论文部分内容阅读
本文提出了一种结合了生物进化和群体智能思想的新型智能算法,并应用于水库群的梯级调度优化研究中。本算法以人工蜂群算法中群体协作的正反馈机制、个体分工的性态多样性思想、优良的全局搜索能力、并行计算性及较强的鲁棒性为基础,进行问题空间的全局寻优;在个体的局部寻优行为中,引入遗传算法的杂交和变异算子来优化侦查蜂路径,避免陷入早熟问题。同时针对梯级调度优化中常见的多维变量约束条件,借鉴模拟退火算法思想,在目标函数中构造的惩罚因子,使得带约束问题转化为了纯粹的优化问题。经实例验证,本算法具有普遍的梯级调度优化解决能力,并与传统的遗传算法及人工粒子群算法相比,具有更好的精度、收敛速度和寻优能力。
In this paper, we propose a novel intelligent algorithm that combines biological evolution and group intelligence, and is applied to the optimization of cascade scheduling in reservoirs. Based on the positive feedback mechanism of group collaboration in artificial bee colony algorithm, the idea of morphological diversity of individual division of labor, excellent global search ability, parallel computing ability and strong robustness, this algorithm performs global optimization of the problem space In the part of individuals, the crossover and mutation operators of genetic algorithm are introduced to optimize the detection of bee paths and avoid the problem of premature convergence. At the same time, aiming at the constraints of multi-dimensional variables commonly used in cascade scheduling optimization, the penalty factor constructed in the objective function is borrowed from the thought of simulated annealing algorithm, which makes the constraint problem into a pure optimization problem. The example shows that this algorithm has the ability of general cascade scheduling optimization and resolution, and has better accuracy, convergence speed and optimization ability than the traditional genetic algorithm and artificial particle swarm optimization algorithm.