BP神经网络应用于散货船空船质量估算

摘要

Based on the data of more than 200 bulk cargo ships, statistical regression model and three kinds of BP neural network models are established in order to estimate light weight of ships. These models are tested through 10 bulk cargo ships and compared to each other. The results of BP neural network models are close to the data of ships and superior to statistical regression. Estimation of light weight of ships by BP neural network is feasible and practical, and can be carried out in different input conditions.%以搜集的200余条散货船为样本,建立了空船质量统计回归模型和三种情况下的BP神经网络模型,并选取10条散货船对各模型进行了测试和比较.三种BP神经网络模型测试结果与实际值很接近,精度优于传统的统计回归模型,表明用BP神经网络进行空船质量估算是可行和实用的.