Showing papers on "Uncertain data published in 1989"

A neural network methodology for process fault diagnosis

Abstract: The ability of knowledge-based expert systems to facilitate the automation of difficult problems in process engineering that require symbolic reasoning and an efficient manipulation of diverse knowledge has generated considerable interest recently. Rapid deployment of these systems, however, has been difficult because of the tedious nature of knowledge acquisition, the inability of the system to learn or dynamically improve its performance, and the unpredictability of the system outside its domain of expertise. This paper proposes a neural-network-based methodology for providing a potential solution to the preceding problems in the area of process fault diagnosis. The potential of this approach is demonstrated with the aid of an oil refinery case study of the fluidized catalytic cracking process. The neural-network-based system successfully diagnoses the faults it is trained upon. It is able to generalize its knowledge to successfully diagnose novel fault combinations it is not explicitly trained upon. Furthermore, the network can also handle incomplete and uncertain data. In addition, this approach is compared with the knowledge-based approach.

A concept for an intelligent and fault-tolerant

Abstract: A concept for the intelligent control of subsystems of a flexible assembly cell is presented. Unknown or uncertain data about the real world may lead towards failure during an assembly task. Therefore, a fault tolerant system must be capable of reacting immediately to error situations. Thus, the major topic of this paper is the dynamic handling of unforeseen situations during realtime activities. This will be achieved by combining sensor guided actions with an advanced autonomous supervision system. Experimental results will be derived from the mobile two-arm robot system KAMRO of the Institute for Real-Time Computer Control Systems and Robotics, University of Karlsruhe, Federal Republic of Germany.

Reasoning from uncertain data: a bit enhancement

Abstract: It is proposed that artificial intelligence (AI) principles, coupled with powerful Bayesian statistical inference techniques, can be successfully applied to built-in-test (BIT) technology and can significantly contribute to the improvement of avionics BIT diagnostic capabilities. The goal is to extract more information from available data provided by the BIT, rather than to expand its testing capabilities. The proposed approach is illustrated by a TACAN (tactical air navigation) example. >

Decision-theoretical formulation of the calibration problem

Abstract: The choice of calibration policy is of basic importance in analytical chemistry. A prototype of the practical calibration problem is formulated as a mathematical task and a Bayesian solution of the resulting decision problem is presented. The optimum feedback calibration policy can then be found by dynamic programming. The underlying parameter estimation and filtering are solved by updating relevant conditional distributions. In this way: the necessary information is specified (for instance, the need for knowledge of the probability distribution of unknown samples is clearly recognized as the conceptually unavoidable informational source); the relationship of the information gained from a calibration experiment to the ultimate goal of calibration, i.e., to the estimation of unknown samples, is explained; an ideal solution is given which can serve for comparing various ways of calibration; and a consistent and conceptually simple guideline is given for using decision theory when solving problems of analytical chemistry containing uncertain data. The abstract formulation is systematically illustrated by an example taken from gas chromatography.

Optimal Design of Solar Energy Systems With Uncertain Data

Abstract: The technical and economic characteristics of optimized solar energy systems can be expressed in a concise mathematical expression and this expression manipulated to show the dependence of total system cost on variables such as collector cost and efficiency, weather data, or the cost of the backup fuel. The lack of certain knowledge of future weather, fuel prices, and system performance implies that a system optimized, with respect to an assumed set of these data, may not be optimal with respect to the realities of the operation. This formulation demonstrates clearly how the life cycle costs of the system increase as a result of errors in estimating various system parameters; in most cases, the costs of errors are not large over their expected ranges. In particular, most solar energy systems designed to satisfy a given load (that is, having a given nameplate capacity) should be optimized for the best weather they may encounter, and this design used wherever the systems are economically viable, as it is likely that the savings due to standardization will outweigh the relatively small costs of being imperfectly optimized.

Failure and Safety Analysis of Widespread Overhead Distribution Systems

Abstract: Decision making in respect of hazard and safety analysis in the face of uncertain data in the management of widespread distribution systems is focused on. The narrow interpretation by Courts of the “foreseeable event” has become the benchmark of liability where power system failures have resulted in loss of life, injury or destruction of property; occurrences, the prediction of which relates more to the probability of extreme event combinations than to the assessment of availability by classically‐oriented reliability statistics. The work outlines a computer‐based inferential procedure which combines adverse weather to a probabilistic assessment of failure resulting from a combination of contributing events. The technique utilises Monte‐Carlo simulation in assigning the adverse weather to common mode data analysis in the derivation of failure state conditional probabilities. Limited sensitivity studies have been performed using Northern Rivers Electricity, emphasis being placed on evaluating preventative maintenance as a function of optimal time intervals, priorities and costs.

An interpretation and solution of ill-conditioned linear equations

Abstract: Data insufficiency, poorly conditioned matrices and singularities in equations occur regularly in complex optimization, correlation, and interdisciplinary model studies. This work concerns itself with two methods of obtaining certain physically realistic solutions to ill-conditioned or singular algebraic systems of linear equations arising from such studies. Two efficient computational solution procedures that generally lead to locally unique solutions are presented when there is insufficient data to completely define the model, or a least-squares error formulation of this system results in an ill-conditioned system of equations. If it is assumed that a reasonable estimate of the uncertain data is available in both cases cited above, then we shall show how to obtain realistic solutions efficiently, in spite of the insufficiency of independent data. The proposed methods of solution are more efficient than singular-value decomposition for dealing with such systems, since they do not require solutions for all the non-zero eigenvalues of the coefficient matrix.