under which we might be prone to making a “type II” error in assessing ecological risk using surrogate species. Finally, we use the derived expression to illustrate two cases studies – one in which we are using several fish species as surrogates for endangered salmonids, and the second in which we are comparing the compatibility of a suite of parasitoid wasps with pesticide use. In both cases we highlight potential pitfalls associated with the use of a “one-size-fits-all” approach to protection of species. We discuss the ramifications of these findings on risk assessment and resource management.

Simulation of multibody systems is associated with deriving equations of motion and finding numerical solution of the equation. The combination of the differential equations and constraints yields index-3 differential-algebraic equations (DAE's) that are not, in general, easily solvable by standard integration schemes. Moreover at singular configurations, some methods can fail. This paper focuses on the discussion of two problems: determining the singular configurations and their neighborhood and overcoming the singularity smoothly. Overcoming the singularity is discussed with using the Principle of Compatibility, so far not well-known. In this principle the equations of motion are rewritten in the form which can be solved by numerical techniques smoothly, even over singular configurations without detector. The idea of this approach is introducing so-called generalized reaction forces which appear in the equations of motion system for the dynamical in comparison with the system without constraints. The formulation is proven to be more stable and accurate under repetitive meeting singular configurations. These generalized reaction forces can be determined by using the ideality condition of constraints which employs the null space of constraints imposed to the system. Some numerical experiments are carried out to verify the efficiency and speed of the approach.

In this paper, we study inverse problems for advection-diffusion equations by means of a new deformation formula. As an application of results, we solve the inverse flux problem for parallel-flow two-fluid heat exchanger process.

By using piezoelectric film the present authors proposed the passive electric potential CT (computed tomography) method and the active pulse echo method for crack identification with the help of the inverse analysis. This paper describes the passive and the active methods using the piezoelectric film and their applications to the identification of cracks. The passive electric potential CT method used the electric potential distribution incurred due to the direct piezoelectric effect on the surface of a piezoelectric film pasted on a cracked body subjected to mechanical load. The present authors proposed a smart layer composed of the piezoelectric film and flexible printed circuit. By applying electric pulse to the smart layer a supersonic wave is emitted due to the inverse piezoelectric effect. The active pulse echo method used the reflected wave observed on the smart layer for the crack identification. The applicability of these methods to the identification of two- and three-dimensional cracks was demonstrated.

In this paper, an efficient forward numerical simulation solver for pulsed eddy current testing (PECT) signals has been proposed and developed based on the Fourier series method combined with an interpolation strategy and the database approach. Initially, the PECT signal simulation method based on the Fourier series scheme and an interpolation approach is described. Second, the database type fast numerical solver for single frequency ECT problem is introduced to the Fourier-series-based PECT simulation to enhance its simulation efficiency. In addition, an inversion algorithm of PECT has also been proposed and validated, based on the developed efficient simulator of PECT signals and a deterministic optimization strategy, for the profile reconstruction of wall thinning in pipes of nuclear power plants (NPPs). Through inverse analysis using the simulated PECT signals and signals with random noise, the efficiency and the robustness of the proposed PECT inversion algorithm have been demonstrated.

In this article, it is shown that inverse analysis using data assimilation is an indispensable problem-solving methodology for structural health monitoring. Forward and inverse analysis for eddy current testing are considered as sizing estimation of material damages in structured systems. First, testing environments are mathematically described based on a three dimensional eddy current analysis. Secondly, an inversion technique based on Markov Chain Monte Carlo method (MCMC) is employed for solving the crack shape reconstruction. Results on computational experiments are demonstrated in order to show the feasibility and applicability of the proposed method.

Neuro-control which adopts neural network architectures to synthesis of control has been summarized and its application to electric vehicle control is developed in this paper. The neuro-control methods adopted here is based on proportional-plus-integral-plus-derivative (PID) control, which has been adopted to solve process control or intelligent control. In Japan about eighty four per cent of the process industries have used the PID control. Using the learning ability of the neural network, we will show the self-tuning PID control scheme (neuro-PID) and the real application to an electric vehicle control.

Recently, multiple mobile robots have been used for exploration and remote monitoring in unknown areas. The required functions for human-friendly remote monitoring and control are (1) Extraction of human intention, (2) Semi-autonomous tele-operation, (3) Multi-robot formation behaviors, (4) Sensor fusion for information extraction, (5) Perception of situation, (6) Decision support based on information visualization and (7) Extraction of human interest. We proposed an intelligent navigation system for human-friendly remote monitoring and control. The intelligent navigation system for multiple mobile robots is composed of simultaneous localization and mapping (SLAM), intelligent path planning based on multi-resolution map, 3D information visualization system, and human navigation system based on touch interface. In this chapter, we focus on the information visualization and map building for remote monitoring and control of multiple mobile robots, and show several experimental results of the proposed methods.

Genetic algorithms (GAs) have been successfully used for data mining thanks to their flexibility. Users easily incorporate their preference into objective functions to be optimized. Although GA-based data mining techniques are useful, there is a serious difficulty in the handling of data sets with a large number of patterns and/or attributes. That is, we need much long computation time for the evaluation of candidate solutions because all the patterns have to be classified by each candidate solution. To reduce the computation time of GA-based data mining, we propose a parallel distributed implementation of genetic rule selection. The main characteristic is to divide not only a population in GA but also a training data set into a number of sub-groups. Then a pair of a sub-population and a training data subset is assigned to a single CPU core. This approach can drastically reduce the computation time with no serious deterioration in the generalization ability of obtained classifiers. We demonstrate the effectiveness of our parallel distributed implementation through computational experiments on large data sets available from the UCI machine learning repository.

This paper gives a comprehensive study on “computational robotics” motivated by the recent rapid development of supercomputer technologies. By applying advanced computational scientific approach, it is expected to realize deeper understanding of human as well as to solve more complicated computational problems for the next generation of robotics. Specifically, three main problems are discussed. The first one is about high dimensional super-redundancy considering the full body human motor functions. With the rapid computation ability, it will be possible to realize simultaneous human movement measurement with dynamic human motion analysis and simulation within real time. The second topic is on the massive computation of a robot's complex full body physical interactions with environment and/or manipulated objects such as a cared person. Here, how to simulate, plan and control the interactions with many kinds of frictions are discussed. The discrete feature of physical contacts together with continuous behaviors of motion/force control makes the problem as a hybrid dynamic system. Systematic approach to solve such problem requires massive computation. The final one simply studies on a robot's cognitive motion in the real world so that to adapt to the unknown and/or uncertain environment. Such problem requires to process huge cognitive information inputs within real time so as to determine the robot's environmental adaptive actions.

The multi-fingered robot hand has much attention in various fields. Many robot hands have been proposed so far. We have also developed a small and five-fingered robot hand in order to carry out various tasks and the robot hand teleoperation system in order to utilize the human help. However, the operator cannot pinch and manipulate an object with complicated shapes stably because the operator cannot feel the pinching force when the robot hand pinches and manipulates the object in the teleoperation system. In this paper, we propose a pinching force stabilization method in the robot hand teleoperation system. Here, the operator does not teleoperate all robot fingers directly, but a part of fingers are autonomously controlled in order to stabilize the pinching force. The effectiveness of the proposed pinching force stabilization method are verified through some pinching experiments with various shapes.

Many multi-fingered robot hands have been developed so far. However, there are no multi-fingered robot hands for actual use. One of the reasons is that the conventional control methods are not used for multi-fingered robot hands. A human manipulates an object by using tactile information (contact points, direction and severity of reaction force). In this paper, the tactile based control method for multi-fingered robot hands is proposed like a human manipulation. In the proposed method, 3 fingers are used for the object manipulation. By using this method, object grasping and manipulating is achieved. The proposed method is implemented in Universal Robot Hand II, and the effectiveness is verified.

We propose a tactile sensor that uses variation of magnetic field. The sensor has a simple structure of two layers, an elastic layer and a substrate layer. The elastic layer is made of an elastic material and houses a cylindrical permanent magnet inside. When an object touches the surface of the elastic layer, the layer deforms. The deformation displaces the magnet and the magnetic field for the substrate layer changes. The substrate layer is a glass epoxy board, and has four giant magneto resistance (GMR) elements and four inductors. These elements detect changes of magnetic field strength from the magnet. It is possible to calculate three-axis displacement and force and detect slips from outputs of the elements. In this paper, we describe a structure of the sensor. Then, we confirm a fundamental performance of the sensor and show response character of the elements.

A saliency-based approach is presented to the transfer of landmark information between robotic vehicles participating in over-the-horizon maneuvering processes. By identifying chromatic diversity of naturally complex scenes with degenerated version of fractal attractors, probe vehicles detect and upload as-is representation of scene specific primaries as annotation of the local geographics. In reference to the annotation, future visitors adaptively restore the saliency pattern to be matched with landmarks in encountered scenes. Through experimental studies, the saliency based geographics annotation was demonstrated to significantly reduce the complexity of noisy background.

This study reports the development of monitoring method and system for underground condition. The method and the system have been developed to test industrial fields by their microwave signals. In this system, microwaves are transmitted from an antenna on the ground surface into the ground. The echo signals returned from underground targets are analyzed by using signal processing methods such as the synthetic aperture processing. The images of underground condition are reproduced. The results of testing using this method and system are attained the horizontal resolution 30 cm for a diameter of 5cm iron pipes buried at a depth of 1.5m.This underground condition monitoring system has been tested in our fields and can be also applied to underground exploration for not only industrial fields but other kinds of fields such as fields for the remains of ancient.

This paper describes quantitative nondestructive evaluations of wall thinning in power plants by electro-magnetic acoustic transducers (EMAT). Among Japan's aging plants, their inspections and costs are increasing. Therefore, Interest is currently focused on continuous monitoring to the plants in operation. Inspection using EMAT is suitable for the continuous monitoring, because EMAT itself is attracted on the pipe. In this paper, we tested corrosion shapes of straight pipes which were machined to simulate flow-accelerated corrosions. Additionally, we measured thicknesses of an elbow pipe. The measured values gave close agreement with the observed values which were measured by an ultrasound thick gauge. We confirm that EMAT has the capability to measure thicknesses of corrosion wastages of the pipes.