基于Levenberg-Marquardt算法的船体型线反向设计研究

摘要

Traditional ship hull optimization method belongs to the method of solving the forward problem. This kind of method has shortcomings of high consumption of time and low efficiency. To solve this problem, an in-depth study has been made, in which Levenberg-Marquardt algorithm based inverse design method is proposed. The relationship between ship parameters and performance indicators is represented with Jacobian matrix, and then a better hull form can be predicted. To verify this method, a 1300TEU container ship is selected as an example ship for the optimization. The results show that the inverse design method has advantages of short time taking and high efficiency compared with traditional forward optimization and has strong engineering value.%传统的船体型线优化属于求解正向问题，其优化时间长、优化效率低。为了解决这个问题，论文对基于LM 算法的反向设计方法进行了深入研究。该方法通过建立雅克比矩阵来表示船型参数和性能指标之间的关系，据此预测出更优的船型。为验证反向设计方法，对1300TEU集装箱船的前体型线进行了反向设计。结果表明：同传统的正向优化设计相比，反向设计方法具有优化时间短、优化效率高的优点，具有很强的工程应用价值。