Rule-based Production Research Articles

A semantics for a class of non-deterministic and causal production system programs

We define a class of function-free rule-based production system (PS) programs that exhibit non-deterministic and/or causal behavior. We develop a fixpoint semantics and an equivalent declarative semantics for these programs. The criterion to recognize the class of non-deterministic causal (NDC) PS programs is based upon extending and relaxing the concept of stratification, to partition the rules of the program. Unlike strict stratification, this relaxed stratification criterion allows a more flexible partitioning of the rules and admits programs whose execution is non-deterministic or causal or both. The fixpoint semantics is based upon a monotonic fixpoint operator which guarantees that the execution of the program will terminate. Each fixpoint corresponds to a minimal database of answers for the NDC PS program. Since the execution of the program is non-deterministic, several fixpoints may be obtained. To obtain a declarative meaning for the PS program, we associate a normal logic program\(\overline {PS} \) with each NDC PS program. We use the generalized disjunctive well-founded semantics to provide a meaning to the normal logic program\(\overline {PS} \) Through these semantics, a well-founded state is associated with\(\overline {PS} \) and a set of possible extensions, each of which are minimal models for the well-founded state, are obtained. We show that the fixpoint semantics for the NDC PS programs is sound and complete with respect to the declarative semantics for the corresponding normal logic program\(\overline {PS} \).

Optimal compression of propositional Horn knowledge bases: complexity and approximation

Horn formulae play a prominent role in artificial intelligence and logic programming. In this paper we investigate the problem of optimal compression of propositional Horn production rule knowledge bases. The standard approach to this problem, consisting in the removal of redundant rules from a knowledge base, leads to an “irredundant” but not necessarily optimal knowledge base. We prove here that the number of rules in any irredundant Horn knowledge base involving n propositional variables is at most n − 1 times the minimum possible number of rules. Therefore, the quadratic time transformation of an arbitrary Horn production rule base to an equivalent irredundant and prime one (presented in [9]) provides a reasonable approximation algorithm. In order to formalize the optimal compression problem, we define a Boolean function of a knowledge base as being the function whose set of true points is the set of models of the knowledge base. In this way the optimal compression of production rule knowledge bases becomes a problem of Boolean function minimization. In this paper we prove that the minimization of Horn functions (i.e. Boolean functions associated to Horn knowledge bases) is NP-complete.

Development of a self-tuning fuzzy logic controller

In developing a Fuzzy Logic Controller, the establishment of boundaries between adjacent fuzzy sets of the function that characterizes these boundaries is a formidable task. Considerable re-examination of the various process state combinations is necessary to ensure that the output distribution changes in a reliable and consistent manner. Rule-based production systems that depend on fuzzy logic can produce vacuums and valleys in the output distributions. These deficiencies must be eliminated. An Expert System has been developed to adjust the boundary conditions of a two-variable fuzzy logic control system using a feedback learning procedure. Membership values are recalculated on each cycle to ensure smooth transfer of output belief from one rule to another as the process states change. This method is performed whenever the confidence level of the definition of a specific linguistic expression is adjusted. The application of this controller to supervice a secondary crushing plant simulator is described.

GTSIM A computer simulation of music perception

GTSIM is a computer simulation model of human perception of rhythm, based on Lerdahl and Jackendoff'sA Generative Theory of Tonal Music. GTSIM, a rule-based model with a neural network component, simulates the metric and grouping stages of Lerdahl and Jackendoff's theory. Using a rule-based production system, GTSIM simulates parallel processing, operating left-to-right without backtracking. GTSIM identifies the key of the piece of music and produces hierarchical representations of the metric and grouping structure which conform to Lerdahl and Jackendoff.

Modeling Pilot Expertise in Air Combat

The purpose of this effort was to model expert pilot performance and decision making in one-versus-one (1v1) air-to-air combat. Several knowledge-elicitation techniques were used to extract air combat expertise from a former fighter pilot, who served as the subject-matter-expert (SME). Unstructured and then structured interviews were used to elicit the goals and sub-goals of air-to-air combat, plus some of the pilot behaviors necessary to accomplish the goals. The SME also flew a number of combat sorties against another former fighter pilot in the Simulator for Air-to-Air Combat (SAAC) to demonstrate pilot performance required to accomplish the goals of air combat. Based on the SME's verbal protocols, a group of air combat rules were developed. A rule-based production system was then designed to incorporate the resulting knowledge base. The production system was also designed to be capable of analyzing an existing data base of air combat engagements. Expert system development required additional input from the SME to identify specific values of flight parameters required by the production system. Upon completion and SME verification of the expert model, it will be validated by comparing its performance to that of our SME in simulated air-to-air combat. If the model can successfully describe expert pilot performance, the model will be used to provide diagnostic performance feedback in conjunction with SAAC training.

A computerized data management and decision support system for gypsy moth management in suburban parks

The GYpsy Moth Management Decision support SYStem (GYMSYS) was developed to address specific data management and decision support needs for gypsy moth management in urban parks and other wooded public lands. These needs include a map-based user interface, simple and reliable data entry, a means to customize the system to a variety of separate and heterogeneous management units and programs, and the use of readily available and inexpensive hardware and software systems. These needs were addressed by incorporating a map image, selected by the user, into the background of the user interface, making it possible for the user to configure the system separately for each individual management unit, and utilizing Apple Macintosh computers and Hypercard software. A knowledge base for making treatment decisions based on egg mass density and size, host tree susceptibility, and defoliation history was encoded into a production rule base.

Rule-based production planning in flexible manufacturing systems

Production planning in flexible manufacturing may require the solution of a large-scale discrete-event dynamic stochastic optimization problem, due to the complexity of the system to be optimized, and to the occurrence of discrete events (new orders and hard failures). The production planning problem is here approached for a multistage multipart-type manufacturing shop, where each work cell can share its processing time among the different types of parts. The solution of this problem is obtained by an open-loop-feedback control strategy, updated each time a new event occurs. At each event time, two coupled problems are solved: 1) a product-order scheduling problem, conditioned on estimated values of the production capacities of all component work cells; and 2) a production-capacity planning problem, conditioned on predefined sequences of the product orders to be processed. In particular, the article aims at defining a production planning procedure that integrates both analytical tools, derived from mathematical programming, and knowledge-based rules, coming from experience. The objective is to formulate a hybrid (knowledge-based/analytical) planning architecture, and to analyze its use for multicell multipart-type manufacturing systems.

Development of an Expert System for the Design Synthesis and Reliability Assessment

An expert system is developed for design synthesis and reliability analysis. It uses a rule-based production system called KOPS, and reliability analysis is composed in computer codes CAT, FTAP, and IMPORTANCE. These two system parts are linked. This expert system is applied to the design of an auxiliary feedwater system with various design alternatives and performs a successful reliability analysis.

Development of An Expert Systems for Ergonomic Workplace Design and Evaluation

Industrial workplace design and evaluation is the outcome of a multi-factored process which requires diverse disciplines and interrelated techniques to consider alternative factors and to achieve an optimal design solution. However, it is still evident that industrial workplace design fails to incorporate ergonomic principles throughout all stages of design and evaluation, One approach to solve the problem is to introduce an expert system for integrating existing analytic models and expertises into a framework which guides the designer along the necessary steps to reach a solution, with explanations on its reasoning process. This paper discusses the framework of the prototype expert system being implemented using VM/PROLOG on IBM VM/CMS mainframe. The rule-based production system was selected as a representation scheme due to its versatility and expressive power. It is consisted of two main parts, First, modularized knowledge bases incorporate multidisciplinary ergonomic factors such as biomechanics, work physiology, and psychophysics. Each module stores knowledge either in the fact base or the rule base, however, massive experimental findings and table-lookups are separately stored in the external database through its interface and retrieved without bothering the main inference mechanism. Secondly, inference mechanism was built as a control mechanism, with a front-end user interface. It has a pattern-directed architecture coupled with a normal forward/backward chaining mechanism. The prototype expert system also incorporates analytical models (usually written in FORTRAN) into the reasoning process so that it is highly flexible to the problem specificity.

I-Test: integrated testing expert system for trunks

A prototype trunk-testing expert system developed by Bell Communications Research (Bellcore) is described. The integrated-testing expert-system for trunks (I-TEST) mechanizes and automates many of the conventional manual testing procedures in a switching control center (SCC) of a Bellcore client company (BCC). I-TEST demonstrates the usefulness of integrating artificial intelligence (AI) techniques into the network operations environment. I-TEST was developed in the INTELEWINDOWS environment. This environment includes a windowing system, graphical kernel system (GKS), and a set of high-performance portable C-based AI tools. I-TEST combines traditional structured procedural knowledge and a typical rule-based production system into one inference mechanism to manage a complex trunk-testing environmental in the SCC. I-TEST is an effort to use one intelligent system to integrate all of the future network testing needs, and provide an automated testing environment. The modular design and generic nature of I-TEST allow it to be easily reconfigured to include test access to additional operations systems or network elements. >

An object-oriented design system shell

We present a design system shell which can be used to experiment with principles of design and be used as a design tool where complex layers of information need to be specified about objects, such as in database design. The shell can be tailored to a variety of application areas. It is object-oriented in its implementation and structure. Objects and messages are used as the specification language. The basic ingredients of a rule-based production system are provided, with rules treated as objects and defined independently of the classes to which they are applied.

Knowledge representation and planning control in an expert system for the creative design of mechanisms

An interactive system, referred to as MECXPERT {Mechanism Expert}, has been designed with the expressed purpose of assisting nonexpert design engineers in creating mechanisms for fulfilling specific motion-conversion and/or power-transmission requirements. The particular knowledge representation scheme chosen for this application comprises a hybrid formulation of a rule-based production system with a frame-based approach. The underlying control strategy is based on a series of special-purpose, domain-specific operators whose function is to move from one problem space to another through various stages or ‘states’ that comprise the mechanism design process.The primary focus of this paper centers on the representation of knowledge and its control within an expert system for creative mechanism design. An overview summarizing the reasons for developing such an expert system is provided, and the formulation of a problem is discussed through an example taken from the design of a variable-stroke internal-combustion engine mechanism.

Strategies for Diagnosis

Diagnosis is among the dominant applications of expert systems technology today. In the past, most diagnostic systems were built using some form of rule-based production system. Recently, however, many new strategies have emerged to support much more complex reasoning for diagnosis. A survey of the architectures for the complex strategies now available for diagnosis is presented.

Execution of OPS5 production systems on a massively parallel machine

In recent years, the development of expert systems implemented by rule-based production systems has emerged as one of the dominant paradigms in the field of artificial intelligence. While production systems offer important advantages in large-scale AI applications, their use in such applications is typically very costly in execution time. In this paper, we describe an algorithm for executing production systems expressed in the OPS5 language on a massively parallel multiple-SIMD machine called NON-VON, portions of which are currently under construction at Columbia University. The algorithm, a parallel adaptation of Forgy's Rete Match, has been implemented and tested on an instruction-level simulator. We present a detailed performance analysis, based on the implemented code, for the averaged characteristics of six production systems having an average of 910 inference rules each. The analysis predicts an execution rate of more than 850 production firings per second using hardware comparable in cost to a VAX 11/780. By way of comparison, a LISP-based OPS5 interpreter running on a VAX 11/780 typically fires 1 to 5 rules per second, while a Bliss-based interpreter executes 5 to 12 rules per second.

Automated reasoning about image motion using a rule-based deduction system

A rule-based production system for the identification and structural description of image motion is presented. The system employs a set of intermediate-level motion description rules of antecedent-consequent form and a set of low-level, incomplete and possibly incorrect local spatiotemporal motion features together with a forward-backward chaining paradigm to produce or verify high-level motion descriptions. The system has been implemented in LISP and provides a basis for more complex designs, perhaps employing motion-specific frames and iterative deduction and feature extraction. Sample results and future research directions are presented.

A Boolean array based algorithm in APL

This paper describes an algorithm for forward chaining in rule-based production expert systems which makes extensive use of array structures and Boolean operations, to which APL lends itself. Additionally, a short introduction is provided for the reader unfamiliar with expert systems.