Ivanov Y. Task scheduling optimization in real-time simulators

The subject of research:computational process in real-time software in the loop multifrequency simulation of dynamic systems. The aim - increase efficiency of digital processing in real-time simulators for complex dynamical multifrequency objects simulation. New approach for scheduling dynamic systems processing was proposed based on scheduling theory and features the problem under consideration. All necessary conditions that describe its feasibility were determined. In the thesis real-time computational period optimization problems were proposed. Author set tasks of synthesis scheduling strategies for common multicore system. For systems with general digital computing unit new approach was proposed to determine optimal schedule parameters using branch and bound method. Also in thesis a corresponding algorithm was developed. Artificial bee colony algorithm was modified and applied for optimization multicore system model. In thesis system for simulation control was developed for simulation multifequency aerodynamic problem - mine ventilation network, using multicore processor and real-time extension for Windows. The investigations showed significant increase effectiveness through reduce CPU time for computing methane equation in 4,5 times. Sphere of application:gas transmission and aerodynamic systems, avionics.