Yikun Yang

Adaptive infinite impulse response system identification using opposition based hybrid coral reefs optimization algorithm

An efficient global adaptive algorithm is required to determine the parameters of infinite impulse response (IIR) filter owing to the error cost surface of adaptive IIR system identification problem being generally nonlinear and non-differentiable. In this paper, a new bio-inspired algorithm, called opposition based hybrid coral reefs optimization algorithm (OHCRO) is applied for the IIR system identification problem. Coral reefs optimization algorithm (CRO) is a novel global algorithm, which mimics the behaviors of corals’ reproduction and coral reef formation. OHCRO is a modified version of CRO, on the one hand utilizing opposition based learning to accelerate global convergence, on the other hand cooperating with rotational direction method to enhance the local search capability. In addition, the Laplace broadcast spawning and power mutation brooding operator are used to maintain the diversity. The simulation studies have been performed for the performance comparison of genetic algorithm, particle swarm optimization and its variants, differential evolution and its variants and the proposed OHCRO for well-known benchmark examples with same order and reduced order filters. Simulation results and comparative studies justify the efficacy of the OHCRO based system identification approach in terms of convergence speed, identified coefficients and fitness values. In conclusion, OHCRO is a promising method for adaptive IIR system identification.

Dynamic/static displacement sensor based on magnetoelectric composites

This study presents a dynamic/static displacement sensor based on both end fixed Terfenol-D/PZT (lead zirconate titanate) magnetoelectric composites. Owing to the fixed boundary condition, the magnetostriction of Terfenol-D under the magnetic field is limited and the stress acts on PZT through mechanical coupling, and finally, an induced voltage is generated due to the piezoelectric effect in PZT. The magnetic field applied on the magnetoelectric composites varies with the displacement of the permanent magnet which represents the displacement to be measured. In that case, the determination of displacement can be transformed into determining the variation of the magnetic field. The experimental results manifest that the frequency of magnetoelectric voltage is the same as the frequency of displacement. The dynamic displacement sensitivity of this sensor increases with the increase in the input frequency in the low frequency range and the decrease in the length of the air gap. The sensitivity is 6.549 mV/μm with a 0.75 mm air gap for dynamic displacement at 10 Hz and 0.84 μV/μm with a 2.5 mm air gap and an external magnetic field at 1 kHz for static displacement. It can be concluded that the displacement measuring mechanism based on the magnetoelectric effect is a promising robust and accurate method.This study presents a dynamic/static displacement sensor based on both end fixed Terfenol-D/PZT (lead zirconate titanate) magnetoelectric composites. Owing to the fixed boundary condition, the magnetostriction of Terfenol-D under the magnetic field is limited and the stress acts on PZT through mechanical coupling, and finally, an induced voltage is generated due to the piezoelectric effect in PZT. The magnetic field applied on the magnetoelectric composites varies with the displacement of the permanent magnet which represents the displacement to be measured. In that case, the determination of displacement can be transformed into determining the variation of the magnetic field. The experimental results manifest that the frequency of magnetoelectric voltage is the same as the frequency of displacement. The dynamic displacement sensitivity of this sensor increases with the increase in the input frequency in the low frequency range and the decrease in the length of the air gap. The sensitivity is 6.549 mV/μm ...