Catálogo de publicaciones - libros

Advances in grid computing have recently sparkled the research and development of Grid problem solving environments for complex design. Parallelism in the form of distributed computing is a growing trend, particularly so in the optimization of high-fidelity computationally expensive design problems in science and engineering. In this paper, we present a powerful and inexpensive grid enabled evolution framework for facilitating parallelism in hierarchical parallel evolutionary algorithms. By exploiting the grid evolution framework and a multi-level parallelization strategy of hierarchical parallel GAs, we present the evolutionary optimization of a realistic 2D aerodynamic airfoil structure. Further, we study the utility of hierarchical parallel GAs on two potential grid enabled evolution frameworks and analysis how it fares on a grid environment with multiple heterogeneous clusters, , clusters with differing specifications and processing nodes. From the results, it is possible to conclude that a grid enabled hierarchical parallel evolutionary algorithm is not mere hype but offers a credible alternative, providing significant speed-up to complex engineering design optimization.

This paper investigates a global search optimisation technique, referred to as the repeated weighted boosting search. The proposed optimisation algorithm is extremely simple and easy to implement. Heuristic explanation is given for the global search capability of this technique. Comparison is made with the two better known and widely used global search techniques, known as the genetic algorithm and adaptive simulated annealing. The effectiveness of the proposed algorithm as a global optimiser is investigated through several examples.

Modularity is a widespread form of organization in complex systems, but its origins are poorly understood. Here, I discuss the causes and consequences of modularity in evolutionary systems. Almost all living organisms engage in sexual exchange of genes, and those that do so are organized into discrete modules we call species. Gene exchange occurs within, but not between, species. This genetic segregation allows organisms to adapt to different niches and environments, and thereby evolve complex and long-lasting ecosystems. The process that generates such modularity, speciation, is therefore the key to understanding the diversity of life. Speciation theory is a highly developed topic within population genetics and evolutionary theory. I discuss some lessons from recent progress in speciation theory for our understanding of diversification and modularity in complex systems more generally, including possible applications in genetic algorithms, artificial life and social engineering.

Small populations are very desirable for reducing the required computational resources in evolutionary optimization of complex real-world problems. Unfortunately, the search performance of small populations often reduces dramatically in a large search space. To addresses this problem, a method to find an optimal search dimension for small populations is suggested in this paper. The basic idea is that the evolutionary algorithm starts with a small search dimension and then the search dimension is increased during the optimization. The search dimension will continue to increase if an increase in the search dimension improves the search performance. Otherwise, the search dimension will be decreased and then kept constant. Through empirical studies on a test problem with an infinite search dimension, we show that the proposed algorithm is able to find the search dimension that is the most efficient for the given population size.

Membrane systems are gaining a prominent role in the modeling of biochemical processes and cellular dynamics. We associate specific reactivity values to the production rules in a way to be able to tune their rewriting activity, according to the kinetic and state-dependent parameters of the physical system. We come up with an algorithm that exhibits a good degree of versatility, meanwhile it gives an answer to the problem of representing oscillatory biological and biochemical phenomena, so far mostly treated with differential mathematical tools, by means of symbolic rewriting. Results from simulations of the Lotka-Volterra’s predator-prey population dynamics envision application of this algorithm in biochemical dynamics of interest.

The essential of Molecular docking problem is to find the optimum conformation of ligand bound with the receptor at its active site. Most cases the optimum conformation has the lowest interaction energy. So the molecular docking problem can be treated as a minimization problem. An entropy-based evolution model for molecular docking is proposed in this paper. The model of molecular docking is based on a multi-population genetic algorithm. Two molecular docking processes are investigated to demonstrate the efficiency of the proposed model.

Based on the concept of coarse-grained description, a new encoding scheme with grouped weight for protein sequence is presented in this paper. By integrating the new scheme with the component-coupled algorithm, the overall prediction accuracy of protein structural class is significantly improved. For the same training dataset consisting of 359 proteins, the overall prediction accuracy achieved by the new method is 7% higher than that based solely on the amino-acid composition for the jackknife test. Especially for + the increase of prediction accuracy can achieve 15%. For the jackknife test, the overall prediction accuracy by the proposed scheme can reach 91.09%, which implies that a significant improvement has been achieved by making full use of the information contained in the protein sequence. Furthermore, the experimental analysis shows that the improvement depends on the size of the training dataset and the number of groups.

In this paper, after a new DNA-based semantic model is theoretically proposed, the preliminary experiment on construction of the small test model is successfully done. This model, referred to as ‘’ (SMC) has the structure of a graph formed by the set of all (attribute, attribute values) pairs contained in the set of represented objects, plus a tag node for each object. Each path in the network, from an initial object-representing tag node to a terminal node represents the object named on the tag. Input of a set of input strands will result in the formation of object-representing dsDNAs via parallel self-assembly, from encoded ssDNAs representing both attributes and attribute values (nodes), as directed by ssDNA splinting strands representing relations (edges) in the network. The proposed model is rather suitable for knowledge representation in order to store vast amount of information with high density. The proposed model will appears as an interaction between AI and biomolecular computing research fields, and will be further extended for several AI applications.

Interest in DNA computing has increased overwhelmly since Adleman successfully demonstrated its capability to solve Hamiltonian Path Problem (HPP). Many research results of similar combinatorial problems which are mainly in the realm of computer science and mathematics have been presented. In this paper, implementation ideas and methods to solve an engineering related combinatorial problem using this DNA computing approach is presented. The objective is to find an optimal path for a complex elevator scheduling problem of an 8-storey building with 3 elevators. Each of the elevator traveled path is represented by DNA sequence of specific length that represent elevator’s traveling time in a proportional way based on certain initial conditions such as present and destination floors, and hall calls for an elevator from a floor. The proposed ideas and methods show promising results that DNA computing approach can be well-suited for solving such real-world application in the near future.

In this paper we propose a procedure for the storage and retrieval of digital signals utilizing DNA. Digital signals are encoded in DNA sequences that satisfy among other constraints the Noise Tolerance Constraint (NTC) that we have previously introduced. NTC takes into account the presence of noise in digital signals by exploiting the annealing between non-perfect complementary sequences. We discuss various issues arising from the development of DNA-based database solutions (i) (in test tubes, or other materials) for short-term storage and (ii) (inside organisms) for long-term storage. We discuss the benefits and drawbacks of each scheme and its effects on the codeword design problem and performance. We also propose a new way of constructing the database elements such that a short-term database can be converted into a long term one and vice versa without the need for a re-synthesis. The latter improves efficiency and reduces the cost of a long-term database.