Journal of Intelligent & Fuzzy Systems - Volume 14, issue 2

Authors: Varela, Leonilde Rocha | Ribeiro, Rita Almeida

Article Type: Research Article

Abstract: Simulated Annealing (SA) is a reasonable algorithm for solving optimization problems, through the selection of the best solution among a finite number of possible solutions. It is a particularly attractive technique to solve fuzzy optimization problems, because it allows finding near-optimal solutions, which, in a fuzzy environment, is usually good enough and without a large computational effort. We present a representative set of problems for testing the SA algorithm suitability and performance. The SA performance is …measured in terms of the objective function values, considering several trade-offs on constraints satisfaction levels, and computational time to achieve a solution. Furthermore, we discuss the parameters that control the SA algorithm to show how easily they can be manipulated. The set of fuzzy optimization problems tested were formulated following the complete fuzzification method proposed by Ribeiro and Moura-Pires (1999). The results obtained show the flexibility and adaptability of the SA algorithm to solve fuzzy optimization problems. Show more

Keywords: fuzzy optimization, simulated annealing, fuzzy constraints, fuzzy coefficients

Citation: Journal of Intelligent & Fuzzy Systems, vol. 14, no. 2, pp. 59-71, 2003

Authors: Ahnlund, Jonas | Bergquist, Tord | Spaanenburg, Lambert

Article Type: Research Article

Abstract: The proper handling of alarms is crucial to any automated process control. In practice, many alarms are only distractive and do not represent a potentially dangerous situation. This paper presents a methodology and a computerized tool that aims to remove such nuisance alarms, a so-called alarm cleanup. This is a general, systematic approach that takes advantage of the control system's built-in functions, and is a first step to an improved overall alarm situation. By the strong …reduction of the alarm count, the efficient construction of fault diagnosis and isolation models becomes feasible. In a typical case study, the number of alarms received at the remote control room of an operational bio-fueled District Heating Plant was effectively reduced by 83%. Show more

Keywords: process control, fault diagnosis, alarm cleanup, expert system, signal processing, multilevel flow model

Citation: Journal of Intelligent & Fuzzy Systems, vol. 14, no. 2, pp. 73-84, 2003

Authors: Boroushaki, Mehrdad | Ghofrani, Mohammad B. | Lucas, Caro

Article Type: Research Article

Abstract: In pressurized water reactors, the fuel reloading problem has significant meaning in terms of both safety and economics. An optimal loading pattern is defined as a pattern in which the local power peaking factor (P_q ) is lower than a predetermined value during one cycle and the effective multiplication factor (k_{eff} ) is maximized to extract the maximum energy. This article presents a specialized genetic algorithm for loading pattern design. The tests on well-researched …cases have shown that the genetic algorithm is capable of finding better loading patterns than solutions found by direct search methods. However, most of the previous researchers have considered simple fitness (cost) functions; therefore the reported solutions cannot minimize the P_q together with maximization of the k_{eff} . To solve this difficulty, we used Fuzzy Nonlinear Programming (FNLP) technique with the genetic algorithm to perform multi-objective optimization (maximizing the k_{eff} together with minimizing the P_q ). The results show that this method improves the genetic algorithm results compared to previous methods. Show more

Citation: Journal of Intelligent & Fuzzy Systems, vol. 14, no. 2, pp. 85-93, 2003

Authors: Smith, Earl B. | Langari, Reza

Article Type: Research Article

Abstract: In this work we apply the notion of fuzzy multi-objective decision making to the problem of navigation of mobile robots in unstructured environments such as lunar and planetary surfaces and military zones. In particular the proposed approach is concerned with multi-objective decision making in a dynamic setting where limited a priori information, say only a coarse image of the robot surroundings may be available. To this end, the proposed approach first constructs a coarse motion plan …and subsequently employs a sensor-based navigation scheme based on fuzzy multi-objective decision making to detect any intervening obstacles, dynamically define its objectives and preferences based on the relative locations of the obstacles, and move towards is goal. This process is comparable to human decision-making where an initial assessment of the surroundings and one's own sensing capability provide the basis for navigating and moving around in unstructured environments. In particular it is shown that the proposed approach offers a viable alternative to existing approaches to mobile robotics by providing a flexible architecture for addressing the complexity of motion planning and navigation in dynamic environments. Show more

Citation: Journal of Intelligent & Fuzzy Systems, vol. 14, no. 2, pp. 95-108, 2003

Authors: Wu, Peitsang | Fang, Shu-Cherng | Nuttle, Henry L.W.

Article Type: Research Article

Abstract: The performance of the "fuzzy-controlled curved search" method for back propagation neural network learning depends heavily upon the membership functions used in the fuzzy controller. Manually tuning the membership functions becomes a tedious job. In this paper, a fuzzy neuron controller with self-tuning capability is introduced to adjust the related membership functions in a self-adaptive manner. Computational results are included to illustrate the potential of this enhanced learning method.

Keywords: neural network learning, curved search algorithm, fuzzy neuron control

Citation: Journal of Intelligent & Fuzzy Systems, vol. 14, no. 2, pp. 109-118, 2003