Catálogo de publicaciones - libros

DNA computing has been proposed to solve the famous “SAT” problem of computer science since Adleman shows that DNA strands can be used to solve an instance of the NP-complete Hamiltonian path problem (HPP). Based on the Adleman-Lipton model and apply fluorescein-labeled operations, this paper put forward several DNA based parallel algorithms to control the running-time of algorithm to constant. In order to achieve this, some effective encode methods are designed to reduce the time complexity. In addition, the DNA parallel algorithms can improve the space complexity to a reasonable spectrum by introducing a trichotomy strategy. The analysis of the computational complexity for algorithm model will show you the advantage of the fluorescence model.

3-vertex-colourability problem is one of computationally hard problems in combination optimization. The paper presented an algorithm via the interaction of PNA molecules and DNA molecules. We initially synthesized DNA memory strands of single molecular variety, and then PNA stickers were gradually annealed to specified regions on DNA strands, followed by magnetic bead separation to separate those memory strands that were not solutions. The novelty of proposed method is that solution to the problem of interest was constructed during the course of computation instead of searched in pre-synthesized vast solution space.

We study the approximation of the integration of multivariate functions classes in the quantum model of computation. We first obtain a lower bound of the n -th minimal query error for integration on anisotropic Sobolev-Slobodezkii classes. Then combining our previous results we determine the optimal bound of n -th minimal query error for anisotropic Hölder-Nikolskii class and Sobolev class. The results show that for these two type of classes the quantum algorithms give significant speed up over classical deterministic and randomized algorithms.

Agent should work collaboratively to solve complex task. Policy plays an important role in agent grid collaboration and improves the collaborative schema reusability and scalability. This paper studies the policy concept model and formal policy representation to describe the action policy, goal policy and utility policy. Then the general agent grid collaborative schema is defined based on policy. Furthermore in a hierarchical view the behavior, agent, agent society and agent grid is introduced. Ultimately, based on policy we illustrate an oil supply system experiment.

One of the major services from a higher education department is to provide up-to-date information for potential students who want to pursue higher education in foreign countries. These information collection processes, if handled by human, are time consuming and tedious. Educational Service Web Database Prototype (ESWD) is a prototype system built for educational counselors whose one of duties is to provide information to aid students in university selection process. It is an automated information collection prototype system. The main purpose is to automatically collect information from the webpages of top universities worldwide. Technologies employed during the development are web search crawling, text processing, and database management.

A searching result can be returned for submitting user query to Search Engine(SE). However, large numbers of Web pages in the searching result are not interested by users. To overcome this problem, from automated reasoning perspective, we put forward a matching model how to match user query with Web pages. According as the basic theory of formal concept lattice, we defined OR-RULE and AND-RULE, put forward AND-MATCH and OR-MATCH. Two reasoning methods based on AND-MATCH and OR-MATCH, and their algorithms of the matching model, are proposed. The experiment demonstrates that these method is efficient.

MANET is a collection of two or more mobile devices equipped with wireless communications and networking capability. The mobile nodes have good portability and flexibility, but they also have energy limitation. If the energy consumption of nodes can not be balanced efficiently through the network, some nodes will be out of commission too early because of running out of energy, which breaks the path and communication. For energy balancing of the entire network, we propose a packet distribution method based on four energy collection criteria; Static-Average, Static-Minimum, Dynamic-Average and Dynamic-Minimum on node-disjoint multipaths between a source and a destination. By adopting the methods, we attempt to achieve the optimal energy balance of the network. The performance of packet distribution methods by the four criteria on the node-disjoint multipath and that of AODV are evaluated by NS2 simulation. The research result proves that Dynamic-Minimum is the best one to accomplish the energy balance of the network.

The traditional qualitative SDG method has a great significance in safety engineering of process industry. But its inherent qualitative ambiguities can usually result in many spurious interpretations about faults because of the absence of quantitative information. On the basis of traditional qualitative SDG, we present a new probabilistic SDG model to describe large-scale systems utilizing probabilistic information. The definition of probabilistic SDG is presented and the inference method is discussed. Fault diagnosis based on the new probabilistic SDG approach can compute the probabilities of each candidate fault given the on-line measured evidence and then sort them by their probabilities. So the faults which will result in the same qualitative symptoms can be distinguished. A real fault diagnose case of the inversion of synthetic ammonia based on probabilistic SDG is given. Experimental results show the validity and advantages of the new probabilistic SDG approach.

In this paper phase space reconstruction technique is used for distinguishing the combustion states in the boiler furnace. The phase space maps and the correlation dimensions under the stable combustion & the unstable combusition are charted and calculated. The simulation and calculation results of the actual flame signals show that the phase space maps and the correlation dimensions under both combustion states are different at all time. Therefore it is presented that the correlation dimension ought to be regarded as the characteristic parameter to distinguish the combustion states of the flame in the boiler furnace. And the approach presented in this paper may be applied to develop the novel optical flame detector.

The proposal is regarding the fault diagnosis method of included angle sine based on theory of cluster analysis. The fault models are established in accordance with the included angle sine among the fault symptom and the fault model sets. Ultimately, the method is used to diagnose network security fault. The results show the effectiveness of the proposed approach.