مريم خسروشاهي

This thesis proposes a Hurst exponent-based index for performance assessment of control processes that can eva‎luate the performance using neither any prior knowledge of system model parameters and delays nor the system optimal operating data. Instead, only system routine operating data are used for calculating this proposed index, and hence, this method is more suitable to be applied on various industrial control processes. The proposed index can perform similarly to the minimum variance index and somehow assesses the performance of the system with the minimum variance criterion. In this thesis, Detrended Fluctuation Analysis (DFA) and Rescaled Range Analysis (RRA) are used to estimate the Hurst exponent, which are powerful and successful methods for determining time series correlations. In estimating the Hurst exponent of different control signals, crossover phenomena are observed in the graphs used, which complicate deciding the Hurst exponents values. To address this problem, in this thesis, the crossover phenomenon is investigated for control signals; then, the Hurst index is modified based on it. Also, to better understand the performance of the Hurst index in control systems, it is necessary to examine the effect of system parameters and dynamics on the Hurst exponent of the signals, which is discussed in this thesis. The proposed method is used to assess the performance of various types of control systems, such as multivariable and cascade control systems. Simulation results illustrate the excellent efficiency and accuracy of the proposed method.

ارزيابي عملكرد كنترل , نماي هرست , پديده كراس‌اور , شاخص واريانس كمينه , آناليز نوسانات روندزدايي شده , آناليز دامنه بازمقياس شده

control performance assessment , Hurst exponent , crossover phenomenon , minimum variance index , detrended fluctuation analysis , rescaled range analysis

Maryam Khosroshahi

Dr. Javad Poshtan