首页    期刊浏览 2024年11月25日 星期一
登录注册

文章基本信息

  • 标题:操縦性解析への最小2乗法の応用
  • 本地全文:下载
  • 作者:小山 健夫 ; 渡辺 高根 ; 渡辺 岩夫
  • 期刊名称:日本造船学会論文集
  • 印刷版ISSN:0514-8499
  • 电子版ISSN:1884-2070
  • 出版年度:1973
  • 卷号:1973
  • 期号:134
  • 页码:173-181
  • DOI:10.2534/jjasnaoe1968.1973.134_173
  • 出版社:The Japan Society of Naval Architects and Ocean Engineers
  • 摘要:

    Since Nomoto proposed the first order system to describe the ship maneuverability, dynamic characteristics of many ships have been measured from the results of zig-zag tests. However, it is gradually realized that we can not describe the ship maneuverability satisfactorily, . especially in the case of full ships, by simple first order equation, and that it is necessary to adopt the second order equation with a non-linear term. To this end, Bech and Wagner-Smitt proposed the reverse spiral test and the phase-plane analysis. Fujino, Nomoto and Kose are also developing the phase-plane analysis. van Leeuwen emphasized the importance of the advance speed effects from another view point. In addition to these proposals, Nonaka et al. tried to obtain as much as 20 derivatives from the free running model test adopting the least square identification method with the data of rudder angle, rate of turn and sway velocity. Kaplan obtained the characteristics of ships and marine vehcles by means of recently developed system identification method. In this report, the authors would like to make some comments to this state of arts. It should be realized that there may not be involved enough information in the free running test data to obtain many parameters for a complicated system model with good accuracy. If this were the case, and nevertheless we tried to obtain many parameters, we talk too much beyond the rational limit. In general, the free running test data involve less information than that of the captive model test, because ships and models are always in the dynamic equilibrium condition. during the free running test. The parameters obtained from experiments are statistic values. Therefore, there should be some rational limit from the view point of statistics to what extent we may talk about the ship's characteristics with a set of experiment data. The authors wish to find a methodology for this purpose, and try the least square identification method applying the likelihood ratio test. An experiment data set for this analysis was obtained from the free running test of a 2. 5M container ship model with the gaussian white rudder input.

国家哲学社会科学文献中心版权所有