javad taghipour

Nonlinear localization approaches are used not only for detecting the exact location ofthe nonlinear elements in mechanical structures, but they are also exploited in order to find any possible flaws such as cracks in Structural Health Monitoring (SHM) applications. This study aims to develop a localization method to determine the location of localized nonlinearities in dynamic structures utilizing the experimentally measured data obtained from the base excitation test. The nonlinear element in the experimental set-up is represented by a pair of permanent magnets placed on both sides on the free end of the cantilever, and a pair of electromagnets placed with equal distances on both sides of thepermanent magnets. The combination of permanent and electromagnets create andapply nonlinear electromagnetic force on the free end of the cantilever beam.Hence, stepped-sine vibration tests are carried out using constant accelerationbase excitation to measure the response of the nonlinear system. The linearresponse of the system obtained from the low amplitude test is used to updatethe Finite Element (FE) model of the underlying linear system of the structure.Then, the developed approach utilizes the updated linear model along with themeasured nonlinear dynamics of the experimental set-up obtained usinghigh-amplitude excitation to determine the location of nonlinearity. The results of the experimental study are demonstrated to show the performance ofthe presented method.

Energy detector is one of the non-coherent receivers for impulse radio ultra-wideband (IR-UWB) systems using on-off keying (OOK) modulation. Energy detection for IR-UWB OOK systems gets accomplished based on a threshold value. In these systems, the threshold value depends on several parameters that should be determined in the receiver. In this paper, we propose a blind approach to estimate proper thresholds for the symbol decision. The performance of the proposed method turns out to be asymptotically comparable to that of the optimal threshold selection. The proposed algorithm was also improved using an optimum integration interval. By using a blind estimating of the channel time delay spread, the integration interval was adjusted. Simulation results show that this method improves the performance about 1.5 dB in BER of 10-4.