Table of Contents


    • Neural network approaches for a facility location problem

      pg(s) 3-6

      This paper examines the possibility to use neural networks for approximately solving the MiniSum problem, a classic facility location problem. For this we first create a set of realistic MiniSum instances, based on the Bulgarian road network. Two standard neural network approaches – Hopfield networks and Boltzmann machines, are then applied to the instances. Since the quality of solutions is not satisfactory, the reasons for the poor performance are discussed. An improved neural network approach is then proposed. This approach has excellent performance on the MiniSum instances. It always finds solutions just several percent worse than the optimum, and is often able to find the exact optimum.

    • Address sequence generator for memory BIST investigation

      pg(s) 7-8

      In this paper a method of construction of a generator for address sequences with given values of switching activity, based on the Idea of Antonov and Saleev and with use of Toeplitz and Hankel matrices, was proposed. The broad possibilities of this approach and the limitations associated with conflicting requirements for the values of the rows of the matrix and their linear independence are shown. Examples of the formation of standard address sequences used for self-testing of storage devices are described

    • Nagata patch interpolation for finite volume mechanical contact simulations

      pg(s) 10-12

      This paper describes a finite volume contact algorithm with surface smoothing using the Nagata patch interpolation. Nagata interpolation is derived from existing discretisation, using the mesh points positions and their calculated normal vectors. The contact between a deformable and a rigid body is analysed, whereas the rigid body is described with Nagata patch interpolation. Such approach allows a more accurate evaluation of the resulting contact stresses and forces


    • Modeling and simulation of forging processes

      pg(s) 13-17

      Forging is an experience-oriented technology. The physical phenomena that describe the forging operations are difficult to express with quantitative relationships. In order to avoid the trail-and-error method, we use numerical simulations for studying the forging process. With the help of these simulations, the engineers are able to uncover the potential defects which may happen during the forging process. Concurrent Engineering (CE) helps in making the forging process more effective. In the CE system, each modification of the product represents a taxonomical relationship between specifications, outputs, and the concept it represents. In the study, the forging process of a disc shaped part is analysed. Thanks to numerical simulations it is determined that the dimensions of the billet are larger than needed. This resulted in overfilling the flash of the tool, thus the simulation was unsuccessful. After correcting the dimensions of the billet, the simulation ran with no interruptions.

    • Buckling analysis of columns made of functionally graded materials via Rayleigh-Ritz method

      pg(s) 18-21

      This paper presents the buckling analysis of functionally graded (FG) beams. To solve diff erential buckling equations of diff erent FG sections the Rayleigh-Ritz method is used. The FG material is supposed to vary continuously over the beam depth according to power low. A set of analytical evaluation is run in order to calculate critical buckling loads in dependents of material power low index. The accuracy of method is confirmed comparing the results with the finite element ones.

    • Mechanical and hydraulical stability of the offshore composit structure

      pg(s) 22-25

      The aim of this research it was the installation of an intake sea water GRP (glass-reinforced plastic) structure in Vlore Bay. A detailed study about the extreme wave condition offshore was carried out by the SWAN (Simulating Waves Nearshore) model. Then the mechanical structure stability calculation was performed with AxisVM8, a finite element modeler & solver. Conservative hypothesis were taken, such as to use the max static pressure value as per energy balance calculation on the sea waves. The program Fluid Flow was used to model the pipe line end with pipe chimney connections. The scope of the calculation was primarily to balance fluid flows in each chimney in order to match the range of 0, 2 – 0, 25 m/s velocity at each chimney entrance. The result of the mechanical and hydraulically stability verification for the sea water intake structure take in consideration it was positive and this structure was installed successfully.

    • Optimization of pneumatic vane motor based on mathematical modeling and computer simulation

      pg(s) 26-30

      Pneumatic vane motors are important actuators in industry, their optimization is difficult without e best mathematic model and computer simulation. In this article we presents a mathematical model for pneumatic vane type motors, construction of a simulation program associated with constructive engine parameters, experimental confirmation , and pneumatic motor performance optimization based on this model. The article begins with the description of the working principle of the vane type pneumatic motors, the geometric parameters that characterize it. The set of mathematical model equations consists in: the equations of geometry, the equation that describe the rotating moment, and equations expressing the mass flow into the motor. For the construction of the simulation program we have used the programming language of the G type LabView. Matching the results obtained from the simulation with the measured ones experimentally indicate that the built mathematical model is accurate and can be used to optimize pneumatic motors.

    • Study the work of specialized car service as queue theory

      pg(s) 31-34

      Study the refusal of requests received in a specialized automotive service workshop in the city of Rousse was investigated. The present work analyses the average monthly requests from customers of the workshop and the number of the actual repairs is determined. The work in the service have been seen as a mass service system with a non-stationary mid-month incoming stream with queries. Under nonstationary conditions, the basic values of the system parameters were calculated and Mat Lab application was created. The proposed approach can serve as a methodology for analysing and optimizing the activity of other universal automotive service

    • Determination of empirical coefficients of heat dissipation characteristics by optimization methods for a spark-ignited car engine

      pg(s) 35-38

      For the development of new engines, as well as to obtain the correct results from mathematical modeling, numerical methods should be used, with the help of which verification of mathematical models takes place. This approach can significantly reduce the material and time costs at the development stage. The influence of the optimization method used in processing the experimental data of engine indexing to obtain semi-empirical coefficients of the form Wiebe is explored. A comparison is made of the rate of calculation of the coefficients of the Wiebe formula using brute force and gradient methods. In both methods, the integral using the modulus of the difference between squares of the experimental and calculated dependences is chosen as the target function. Analysis of the rate of calculation of the coefficients of the Wiebe formula showed the need to use the combined use of both methods: the global minimum of the objective function by the method of brute force with a large step and near the optimal point – the gradient method


    • On the distribution of the incentive fund at the university

      pg(s) 39-41

      The article highlights two approaches to solving the problem of the distribution of the incentive fund, one of which is based on the calculation of the cost of one point; the second one is based on elementary incentive payment in the system with weight indices. The latter approach makes it possible to single out a small number of clusters of successful personalities. Comparison of the considered approaches at the stage of their implementation allows us to draw the conclusion about the different aspirations of teachers in the performance assessment systems. A mathematical solution is proposed for the distribution of the incentive fund in the framework of the second approach. The proposed method allows dividing fund resources in an extremely short period of time.