种子优化算法及其在动态优化问题求解中的研究

There are many Dynamic Optimization Problems(DOPs) in real-world applications. Studying and solving these DOPs have realistic significance. Inspired by the transmission mode of seeds, a novel evolutionary algorithm named Bean Optimization Algorithm (BOA) is proposed, which can be used to solve complex optimization problems by simulating the adaptive phenomenon of plants in the nature. BOA is the combination of nature evolutionary tactic and limited random search. BOA has stable robust behavior on explored tests and stands out as a promising alternative to existing optimization methods for engineering designs or applications. This project will continue to research on the Bean Optimization Algorithm. We will focus on the construction of population distribution evolution model by learning from the natural discipline of biological population distribution evolution. The relationship between the distribution evolution model and the performance of BOA algorithm will be established. Parameters and convergence greatly influence the performance and efficiency of BOA. A chaos bean optimization algorithm algorithm (CBOA) is introduced to overcome the problem of premature convergence. CBOA uses the properties of ergodicity, stochastic property, and regularity of chaos to improve the population initialization and algorithm optimization process. Focusing on Dynamic Optimization Problems, we will firstly analyze the characteristics of DOPs and Bean Optimization Algorithm. Then selection and improvement will be made on the population distribution evolution models. A kind of collaborative computing framework based on multi-distribution models will be built to improve the capability of tracking down the problem optimal solutions or Pareto optimal solution. Finally, algorithm experiments will be carried out by solving some typical single-objective and multi-objective dynamic optimization problems. We will also analyze the effectiveness of BOA based on the theoretical proof and experiments. One of the purposes of the research is to apply BOA to solve problems in the real world as soon as possible. We also hope that our idea can inspire others to solve DOPs better.

现实世界中存在大量的动态优化问题，研究动态优化问题求解具有重要的现实意义。种子优化算法是本申请人受自然界种子传播方式启发提出的一种群体智能优化算法，它通过模拟植物生存的宏观自适应现象，来解决优化计算问题。在前期的研究中，该算法已经表现出了优秀的寻优性能。本课题将继续深入研究种子优化算法，借鉴生物种群分布演化规律，重点完善种群分布演化模型的构建，明确分布演化模型与算法性能的关系；开展算法与混沌系统的结合研究，构建基于混沌思想的种子优化算法，改善种群初始化和算法寻优过程，进一步提高算法性能；重点针对动态优化问题的求解，在分析动态优化问题和种子优化算法特点的基础上，对种群分布演化模型进行选择和改进，并构建基于多分布模型的种群协同计算框架，提高算法跟踪问题最优解或Pareto最优解的能力，对典型的单目标和多目标动态优化问题进行求解实验，分析算法的有效性，尽快实现种子优化算法在真实环境中的应用。

本课题延续之前的研究，对种子优化算法开展了后续内容的研究，借鉴了生物种群分布演化规律，重点完善了种群分布演化模型的构建，构建了性能更好的组合分布模型，进一步明确了分布演化模型与算法性能的关系；开展了算法与混沌系统的结合研究，构建了基于混沌思想的种子优化算法，改善了种群初始化和算法寻优过程，进一步提高了算法性能，并开展了算法对比实验和分析；重点针对动态优化问题的求解，在分析动态优化问题和种子优化算法特点的基础上，对种群分布演化模型进行了选择和改进，并构建了基于多分布模型的种群协同计算框架，提高算法跟踪问题最优解或Pareto最优解的能力，对典型的单目标和多目标动态优化问题进行了求解实验，分析了算法的有效性，为算法今后的应用奠定了理论基础。