【作者】 王琳；

【导师】 施荣华；

【作者基本信息】 中南大学 ， 工程（专业学位）， 2022， 硕士

【摘要】 现实世界中存在的许多复杂的工程问题和科学研究都可以抽象为优化问题。传统的优化方法在应用于复杂和困难的优化问题时具有严重的局限性,多年来,随着大量基于人工智能、生物群群社会性或自然现象规律的算法的发展,优化算法对于众多优化应用至关重要。其中,蜻蜓算法是一种以蜻蜓飞行行为为基础的新型群智能优化算法,由于其参数简单及易于实现的特点受到了许多研究者的青睐。不过,目前相关的研究有限,算法的优化能力尚待提高。因此,本文重点研究如何提高蜻蜓算法的求解精度及收敛速度,增强算法的全局搜索能力,并最终用来实现天线实例优化。本文主要研究内容如下:1、本文针对传统二进制算法使用传递函数所产生的搜索空间离散化、精度损失和维数诅咒等问题,根据蜻蜓算法的特点,提出了一个改进的角度调制蜻蜓算法(IAMDA),用于改善算法求解精度及加快算法收敛速度。在所提改进的角度调制蜻蜓算法中,引入了角度调制机制,将复杂的二进制优化问题转换为更简单的连续问题优化,且为了进一步提高算法性能,从角度调制机制中的生成函数入手,引入了新的系数控制生成函数扰乱程度,有益于改善算法的全局寻优能力及算法稳定性。然后使用基准函数和0-1背包问题进行测试,并用统计学检验进行了验证,数值结果显示IAMDA表现优越。此外,利用IAMDA进行平面单极天线的拓扑优化设计,使得在指定频段内回波损耗值符合设计要求,仿真实例验证了IAMDA在天线拓扑优化问题上的优越性。2、本文提出了一种基于竞争机制的多目标蜻蜓算法(CMODA)。该算法通过在多目标蜻蜓算法中引入精英竞争机制,从而挑选出蜻蜓种群中有潜力的蜻蜓个体,引导群体更新,加强精英个体的优势,增加蜻蜓种群多样性,有益于在提升算法求解精度的同时,保证解的良好分布性。利用CMODA在多目标测试问题ZDT及DTLZ上进行测试,并与另外三个多目标算法进行比较以验证算法性能,实验结果表明所提算法在解决多目标问题上的有效性。此外,将CMODA用于双频高增益平面单极天线的多目标优化设计,获得的天线结构不仅满足在指定频段内的回波损耗值要求,在这些频段上也获得高增益,并且计算成本也较小,验证了CMODA在多目标天线拓扑优化上的有效性。图30幅,表18个,参考文献117篇更多还原

【Abstract】 Many complex engineering problems and scientific research existing in the real world can be abstracted into optimization problems.Traditional optimization methods have serious limitations when applied to complex and difficult optimization problems.Over the years,with the development of a large number of algorithms based on artificial intelligence,biota sociality,or the laws of natural phenomena,optimization algorithms have become critical for many optimization applications.Among them,the dragonfly algorithm is a new swarm intelligence optimization algorithm based on the flight behavior of dragonflies,which has been favored by many researchers due to its simple parameters and easy implementation.However,the current related research is limited,and the optimization capability of the algorithm has yet to be improved.Therefore,this thesis focuses on how to improve the solution accuracy and convergence speed of the dragonfly algorithm,enhance the global search ability of the algorithm,and finally use it to achieve antenna instance optimization.The main research contents of this thesis are as follows:Firstly,this thesis addresses the problems of search space discretization,accuracy loss and the curse of dimensionality generated by the traditional binary algorithm using the transfer function,and proposes an improved angle modulated dragonfly algorithm(IAMDA)according to the characteristics of the dragonfly algorithm,which is used to improve the solution accuracy and speed up the convergence speed.In the proposed improved angle modulated dragonfly algorithm,an angle modulation mechanism is introduced to convert the complex binary optimization problem into a simpler continuous problem optimization.In order to further improve the performance of the algorithm,a new coefficient is introduced to control the degree of disturbance of the generating function in the angle modulation mechanism,which is beneficial to improve the global optimization ability and the stability of the algorithm.Then the performance is tested using the benchmark function and the 0-1 knapsack problems,and is validated with statistical tests,the numerical results show the superior performance of IAMDA.In addition,the topology optimization design of planar monopole antenna is carried out by IAMDA,in order to make the return loss value meets the design requirements in the specified frequency band,and the simulation example verifies the superiority of IAMDA in the antenna topology optimization problem.Secondly,this thesis proposes a multi-objective dragonfly algorithm based on competition mechanism(CMODA).By introducing an elite competition mechanism into the multi-objective dragonfly algorithm,so as to select potential dragonfly individuals in the dragonfly population,guide the population renewal,strengthen the advantages of elite individuals,and increase the diversity of the dragonfly population,which is beneficial to ensure a good distribution of solutions while improving the algorithm solution accuracy.CMODA is tested on the multi-objective test problems ZDT and DTLZ,and compared with three other multi-objective algorithms to verify the performance of the algorithm.The experimental results show the effectiveness of the proposed algorithm in solving multi-objective problems.In addition,CMODA is used for the multi-objective optimization design of a dual-band high-gain planar monopole antenna,the the obtained antenna structure not only satisfies the requirement of return loss value in the specified frequency bands,but also obtains high gain in these bands.It is also less computationally expensive,which verifies the effectiveness of CMODA in multi-objective antenna topology optimization.更多还原