PIETRIKOVÁ, A., BANSKÝ, J., ĎURIŠIN, J.:
Measurement of Nonlinearity of LeadFree Vapour Phase Reflowed Solder Joints [full paper]
The goal of the study is to compare quality of vapour phase reflowed solder joints soldered on different printed circuit board (PCB) surface finishes. Vapour phase reflow soldering guarantees the highest reliability, quality of processing and process safety. The big advantage of heat transfer in vapour phase soldering (VPS) is the extremely efficient transfer of latent heat through the condensing vapour of a chemically inert liquid. The heat transfer process is about ten times faster than hot air and about eight times faster than infra red heat. The condensation mediums used are perfluorocarbon, perflouropolyeter, flouriertes phenandren or perflouramin liquids, which are clear and colourless, have no odour, no taste, are nonflammable, noncorrosive and nonozone depleting. For lead free soldering a boiling point of 230 °C of these liquids is recommended. The temperature of any saturated vapour cannot exceed the temperature of the boiling liquid, therefore the temperature of the PCBs and components can never be higher than that of the vapour in the machine. As the liquid boils, a layer of saturated vapour forms above the fluid creating a 100 % inert atmosphere – it allows use of less aggressive flux than in other forms of reflow. The reflow in inert gas atmosphere defines an own unique standard. As there is no danger of overheating, there is no need to set temperature profiles. Vapour phase provides probably the best possible solder reflow conditions with only minimal temperature differences across the PCB. Vapour phase is independent of size, shape or geometry with the maximum surface area exposed and every component part seeing precisely the same temperature. Comparison of quality of the solder joints is based on measurement of the solder joints nonlinearity of the currentvoltage (CV) characteristics.

VOKOROKOS, L., ÁDÁM, N., BALÁŽ, A.:
Training Set Parallelism in PAHRA Architecture [full paper]
Multilayered feedforward neural networks trained with backpropagation algorithm are one of the most popular “online” artificial neural networks. These networks are showing strong inherit parallelism because of the influence of high number of simple computational elements. So it is natural to try to implement this kind of parallelism on parallel computer architecture. The Parallel Hybrid Ring Architecture (PAHRA), which is described in this article, provides flexible platform for simulation of multilayered feedforward neural networks trained with backpropagation algorithm. The computational model of given architecture, bound to the modified error backpropagation algorithm, allows to describe the formal elements of parallel implementation of multilayered feedforward neural network. It also allows the mathematical tool for verification of performance, which is used in simulation experiments of multilayered feedforward network on specific hardware platform.

DŽURINA, J., KOTOROVÁ, R.:
Zero Points of the Solutions of a Differential Equation [full paper]
Our aim in this paper is to count the zero points of the solutions of the second order differential equation
(r(t)u'(t))'+ p(t)u(t) = 0. (E+)
The sufficient conditions for the equation (E+) to be oscillatory are also presented.

PLAVKA, J., GAVALEC, M.:
Fast Algorithm for Extremal Biparametric Eigenproblem [full paper]
Please find the abstract in the attached paper

PEŤKO, M., KOVÁČ, D.:
PSPICE Program Model of Elastomagnetic Sensors of Pressed Force [full paper]
An article says about the basic magnetic features of ferromagnetic material, which need to be taken to the consideration when designing model of elastomagnetic sensor. Mentioned is an intelligent calculator, which is used for calculation of change of sensor permeability in cubic grate in dependence of the lay of vectors of magnetostriction and external field. In the article, the sensor model in program PSPICE is described. Model in program PSPICE simulates a real hysterezis loop of sensor, his loses, his dependency from the frequency and the entry signal. Realized is by passive elements and controlled current sources the way it could simulates real entry signal and also the shape of the hysterezis loop. In the conclusion of the article, the hysterezis loop of real sensor of a torque is showed. The basic requirements for the sensor model were simplicity, accuracy, comprehensibility, open system and further utility for similar types of sensors.

KANÁLIK, M., KOLCUN, M.:
Computation of Harmonic Flows in ThreePhase Systems [full paper]
This paper deals with computation of harmonic current flows and voltages in three phase systems using by a nodal method. The nodal method is useful to determinate unknown branch currents and nodal voltages by known source voltages or currents and nodal admittance matrix of the network. To obtain the correct solution it is necessary to correctly determinate the nodal admittance matrix, which contains admittances of threephase elements. This article is also describing the ways how to model some threephase elements for the purpose of nodal method analysis.

JADLOVSKÁ, A., KABAKOV, N., ZOLOTOVÁ, I.:
Predictive Control Algorithm Based on State Space Models of Dynamic Systems – Internet Approach [full paper]
The paper is focused on Predictive Control Algorithm design for timeinvariant state space models of dynamic systems with various dynamics, (mechanical systemhelicopter, hydraulical system  three tanks with interaction). The predictive control algorithm based on state space models is verified by simulation schemes in language Matlab/Simulink using architecture of Sfunctions of the library PredicLib. The results are presented by created Internet applications using the technology Matlab Web Server. Internet applications enable to simplify the usage of designed predictive algorithms for testing their properties at various dynamics of MIMO nonlinear systems without the proper installation language Matlab/Simulink on client’s PC.

KOLLÁR, J., FORGÁČ, M., PORUBÄN, J.:
Adaptiveness of Software Systems Using Reflection [full paper]
This paper deals with basic principles of metaprogramming and reflection with connection to aspectoriented programming (AOP). Metaprogramming is about writing programs that represents and manipulate other programs or themselves, i.e. metaprograms are programs about programs. The impact of metaprogramming is obvious in traditional development processes, by sorting existing programs as transformational processes with inputs and outputs. Open implementation and metalevel architectures are related to their reflective properties. Reflection is an entity’s integral ability to represent, operate on and otherwise deal with itself in the same way that it represents, operates on, and deal with its primary subject matter. Structural reflection represents the ability of a program to access a representation of its structure, as it is defined in the programming language. Behavioural reflection represents the ability of a program to access a dynamic representation of itself, that is to say, of the operational execution of the program as it is defined by the programming language implementation (processor). AOP allows development of required application using principle of separation of concerns. Reflection and AOP share many similarities in concepts, possibilities and applied techniques. There are several solutions to provide a reflective system among which belong following approaches: MetaclassTalk, Geppetto Reflex and Iguana/J. The former two are systems based on Smalltalk  Squeak; the latter two are based on Java.

NOVITZKÁ, V., SLODIČÁK, V.:
On Applying Stochastic Problems in HigherOrder Theories [full paper]
In the article presented we deal with one method of stochastic programming – probabilistic programming. It is a programming in which the probabilities of the values of the variables are of interest. Our approach is that solving problem we can construct in logical reasoning over some mathematical theories. In this approach we use category theory for construction of type theory and of the logical system. Then we formulate the logical theory to enclose the solution of the problem. These steps we show at wellknown examples: the random number generator and the Blackjack game.

VINCE, T., KOVÁČOVÁ, I.: Distance Control of Mechatronic Systems via Internet [full paper]
The article presents possibilities of distance remote via Internet. The Internet protocols divided by OSI model that can be used for controlling and regulations are discussed with final focusing on TCP and UDP protocols and comparing these protocols from the controlling point of view. The paper is analyzing advantages and disadvantage of using the Internet in different level of the existing levels in the information hierarchy and the delay time problem in communication via Internet. The item compares more buses used in distance remote. It contains also some distance remote control system proposal.

SEMANČÍK, P., CIMBALA, R., KOLCUNOVÁ, I.:
Dielectric Analysis of Natural Oils [full paper]
The study of the Power
flow is very significant, because it makes it possible to provide
information of large importance; in order to better plan, to control and
ensure finally the correct operation of the electrical supply networks.
In our article, we present a method of calculation of the flow of powers
in a mixed electrical supply network in high voltage (AC/DC). This
method called also method SPI: (Substituted Power Injection) makes a
small modification to the level of the bus of connection DC. The
calculation of the active and reactive powers at the final bus of system
ACDC, are obtained by the subtraction of the powers of the converters
and those which are specified with the bus of connection DC. This method
will be applied to the 14 bus networks.

TOMČÍKOVÁ, I., ŠPALDONOVÁ, D.:
Determination of Critical Atmospheric Electric Field around Franklin's Lightning Protection Rod that Leads to BreakDown [full paper]
In the presented paper we deal with the determination of magnetic field in the elastomagnetic sensor core. The elastomagnetic sensor core consists of the ferromagnetic plates which thickness is negligible to the rest of the core dimensions so it is sufficient to solve the magnetic field in one core plate. The dimension of the plate is designed in such a way that the plate can be divided into the squares a number of which is equal to a number of the holes in each plate. Each of the square is called as an integrative element of the sensor and the magnetic field in such one is solved.
The magnetic field in this integrative element is computed by using Partial Differential Equation Toolbox of MATLAB that provides effectual tool for the solution of partial differential equations in two space dimensions and time. The use of this toolbox requires only a basic level of the partial differential equation knowledge what is considered to be the most important advantage of this toolbox. The magnetic field in the integrative sensor element for the plane case can be expressed by the elliptic partial differential equation for the magnetic vector potential. The partial differential equation coefficients are magnetic permeability and the current density for the domain in which the magnetic vector potential field is solved. The domain consists of three subdomains so the coefficients in each of these regions have to be determined. For the subdomain that represents ferromagnet the functionality between the magnetic permeability and the magnetic flux density is needed because ferromagnet permeability is dependent on the field strength. The boundary conditions specifying for the outer boundaries and the interior ones is the necessary part of the PDE problem formulation too.
The toolbox provides execution the most of hard work such as the mesh generation, the partial differential equation transformation to the discrete form by the finite element method and the approximation to the solution itself. The results can be visualized in several ways. In our case the equipotential lines of magnetic vector potential in the integrative sensor element are selected.

