Browsing by Author "WANG, JD"

Now showing 1 - 5 of 5

A COMPONENT BEHAVIORAL-MODEL FOR AUTOMATIC FAULT-TREE CONSTRUCTION
(1993)
WANG, JD
;
LIU, TS
;
機械工程學系
;
Department of Mechanical Engineering
This study presents a new definition of events, an extended decision table, a virtual transfer component, and two pruning procedures for use in systematically constructing fault trees. The component states are modelled by events, which are defined by both qualitative event descriptions and quantitative event types. The quantitative event types are proposed to clarify the vagueness inherent in qualitative event descriptions. Moreover, decision tables, which are employed in the literature to model the behaviour of components, are extended in this study to generate succinct fault trees. A virtual transfer component is defined to deal with redundancy, such as k-out-of-n and standby modules. The construction of fault trees begins with the TOP event and traces backward to the primary and undeveloped events of components based on the structural configuration of the system and extended decision tables stored in a database. Procedures for producing transfer symbols and combining successive logic gates with gates of the same type are proposed to prune the generated fault tree during and after the tracing process, thereby reducing the construction time and the amount of computer memory needed. A hydraulic brake system and a pressure tank system are employed as examples to demonstrate the proposed method. The results of the examples show that the proposed method efficiently generates succinct fault trees for systems with redundancy.
A DISCRETE STRESS STRENGTH INTERFERENCE MODEL FOR UNRELIABILITY BOUNDS
(1994)
WANG, JD
;
LIU, TS
;
機械工程學系
;
Department of Mechanical Engineering
This study presents a multiple-line-segment method of implementing the stress-strength interference model when only discrete interval probabilities of stress and strength inside an interference region are available. Based on the discrete interval probabilities, the probability density functions are represented by line segments, which are in turn expressed by piecewise linear polynomials. Thus, unlike in existing empirical methods, when both stress and strength fall into the same subinterval, one can calculate the probability of failure by directly integrating the joint probability of stress and strength over the subinterval. Quadratic programming optimization is implemented to determine the upper and lower bounds on the failure probability, and the average value of the bounds is treated as the point estimate of the failure probability. Finally, numerical examples are carried out to demonstrate the proposed method. Three quantities are used to evaluate the accuracy and effectiveness of the proposed method-the relative error of the point estimate of the unreliability, the width of the bounds on the unreliability, and the minimum number of subintervals needed to generate the bounds that contain the exact unreliability. These three quantities are compared with results in the literature.
A RELIABILITY APPROACH TO EVALUATING ROBOT ACCURACY PERFORMANCE
(1994-01-01)
LIU, TS
;
WANG, JD
;
交大名義發表
;
機械工程學系
;
National Chiao Tung University
The capability of a robot to position its end-effector with accuracy is an important attribute for advanced robots in industrial automation technology. This study aims to assess positioning accuracy of robot end-effectors using reliability approaches. The reliability provides a statistical means for representing robot accuracy. Both the linear regression and iterative Taylor series methods are used to deal with limit states in reliability calculation. The proposed iterative Taylor series method improves computation accuracy compared with the prior literature. Both computer simulations and experiments were conducted to verify the performance of the accuracy assessment method. Examples show that the positioning accuracy is sensitive to tolerances of arm length and actuating accelerations. More important, dynamic errors, in comparison to geometric and kinematic errors, influence accuracy significantly.
AUTOMATIC FAULT-TREE CONSTRUCTION USING A NEW COMPONENT BEHAVIORAL-MODEL
(1994)
LIU, TS
;
WANG, JD
;
交大名義發表
;
National Chiao Tung University
TRAINING SIGNAL AND RECEIVER DESIGN FOR MULTI-PATH CHANNEL CHARACTERIZATION FOR TV BROADCASTING
(1990-11-01)
WANG, JD
;
CHAO, THS
;
SALTZBERG, BR
;
電信工程研究所
;
Institute of Communications Engineering
The problem of multi-path propagation in television terrestrial broadcasting can be solved by channel equalization. The use of such equalization is needed for an enhanced NTSC system and is particularly critical in most HDTV proposals. We propose techniques for precisely characterizing a multi-path channel. These characterizations can be used to reduce hardware complexity and to speed up equalizer convergence.