1000 Ñîôèÿ, óë. “Ðàêîâñêè” 108, òåë. 029876169, sai.bg.office@gmail.com, www.sai.bg

MODELING OF THE 2D ALTERNATING FREEZING AND THAWING OF LOGS AT AMBIENT AIR TEMPERATURE M. Hadjiski, N. Deliiski, N. Tumbarkova Key Words 2D mathematical models; beech logs; freezing; thawing; icing degree; ambient air temperature. Abstract This paper presents a methodology for mathematical modeling and research of two mutually connected problems: 2D non-stationary temperature distribution in logs subjected to many days and nights alternating freezing and thawing at periodically changing air temperature near them and change in the icing degree of the logs during these processes. Mathematical descriptions of the periodically changing ambient air temperature and of the icing degree of the logs under influence of that temperature have been carried out. These descriptions are introduced in our mutually connected 2D non-linear mathematical models of the 2D temperature distribution in logs during their freezing and thawing at convective boundary conditions. The paper presents solutions of the models with explicit form of the finite-difference method in the calculation environment of Visual FORTRAN. Results from a simulative investigation of 2D nonstationary temperature distribution and icing degree of beech logs with a diameter of 0.24 m, length of 0.48 m, moisture content of 0.6 kg·kg-1 , and initial temperature of 0 ?C during their 5 days and nights alternating freezing and thawing at sinusoidal change of the air ... Article PDF Download article (PDF)

TRAFFIC MANAGEMENTOF URBAN NETWORK BY BI-LEVEL OPTIMIZATION K. Stoilova, T. Stoilov, K. Pavlova Key Words Traffic control; transportation behavior; urban network of crossroad traffic lights; hierarchical approach; bi-level optimization. Abstract A city network of crossroad sections is under consideration in order to reduce the traffic jams, the traffic queue lengths in front of the junctions, and to increase the outgoing traffic flows. The implementation of these goals is achieved by application of hierarchical approach. A bi-level optimization is applied for finding the optimal control parameters as solutions of appropriate optimization problems, hierarchically interconnected. The numerical simulations’ results show improvement of the traffic’s characteristics Article PDF Download article (PDF)

LEARNABILITY AS AN INDICATOR FOR PLANNING AND CONTROL OF LEARNING SYSTEMS M. Hadjiski, R. Kaltenborn Key Words Intelligent learning systems; learnability; learning process; planning; reachability. Abstract The paper discusses a generalized approach to determining the indicator “learnability” for an individual student or a cluster of students in intelligent learning systems. Learnability is treated as a complex indicator that takes into account the impact of all key elements in the overall learning process. Learnability is a reliable indicator in planning and managing the learning process in all its aspects, taking into account its current state and forecasts for short and long time horizons Article PDF Download article (PDF)

SCANNING OF ELONGATED CONSTRAINED AREAS IN MIXTURE SPACES K. Stoyanov Key Words Scanning algorithm; grid search optimization algorithm; experiments with mixtures; constrained mixture spaces. Abstract The scanning of factor spaces (both process and mixture), usually is referred to as a technique for creation of a set of points. This set can be applied as the domain, where the values of some response function are calculated. Alternatively this can be the set of candidate points, used for searching of optimal experimental designs. In many practical applications, where some constraints over the components are imposed, the width of the subregion along the components, can differs significantly. In such cases the regions of interest obtain complicated, elongated shape. A scanning algorithm, which provides options to choose different steps over the components, will save both computer resources for multicomponent (multifactor) tasks as well will provide regular (or at least controlled) distribution of the points across the scanned region. While the scanning, with different step lengths in an Euclidean coordinate space is trivial, the scanning of constrained mixture regions rises some algorithmic difficulties ... Article PDF Download article (PDF)

MODELING OF A HUMANOID BIPED ROBOT THROUGH DIFFERENT APPROACHES Sh. Sherif, J. Kralev, Ts. Slavov Key Words Humanoid robot; modelling; biped robot; nonlinear model; walking algorithm; multivariable identification Abstract In this article, the authors present two differentapproaches to modeling a biped humanoid robot. The first approach to modeling is based entirely on geometric and physical parameters, using blocks from SimMechanicsTM library in Simulink®, as well as personally constructed blocks by the authors. The other modeling approach consists of several stages: first a basic walking algorithm is implemented on a biped humanoid robot through the programming language C with purpose to record a set of experimental data necessary for the identification procedure. An initial identification with a multivariable ARX model is then made, and final model is in state-space and is obtained by identification the initial one by the method of the predicted error Article PDF Download article (PDF)