摘要
本文研究了基于遗传算法的系统辨识的可靠性。首先,分析了基于响应曲线拟合的系统辨识方法及各种模型的拟合精度对模型参数摄动的敏感度。然后,对比分析了几种遗传算法应用于此类系统辨识方法的可靠性及效率,并给出了提高系统辨识效率及可靠性的改进遗传策略。实验结果表明基于改进遗传算法的系统辨识方法的可靠性乃有效性。
The credibility of system identification based on genetic algorithms (GA) is studied in this paper. First, system identification based on response curve fitting and the sensitivity of various model fitting precision to the model parameter perturbation is analyzed. Then, the credibility and efficiency of the kind of system identification based on various GA is compared, and an improved GA to raise the efficiency and credibility of system identification is proposed. Lastly, the experiment result is given to show the credibility and validity of system identification based on the improved GA.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2000年第4期447-450,共4页
Pattern Recognition and Artificial Intelligence
关键词
系统辨识
遗传算法
可靠性
曲线拟合
System Identification, Genetic Algorithm, Credibility
