Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Civil-Comp Proceedings
ISSN 1759-3433
CCP: 82
PROCEEDINGS OF THE EIGHTH INTERNATIONAL CONFERENCE ON THE APPLICATION OF ARTIFICIAL INTELLIGENCE TO CIVIL, STRUCTURAL AND ENVIRONMENTAL ENGINEERING
Edited by: B.H.V. Topping
Paper 3

A Prototype Hybrid Rule-Object System for Structural Design

F. Oudjehane+ and H. Mili*

+Department of Software Engineering, École de Technologie Supérieure, Montréal, Québec, Canada
*Latece Laboratory Centre, Université du Québec à Montréal, Québec, Canada

Full Bibliographic Reference for this paper
F. Oudjehane, H. Mili, "A Prototype Hybrid Rule-Object System for Structural Design", in B.H.V. Topping, (Editor), "Proceedings of the Eighth International Conference on the Application of Artificial Intelligence to Civil, Structural and Environmental Engineering", Civil-Comp Press, Stirlingshire, UK, Paper 3, 2005. doi:10.4203/ccp.82.3
Keywords: structural steel design, open-web-steel joints, object-oriented software development, Java, rule-based development, rule engines.

Summary
The design of steel structures is a fairly complex process with many inter-dependant parameters. The design space is not a continuous one as steel structures are built from standardized components with codified static characteristics and dynamic behaviour. For that reason, there is no single solution to any design problem, and designers must integrate non-physical criteria in their design choices. Structural designers typically rely on component catalogues and procedural manuals which show the properties of the individual components, their usage patterns, and the set of validations that need to be performed once a set of components and the way to connect them has been chosen. The process can be overwhelming for the novice designer. In this paper, we show how a structural steel design software can be developed using a hybrid object-rules approach using ILOG JRulesTM and Java. The object-oriented paradigm, pioneered with the Simula language, is appropriate for representing and simulating physical objects. The rules-based approach is appropriate for codifying and enacting domain knowledge that is modular and incremental. Our system, built with ILOG Rules [1] for Java (JRulesTM), is intended to assist the designer in designing open-web steel joist component for conformance with the Canadian steel design standard and manufacturer catalogue.

In civil engineering, conventional structural software uses algorithmic treatment and numerical analysis and cannot take into account the designer's previous experience, their judgement and the decisions about the structural design or system selection. For each design verification the designer must provide the geometric parameters of the structural elements, their structural resistance, the stiffness, and the constructive dispositions. This approach is simple but not very efficient, because the parameters must be redefined for every verification.

In the past decades, the software field has witnessed a number of advances in languages, tools, and methods for developing effective software applications efficiently. However, maintenance remains a challenge. Maintenance is inevitable as requirements change: both the deployment environment and the underlying business rules will change during the lifetime of a software application, and the application needs to change accordingly. Separating the business rules from application code and implementing them, as a rule engine component, can make applications adaptable and maintainable. The rules can be changed independently and redeployed to the running application, without requiring to recompile the application or even to shut it down. The business rules approach, as it is called, is being used successfully in a variety of application domains (banking, manufacturing, customer relationship management, border control, etc.). A number of rule products are being offered. The ILOG JRules product integrates into Java applications, and JRules components can be easily embedded in Web, legacy and traditional applications

Prototype knowledge-based rules for structural steel design systems reported are based on blackboard architecture and implemented with Lisp or Prolog and FORTRAN. The structural design entities are modelled according to functional decomposition, as opposed to object-oriented systems. The approaches that focus on knowledge representation are typically production rules and frames systems.

Our system is a small-scale proof-of-concept using ILOG business rule technology. ILOG JRules allows combining rule-based and object-oriented programming. The methodology involves the building of the object model by identifying relevant classes that represent the different entities involved in the structures, the rules classification contained in the CSA standard S16-01 [2] and manufacturer catalogue [3] and the implementation in which the rules are implemented using JRules and the methods are implemented in Java. The structure design process has been defined using a JRules rule flow. A simplified ruleflow composed of a set of rules to be executed is shown.

References
1
ILOG, "JRules 4.5", ILOG Documentation, 2003.
2
Canadian Institute of Steel Construction, "Handbook of steel construction", Willowdale, Ontario, 2000.
3
Canam Manac Group, "Joist Catalogue", Canada,2003

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £80 +P&P)