Motion control of a novel planar biped with nearly linear dynamics

In this paper, we propose the design of a planar biped for which the model is nearly linear, i.e., the inertia matrix is a constant and the gravity terms in the equations of motion are still nonlinear, but simplified. The legs are designed such that the inertia matrix is independent of the joint variables. As a result, the nonlinear terms in the centrifugal and Coriolis terms disappear. In this design, each leg has two links that are connected by a revolute joint at the knee. The two legs are connected to each other at the hip. The center of mass of each leg is located at the hip, using counterweights. We assume that the stance leg is locked at the knee during the support phase. For this system, dynamic model for complete walking, i.e., swing phase, knee lock, and foot impact are considered. Motion control for trajectory following of this design is studied using a nonlinear controller. The paper discusses the issue of tracking of desired trajectories during the full cycle motion.     

Explanation-Based Learning for Diagnosis

We present explanation-based learning (EBL) methods aimed at improving the performance of diagnosis systems integrating associational and model-based components. We consider multiple-fault model-based diagnosis (MBD) systems and describe two learning architectures. One, EBLIA, is a method for learning in advance. The other, EBL(p), is a method for learning while doing. EBLIA precompiles models into associations and relies only on the associations during diagnosis. EBL(p) performs compilation during diagnosis whenever reliance on previously learned associational rules results in unsatisfactory performance—as defined by a given performance threshold p. We present results of empirical studies comparing MBD without learning versus EBLIA and EBL(p). The main conclusions are as follows. EBLIA is superior when it is feasible, but it is not feasible for large devices. EBL(p) can speed-up MBD and scale-up to larger devices in situations where perfect accuracy is not required.     

Low-sensitivity SC realization of brune and C-type reactance sections

The SC implementation of doubly terminated reactance reference filters containing Brune or C-type sections is presented. It is shown that the one-to-one correspondence between the elements of the reactance reference network and its simulated SC circuit is still maintained. Computer evaluations as well as experimental verifications have shown that the resulting structure is less sensitive with respect to parasitics. Illustrative examples are given.     

Sentence alignment using P-NNT and GMM

Parallel corpora have become an essential resource for work in multilingual natural language processing. However, sentence aligned parallel corpora are more efficient than non-aligned parallel corpora for cross-language information retrieval and machine translation applications. In this paper, we present two new approaches to align English–Arabic sentences in bilingual parallel corpora based on probabilistic neural network (P-NNT) and Gaussian mixture model (GMM) classifiers. A feature vector is extracted from the text pair under consideration. This vector contains text features such as length, punctuation score, and cognate score values. A set of manually prepared training data was assigned to train the probabilistic neural network and Gaussian mixture model. Another set of data was used for testing. Using the probabilistic neural network and Gaussian mixture model approaches, we could achieve error reduction of 27% and 50%, respectively, over the length based approach when applied on a set of parallel English–Arabic documents. In addition, the results of (P-NNT) and (GMM) outperform the results of the combined model which exploits length, punctuation and cognates in a dynamic framework. The GMM approach outperforms Melamed and Moore’s approaches too. Moreover these new approaches are valid for any languages pair and are quite flexible since the feature vector may contain more, less or different features, such as a lexical matching feature and Hanzi characters in Japanese–Chinese texts, than the ones used in the current research.     

Towards a unified visual framework in a binocular active robot vision system

This paper presents the results of an investigation and pilot study into an active binocular vision system that combines binocular vergence, object recognition and attention control in a unified framework. The prototype developed is capable of identifying, targeting, verging on and recognising objects in a cluttered scene without the need for calibration or other knowledge of the camera geometry. This is achieved by implementing all image analysis in a symbolic space without creating explicit pixel-space maps. The system structure is based on the ‘searchlight metaphor’ of biological systems. We present results of an investigation that yield a maximum vergence error of ～6.5 pixels, while ～85% of known objects were recognised in five different cluttered scenes. Finally a ‘stepping-stone’ visual search strategy was demonstrated, taking a total of 40 saccades to find two known objects in the workspace, neither of which appeared simultaneously within the field of view resulting from any individual saccade.     

Optimizing Struvite Production for Phosphate Recovery in WWTP

Struvite, a crystalline structure comprised of ions of magnesium (Mg+2), ammonium (NH4–N), and phosphate (PO4–P), is commonly encountered in wastewater treatment plants (WWTPs) through struvite encrustation. The gradual accumulation of this crystal in pipes and fittings leads to high costs due to downtime and replacement of parts. Technologies that are used to reduce this problem are ideally located in biological nutrient removal plants downstream of anaerobic digesters, as high levels of NH4–N and PO4–P typically characterize anaerobic digester supernatants. In 2003–2004, two technical-scale, struvite recovery studies were conducted on-site at the City of Penticton, B.C., Canada and the City of Richmond, B.C., Canada using a novel technology developed by the environmental engineering group at the University of British Columbia. The results of these studies showed an average reduction of 80% in phosphate and a dense, spherical product, 5–10 times larger than any commercially available struvite to date. Overall, the recovery of struvite has enormous commercial potential, which results in WWTP becoming more sustainable, concurrently reducing problems typically associated with their operation.     

Predictive models for the viscosities of multicomponent liquid n‐Alkane and regular solutions

The pseudo‐binary model developed by Wu and Asfour (1992) has been used to modify some existing viscosity predictive models to that they can predict the viscosities of multicomponent liquid n‐alkane and regular mixtures. The McAllister three‐body interaction model and the Grunberg‐Nissan viscosity equation were both employed, after modification, to successfully predict the viscosity of multicomponent liquid n‐alkane and regular solutions. The modified Generalized Corresponding States Principel (MGCSP) reported by Wu and Asfour (1992), for only predicting the viscosities of n‐alkane mixtures, has been extended to predit the viscosities of multicomponent regular solutions. The predictive capabilities of several predictive models were analyzed. The obtained results showed that the pseudo‐binary McAllister model predicts the viscosity data better than the other existing predictive models.     

Towards automation of color/weave selection in Jacquard design: Model verification

Jacquard woven fabrics are made from colored yarns and different weaves for designing complex pictorial and other patterning effects. The final visualized color effect is the result of assigning weave designs to different areas of the pattern to be created. The current practice in creating Jacquard woven fabric designs is to produce many samples in a trial‐and‐error attempt to match artwork colors. An ability to simulate accurately the appearance of a design prior to manufacture is highly desirable to reduce trial‐and‐error sample production. No automated accurate digital color methodology is yet available to assist designers in matching the patterned woven fabric to the desired artwork. To achieve this, we developed a geometrical model to predict the color contribution of each yarn on the face of the fabric. The geometrical model combined with a Kubelka‐Munk based color mixing model allowed the prediction of the reflectance properties of the final color for a given design. We compared the predicted and experimental values of the reflectance properties for a range of fabrics using the same geometric model with three separate color mixing models. The geometrical model combined with a log‐based color mixing model produced reasonable agreement between predicted and measured ΔE_ab, with an average ΔE_ab of approximately five. © 2009 Wiley Periodicals, Inc. Col Res Appl, 34, 225–232, 2009     

Assessment of motion of a swing leg and gait rehabilitation with a gravity balancing exoskeleton.

The gravity balancing exoskeleton, designed at University of Delaware, Newark, consists of rigid links, joints and springs, which are adjustable to the geometry and inertia of the leg of a human subject wearing it. This passive exoskeleton does not use any motors but is designed to unload the human leg joints from the gravity load over its range-of-motion. The underlying principle of gravity balancing is to make the potential energy of the combined leg-machine system invariant with configuration of the leg. Additionally, parameters of the exoskeleton can be changed to achieve a prescribed level of gravity assistance, from 0% to 100%. The goal of the results reported in this paper is to provide preliminary quantitative assessment of the changes in kinematics and kinetics of the walking gait when a human subject wears such an exoskeleton. The data on kinematics and kinetics were collected on four healthy and three stroke patients who wore this exoskeleton. These data were computed from the joint encoders and interface torque sensors mounted on the exoskeleton. This exoskeleton was also recently used for a six-week training of a chronic stroke patient, where the gravity assistance was progressively reduced from 100% to 0%. The results show a significant improvement in gait of the stroke patient in terms of range-of-motion of the hip and knee, weight bearing on the hemiparetic leg, and speed of walking. Currently, training studies are underway to assess the long-term effects of such a device on gait rehabilitation of hemiparetic stroke patients.     

Ultrasonic investigation of some rubber blends

Compression ultrasonic velocity, longitudinal ultrasonic absorption, and heat of mixing for NR-BR, NR, NBR, and NR-IIR rubber blends were measured. Results show the applicability of examining the compatibility degree by using longitudinal ultrasonic absorption techniques. Results showed that NR-BR blends are compatible, whereas NR-NBR and NR-IIR blends are semicompatible and incompatible, respectively.