Knowledge Engineering Methods Research Articles

Knowledge-based methods and smart algorithms in computational mechanics

Effective methods leading to automated, computer-based solution of complex engineering design problems are studied in this paper. In particular, methods of automation of the finite element analyses are of primary interest here. These include algorithmic approaches, based on error estimation, adaptivity and smart algorithms, as well as heuristic approaches based on methods of knowledge engineering. A computational environment, which interactively couples h- p adaptive finite element methods with object-oriented programming and expert system tools, is presented. Several examples illustrate the merit and potential of the approaches studied here.

Towards automatic adaptive geometrically non‐linear shell analysis. Part I: Implementation of an h‐adaptive mesh refinement procedure

AbstractEffective methods leading to automated adaptive numerical solutions to geometrically non‐linear shell‐type problems are studied in this work. In particular, procedures for improving the accuracy, the reliability and the computational efficiency of the finite element solutions are of primary interest here. This is addressed using h‐adaptive mesh refinement based on a posteriori error estimation, self‐adaptive methods in global incremental/iterative processes, as well as smart algorithms and heuristic approaches based on methods of knowledge engineering. Seemless integration of h‐adaptive finite element methods with adaptive step‐length control makes it possible to maintain a prescribed accuracy while maintaining the solution efficiency without user intervention throughout the process of a non‐linear analysis. Several examples illustrate the merit and potential of the approach studied herein and confirm the feasibility of developing an automatic adaptive environment for geometrically non‐linear analysis of shell structures.

A revitalization of chemical engineering knowledge with fuzzy calculus

Realistic conventional chemical engineering data (usually in the form of tables, simple correlations, graphs) are partially inconsistent and sparse. The revitalization involves application of methods of knowledge engineering and integration of specific and rather isolated information items based on systematic identification of all inconsistencies, and formalizing and consequently incorporating such information which is too vague for conventional formal tools. Any revitalization is a highly subjective and ad hoc process. The objectivity of the revitalization can be increased by revitalization algorithms. One of these, discriminative power evaluation, is described. The discriminative power quantifies the ability of the information to return distinguishably different answers to similar queries. A case study (a 13-dimensional decision-making problem, solid—liquid separation) is presented in detail.

Applied terminology:

Two recently published collections of articles that provide an overview of the state of the art in applied terminology are discussed from the point of view of two themes recurring throughout them: comparative terminology and knowledge engineering. Comparative terminology is one of the most traditional forms of terminology and is still among the most prevalent practiced today. New necessary approaches to handling problems in this area are outlined. One of the newest areas of research is knowledge-based terminology. The growing interest in exploring the potential benefits of a relationship between terminology and knowledge-engineering methods is shown by the discussion of various articles on this subject.

An Expert System for Evaluating Exams

AbstractIn order to simplify the formal evaluation procedures for matriculation examination forms proposed by high school teachers in Israel, an expert system was constructed. The system serves both the examination authors and the evaluators. Consequently, both sides are familiar with the tradeoffs involved in using the expert system without having to resort to lengthy explanations or complex guidebooks. Because many evaluators were involved in the process, conventional knowledge-engineering methods were found to be inadequate. Instead, a learning-by-example paradigm called the Ordinal Learning Model was utilized. Responses from the system’s users during a two-year beta test has been very encouraging.

Constructing knowledge-based systems

Vital, a four-and-a-half-year ESPRIT II research and development project that involves nine organizations in five countries is discussed. It addresses the problems of effective process modeling for knowledge-based systems, providing guidelines on when to use various knowledge-engineering methods and techniques, and reducing the bottleneck in acquiring expert knowledge by providing both methodological and software support for developing large, industrial, knowledge-based system applications. The project goals, approach, and workbench are outlined, and a case study is described. >

Towards an automated environment in computational mechanics

Effective methods leading to an automated, computer-based solution of complex engineering design problems are studied in this paper. In particular, methods of automation of the finite element analyses are of primary interest here. This includes algorithmic approaches, based on error estimation. adaptivity and smart algorithms, as well as heuristic approaches based on methods of knowledge engineering. A computational environment, which interactively couples h-p adaptive finite element methods with object oriented programming and expert system tools, is presented. Several examples illustrate the merit and potential of the approaches studied here and confirm the feasibility of developing fully automated design environments.

Fuzzy Control of an Anaerobic Digester for the Treatment of the Organic Fraction of Municipal Solid Waste (MSW)

This paper deals with an application of the fuzzy approximate reasoning to the management and control of a pilot scale anaerobic digester. The organic fraction of MSW was used as substrate of the digester operating under semi-dry, thermophilic conditions. The fuzzy control actions suggested by the controller are taken on the basis of an emulation of expert reasoning instead of a mathematical representation of the digester behaviour. This approach is realized by capturing the expert's knowledge of the process through knowledge engineering methods. The controller performance was checked against the expert's behaviour for a set of experimental data. The experimentation has thus far proven that a controller for anaerobic reactors based on fuzzy logic is feasible and that the fuzzy approach provides a relatively simple description of human behaviour from different points of view.

Computer aided reliability modeling and applications in semiconductor manufacturing

We present a computerized reliability model based on evolving Markov processes. The model has the capability to self-organize and adopt. The differences between knowledge engineering methods and the proposed model are also discussed.

Intelligent Fault Diagnosis of the ICATS

Abstract In this paper, we introduce the structure and function of the inertial component automatic test system first .And then we analyse various methods which are used to dispose the system failure by the experts of the test system in detail . A fault diagnosis system of the inertial component automatic test system is founded by using the theories .techniques and methods of knowledge engineering. The diagnosis system can improve the diagnosis rate by using other kinds of knowledge on the baiss of the hierarchy diagnosis model. The establishment of the protomodel system verifies the feasibility and correctness of the system .

Extraction of Ocean Conditions by Applying Knowledge Engineering Method

Extraction of Ocean Conditions by Applying Knowledge Engineering Method

Challenges and promises of integrating knowledge engineering and qualitative methods

Our goal is to expose some of the close ties that exist between knowledge engineering (KE) and qualitative methodology (QM). Many key concepts of qualitative research, for example meaning, commonsense, understanding, and everyday life, overlap with central research concerns in artificial intelligence. These shared interests constitute a largely unexplored avenue for interdisciplinary cooperation. We compare and take some steps toward integrating two historically diverse methodologies by exploring the commonalities of KE and QM both from a substantive and a methodological/technical perspective. In the second part of this essay, we address knowledge acquisition problems and procedures. Knowledge acquisition within KE has been based primarily on cognitive psychology/science foundations, whereas knowledge acquisition within QM has a broader foundation in phenomenology, symbolic interactionism, and ethnomethodology. Our discussion and examples are interdisciplinary in nature. We do not suggest that there is a clash between the KE and QM frameworks, but rather that the lack of communication potentially may limit each framework's future development.

A framework for knowledge-based control

A framework for knowledge-based control is proposed. The approach presented is suitable for control systems and control support of systems which have no adequate mathematical models. Thus, the control is performed by using knowledge engineering methods rather than pure mathematical control methods. The domain expert's knowledge is assumed to be encoded in the form of simple statements (facts) and special reasoning rules, which form the core of the Knowledge-Based Control System (KBCS). The control system reads the input information, and on the basis of the current state of its knowledge base, together with the application of supplied inference rules updates the knowledge base and performs the required control actions. Moreover, some inference control knowledge, reflecting the expert's way of reasoning, is to be incorporated in the KBCS. The main idea of the system consists of selecting an appropriate set of actions to be executed, with regard to the current state specification and the control goal given. An abstract mathematical model of the control process is formulated and a suitable language for knowledge representation is proposed. The reasoning scheme is discussed and the structure of the control system is outlined. A representative application example is provided.

An expert system prototype for event diagnosis and real-time operation planning in power system control

A prototype system called EKA, which is an attempt to apply sophisticated knowledge-engineering methods to power systems, is described. EKA is based on an object-oriented model and model-based reasoning, including multiple ways to represent knowledge: rules, facts, methods, daemons, procedures, logic programs, etc. The EKA system consists of three separate but structurally similar systems: an operation planning system, an event analysis system, and a training simulator. As a development environment, hybrid software and a Lisp computer have been used. The tasks and methods involved are described with and without an expert system. The application environment and prototype implementation are outlined. >

Strategic Issues in Knowledge Engineering

Expert systems should find widespread application in human factors, provided the problem of building expert systems (knowledge engineering) is dealt with satisfactorily. The task of knowledge engineering is outlined and five key strategic issues are identified. These strategic issues are discussed from a human factors perspective, and it is shown how current knowledge engineering methods can be extended to include the concerns of human factors. The potential contribution of human factors to knowledge engineering methodology is illustrated with two examples. The first shows how ranking techniques can be adapted to meet the requirements of knowledge elicitation; the second shows how protocol analysis can be used to validate expertise.

Computer environments for design and designers

The mode of use of expert systems in design applications is significantly different from other application areas. First experience in the design domain suggests that expert systems will be employed on a design problem in a variety of functions. For example, they will be employed as consultants in generating architectural programs. They will also find an important critiquing role in the appraisal of proposals by the designer. There is evidence also that design automation, especially in design detailing, will be an important application.This paper reports on the work undertaken as part of the research programme ‘Designing and Design Knowledge in Architecture and Engineering’ (NSF Grant 98348). It investigated the application of knowledge engineering methods in a general computational environment for designers. A representative selection of four knowledge-based system tools or ‘shells’ were used in the study. The primary interest was to determine the performance of these software packages as the core element in a new genre of integrated KBDS (Knowledge-Based Design Systems). The final sections of the paper described the further work that remains to be done, including a methodology for acquisition of design knowledge.

Integration of Hardware and Software in New Generation Workstations

Implementation of knowledge engineering methods on workstations and some system design problems are discussed. It is shown that for efficient problem solving in a man-machine system both, conceptual and expert knowledge are needed. The efficiency of dedicated hardware and its dependence on system size is investigated. A general de scription of an object-oriented workstation and its programming environment are given.

Knowledge engineering in quantitative histopathology

Ongoing research on a system for the automated evaluation of digitized images of histopathologic sections is described. The system comprises an ultrafast laser scanner microscope capable of recording image data at 64 Mhz in each of two wavelength channels, and a multiprocessor computer containing thirty-six Motorola 68000 processing elements. The operating system allows an image-data-driven dynamic reconfiguration of the computer architecture; this reconfiguration is based on prior knowledge of the histopathologic sections to be processed. Knowledge engineering methods are used extensively as development tools in the development of an expert system-controlled scene segmentation, processing task scheduling, in the diagnostic expert system module, and its validation procedures.

An Expert System Approach to Analysis and Control in Certain Complex Systems

An expert system approach to analysis and control in certain complex systems is proposed. It is assumed that a system under consideration can be described by means of knowledge engineering methods, i.e. the specific domain knowledge can be encoded in the form of inference rules. The particular knowledge about certain states of the system is represented as sets of facts true in a certain state, so-called state-sets. A language for knowledge representation, based on First Order Predicate Calculus, which offers an easy way for representation of constraints imposed on certain variables is proposed and a uniform inference rule scheme for both forward and backward- chained reasoning is given. Moreover, an easy and natural way of representing and dealing with the uncertainty of domain knowledge is proposed, both for probabilistic and fuzzy approaches. Unlike some other expert systems, in the presented approach it is possible to express some general knowledge with the use of variables, constraints and even functional dependences.

The Roles of Expert Systems and Biomechanical Models in Eye Muscle Surgery

Misalignment of the eyes (strabismus) is commonly treated by surgical manipulation of the extraocular muscles. The high failure rate of such surgery has led to the application of engineering methods in diagnosis and surgical planning. Computer models of orbital statics can now aid in diagnosis, and can help predict the outcome of surgery. Methods have been developed to collect biophysical data to customize the models for particular patients. Such biomechanical models will become increasingly important in clinical practice. Nevertheless, the heuristics employed by expert strabismus surgeons will not soon be reduced to algorithms Knowledge engineering methods are needed to explicate and disseminate the reasoning of experts in their own terms. Such an expert system could make use of a biomechanical model to explore hypotheses, much as human experts do now. Such a hybrid system might effectively advise surgeons on how to treat individual cases.