Abstract
Artificial intelligence systems are one of the important machines in performing operations that are difficult to perform traditionally. Optimization is one of the difficult and delicate processes that AI can be used to accomplish, especially if the optimizations are too small for antennas like microstrip patch antenna. A Microstrip patch antenna is considered one of the most widely used antennas that vary from lightweight wireless devices to airplanes and airspaces applications. One of the most attractive points about those antennas is their lightweight, small size, and ease of fabrication process. Although this antenna has many advantages, it suffers from some drawbacks like low gain and limited bandwidth. In this paper, we are presenting an optimization process by using the Nelder-Mead algorithm to achieve a new design of patch antenna that offers a broader bandwidth and higher gain. This design is achieved by optimizing the dimensions of the width and the frequency of the antenna. The results show that this device is responding perfectly at 1.471GHz and the ranges of substrate dimensions and relative permittivity affect the device performance and behavior.
Highlights
The Nelder–Mead approach is a widely applied computational method of finding the minimum or maximum optimization problem in a multidimensional space[1]
Optimization consists of locating the ranges of real functions that describe the current state of the device in this case to obtain the best potential approximation of antenna gain for such a given frequency[3]
The optimal solution is the n-dimensional real function of the real variables labeled with f (x1, ..., xn)[4, 3]
Summary
The Nelder–Mead approach is a widely applied computational method of finding the minimum or maximum optimization problem in a multidimensional space[1]. Microstrip antennas are fabricated on a printed circuit board (PCB), (Figure 1). When the reflected points are best as the second point worse, worst as the best, f(x1) ≤ f(xr) < f(x1), later on, find a newer simplex by changing the not better points (xn+1) by the mirror points (xr) return step. Find the new simplex by switching the not best point xn+1by the back points xr and return step 1. F(xc) < f(xn+1), that find a newer simplex by switching not better points (xn+1) by a contract point xc return step1; 6. Since the simplex xn+1represents the vertex with a high associated value between vertex, an exception can be done for finding the lowest value at xn+1 in the other side which formed by each vertex xiexcept xn+1 [13]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
