Yazar "Yilmaz, Bayram Melih" seçeneğine göre listele

Listeleniyor 1 - 2 / 2
  • Küçük Resim Yok
    Öğe
    Luenberger based Observer Design in Time Delay Systems
    (IEEE, 2016) Yilmaz, Bayram Melih; Canevi, Mehmet; Dilaver, Kamil Fatih
    Observers are necessary to use in case of an urge to control a linear single input-single output system with feedback control or in a necessity to measure states of the system which are not avaliable for measurement or the high cost of the state measurements. Even though simple electrical system states are not effected very much from time delay, the mechanical systems and electrical transmission lines are more effected. To observe the states of a time delay system a time delay observer is necesarry. In this paper we have proposed a new observer for time delay systems via linearizing the delay term in the Laplace domain and we have compared the results with previous observer designs with simulations.
  • Küçük Resim Yok
    Öğe
    Proportional-Derivative Observer Design via BMI Approach using BARON Optimization
    (IEEE, 2016) Canevi, Mehmet; Yilmaz, Bayram Melih; Dilaver, Kamil Fatih
    This work begins with LMI formulations of Proportional and Derivative observer designs and continues with derivation of the error dynamics of the Proportional-Derivate observer. Proportional observer and Derivative observer are combined to acquire the Proportional-Derivative observer. The observers which are combined can be designed via LMI methods but the new observer can only be designed via BMI formulations. A substitution is used for a term in the error dynamics in order to obtain the BMI formulation of the proposed observer. The performance is optimized via gain minimization, decay rate maximization and L-2 gain minimization. BARON is used to solve the BMI problem to design a PD observer for the given example. BARON is a software that solves nonconvex optimization problems for global optimality via Branch and Bound optimization algorithm. A comparision of three different designs are given to verify that the optimization criteria are valid.