Yosza Dasril

An alternative hybrid search direction for unconstrained optimization

Abstract Quasi-Newton is one of the most popular iterative method for solving unconstrained optimization and recently, there are several researchers who work on hybrid quasi-Newton and steepest descent method and this method namely QN-SD. In this paper, we proposed an alternative search direction for this method and we proved its global convergence. Several standard optimization problems had been tested on this method and the result analysis was displayed at the end of this paper.

Investigation of symmetrical split ring resonator (SSRR) couplings for high sensitivity material characterization

In this paper an investigation of enhanced coupled methods based on planar symmetrical split ring resonator (SSRR) is presented for characterizing the properties of materials. The designed SSRR sensor demonstrates significant improvement in resonant frequency and reflection coefficients when the coupling gap between the feedlines and rings of the sensors is enhanced. The simulated results show that the SSRR achieves narrow resonance with higher Q value of 407. The SSRR sensor can be used for bio -sensing, food industry, quality control, and medicine and pharmacy applications. Experimentally, there is a good agreement in the results of simulation and measurement with possibility of 1.4 % typical error in shifting of the resonance frequency.

Modified Artificial Bees Colony algorithm with Nelder-Mead search algorithm

In this paper, a new Artificial Bee Colony (ABC) called NM-ABC which integrates NM into original ABC has been developed. ABC algorithm is one of the most recent biological-inspired optimization algorithm proposed in year 2005. ABC is inspired by the foraging behavior of honey bee swarm. The exploration and exploitation are two important mechanisms in ABC. Exploitation process starts when the employed bees approach to the food sources. However, ABC can perform better during exploration stage but weaker at exploitation stage. In this study, the exploitation search stage has been integrated with Nelder-Mead Simplex Method (NM). The numerical results shown that the NM-ABC can perform better than the original ABC with less colony size and number of cycle allocated

Analysis of Matching Circuit to Improve the Efficiency of RF to DC Conversion for Ambiet RF Energy Harvesting

This paper presents an analysis of matching circuit with rectifier for RF energy harvesting. The matching circuit has a significant influence to improve the efficiency of RF-DC conversion. In this paper, comparative analysis of rectifier efficiency is performed, particularly the rectifier with-and without-matching circuit. The analysis is conducted using ADS simulation version 2011, to determine the output power as well as the efficiency of the rectifier. The results show the rectifier with-matching circuit produced higher efficiency about 72%, compared to the rectifier without-matching circuit which is only achieved 20% of the efficiency. The effects of tuning the matching circuit upon the rectifier efficiency had been studied. In this paper, the parametric studies of the effect of short circuit and open circuit stubs tuning are performed to determine the efficiency of the rectifier. The results show matching circuit after tuning produced higher efficiency compared to matching circuit before tuning. This study is useful to provide an insight on the optimum technique to design RF energy harvesting.