Catálogo de publicaciones - libros

The traveling salesman problem concerns the best way to visit a set of customers located in some metric space and to minimize the length of the tour over all customer subsets. The problem is a typical NP-hard combinatorial optimization problem, which is of major importance in real world applications. An effective hybrid genetic algorithm-ECOGA is proposed for the problem in this paper, which combines 2-exchange crossover heuristic operator and improved 2OPT of neighbor search algorithm absorbing K-Nearest Neighbor. What’s more, the rule of 5 is applied to the proposed algorithm to guide the search direction. On a set of standard test problems with symmetric distances, the proposed ECOGA found the solutions that were optimal in every case and some of them are superior to the optimality found in TSPLIB. The ECOGA is completive with other genetic algorithm published to date in both solution quality and computation time.

This paper presents a GEP-based simulation – data mining approach for obtaining closed-form reliability formulas of consecutive-k-out-of-n: F systems with identical component reliabilities. This work proves to be GEP’s first exploration into the reliability realm and also provides a new perspective for the reliability community to solve for complex reliability formulas. Experimentation has shown the feasibility and effectiveness of the proposed framework, although further revisions and developments must be made to the model in order to solve larger scale problems.

This paper presents a rescheduling model of a vehicle routing problem when a disruption occurs at a particular time and lasts for a period of time after a subset of the customers has been visited. In such cases, continuing with the original schedule is likely to be infeasible. The rescheduling model taken here is significantly different from the original one due to the fact that the objective is to find a new schedule that minimizes total distance and deviations from the original plan, and that the different neighborhood size and several new constraints must be considered during the recovery procedure. A hybrid algorithm, which is to hybridize ant colony optimization (ACO) with scatter search, is adopted to determine good approximate solutions. Computational experiments were also tested to determine the effects of factors affect the recovery procedure, and our studies will be helpful to disruption management for the vehicle routing problem.

The paper presents the representation of solutions and design of genetic operators for solving of flexible job shop problem (FJSP). Effective methods of representation of problems with the object of further use of genetic operators are considered. Permutation with repetitions are connected with genes which code foreground location of machines for seperate operations. Repairing of chromosomes after crossover for obtaining feasible schedules, as well as the two auxiliary functions IsValid and FindValid are describes. Our computional tests shows that the proposed representation is effective in improving solution quality.

Conventional computation is usually based on the manipulation of numbers and symbols. However, novel methodologies are needed to deal with the words and propositions drawn from natural languages (NL), because a lot of real-world problems can be characterized only by linguistic measurements. In this paper, we give an overview on the recent progresses in the study of the Computing with Words (CW). The applications of the CW technique in data mining and pattern recognition are also briefly discussed.

This paper describes a novel fuzzy rule-based modeling approach for some slow industrial processses. Structure identification is realized by clustering and support vector machines. When the process is slow, fuzzy rules can be obtained automatically. Parameters identification uses the techniques of fuzzy neural networks. A time-varying learning rate assures stability of the modeling error.

It is well known that classical computing is strict, while soft computing is soft. But how soft is soft computing? Like water, cotton, or spring? Can the soft be at will? If can not be, how soft it should be? Is there a proper principle or restriction? Soft computing is a partnership of many solutions, and fuzzy logic is one of the most important members of it. On the basis of fuzzy logic, this paper analyzes the problems and weaknesses of the two ording methods– the ording in generalizing sequential relation and the ording method according to average with weight, which are two of the existing fuzzy matching and fuzzy ording methods. We propose several significant principles, which are not only effective to fuzzy logic but also to the whole soft computing. The principles remind us that we should hold the ’degree’ in soft computing, or else soft computing may lose its scientific character and soundness.

In this paper, intelligent filtering methodology for masterarm translation signal is proposed. Fidelity and stability are contradicting factors in teleoperation. Human hand trembling filtering is one of the problems in telemanipulation field. During every operation the hand has a certain vibration that can affect the quality of teleoperation, especially in telesurgery, nanomanipuation and other precise tasks. It is very important to study the kinesthetic perception of the human and to optimize the teleoperation system accordingly. To cancel out the influence of human’s hand vibration the signal from the masterarm should be filtered. One of the feasible solutions is to use an intelligent filter, which is a very flexible instrument. Applying intelligent filtering methodology, we can use some heuristic methods to solve the filtering problem.

This paper studies the problems of fuzzy multi-attribute group decision making. First of all, it gives a new method for ranking trapezoidal fuzzy numbers based on the credibility distribution, then it presents the TOWA operator to aggregate the trapezoidal fuzzy numbers. Furthermore, a new model is presented for the problems of fuzzy multi-attribute group decision making via TOWA operator and grey relative degree. Finally, a decision making example is given to demonstrate the feasibility and rationality of this new model.

To treat the difficulties in the design of MIMO fuzzy controller, which arise as high dimensional rule-bases and the acquirement of the membership functions and rules. Based on the importance of each input is different, a multi-dimensional controller is decomposed into a lot of one-dimensional controller that is presented in this paper. The total number of rules is drastically decreased. For an inverted pendulum, The simulation results show the controller has better dynamic performance and stability than the conventional MIMO fuzzy controller.