摘要
以某单桨大型船舶在海上的回转性能为例,探讨了应用改进的BP神经网络(Back-propagationNeuralNetwork)建立船舶操纵性预报数学模型的方法,并利用matlab语言对其进行了仿真。研究结果表明,改进的BP算法有更快的收敛速度和更好的计算精度。
Taking the turning performance of a single screw ship as example, the method for establishment of the mathematical model of ship maneuverability by using modified BP neural algorithm is investigated in the paper and its simulation is carried out with MATLAB language. The rsults of the research indicate that the modified BP networks have faster convergence speed and better prediction precision than normal BP algorithm.
出处
《上海船舶运输科学研究所学报》
2006年第1期48-51,共4页
Journal of Shanghai Ship and Shipping Research Institute
关键词
船舶、舰船工程
船舶操纵
预报
BP神经网络
ship, naval engineering
ship maneuverability
prediction
BP neural network
