Visual feedback control of a robot in an unknown environment (learning control using neural networks)

In this paper, a visual feedback control approach based on neural networks is presented for a robot with a camera installed on its end-effector to trace an object in an unknown environment. First, the one-to-one mapping relations between the image feature domain of the object to the joint angle domain of the robot are derived. Second, a method is proposed to generate a desired trajectory of the robot by measuring the image feature parameters of the object. Third, a multilayer neural network is used for off-line learning of the mapping relations so as to produce on-line the reference inputs for the robot. Fourth, a learning controller based on a multilayer neural network is designed for realizing the visual feedback control of the robot. Last, the effectiveness of the present approach is verified by tracing a curved line using a 6-degrees-of-freedom robot with a CCD camera installed on its end-effector. The present approach does not necessitate the tedious calibration of the CCD camera and the complicated coordinate transformations.

Multi-agent model for robotic assembly system

In this paper, a multi-agent based model for a robotic assembly system is presented. Firstly, an organization model is used to construct the multi-agent model. Secondly, a dynamic self-organizing method is then put forward for the multi-agent robotic system to bid and contract the operations. Thirdly, a real multi-agent robotic system is built and assembly experiments are carried out. Finally, the experimental results confirm that the present multi-agent robotic system has flexibility, adaptation and stability.

An active stereo vision-based learning approach for robotic tracking, fixating and grasping control

In this paper, an active stereo vision-based learning approach is proposed for a robot to track, fixate and grasp an object in unknown environments. First, the functional mapping relationships between the joint angles of the active stereo vision system and the spatial representations of the object are derived and expressed in a three-dimensional workspace frame. Next, the self-adaptive resonance theory-based neural networks and the feedforward neural networks are used to learn the mapping relationships in a self-organized way. Then, the approach is verified by simulation using the models of an active stereo vision system which is installed in the end-effector of a robot. Finally, the simulation results confirm the effectiveness of the present approach.

A reinforcement learning approach for robot control in an unknown environment

In this paper, a control approach based on reinforcement learning is present for a robot to complete a dynamic task in an unknown environment. First, a temporal difference-based reinforcement learning algorithm and its evaluation function are used to make the robot learn with its trials and errors as well as experiences. Second, the simulation are carried out to adjust the parameters of the learning algorithm and determine an optimal policy by using the models of a robot. Last, the effectiveness of the present approach is demonstrated by balancing an inverse pendulum in the unknown environment.

Assessment of delamination in composite beams using shear horizontal (SH) wave mode

Interaction of Lamb waves with structural damage can lead to wave mode conversion. In this study, the shear horizontal (SH) wave, from the mode conversion of the fundamental symmetric Lamb wave (S0), was used for quantitative identification of delamination in composite beams, based on advanced signal processing using an inverse approach. SH wave propagation under various delamination conditions in CF/EP beams made of orthotropic plain woven fabrics was simulated, and signal characteristics were extracted in terms of the concept of digital damage fingerprints (DDF). With the aid of an artificial neural algorithm, the relation between the DDF of delamination-scattered SH mode and damage parameters was calibrated, whereby the occurrence, location and size of delamination in the composite beams were assessed. The approach was experimentally validated, and the results demonstrate the effectiveness of SH mode for quantitative damage evaluation of composite structures.

Preliminary study of the effect of low-intensity home-based physical therapy in chronic stroke patients

This study was a preliminary examination of the effect of low-intensity home-based physical therapy on the performance of activities of daily living (ADL) and motor function in patients more than 1 year after stroke. Twenty patients were recruited from a community stroke register in Nan-Tou County, Taiwan, to a randomized, crossover trial comparing intervention by a physical therapist immediately after entry into the trial (Group I) or after a delay of 10 weeks (Group II). The intervention consisted of home-based physical therapy once a week for 10 weeks. The Barthel Index (BI) and Stroke Rehabilitation Assessment of Movement (STREAM) were used as standard measures for ADL and motor function. At the first follow-up assessment at 11 weeks, Group I showed greater improvement in lower limb motor function than Group II. At the second follow-up assessment at 22 weeks, Group II showed improvement while Group I had declined. At 22 weeks, the motor function of upper limbs, mobility, and ADL performance in Group II had improved slightly more than in Group I, but the between-group differences were not significant. It appears that low-intensity home-based physical therapy can improve lower limb motor function in chronic stroke survivors. Further studies will be needed to confirm these findings.

Light emission and excitonic effect of boron nitride nanotubes observed by photoluminescent spectra

Photoluminescent (PL) and optical absorption spectra of high-yield multi-wall BN nanotubes (BNNTs) were systematically investigated at room temperature in comparison with commercial hexagonal BN (h-BN) powder. The direct band gap of the BNNTs was determined to be 5.75 eV, just slightly narrower than that of h-BN powder (5.82 eV). Two Frenkel excitons with the binding energy of 1.27 and 1.35 eV were also determined. However, they were not a distinctive characteristic of the BNNTs as reported previously. Observed broad UV–visible–NIR light emission demonstrates the potential of the BNNTs as a nano light source.

Partial purification and characterization of limonoate dehydrogenase from Rhodococcus fascians for the degradation of limonin

An extracellular limonoate dehydrogenase was purified 10-fold from a cell-free extract of Rhodococcus fascians by ammonium sulfate precipitation, dialysis, and ultrafiltration. This purified dehydrogenase catalyzed the
conversion of limonoate to 17-dehydrolimonoate. The enzyme showed optimum activity at pH 8.0 and 40oC, with K_m value of 0.9 µM, and requires Zn ions and sulfhydryl groups for catalytic action. The enzyme activity was inhibited by Hg²⁺ and NaN₃ ions. The degradation of limonin (66%) in Kinnow mandarin juice was successfully demonstrated with partially
purified limonoate dehydrogenase. With scale-up preparation of limonoate dehydrogenase, a successful debittering operation of fruit juices appears feasible.

Aligning school reform and teacher education reform in the middle years : an australian case study

This paper discusses the development of a new Bachelor of Education (Middle Years of Schooling) at The University of Queensland. The middle years of schooling have increasingly been the focus of education reform initiatives in Australia, but this has not been accompanied by significant increases in the number of teacher education institutions offering specialised middle schooling-level teacher preparation programmes. Considering the rapidly changing social and economic context and the emergent state of middle schooling in Australia, the programme represented a conceptual and practical opportunity and challenge for The University of Queensland team. Working collaboratively, the team sought to design a teacher education preservice programme that was both responsive and generative: that is, responsive to local school contexts and to current educational research and reform at national and international levels; and generative of cutting-edge theories and practices associated with middle schooling, teachers' work, and teacher education. This paper focuses on one component of the Middle Years of Schooling Teacher Education programme at The University of Queensland; namely, the practicum. We first present the underlying principles of the practicum programme and then examine "dilemmas" that emerged early in the practicum. These issues and tensions were associated with the ideals of "middle years" philosophy and the pragmatics of school reform associated with that new approach. In this paper, and within this context, we explore what it means to be both responsive and generative, and describe how we as teacher educators negotiated between the extremes these terms implied.

Warrock collection: original line drawings by fourth year Deakin University architecture students of the Warrock Homestead, a local heritage listed property in Victoria

This collection includes the original line drawings by fourth year Deakin University architecture students of a well known local heritage listed property, the Warrock homestead. The Warrock homestead consisted of detailed original timber structures of the 19th century. The drawings are the result of a conservation project funded by the Commonwealth Government of Australia National Estate Grants program. In 1999 a further deposit of original reports relating to individual buildings on the property was received. The collection consists of monographs, photographs and photograph negatives, architectural drawings, VHS tapes and ephemera.