Catálogo de publicaciones - libros

In order to resolve existing problems such as low efficiency, high cost and lack of technical resource in current maintenance, it is necessary to realize remote diagnosis for hydroelectric generating sets (HGSs). In this work, basing on the Service-Oriented Architecture (SOA) and Web Services technology, a SOA-oriented E-diagnostics system for HGSs (HGS-SES) is proposed, the framework of HGS-SES is constructed, a layout of the system’s hardware settings is described, the key modules of the system and a specific diagnostic procedure are given. HGS-SES makes rapid and convenient information transmission for services function and diagnosis decision-making, develops a dynamic network diagnostic platform for HGSs, and has broad prospects for farther research.

As the new era of RFID (Radio Frequency Identification) has come, it makes the visualizing of a plane possible. The proposed research focuses on how to plan the locations of RFID readers in a non-rectangular plane such that regions with or without forbidden blocks can be fully monitored by using a well developed RFID system under the spending of a minimum cost. An algorithm constitutes of three phases is used to obtain the optimal devices’ layout planning. In the beginning, we scrutinize a non-rectangular plane and its forbidden blocks so that a general grid-scheme can be applied. Secondly, the linear programming (LP) approach is applied to decide the number of RFID readers needed. At last, a hybrid genetic algorithm (GA) is used to find the appropriate locations for these designated number of RFID readers. The overall cost of deploying RFID readers and the total monitored region of the proposed RFID system are recorded. They are two key performance indexes to evaluate the efficiency of the proposed method. Simulation results show the proposed method has high efficiency on dealing RFID readers’ planning problems in visualizing a non-rectangular plane.

Performance of a genetic algorithm for function optimization, often appeared in real-world applications, depends on its crossover operator strongly. Existing crossover operators are designed for intensive search in certain promising regions. This paper, first, discusses where the promising search regions are on the basis of some assumptions about the fitness landscapes of objective functions and those about a state of a population, and this discussion reveals that existing crossover operators intensively search some of the promising regions but not all of them. Then, this paper designs a new crossover operator for searching all of the promising regions. For utilizing the advantageous features of this crossover operator, a new selection model considering characteristic preservation is also introduced. Several experiments have shown the proposed method has worked effectively on various test functions.

Constructing a fingerprint database is important to evaluate the performance of automatic fingerprint recognition systems. Because of the difficulty in collecting fingerprint samples, there are only few benchmark databases available. Moreover, various types of fingerprints are required to measure how robust the system is in various environments. This paper presents a novel method that generates various fingerprint images automatically from only a few training samples by using the genetic algorithm. Fingerprint images generated by the proposed method include similar characteristics of those collected from a corresponding real environment. Experiments with real fingerprints verify the usefulness of the proposed method.

This paper proposes a new hybrid heuristic to a difficult but frequently occurring problem in the apparel industry: the cut order planning (COP). This problem consists of finding the combination of ordered sizes on the material layers that minimizes total material utilization. The current practice in industry solves COP independently from the two-dimensional layout (TDL) problem; i.e., COP estimates the length of the layout required to cut a particular combination of sizes instead of packing the pieces on the fabric and determining the actual length used. Evidently, this results in a build up of estimation errors; thus increased waste. Herein, COP and TDL are combined into a single problem CT. The resulting problem is modeled and solved using a hybrid heuristic which combines the advantages of population based approaches (genetic algorithms) with those of local search (simulated annealing). The experimental results show the validity of the proposed model, and the sizeable savings it induces when solved using the proposed hybrid heuristic.

In spite of the prosperous literature in adaptive control, application of this promising control strategy has been restricted by the lack of assurance in closed-loop stability. This paper proposes an adaptive control architecture, which is augmented by a supervising controller, to enhance the robustness of an adaptive PID control system in the face of exaggerated variation in system parameters, disturbances, or parameter drift in the adaptation law. Importantly, the supervising controller is designed based on an on-line identified model in a fuzzily blended time-varying canonical form. This model largely simplified the identification process, and the design of both the supervising controller and the adaptation law. Numerical studies of the tracking control of an uncertain Duffing–Holmes system demonstrate the effectiveness of the proposed control strategy.

In this paper, a novel multi-agent control system incorporating hybrid intelligence and its physical testbed are presented. The physical testbed is equipped with a large number of embedded devices interconnected by three types of physical networks. It mimics a ubiquitous intelligent environment and allows real-time data collection and online system evaluation. Human control behaviours for different physical devices are analysed and classified into three categories. Physical devices are grouped based on their relevance and each group is assigned to a particular behaviour category. Each device group is independently modelled by either fuzzy inference or neural network agents according to the behaviour category. Comparative analysis shows that the proposed multi-agent control system with hybrid intelligence achieves significant improvement in control accuracy compared to other offline control systems.

With the explosive growth of data available on the Internet, a recent area of investigation called Web Mining has arise. In this paper, we will study general aspects of this area, principally the process of Web Usage Mining where log files are analyzed. These files register the activity of the user when interact with the Web. In the Web Usage Mining, different techniques of mining to discover usage patterns from web data can be applied. We will also study applications of Fuzzy Logic in this area. Specially, we analyze fuzzy techniques such as fuzzy association rules or fuzzy clustering, featuring their functionality and advantages when examining a data set of logs from a web server. Finally, we give initial traces about the application of Fuzzy Logic to personalization and user profile construction.

This paper considers a fuzzy Lyapunov method for stability analysis of nonlinear systems subjected to external forces. The nonlinear systems under external forces can be represented by Tagagi-Sugeno (T-S) fuzzy model. In order to design a nonlinear fuzzy controller to stabilize this nonlinear system, the parallel distributed compensation (PDC) scheme is used to construct a global fuzzy logic controller. We then propose the robustness design to ensure the modeling error is bounded and some stability conditions are derived based on the controlled systems. Based on the stability criterion, the nonlinear systems with external forces are guaranteed to be stable. This control problem can be reformulated into linear matrix inequalities (LMI) problem.

Although reverse logistics has been disregarded for many years, pressures from both environmental awareness and business sustainability have risen. Reverse logistical activities include return, repair and recycle products. Traditionally, since the information transparency of the entire supply chain is restricted, business is difficult to predict, and prepare for these reverse activities. This study presents an agent-based framework to increase the degree of information transparency. The cooperation between sensor and disposal agents helps predict reverse activities, avoid return, speed up repair and prepare for recycling behaviors.