摘要
为提高延时控制系统现有决策算法的精度,提出了基于遗传算法的差异并联决策算法,利用遗传算法对各个延时电路的决策增量进行优化求解;为改善系统决策算法的鲁棒性,提出一种自适应加权融合算法,在对实测数据进行分析的基础上确定合理的权值分配策略。利用数值算例对精度、可靠度、鲁棒性和实时性的对比表明:与原有算法相比,所提两种算法具有更优的综合性能,其中自适应加权融合算法性能较为均衡,而差异并联算法具有较强的抑制异常测量值能力,适用于对鲁棒性要求较高的场合。
In order to improve the accuracy of the existing decision algorithm for delay control system,an increment optimization method based on genetic algorithm was proposed,genetic algorithm was applied in which for optimized solution.And an adaptive weighted fusion method was proposed to improve system robustness.The weight assignment method was decided based on the analysis of the measurement data.Compared with the conventional algorithm,the comprehensive performance of adaptive weighted method was more balanced,and increment optimization method could restrain measurement outliers efficiently,which were demonstrated by numerical results.
出处
《探测与控制学报》
CSCD
北大核心
2011年第3期43-46,共4页
Journal of Detection & Control
基金
"十一五"行业预先研究项目资助(426010501)
"十一五"跨行业预先研究项目资助(51305060204)
中国工程物理研究院重大预先研究(06-0401)
关键词
延时控制系统
决策算法
加权融合
综合性能
delay control system
decision algorithm
weighted fusion
comprehensive performance
