[author_cn_name].[cn_title][J].空军工程大学学报:自然科学版,[year_id],[volume]([issue]):[start_page]-[end_page] 平均差分进化算法在等效系统拟配中的应用-An Application of Average Differential Evolution Algorithm in Equivalent System Fitting
文章摘要
田宏峰1,薛源1,2,徐浩军1,杨梓鑫1,吴嘉杰3.平均差分进化算法在等效系统拟配中的应用[J].空军工程大学学报:自然科学版,2020,21(1):9-14
平均差分进化算法在等效系统拟配中的应用
An Application of Average Differential Evolution Algorithm in Equivalent System Fitting
  
DOI:
中文关键词: 平均差分进化算法  等效系统拟配  最小二乘法  极大似然法
英文关键词: average differential evolution algorithm  equivalent system fitting  least squares method  maximum likelihood method
基金项目:国家自然科学基金(61873351)
作者单位
田宏峰1,薛源1,2,徐浩军1,杨梓鑫1,吴嘉杰3 1.空军工程大学航空工程学院 西安 710038 2.西北工业大学航空学院 西安 710129 3.95960部队 西安 710089 
摘要点击次数: 100
全文下载次数: 93
中文摘要:
      等效系统方法是评价飞机飞行品质的重要方法之一,最小二乘法和极大似然法在等效系统拟配中应用较为广泛,但是二者在使用时对初值有较高的要求且容易陷入局部最优解,因而导致拟配精度不高。针对这一问题,对传统差分进化算法的核心公式进行了改进,提出了基于平均差分进化算法进行等效系统拟配的新思路。此外,由于数据中存在一定比例的噪声干扰,所以分别以理想3 2 1 1信号和混入10%、30%白噪声的3 2 1 1信号作为系统输入,对比研究了平均差分进化算法和传统等效拟配方法的差异,发现平均差分进化算法具有较强抗噪声干扰能力和良好的鲁棒性,因此,平均差分进化算法应用于等效系统拟配具有较好的辨识效果,同时对于噪声的处理能力较强。
英文摘要:
      The equivalent system method is one of the important methods for evaluating the flight quality of aircraft. The least square method and the maximum likelihood method are widely applied in the formulation of equivalent systems, but both of them require highly of the initial value and are easy to fall into a local optimal solution leading the fitting accuracy to be low. In view of the above mentioned facts, the core formula of traditional differential evolution algorithm is improved, and a new idea of equivalent system fitting based on average differential evolution algorithm is proposed. In addition, because there is a certain percentage of noise interference in the data, the ideal 3 2 1 1 signal and the 3 2 1 1 signal mixed with 10% and 30% white noise are used as system inputs, respectively. The difference between the differential evolution algorithm and the traditional equivalent matching method is studied, and the results show that the average differential evolution algorithm has a strong anti noise interference ability, good robustness, good identification effect, and at the same time the processing ability of noise is strong.
查看全文   查看/发表评论  下载PDF阅读器
关闭