ISSN:0160-5682    

DOI:10.1057/jors.1990.71

出版商:Taylor&Francis    

年代:1990

数据来源: Taylor

摘要:

AbstractOne of the most important difficulties when developing knowledge based systems in manufacturing scheduling or control, is finding the required knowledge. We address here the problem of acquiring knowledge about the behavior of manufacturing systems. Learning algorithms are proposed to generate, from simulation experiments, a set of production rules. This set may be considered as a simulation meta-model, and may be used either directly by the shop manager, or inserted into a knowledge base. This approach is illustrated by the use of the learning program GENREG. It generates rules related to the behavior of a simplified flow shop when different dispatching rules are applied.

ISSN:0160-5682    

DOI:10.1057/jors.1990.72

出版商:Taylor&Francis    

年代:1990

数据来源: Taylor

摘要:

AbstractResearch developments leading to implementation of an intelligent software environment supporting system design and simulation are presented. Knowledge-based system design and multifaceted simulation methodologies are a foundation for the system realization. The paper describes the major theoretical concepts and processes employed to develop and simulate design models. The environment implementing these concepts and methods consists of two basic components: one serves as a front end supporting the model construction processes; the other is an object-oriented, discrete-event simulator supporting evaluation of hierarchical, multi-component models. Current state of the system implementation and future work are discussed.

ISSN:0160-5682    

DOI:10.1057/jors.1990.73

出版商:Taylor&Francis    

年代:1990

数据来源: Taylor

摘要:

AbstractIntelligent tutoring systems (ITS) have recently found a wide range of potential applications across a multitude of disciplines and subject areas. This paper investigates the suitability and feasibility of developing ITS in the general context of OR as well as in a specific OR activity, simulation. It also examines how a current teaching aid for simulation model debugging can be transformed into an ITS and proposes a specification for the transformation process. The paper concludes that ITS could be useful in OR either for use as prime training systems or integrated in existing OR teaching packages. However, the field of ITS is in its infancy and development efforts are likely to be labour-intensive.

ISSN:0160-5682    

DOI:10.1057/jors.1990.74

出版商:Taylor&Francis    

年代:1990

数据来源: Taylor

摘要:

AbstractVehicle scheduling is one of the most commonly occurring problems of transport management. Traditionally, human schedulers tackle these problems, but over the past 30 years a considerable effort has been put into developing computer systems to replace or assist them. Despite this effort, very few organizations routinely use computerized vehicle scheduling. There are several reasons for this, including the difficulty in dealing with real complexity and uncertainty in algorithmic solutions. In recent years, interactive systems have tried to overcome these difficulties by using humans to play a part in the scheduling. A next step would be to include the schedulers' skills in an expert system. Unfortunately, there is considerable diversity in scheduling problems and little agreement about the facts or rules needed for a knowledge base. At present, some characteristics of a general expert system for vehicle scheduling can be suggested, but several practical difficulties must still be overcome.

ISSN:0160-5682    

DOI:10.1057/jors.1990.75

出版商:Taylor&Francis    

年代:1990

数据来源: Taylor

摘要:

AbstractThis paper shows how tools and techniques of artificial intelligence can be successfully integrated into a computer system working in the vehicle routing domain. The aim of this system, called ALTO, is to facilitate the development of routing algorithms for transportation vehicles. In this paper, we describe the general algorithmic framework and the rich interface provided by the system to the expert algorithm designer. We also introduce a methodology for acquiring useful knowledge in the domain, based on examples of successful and unsuccessful problem-solving strategies. With such knowledge, ALTO would then be capable of actively supporting the algorithm designer by suggesting good candidate algorithms for solving new problems.

ISSN:0160-5682    

DOI:10.1057/jors.1990.76

出版商:Taylor&Francis    

年代:1990

数据来源: Taylor

摘要:

AbstractThe assignment of empty railcars by the SNCF (French Railways) distributors according to supply and demand from the regions could be considered as a conventional resource allocation problem. As such, it would be suitable for processing by integer programming. However, as the problem is currently solved by the reasoning process of the distributors, it involves much more strategic knowledge than linear programming. Based on this observation, we developed an allocation system centred on an expert system designed to reproduce the reasoning processes of the distributors. This paper describes the system, and the reasons why the ES approach seems more adapted to the problem than the OR pure solution.

ISSN:0160-5682    

DOI:10.1057/jors.1990.77

出版商:Taylor&Francis    

年代:1990

数据来源: Taylor

摘要:

AbstractPractical solutions to the production scheduling problem must provide two broad capabilities:an ability to efficiently generate schedules that reflect the actual constraints and objectives of the manufacturing environment, andan ability to incrementally revise these schedules over time in response to unexpected executional circumstances. In this paper, we advocate a common view of predictive and reactive scheduling as an incremental problem solving process that is opportunistically focused by characteristics of the current solution constraints.We describe the architecture of OPIS (opportunistic intelligent scheduler), which defines a general framework for configuring scheduling systems according to this view. We then examine the scheduling knowledge (e.g. analysis and scheduling methods, schedule generation or revision strategies) that is exploited within this architecture by the current OPIS scheduler. Experimental studies with the OPIS scheduler have demonstrated the potential of this constraint-directed scheduling methodology in both predictive and reactive scheduling contexts.

ISSN:0160-5682    

DOI:10.1057/jors.1990.78

出版商:Taylor&Francis    

年代:1990

数据来源: Taylor