Browsing by Author "Hadjiski M."

Now showing 1 - 5 of 5
  • No Thumbnail Available
    Item
    Cost oriented redesign of condition-based maintenance of copper converting process
    (2013-01-01) Hadjiski M.; Boshnakov K.; Ginchev T.
    The presented investigation is focused on one important maintenance problem - tuyeres wear and their free section reduction solidification of melt. The results of experimental data show that this is nonstationary stochastic process. Legacy SCADA system provides data available only for the failure based maintenance. This results in a big losses in productivity of the metallurgical plant. In the present paper legacy hard instructions based maintenance system is re-designed in order to realize more effective condition based maintenance (CBM). Decision making process is based on adaptive forecasting and CBR procedure. The proposed re-engineering is cost oriented. The re-designed system gives competitive advantages, keeps the risk up to chosen settings, improves the operator's performance, and reduces the maintenance expenses. Copyright © 2013 IFAC.
  • No Thumbnail Available
    Item
    Diagnosis of metallurgical ladle refractory lining based on non-stationary on-line data processing
    (2013-08-29) Petkov V.; Hadjiski M.; Boshnakov K.; Mihailov E.
    A new method for diagnosis of the remaining thickness of metallurgical ladle refractory lining working in a batch mode is proposed, where directly measured by a thermovision camera data about the amplitude and phase delay of the maximum surface temperature is used. Regression equations are obtained to determine the remaining thickness of the ladle refractory lining depending on the parameters of steady auto oscillations of the maximum surface temperature. It is shown that the data from nonstationary auto oscillations could be used for adaptation of the regression equations, predicting the remaining thickness of the ladle refractory lining in case of a local damage.
  • No Thumbnail Available
    Item
    Modeling of the dynamic behaviour and energy consumption of the heating process of wooden houses
    (2006-01-01) Deliiski N.; Hadjiski M.; Boshnakov K.
    A summarized mathematical model has been created, solved, and verified for the transient non-linear heat conduction and energy consumption in wooden houses at arbitrary, encountered in the practice initial and boundary conditions. The paper presents the creation and the solution of the model, and the simulative investigation of the impact of the surrounding and room's medium parameters and the initial wood temperature on the change of the heat energy consumption, needed for reaching of different set temperatures in the rooms of the house. The obtained results are used in other papers of the authors for the creation of an algorithm and a system for optimized model predictive control of the energy consumption, required for heating of wooden houses. Copyright © 2006 IFAC.
  • No Thumbnail Available
    Item
    Multi agent intelligent control of centralized HVAC systems
    (2006-01-01) Hadjiski M.; Sgurev V.; Boishina V.
    The perspective of Multi Agent Systems (MAS) implementation in thermal control of centralized HVAC systems is considered. MAS are implemented in all functional levels of hierarchical HVAC control system: estimation of representative control variables by agent based softsensing, evaluation of Predicted Mean Vote (PMV) index using intelligent agents, local control applying agentbased Ant Colony Optimization (ACO). A non-square control problem of realization comfort is solved via ACO. Software agents are realized into the framework of JADE - plathform in JAVA. Simulation results are presented for single input-two output (SITO) control system. Copyright © 2006 IFAC.
  • No Thumbnail Available
    Item
    Rules based autotuning approach of conventional smith predictor
    (2004-01-01) Hadjiski M.; Ivanova E.
    The problem of automation of the system with large time delay take wide place in a series of contemporary researches. The main interest takes the situation when the value of system time delay considerable changes in time. In the present research is presented Smith predictor based retuning approach, as an algorithm for control of plants with large time delay, namely by such a system changes.