摘要
分析了二进制粒子群优化算法和基于二进制粒子群优化的属性约简方法,提出了一种基于退火选择的二进制粒子群算法,在选择粒子更新位置时引入模拟退火算法的策略,通过调整退火速度,控制算法收敛,当温度下降的足够慢时,粒子不会轻易跳出有"希望"的搜索区域,从而增强了粒子的局部搜索能力,使优化算法具有更高的效率.将该算法应用到电力负荷预测的属性约简中,使原来65个属性下降为12个,显著降低了后续处理的复杂度.
This paper analyzes binary PSO algorithm attribute and the reduction methods of attributes based on BPSO, proposes a binary particle swam optimization method based on simulated annealing. The simulated annealing is introduced when particles updated their position. The algorithm convergence is controlled by adjusting the speed of annealing. The particles would not easily jump out of the "expected" search area when the fall of temperature is slow enough, which improved the particles’ local search capability and made the optimization algorithm more efficient. This algorithm is applied to the attribute reduction of casing damage prediction attributes are reduced from original 62 to 12. The complexity of aftermath processing is significantly reduced.
出处
《哈尔滨商业大学学报(自然科学版)》
CAS
2009年第6期715-717,共3页
Journal of Harbin University of Commerce:Natural Sciences Edition
基金
吉林省教育厅"十一五"科学技术研究项目(2008410)