摘要
航行体水动力参数辨识最常用的方法是采用极大似然准则和牛顿—拉夫逊算法。但是牛顿—拉夫逊算法需要计算复杂的导数矩阵及其逆矩阵,而且对初值非常敏感。文章利用最小二乘准则结合智能优化算法对航行体水动力参数进行了辨识仿真,仿真结果表明新方法要好于常用的方法,在不同噪声水平下辨识结果的误差小于7%,而且以往无法辨识的附加质量也能辨识,同时新方法对初值几乎没有要求,也无须计算导数,极大地简化了辨识计算,有较大的工程应用价值。
The traditional method of identification of hydrodynamic parameters for navigating body commonly is based on Maximum likelihood criterion and Newton-Raphson's algorithm (N-R). However, it is very difficult and comlex to calculate derivative matrix and its inverse, and moreover N-R is very sensitive to initial values of parameters. In this paper, instead of N-R, the least-square criterion together with Dif- ferencial Swarm Intelligent (DS) algorithm is employed to identify hydrodynamic parameters. The results of simulating identification show that this approach has better performance than the traditional method.The results error less than 7% and added mass that could not be identified ago can be identified now.Moreover, this appraoch has no requirements on the initial values, need not calculate derivative, consequently remarkably simplifies the identification, so it has a great potential for application in engineering.
出处
《船舶力学》
EI
北大核心
2011年第4期359-363,共5页
Journal of Ship Mechanics
关键词
水下航行体
水动力参数
辨识
牛顿—拉夫逊算法
最小二乘准则
智能优化算法
underwater vehicle
hydrodynamic parameters identification
Newton-Raphson's algorithm
the least-square criterion
swarm intelligent