对粒子群算法改进及在稀布阵方向图上的应用研究

发布时间：2018-06-02 06:38

本文选题：粒子群算法 + 稀布阵列天线方向图　；参考：《成都理工大学》2017年硕士论文

【摘要】：稀布阵列天线不仅在主瓣波束,旁瓣,零陷等方向图上有突出的优势,而且还能降低天线系统的建造成本及复杂度。研究阵列天线方向图的综合技术,旨在确定阵列天线的激励参数,使天线阵的某些辐射特性满足给定的指标要求,或者使阵列的辐射方向图尽可能地逼近所期望的方向图。在实际工程中,可能需要低旁瓣,窄波束,或者在某一指定位置具有深零点等方向图。基于此,研究更为高效的粒子群算法在稀布阵方向图上的应用,解决上述问题,具有重要的现实意义和应用价值。本文在深入探索和研究粒子群算法理论及特性的基础上,围绕粒子群算法在求解不同稀布阵列天线方向图问题时的需求,提出了几种改进粒子群算法的方法,并进行相关实验及数据分析。本文主要研究内容如下:(1)本文将粒子群算法与遗传算法相结合,提出了杂交粒子群算法。本文将遗传算法中杂交变异的特点引入到粒子群算法中,改变粒子种群的多样性,使粒子更容易跳出局部最优解,寻找全局最优解,同时也提高了搜索能力,改变了算法的性能。(2)本文将粒子群算法与模拟退火算法相结合,提出了退火粒子群算法。本文运用模拟退火算法的方法初始化粒子群,使粒子群算法初始种群能够均匀覆盖整个搜索空间,避免了传统初始化方法在解决高维空间优化问题时易于向边缘聚集的现象,有利于粒子群算法在高维空间中的寻优。同时将模拟退火思想引入到粒子群算法中,结合了粒子群算法的快速寻优能力和模拟退火的概率突跳特性,使算法可以跳出局部最优从而实现全局最优,达到更好的收敛精度。(3)为了使粒子群算法更有效地解决稀疏阵方向图问题,结合混沌算法的优点,本文提出了混沌粒子群算法。本文首先利用混沌序列初始化粒子的速度和位置,提高整个种群搜索的遍历性。其次,根据当前整个种群搜索到的最优位置产生混沌序列,将新产生的最优位置代替当前种群中的一个粒子的位置。引入混沌序列的搜索算法可在进化过程中产生局部最优解的许多邻域点,以此帮助惰性粒子逃离局部极小点,并快速搜寻到最优解,改善算法的搜索能力。(4)本文采用改进后的粒子群算法,分别求解不同的稀布阵列天线方向图问题。首先,本文将杂交粒子群算法应用到稀布阵旁瓣方向图中,为了验证该算法的性能,将该算法用于两个典型的稀布阵优化布阵设计中,并将求解结果同粒子群算法和遗传算法的最优解进行比较,得到该方法的求解精度和速度都优于粒子群算法和遗传算法。其次,本文将退火粒子群算法应用到稀布阵零陷方向图中,设计在某一指定位置有深零点的稀布阵,经过算法的迭代优化,得到比较好的阵元分布,通过与其他算法优化过后的结果进行对比分析,体现了退火粒子群算法的优点。最后,本文将混沌粒子群算法应用到稀疏阵方向图中,运用混沌粒子群算法设计不同的稀疏直线阵,并与已有文献结果进行比较,显示了混沌粒子群算法在求解此类问题的有效性。
[Abstract]:The sparse array antenna has a prominent advantage not only in the direction of the main lobe, side lobe and zero sink, but also in reducing the construction cost and complexity of the antenna system. The comprehensive technique of the antenna array is studied to determine the excitation parameters of the array antenna so that some radiation characteristics of the antenna array can meet the given index requirements or make the antenna array meet the requirements of the given index. The radiation pattern of the array is as close to the desired direction as possible. In practical engineering, it may require low side lobe, narrow beam, or a deep zero point in a certain position. Based on this, it is of great practical significance to study the more efficient particle swarm optimization in the dilute array direction and solve the above problems. On the basis of the deep exploration and study of the theory and characteristics of particle swarm optimization, this paper puts forward several methods to improve particle swarm optimization, and carries out related experiments and data analysis. The main contents of this paper are as follows: (1) in this paper, the main contents of this paper are as follows: (1) the particle swarm optimization (PSO) Hybrid particle swarm optimization (PSO) is proposed in this paper. In this paper, hybrid particle swarm optimization (PSO) is proposed in this paper. In this paper, the characteristics of hybrid mutation in the genetic algorithm are introduced into particle swarm optimization (PSO), and the diversity of the particle population is changed to make the particle more easily jump out of the local optimal solution and find the global optimal solution. At the same time, the search capability is improved and the performance of the algorithm is changed. (2) (2) The particle swarm algorithm is combined with simulated annealing algorithm, and the annealing particle swarm optimization algorithm is proposed. This paper uses the simulated annealing algorithm to initialize the particle swarm, so that the initial population of the particle swarm optimization algorithm can cover the whole search space evenly, and avoids the traditional initialization method which can easily aggregate to the edge when it solves the problem of high dimensional space optimization. It is beneficial to the optimization of particle swarm optimization in high dimensional space. At the same time, the simulated annealing idea is introduced into particle swarm optimization (PSO), the fast optimization ability of particle swarm optimization and the probability jump characteristic of simulated annealing is combined, so that the algorithm can jump out of the local optimal and achieve the best global optimization. (3) in order to make the particle the particle swarm optimization, the particle swarm optimization algorithm can achieve better convergence accuracy. Group algorithm is more effective to solve the problem of sparse array pattern. Combining the advantages of chaos algorithm, this paper proposes a chaotic particle swarm optimization algorithm. Firstly, the velocity and location of particles are initialized by chaotic sequence, and the ergodicity of the whole population search is improved. Secondly, the chaotic sequence is generated according to the optimal location of the current whole species group, and the new chaotic sequence will be generated. The optimal position is replaced by the position of a particle in the current population. The search algorithm introduced into the chaotic sequence can generate a number of neighborhood points of the local optimal solution in the evolutionary process, in order to help the inert particles escape from the local minima, and quickly search for the optimal solution and improve the searching ability of the algorithm. (4) the improved particle swarm is adopted in this paper. In order to verify the performance of the algorithm, the hybrid particle swarm optimization (PSO) algorithm is applied to the sparse array sidelobe pattern. In order to verify the performance of the algorithm, the algorithm is applied to the design of two typical sparse array arrays, and the solution results are in the same way as the optimal solution of particle swarm optimization and genetic algorithm. In comparison, the accuracy and speed of the method are better than particle swarm optimization and genetic algorithm. Secondly, this paper applies the annealing particle swarm algorithm to the sparse array zero subsidence direction map, designs the sparse array with deep zero in a certain position, and through the iterative optimization of the algorithm, the better array element distribution is obtained, through which the algorithm is superior to other algorithms. The results are compared and analyzed, which embodies the advantages of the annealing particle swarm optimization. Finally, the chaotic particle swarm optimization algorithm is applied to the sparse array direction map, and the chaotic particle swarm algorithm is used to design different sparse linear arrays, and compared with the existing literature results, the chaotic particle swarm optimization algorithm is shown to solve such problems. Efficiency.
【学位授予单位】：成都理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;TN820

【参考文献】