Invasive Weed Optimization Algorithm For The Synthesis Of Antenna Arrays | Posted on:2016-08-14 | Degree:Doctor | Type:Dissertation | Country:China | Candidate:Y Liu | Full Text:PDF | GTID:1108330464962884 | Subject:Electromagnetic field and microwave technology | Abstract/Summary: | PDF Full Text Request | Pattern synthesis of array antennas refers to making radiation pattern approach the desired pattern as far as possible through determination of several parameters of antenna arrays.Compared to most traditional evolutionary algorithms, Invasive Weed Optimization(IWO)algorithm performs strong robust and adaptation, can effectively overcome the premature convergence because of its unique evolutional mechanism. IWO algorithm has been successfully used in the area of antenna design. Therefore, study of more effective modified IWO algorithms are of great theoretical significance and potential practical value.This paper aims to explore the theories and features of IWO algorithm, to design modified IWO algorithms for different synthesis problems, and to do the corresponding numerical experimental analyses. The main research works in this dissertation consist of the following aspects:1. Adaptive Invasive Weed Optimization(AIWO) algorithm is designed to pattern synthesis of array antennas. An adaptive standard deviation is designed to improve the performance of IWO algorithm, which changes not only with the iterations but also with fitness function value of each individual in the generation. The design of adaptive standard deviation improves evolutionary speed and accuracy of the original IWO algorithm, which makes better balance between global and local searching capabilities effectively at the same time.By synthesizing of antenna arrays with sidelobe reduction, nulls and notches, the advantages of AIWO algorithm over original IWO algorithm and other evolutionary algorithms are shown.2. By combining IWO algorithm and local search operator, a hybrid algorithm called IWOSQA algorithm is proposed and used in pattern synthesis of sidelobe reduction, mainlobe shaped and phase-only pattern reconfigurable arrays. In the hybrid algorithm, AIWO algorithm is used as global algorithm, Simplified Quadratic Approximation(SQA) is embedded into IWO algorithm as a local search operator to enhance the local searching capability of the algorithm, to better balance the global and local searching capabilities, and make the algorithm converge to the optimal solution quickly and effectively. Simulated results for six benchmark functions show that IWO-SQA algorithm performs better than original IWO algorithm in convergence rate and accuracy. To test and verify the effectiveness of IWO-SQA algorithm in antenna design, the algorithm is used to synthesis of different antenna arrays, and obtains better results than exciting literatures.3. A Modified IWO(MIWO)algorithm is proposed for synthesizing of thinned arrays.Iterative Fourier Technique(IFT) is executed repeatedly to produce a set of solutions as the initial population of AIWO algorithm, in which more excellent individuals are included,which guide the algorithm to find optimal solutions quickly. The MIWO algorithm can improve convergence speed and accuracy at the same time. Through synthesizing of antenna arrays with a variety of aperture sizes and fill factors of thinned linear arrays and planer arrays, comparisons are made with exciting literatures to verify the effectiveness of MIWO algorithm in solving such problems.4. Synthesis of antenna arrays is formulated as a multiobjective optimization problem and the AIWO algorithm is improved to integrated in the framework of Multi-objective Evolutionary Algorithm Based on Decomposition(MOEA/D), to propose Multiobjective Invasive Weed Optimization Based on Decomposition(MOEA/D-IW O) algorithm. The new algorithm makes good use of the powerful searching and colonizing ability of invasive weeds and maintains the advantages of original MOEA/D. When using MOEA/D-IWO in pattern synthesis of low sidelobe beam with nulls or notches and phase-only pattern reconfigurable arrays, results obtained by MOEA/D-IWO algorithm are better than these obtained by exciting MOEA/D-DE algorithm. Through analyzing of important metrics, the superiority of MOEA/D-IWO algorithm is confirmed; solutions obtained by MOEA/D-IWO algorithm have better overall performance with higher accuracy, convergence rate and diversity. | Keywords/Search Tags: | Invasive Weed Optimization, Antenna Arrays, Pattern Synthesis, Thinned Linear Arrays, Thinned Planer Arrays, Pattern Reconfigurable Arrays, Multiobjective Optimization, Simplified Quadratic Approximation | PDF Full Text Request | Related items |
| |
|