Efficient Global Optimization (EGO) is a competent evolutionary algorithm suited for problems with limited design parameters and expensive cost functions [1]. Many electromagnetics problems, including some antenna designs, fall into this class, as complex electromagnetics simulations can take substantial computational effort. This makes simple evolutionary algorithms such as genetic algorithms or particle swarms very time-consuming for design optimization, as many iterations of large populations are usually required. When physical experiments are necessary to perform tradeoffs or determine effects which may not be simulated, use of these algorithms is simply not practical at all due to the large numbers of measurements required. In this paper we first present a brief introduction to the EGO algorithm. We then present the parasitic superdirective two-element array design problem and results obtained by applying EGO to obtain the optimal element separation and operating frequency to maximize the array directivity. We compare these results to both the optimal solution and results obtained by performing a similar optimization using the Nelder-Mead downhill simplex method. Our results indicate that, unlike the Nelder-Mead algorithm, the EGO algorithm did not become stuck in local minima but rather found the area of the correct global minimum. However, our implementation did not always drill down into the precise minimum and the addition of a local search technique seems to be indicated.
IntroductionLarge phase scanned arrays have instantaneous bandwidth limited by beam displacement (squint) as frequency is varied, and that displacement can only be eliminated by introducing time delay at the subarray level. Unfortunately the use of rectangular subarrays creates a periodic grid with relatively large spacings between subarrays, and this causes the radiation of large quantization lobes. Our recent studies have shown that aperiodic grids made of polyomino subarrays are free of the quantization lobes of rectangular subarray grids and do not introduce any significant pattern distortion or additional loss. Figure 1 shows an array of 64 x 64 (4096) elements, using eight element polyomino-shaped subarrays. The basic L-octomino subarray shown in the figure is rotated and/or flipped to form eight different shapes, but needing only two basic power divider configurations. Throughout this discussion, the array has center frequency at f o , and all the data was computed at 1.3 f o , so the array is meant to operate over 60% bandwidth at u o =v o =0.5 (a 45 degree conical scan area). Figure 2 shows the projection of the 3-dimensional plots of the radiated patterns from an array of 64 x 64 (4096) elements using eight element rectangular subarrays as compared to those for the irregular subarrays. The use of irregular subarrays eliminated the well-defined quantization lobes and reduced the peak sidelobe level from approximately -11.4 to -26.3, a reduction of 14.9 dB.In our studies we have imposed several limitations that are necessary for the arrays to be practical. One is that we require the subarrays to have 2 n elements (ie 2,4,,8…) to avoid power divider loss. The second is that, for ease of fabrication, we use only one basic type of subarray per array, and then rotate that subarray through 360° by 90° increments. This last restriction gave rise to the concern that the duplication of shapes within any aperture would lead to some large residual sidelobes. Our published results [1] have shown that if the arrays are assembled with "randomly" rotated polyominos, then large peak sidelobes are not present. Still we have suspected that at some design sidelobe level this problem may still become an issue. 129 1-4244-0878-4/07/$20.00 ©2007 IEEEThe question remains: Is there some lower bound on the peak sidelobe level, and if so, what is it? The present study has been undertaken to investigate the behavior of arrays of irregularly shaped polyominos as a function of array size. In this way we expect to be able to make estimates of bandwidth, gain, peak, and average sidelobes for larger arrays. Array GeometryAlthough the presentation may include some data for an even larger array, the study described here entailed using a single array of L-shaped octominos. To investigate the size dependence issues, we looked at this single array of 64 x 64 elements, and cut off sections of the array to make five different arrays with progressively larger dimensions, as shown in Figure 3. These arrays were separately excited with -40 dB T...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.