Journal of Intelligent & Fuzzy Systems - Volume 1, issue 3

Authors: Trappey, Charles V. | Trappey, Amy J.C.

Article Type: Research Article

Abstract: This article presents a model of retail buyers investing in merchandise for resale to consumers. Given a selection or portfolio of merchandise to buy for a store, the retailer must determine the optimal percentage investment in each item. Depending upon the buyer's investment objectives and tolerance of risk, the profit of the portfolio is maximized while the risk is simultaneously minimized. We first model this investment decision using mean-variance optimization and then reformulate the nonlinear programming model to include a fuzzy objective function and fuzzy constraints. Our results indicate that the fuzzy approach increases the flexibility and profitability of the …model, assisting the retailer to build portfolios of merchandise that satisfy consumer demand under conditions of uncertainty. Show more

DOI: 10.3233/IFS-1993-1301

Citation: Journal of Intelligent and Fuzzy Systems, vol. 1, no. 3, pp. 189-197, 1993

Authors: Parkinson, W.J. | Shalek, P.D. | Duerre, K.H. | Luger, G.F. | Jamshidi, M.

Article Type: Research Article

Abstract: Silicon carbide whiskers, a strong material that resembles cat's whiskers, are produced, primarily, for strengthening ceramics and metals. The current method for producing these whiskers is to use a semibatch process, extremely difficult to model mathematically. An alternative model utilizes the expertise of operators accumulated from years of experience in initializing and running the process. We designed two expert systems in which the process can be set up and run by nonexperts. One system uses crisp logic and the other fuzzy logic. In this article, we compare the two systems.

DOI: 10.3233/IFS-1993-1302

Citation: Journal of Intelligent and Fuzzy Systems, vol. 1, no. 3, pp. 199-214, 1993

Authors: Yager, Ronald R.

Article Type: Research Article

Abstract: Consonant belief structures provide a representation for fuzzy sets because their plausibility measures are possibility measures. However, in general the aggregation of these consonant belief structures are not consonant, not fuzzy sets. In this article, we attempt to overcome this problem. We first note that two belief structures are equivalent if their plausibility and belief functions are equal. This observation allows us to provide different equivalent representations for any belief structure. This allows us to induce for different consonant belief structures commensurate representations. We show that if we represent two consonant belief structures in a commensurate manner their aggregations are …also consonant if we impose the additional requirement that the underlying probability distributions satisfy the condition of synonyminity, that is, they are completely correlated. The results of this work allow us to use belief structure representations to manipulate fuzzy subsets. Show more

DOI: 10.3233/IFS-1993-1303

Citation: Journal of Intelligent and Fuzzy Systems, vol. 1, no. 3, pp. 215-224, 1993

Authors: Scarpelli, Heloisa | Gomide, Fernando

Article Type: Research Article

Abstract: This article presents a net-based structure to model approximate reasoning using fuzzy logic, the fuzzy Petri net model. The knowledge bases to be considered here are assumed to be fuzzy production rules. After a brief introduction of knowledge representation and fuzzy reasoning, we give a new definition of the Fuzzy Petri net model. Next, the basic net structures for complex forms of rules such as rules with conjunction in the antecedent, rules with linguistic quantifiers, and rules with certainty factors are presented. Typical rules sets like parallel rules and conflicting rules are addressed as well. Design techniques to be used …when the basic structures are mixed in the context of a large knowledge base are also included. A fuzzy reasoning algorithm is provided. Finally, an application example concerning manufacturing cells modeling is introduced to illustrate the usefulness of the approach proposed. Show more

DOI: 10.3233/IFS-1993-1304

Citation: Journal of Intelligent and Fuzzy Systems, vol. 1, no. 3, pp. 225-241, 1993

Authors: Nedungadi, Ashok

Article Type: Research Article

Abstract: Conventional robot controllers rely on an exact inverse kinematic model to convert user-specified Cartesian trajectory commands to joint set points. These inverse kinematic equations are robot specific and form a set of highly coupled nonlinear equations that pose a considerable computational bottleneck. It is not uncommon for the host CPU to dedicate as much as 80% of its computational resources to solve the inverse kinematic equations alone. Further, the strictly mathematical nature of the inverse kinematic equations results in kinematic singularities in which the robot is uncontrollable. These drawbacks have resulted in current robot controllers being limited in flexibility with …respect to adaptive control, kinematic redundancy, and obstacle avoidance. This observation provided the impetus to investigate the applicability of fuzzy logic to resolve the computational bottleneck of the inverse kinematic module. In the following, an approach is presented based on fuzzy logic, which replaces the computationally intensive inverse kinematic equations and lays the foundation for a robot controller with a simpler architecture. A hardware implementation of a fuzzy logicbased robot controller is described in which simple fuzzy rules replace the complex inverse kinematic equations. The implemented fuzzy robot controller was tested on a planar four-degree-of-freedom lab robot. Hardware test results of the lab robot tracking arbitrary trajectories are discussed. Show more

DOI: 10.3233/IFS-1993-1305

Citation: Journal of Intelligent and Fuzzy Systems, vol. 1, no. 3, pp. 243-251, 1993

Authors: Vadiee, Nader | Jamshidi, Mohammad

Article Type: Research Article

Abstract: Fuzzy rule-based expert systems (FRBES) models, described by a set of fuzzy conditional statements in the canonical form of (If x is Ak then yk is Bk for k = 1, 2, … , r) or a system of conjunctive or disjunctive fuzzy relational equations in the form of (yk = x · Rk for k = 1, 2, … , r) were discussed in part I of this tutorial. In Part II provides detailed descriptions of some interesting special cases of FRBES models of dynamic systems along with a number of commonly used formats …for FRBES models and their solutions. The fundamental computational processes of FRBES models such as partitioning input and output spaces into fuzzy partitions and assigning them membership functions, fuzzification, inferencing, defuzzification processes are illustrated. A number of interesting graphical techniques for the above-mentioned computational processes are also given. Finally, fuzzy associative memories tables are introduced. Show more

DOI: 10.3233/IFS-1993-1306

Citation: Journal of Intelligent and Fuzzy Systems, vol. 1, no. 3, pp. 253-264, 1993