论文部分内容阅读
为了提高细菌觅食优化(BFO)算法的收敛速度,降低它的计算复杂性,从结构设计的新思路提出了结构重组的细菌觅食优化(SRBFC)算法。借鉴粒子群算法(PSO)中单循环结构的优点,对细菌觅食算法的结构重新设计,采用复制频数、驱散-死亡频数、趋化次数三个参数来判断复制、驱散-死亡操作的进行,在不违背原始算法的基本思想的前提下,将嵌套循环结构简化为新的执行结构。为了验证该算法的有效性以及可行性,首先将SRBFO算法与BFO算法在四个常见的测试函数上进行仿真,结果表明SRBFO算法既保持了原算法的优点,又能大大降低计算时间。接着将SRBFO算法应用于求解带有交易费用和不允许卖空的投资组合优化问题。考虑了三种对风险态度的投资者,选择了较为困难的8个资产的投资问题进行仿真,并将结果与PSO和SPSO算法对比,再一次证明了SRBFO算法在寻找最优投资组合的有效性以及可行性。
In order to improve the convergence rate of BFO algorithm and reduce its computational complexity, a structural recombination bacterial search optimization (SRBFC) algorithm was proposed from the perspective of structural design. Referring to the advantages of single-cycle structure in Particle Swarm Optimization (PSO), the structure of bacterial foraging algorithm was redesigned. The replication, disperse-death frequency and chemotaxis frequency were used to judge the replication, disperse-death operations Under the premise of not violating the basic idea of the original algorithm, the nested loop structure is simplified as a new execution structure. In order to verify the effectiveness and feasibility of the algorithm, SRBFO algorithm and BFO algorithm are simulated on four common test functions. The results show that the SRBFO algorithm not only retains the advantages of the original algorithm, but also greatly reduces the computation time. The SRBFO algorithm is then applied to solve the portfolio optimization problem with transaction costs and no short sales. Three investors who consider the attitude of risk choose the more difficult investment problem of eight assets to simulate, and compare the result with the PSO and SPSO algorithms, proving again the effectiveness of the SRBFO algorithm in finding the optimal investment portfolio And feasibility.