论文部分内容阅读
进化计算实现时选用的进化算子不同、其计算的时间复杂度和寻优能力也不同。为了对进化计算优化效率进行定量评价,本文提出“优化平均截止时间”和“优化截止时间分布熵”两个概念,并以此作为评价准则。通过选用有代表性的基准测试函数,应用本文提出的评价准则对实数编码中典型的交叉算子进行大量测试,评价结果与理论分析完全一致,不仅证实本文提出的评价准则是正确有效的,而且为科学合理地选择算子及其算法提供理论依据。同时本文构造的算法可直接用于求解工程优化问题。
Evolutionary computation is different when the evolutionary operator is used, and its computational time complexity and optimization ability are also different. In order to quantitatively evaluate the optimization efficiency of evolutionary computation, this paper proposes two concepts of “optimizing the average cut-off time” and “optimizing the entropy of distribution of deadline” as evaluation criteria. By selecting representative benchmark functions and using the evaluation criteria proposed in this paper, a large number of typical crossover operators in real numbers are tested. The evaluation results are in good agreement with the theoretical ones. This not only confirms that the evaluation criteria proposed in this paper are correct and valid, It provides a theoretical basis for selecting operators and their algorithms scientifically and rationally. At the same time, the algorithm constructed in this paper can be directly used to solve engineering optimization problems.