Document Meta-Stream Technology in Networked E-engineering and E-science
I. Georgiev
Key Words:
Document management; XML stream of metadata; web services; web arrays; classified data filtering.
Abstract:
Web engineering and scientific applications span a large amount of document distribution driven by an enormous diversity of standards, protocols, and data structures. Two trends are dominating: a. presenting the scientific and engineering data in XML documents; b. exchanging the documents by web services. This paper expands these trends by higher degree of document presentation and supporting technologies. Document presentation and management are based on document meta-stream that is a XML document, constructed by different section. Every section contains references to specific document repository. The sections are individually encrypted and provide information to classified users. The following technologies for the meta-stream processing are applied: a. XML streaming that does not need storage of the whole document; b. web arrays that are free-form collection of relevant web resources; c. cryptography classified filtering of the document stream that allows different access policies.
Adaptive intelligence - Essential aspects
K. Penev
Key Words:
Adaptive Intelligence; Free Search; Differential Evolution; Swarm Intelligence; heuristics; No Free Lunch Theorem.
Abstract:
The article discusses essential aspects of Adaptive Intelligence. Experimental results on optimisation of global test functions by Free Search, Differential Evolution, and Particle Swarm Optimisation clarify how these methods can adapt to multi-modal landscape and space dominated by sub-optimal regions, without supervisors’ control. In addition Free Search separately is evaluated on hard constraint global optimisation problem with unknown solution. It illustrates Free Search abilities for adaptation to unknown space. The achieved results are compared and analysed.
Polynomial Algorithms for Solving the Most Reliable Route Problem
A. Hossain, G. Gatev
Key Words:
Networks; possibility; interval possibility; most reliable route; interval most reliable route.
Abstract:
The uncertainty about the reliability of a route is represented in a possibilistic setting. The concept of interval possibility is introduced, as a generalization of fuzzy set concept of possibility, to deal with a higher degree of uncertainty. Four simple algorithms are proposed for solving the Most Reliable Route Problem under parametric uncertainty. The aim is to find the most reliable route on a network that maximizes the possibility of not being stopped on the route. The possibilities on the route segments are uncertain. The applicability of the results is demonstrated by considering several examples.
New Optimality Criteria for Movement Planning in a Group of Mobile Robots Navigated via Trilateration
S. Genchev
Key Words:
Motion planning; trilateration; cooperative robotics
Abstract:
This paper presents an algorithm for motion planning in a team of cooperating mobile robots navigated using time-of-flight trilateration. The method uses some of the robots as stationary beacons guiding the robots in motion, thus enabling long-term working in unstructured environments. The main purpose of the planning is not building collision-free paths, but maintaining the positioning accuracy during the motion. Two important optimality criteria are considered, related to specific aspects of the common motion how to plan trajectories with good movement precision, how to choose which robots to use as beacons and how to position them, in order to form appropriate geometrical arrangements.
Lyapunov Stability and Robustness of Fuzzy Process Control System with Parallel Distributed Compensation
S. Yordanova
Key Words:
Linear matrix inequalities; Lyapunov stability; MATLABTM real time temperature control; parallel distributed compensation PI process controller design; robust performance; time delay.
Abstract:
The Parallel Distributed Compensation (PDC) established itself as a general systematic approach for fuzzy control systems design from stability criteria, using Lyapunov direct method to define analytically and the Linear Matrix Inequalities (LMIs) technique to solve numerically. Its main applications target is the mechanical systems. To spread it to process control some problems need engineering solutions. The aim of the investigation is to adapt the PDC-LMI approach for processes with time delay and uncertainty and PI fuzzy controllers in order to improve system performance and facilitate the PDC-LMI industrial application. The main results are: 1) development of identification based T-S modeling methodology employing Ziegler-Nichols approximation and PI local controllers; 2) derivation of Lyapunov fuzzy system stability conditions and the corresponding LMIs; 3) development of PDC PI controller design method for plants with time delay, ensuring system stability and robust performance; 4) its implementation for temperature control. Measurable improvements achieved by the PDC control in comparison with an ordinary PI control are the decreased by 30% overshoot and settling time.