Browsing by Issue Date, starting with "1998-10-01"

Now showing 1 - 20 of 81854

An efficient transient modeling for 3-D multilevel interconnections in a stratified dielectric medium
(1998-10-01)
Wu, CY
;
Hou, HM
;
電子工程學系及電子研究所
;
Department of Electronics Engineering and Institute of Electronics
An efficient method is presented to model the transient characteristics of distributed resistor-capacitor of ULSI multilevel interconnections on complex topography, in which the reformulation of the boundary-element method (BEM) associated with the Padi-via-Lanczos (PVL) algorithm can avoid the redundant works on both volume mesh and transient analysis associated with the finite-difference method. An adaptive multilayer closed-form spatial Green's function for BEM is developed to examine the voltage and current responses of the multilevel conductor system by using the boundary-element method of integral formulation, in which arbitrary triangular elements on the surface of conductors are used to efficiently calculate the free-charge distributions of complex structure based on actual topography/processes. Applying the Galerkin principle over boundary elements, all of the surface integrals of charge distribution have been discretized and evaluated analytically for constant element. To improve the timing-analysis efficiency of the finite-difference method, the dominant poles are obtained by introducing the Pade-via-Lanczos (PVL) algorithm for model-order reduction. Hence, it is easy to calculate the transient characteristics of both parallel conductors and complicated configurations such as crossing lines, corners, contacts, multilayers and their combinations. Therefore, a simple and more general method is proposed for solving the combinations of complex structures based on actual topography/processes and arbitrary geometric configurations of multilevel interconnection lines in order to link with the present CAD tools. (C) 1998 Published by Elsevier Science Ltd. All rights reserved.
Fast integration of low orbiter's trajectory perturbed by the earth's non-sphericity
(1998-10-01)
Hwang, C
;
Lin, MJ
;
土木工程學系
;
Department of Civil Engineering
A fast algorithm is proposed to integrate the trajectory of a low obiter perturbed by the earth's nonsphericity. The algorithm uses a separation degree to define the low-degree and the high-degree acceleration components, the former computed rigorously, and the latter interpolated from gridded accelerations. An FFT method is used to grid the accelerations. An optimal grid type for the algorithm depends on the trajectory's permissible error, speed, and memory capacity. Using the non-spherical accelerations computed from EGM96 to harmonic degree 360, orbit integrations were performed for a low orbiter at an altitude of 170 km. For a separation degree of 50, the new algorithm, together with the predict-pseudo correct method, speeds up the integration by 145 times compared to the conventional algorithm while keeping the errors in position and velocity below 10(-4) m and 10(-7) mis for a 3-day arc.
Pattern formation and spatial chaos for cellular neural networks with asymmetric templates
(1998-10-01)
Shih, CW
;
應用數學系
;
Department of Applied Mathematics
The cellular neural networks with two kinds of two-parametered asymmetric templates are considered. The parameter space is partitioned into finitely many disjoint regions. In each region, the basic mosaic patterns are characterized. The feasible mosaic patterns corresponding to the parameters in each region can then be determined. To exhibit the spatial complexity of the system, we estimate the entropy of mosaic patterns for parameters in the regime of spatial chaos. In light of this characterization, the effect from the symmetry of the template on pattern formation properties can be seen in detail. We also discuss the existence of some fundamental class of feasible local defect patterns. It is shown that the feasible local k-defect patterns of vertical type cannot exist if the connection weights are anti-symmetric in the vertical direction. Same conclusion also holds for the ones of horizontal type and horizontal direction.
Magnetically charged black holes in induced gravity
(1998-10-01)
Dai, WB
;
Kao, WF
;
Wang, SY
;
物理研究所
;
Institute of Physics
It is shown that the regular SO3 non-abelian spherically symmetric monopole solution does not exist in an induced Einstein-Yang-Mills-Higgs model. It is also shown that Higgs scalar hair cannot exist in the presence of a black hole event horizon in this model. The corresponding global behavior of the gauge field is analyzed in a black hole background. In particular, it is shown that a non-trivial non-abelian monopole solution exists only if the radius of the event horizon is smaller than the characteristic radius of the classical monopole.
Reinforcement learning and robust control for robot compliance tasks
(1998-10-01)
Kuan, CP
;
Young, KY
;
電控工程研究所
;
Institute of Electrical and Control Engineering
The complexity in planning and control of robot compliance tasks mainly results from simultaneous control of both position and force and inevitable contact with environments. It is quite difficult to achieve accurate modeling of the interaction between the robot and the environment during contact. In addition, the interaction with the environment varies even for compliance tasks of the same kind. To deal with these phenomena, in this paper, we propose a reinforcement learning and robust control scheme for robot compliance tasks. A reinforcement learning mechanism is used to tackle variations among compliance tasks of the same kind. A robust compliance controller that guarantees system stability in the presence of modeling uncertainties and external disturbances is used to execute control commands sent from the reinforcement learning mechanism. Simulations based on deburring compliance tasks demonstrate the effectiveness of the proposed scheme.
Contact friction compensation for robots using genetic learning algorithms
(1998-10-01)
Liaw, DC
;
Huang, JT
;
電控工程研究所
;
Institute of Electrical and Control Engineering
In this paper, the issues of contact friction compensation for constrained robots are presented. The proposed design consists of two loops. The inner loop is for the inverse dynamics control which linearizes the system by canceling nonlinear dynamics, while the outer loop is for friction compensation. Although various models of friction have been proposed in many engineering applications, frictional force can be modeled by the Coulomb friction pins the viscous force. Based on such a model, an on-line genetic algorithm is proposed to learn the friction coefficients for friction model. The friction compensation control input is also implemented in terms of the friction coefficients to cancel the effect of unknown friction. By the guidance of the fitness function, the genetic learning algorithm searches for the best-fit value in a way like the natural surviving laws. Simulation results demonstrate that the proposed on-line genetic algorithm can achieve good friction compensation even under the conditions of measurement noise and system uncertainty. Moreover, the proposed control scheme is also found to be feasible for friction compensation of friction model with Stribeck effect and position-dependent friction model.
An incremental-learning-by-navigation approach to vision-based autonomous land vehicle guidance in indoor environments using vertical line information and multiweighted generalized hough transform technique
(1998-10-01)
Chen, GY
;
Tsai, WH
;
資訊工程學系
;
Department of Computer Science
An incremental-learning-by-navigation approach to vision-based autonomous land vehicle (ALV) guidance in indoor environments is proposed. The approach consists of three stages: initial learning, navigation, and model updating. In the initial learning stage, the ALV is driven manually, and environment images and other status data are recorded automatically, Then, an off-line procedure is performed to build an initial environment model. In the navigation stage, the ALV moves along the learned environment automatically, locates itself by model matching, and records necessary information for model updating. In the model updating stage, an off-line procedure is performed to refine the learned model. A more precise model is obtained after each navigation-and-update iteration. Used environment features are vertical straight lines in camera views. A multiweighted generalized Hough transform is proposed for model matching. A real ALV was used as the testbed, and successful navigation experiments show the feasibility of the proposed approach.
Purification, characterization and mechanistic study of beta-glucosidase from Flavobacterium meningosepticum (ATCC 13253)
(1998-10-01)
Li, YK
;
Chu, SH
;
Sung, YH
;
應用化學系
;
Department of Applied Chemistry
A beta-glucosidase (EC 3.2.1.21) from Flavobacterium meningosepticum has been purified and characterized. Purity was enhanced at least 530-fold from crude cell extract with 16.6% yield. The estimated molecular mass of the purified enzyme is 150 kDa by gel filtration and 78 kDa by SDS-PAGE. This dimeric enzyme has a pI=9.0 and an optimal activity at pH 5.0 and temperature of 50 degrees C. Divalent metal ions (Hg2+ CU2+, Ca2+, Mg2+) and EDTA have negligible effect on the enzyme activity. The enzyme exhibited a high specificity on the glycon portion of aryl-P-D-glycosides. NMR spectroscopy revealed the enzyme catalyzed hydrolysis of p-nitrophenyl-beta-D-glucopyranoside with the retention of anomeric configuration, indicating that a double displacement mechanism was involved. A preliminary study of the Bronsted relationship showed a concave-downward plot, which is consistent with the two-step mechanism.
Epitaxial growth of the GaN film by remote-plasma metalorganic chemical vapor deposition
(1998-10-01)
Lai, WC
;
Chang, CY
;
Yokoyama, M
;
Guo, JD
;
Tsang, JS
GaN films were successfully grown by the remote-plasma metalorganic chemical vapor deposition (RPMOCVD) system. The composition of the GaN films could be tuned from nitrogen-rich to stoichiometric growth by varying the mole flow rate of trimethylgallium (TMGa). A hypothesis concerning the collisions between excited nitrogen and TMGa was also brought up. The collision between excited nitrogen and TMGa influences the characteristics of surface morphology, composition, growth rate, and growth mechanism. The characteristics of the GaN film were optimized by changing the growth conditions. The narrowest FWHM of the double-crystal X-ray rocking curve is about 0.2 degrees. Under optimized conditions, the composition of the GaN film is almost the same as that of the reference GaN film grown by MOCVD.
A physical simulation model for field emission triode
(1998-10-01)
Lu, CW
;
Lee, CL
;
電子工程學系及電子研究所
;
Department of Electronics Engineering and Institute of Electronics
A simple but accurate physical model, which can be incorporated into circuit simulation programs such as SPICE for the field emission triode (FET), is developed. The model is based on the Fowler-Nordheim (F-N) current density-electric field (J-E) relationship. An electric field form is adopted to calculate the current density distribution along the surface of the sphere-shape tip. The cathode current is obtained by integration of the current density over the emission surface. The gate current is derived by the same integration, but over part of the emission area. A procedure to extract the values for the parameters of the model is also given. The model and the procedure has been applied to experimental devices to demonstrate its accuracy.
Method of solving triplets consisting of a singlet and air-spaced doublet with given primary aberrations
(1998-10-01)
Chen, CH
;
Shiue, SG
;
Lu, MH
;
光電工程學系
;
Department of Photonics
An effective algebraic algorithm is proposed as a computational tool for solving the thin-lens structure of a triplet which consists of a singlet and an air-spaced doublet. The triplet is required to yield specified amounts of lens power and four primary aberrations: spherical aberration, coma, longitudinal chromatic aberration and secondary spectrum. In addition, the air spacing is used to control the zonal spherical aberration and spherochromatism. The problem is solved in the following manner. First, the equations for power and chromatic aberration are combined into a quartic polynomial equation if the object is at a finite distance, or combined into a quadratic polynomial equation if the object is at infinity. The roots give the element powers. Second, the lens shapes are obtained by solving the quartic polynomial equation which is obtained by combining the equations of spherical aberration and coma. Since quartic and quadratic equations can be solved using simple algebraic methods, the algorithm is rapid and guarantees that all the lens forms can be found.
Effects of CoSi2 on p(+) polysilicon gates fabricated by BF2+ implantation into CoSi amorphous Si bilayers
(1998-10-01)
Cheng, HC
;
Lai, WK
;
Liu, HW
;
Juang, MH
;
電子工程學系及電子研究所
The integrity of thin gate oxide structures fabricated by implanting BF2+ ions into bilayered CoSi/amorphous silicon films and subsequent annealing has been studied as a function of cobalt silicide thickness and implantation energy. Significant degradation of gate oxide integrity and flatband voltage shifts were found with increasing cobalt silicide thickness and annealing temperature. It is shown that although thinner cobalt silicide can result in excellent gate dielectric integrity it also leads to worse thermal stability at a high annealing temperature. Moreover, shallower implantation depth and lower annealing temperature can reduce the boron penetration, but depletion effects in polycrystalline silicon gates are caused accordingly. Hence, appropriate process conditions, involving trade-offs among CoSi2 thickness, implantation energy and annealing temperature, must be used to optimize the device performance while retaining the thin dielectric reliability.
A 723-MHz 17.2-mW CMOS programmable counter
(1998-10-01)
Chang, HH
;
Wu, JC
;
電子工程學系及電子研究所
;
Department of Electronics Engineering and Institute of Electronics
A high-speed complementary metal-oxide-semiconductor (CMOS) programmable divide-by-N frequency divider was proposed, Using a new end-of-count (EOC) detecting and reloading algorithm, the reloading delay is distributed over three clock cycles, which increases the operating frequency. The simulated operating frequency of the new counter is 581 MHz, which is 2.2 times higher than that of a conventional programmable counter. The new programmable counter was implemented in a 0.8-mu m CMOS technology. The active die area is 480 x 100 mu m. The counter was measured to operate at 723 MHz with 5 V power supply and dissipates 17.12 mW.
Propagation mechanisms of UHF radiowave propagation into multistory buildings for microcellular environment
(1998-10-01)
Tarng, JH
;
Chang, YC
;
Chen, CM
;
電信工程研究所
;
Institute of Communications Engineering
Mechanism of UHF radiowave propagation into multistory office buildings are explored by using ray-tracing based models, which include a three-dimensional (3-D) ray-tracing model and a direct-transmitted ray (DTR) model. Prediction accuracy of the models is ascertained by many measured data and the measurements are carried out at many specific sites with different propagation scenarios. Their measured results also demonstrate some important propagation phenomena. II is found that (1) the direct transmitted wave may be the dominant mode; (2) the path loss neither increases nor decreases monotonically as a function of increasing floor level; and (3) there is not much difference of the average path loss among the receiving positions in the same room.
Seismic response analysis of bridges isolated with friction pendulum bearings
(1998-10-01)
Wang, YP
;
Chung, LL
;
Liao, WH
;
土木工程學系
;
Department of Civil Engineering
A systematic method is developed for the dynamic analysis of the structures with sliding isolation which is a highly non-linear dynamic problem. According to the proposed method, a unified motion equation can be adapted for both stick and slip modes of the system, Unlike the traditional methods by which the integration interval has to be chopped into infinitesimal pieces during the transition of sliding and non-sliding modes, the integration interval remains constant throughout the whole process of the dynamic analysis by the proposed method so that accuracy and efficiency in the analysis of the non-linear system can be enhanced to a large extent. Moreover, the proposed method is general enough to be adapted for the analysis of the structures with multiple sliding isolators undergoing independent motion conditions simultaneously. The superiority of the proposed method for the analysis of sliding supported structures is verified by a three-span continuous bridge subjected to harmonic motions and real earthquakes. In addition, the side effect of excessive displacement of the superstructure induced by the sliding isolation is eliminated by replacing one of the roller supports on the abutments with hinge support. Therefore, both reductions in the forces of the substructure and the displacements of the superstructure can be achieved simultaneously. (C) 1998 John Wiley & Sons, Ltd.
Contact friction compensation for robots using genetic learning algorithms
(1998-10-01)
Liaw, DC
;
Huang, JT
;
電控工程研究所
;
Institute of Electrical and Control Engineering
In this paper, the issues of contact friction compensation for constrained robots are presented. The proposed design consists of two loops. The inner loop is for the inverse dynamics control which linearizes the system by canceling nonlinear dynamics, while the outer loop is for friction compensation. Although various models of friction have been proposed in many engineering applications, frictional force can be modeled by the Coulomb friction pins the viscous force. Based on such a model, an on-line genetic algorithm is proposed to learn the friction coefficients for friction model. The friction compensation control input is also implemented in terms of the friction coefficients to cancel the effect of unknown friction. By the guidance of the fitness function, the genetic learning algorithm searches for the best-fit value in a way like the natural surviving laws. Simulation results demonstrate that the proposed on-line genetic algorithm can achieve good friction compensation even under the conditions of measurement noise and system uncertainty. Moreover, the proposed control scheme is also found to be feasible for friction compensation of friction model with Stribeck effect and position-dependent friction model.
Pattern formation and spatial chaos for cellular neural networks with asymmetric templates
(1998-10-01)
Shih, CW
;
應用數學系
;
Department of Applied Mathematics
The cellular neural networks with two kinds of two-parametered asymmetric templates are considered. The parameter space is partitioned into finitely many disjoint regions. In each region, the basic mosaic patterns are characterized. The feasible mosaic patterns corresponding to the parameters in each region can then be determined. To exhibit the spatial complexity of the system, we estimate the entropy of mosaic patterns for parameters in the regime of spatial chaos. In light of this characterization, the effect from the symmetry of the template on pattern formation properties can be seen in detail. We also discuss the existence of some fundamental class of feasible local defect patterns. It is shown that the feasible local k-defect patterns of vertical type cannot exist if the connection weights are anti-symmetric in the vertical direction. Same conclusion also holds for the ones of horizontal type and horizontal direction.
Full control of a PWM DC-AC converter for AC voltage regulation
(1998-10-01)
Tzou, YY
;
Jung, SL
;
電機工程學系
;
Department of Electrical and Computer Engineering
The design and implement action of a DSP-based fully digital-controlled single-phase pulsewidth modulated (PWM) dc-ac converter for ac voltage regulation is described. The proposed multiloop digital controller (MDC) consists of a current controller, a voltage controller, and a feedforward controller. This MDC was realized using a single-chip digital signal processor (DSP), The PWM gating signals are determined at Every sampling instant by the proposed multiloop digital control scheme using a set of detected feedback signals. A software current control scheme has been developed to achieve fast current control of the PWM inverter and decouple the inductor of the output filter Experimental results have been given to verify the proposed digital control scheme. The constructed DSP-based PWM dc-ac converter system can achieve fast dynamic response and with low total harmonic distortion (THD) for rectifier type of loads.
Multilinguistic handwritten character recognition by Bayesian decision-based neural networks
(1998-10-01)
Fu, HC
;
Xu, YY
;
資訊工程學系
;
Department of Computer Science
In this paper, we present a Bayesian decision-based neural network (BDNN) for multilinguistic handwritten character recognition, The proposed self-growing probabilistic decision-based neural network (SPDNN) adopts a hierarchical network structure with nonlinear basis functions and a competitive credit-assignment scheme, Our prototype system demonstrates a successful utilization of SPDNN to the handwriting of Chinese and alphanumeric character recognition on both public databases (CCL/HCCR1 for Chinese and CEDAR for the alphanumerics) and in-house database (NCTU/NNL), Regarding the performance, experiments on three different databases all demonstrated high recognition (86-94%) accuracy as well as low rejection/acceptance (6.7%) rates. As for the processing speed, the whole recognition process (including image preprocessing, feature extraction, and recognition) consumes approximately 0.27 s/character on a Pentium-100 based personal computer, without using a hardware accelerator or coprocessor.
Improvement of reliability of metal-oxide semiconductor field-effect transistors with N2O nitrided gate oxide and N2O polysilicon gate reoxidation
(1998-10-01)
Lai, CS
;
Chao, TS
;
Lei, TF
;
Lee, CL
;
Huang, TY
A systematic study of the impact of the reliability of metal-oxide semiconductor field-effect transistors (MOSFETs) with N2O nitrided gate oxides and N2O poly-reoxidation was conducted. The improvements include hot carrier immunity, suppression of reverse short channel effect and suppression of plasma-induced degradation. These improvements are due to nitrogen incorporation at the oxide/Si-substrate interface. All data presented in this study involve substantially low thermal budget (850 degrees C), making this N2O-related technology extremely attractive and promising for future scaled devices.