Due to their monolithic construction and superior wear and loss properties, flexure joints have been used to reduce the mechanism size and increase the positioning accuracy. The compliance of flexure joints, however, ...
详细信息
Due to their monolithic construction and superior wear and loss properties, flexure joints have been used to reduce the mechanism size and increase the positioning accuracy. The compliance of flexure joints, however, can affect the static and dynamic characteristics of the overall mechanism. To design mechanisms containing flexure joints, we have proposed a multi-objective optimization approach to take into account the multitude of performance metrics and design constraints. A Pareto frontier is first computed, and secondary design criteria, such as sensitivity and dynamic characteristics, are then applied to select the final design. To reduce the computation load and facilitate design iteration, a lumped spring approximation, the Paros-Weisbord model, is used to characterize the flexure joints and the pseudo-rigid-body model is used as an approximate description of mechanisms. This paper presents this approach applied to the design of a micro-gripper. The performance metrics are chosen to be the manipulability of the gripper opening and the decoupling of a stiffness matrix (reflecting the remote center of compliance criterion). Different design and initial fabrication results are included
Multi-FPGA systems (MFSs) are used as custom computing machines, logic emulators, and rapid prototyping vehicles. A key aspect of these systems is their programmable routing architecture which is the manner in which w...
详细信息
Multi-FPGA systems (MFSs) are used as custom computing machines, logic emulators, and rapid prototyping vehicles. A key aspect of these systems is their programmable routing architecture which is the manner in which wires, FPGAs, and Field-Programmable Interconnect Devices (FPIDs) are connected. Several routing architectures for MFSs have been proposed and previous research has shown that the partial crossbar is one of the best existing architectures. A new routing architecture, called the Hybrid Complete-Graph and Partial- Crossbar (HCGP), was proposed by Khalid and was shown to provide superior speed and cost compared to partial crossbar. In this paper we address the issue of scalability of the HCGP routing architecture. The motivation for this work was to evaluate the suitability of the HCGP architecture for a future rapid prototyping system product that was being developed at Cadence. Experimental results show that the HCGP architecture is scalable and can be used with the state-of-the-art, high gate count FPGAs.
This paper presents an extension of the latency insertion method (LIM) to the treatment of circuits with frequency-dependent parameters. In particular, the skin effect of conductors is modeled in the formulation and t...
详细信息
This paper presents an extension of the latency insertion method (LIM) to the treatment of circuits with frequency-dependent parameters. In particular, the skin effect of conductors is modeled in the formulation and the simulation results are compared with actual measurements
This paper proposes a modular-based classifier for the problem of phoneme recognition. This is carried out by the use of a two-level classification approach including, high and low levels. We propose a new concept cal...
详细信息
This paper proposes a modular-based classifier for the problem of phoneme recognition. This is carried out by the use of a two-level classification approach including, high and low levels. We propose a new concept called phoneme family. To obtain phoneme families, we employ k-mean clustering method. A given unknown phoneme is first classified into a phoneme family at high level classification. Then, the exact label of the phoneme is determined at low level classification. We have used a combined framework of statistical and neural network based classifiers. Encouraging results are obtained by applying the proposed method on TIMIT database and its performance is compared against other methods
The problem of delayed reward in reinforcement learning is usually tackled by implementing the mechanism of eligibility traces. In this paper we introduce an extension of eligibility traces to solve one of the challen...
详细信息
The needs of vision impaired students are quite different to sighted students. The increasing use of e-learning means higher education must move to multi-modal user interfaces in order to make e-learning materials acc...
详细信息
The needs of vision impaired students are quite different to sighted students. The increasing use of e-learning means higher education must move to multi-modal user interfaces in order to make e-learning materials accessible to all students. E-Learning materials (particularly in the sciences and technology) are predominantly visual, presented via computer keyboard and screen. Software and devices designed to aid the vision impaired are unable to decipher most images and visual- centric objects contained in e-learning materials. This paper discusses a project undertaken over the past two years to modify the content and presentation of Cisco certification e-learning courses to enable accessibility by vision impaired and blind students. These modifications necessitated rewriting the learning materials so they could be effectively presented via multi-modal user interfaces to vision impaired students, involving speech, audio, haptics and force-feed devices and methods. Evaluation of sections of the project by the vision impaired students using a model based upon Stufflebeam's CIPP model and Kirkpatrick's Four-Level training program evaluation model has been carried out and the results are presented.
Although multi-finger MOS structures have demonstrated very attractive behaviors for high-frequency analog circuit applications, the shared drain and source regions for adjacent gate fingers have led to current crowdi...
详细信息
The problem of delayed reward in reinforcement learning is usually tackled by implementing the mechanism of eligibility traces. In this paper we introduce an extension of eligibility traces to solve one of the challen...
详细信息
The problem of delayed reward in reinforcement learning is usually tackled by implementing the mechanism of eligibility traces. In this paper we introduce an extension of eligibility traces to solve one of the challenging problems in reinforcement learning. The concept of opposition traces is proposed in this work to deal with large state space problems in reinforcement learning applications. We combine the idea of opposition and eligibility traces to construct the opposition-based Q(lambda). The results are compared with the conventional Watkins' Q(lambda) and reflect a remarkable performance increase.
Experimental results are presented to indicate that the widely used power-law models for lifetime estimation are questionable for deep submicron (< 0.25 μm) MOS devices, particularly for the case of large substrat...
Experimental results are presented to indicate that the widely used power-law models for lifetime estimation are questionable for deep submicron (< 0.25 μm) MOS devices, particularly for the case of large substrate current stressing. This observation is attributed to the presence of current components, such as the gate tunneling current and base current of parasitic bipolar transistor, that do not induce device degradation. A more effective extrapolation method is proposed as an alternative for the reliability characterization of deep-submicron MOS devices.
This paper presents spectral characterization for efficient image detection using hyperspectral processing technique. We investigate relationship between the number of bands used in the detection so that the minimum n...
详细信息
This paper presents spectral characterization for efficient image detection using hyperspectral processing technique. We investigate relationship between the number of bands used in the detection so that the minimum number of bands are used in the processing. This approach reduces computation and implementation complexity of the algorithms. Specifically, we considered contribution coefficient by each band and based on this coefficient, we heuristically select the bands for processing. We have shown that for effective detection, only small number of bands are needed.
暂无评论