点击次数：

发表时间：2022-10-10

发表刊物：Harbin Gongcheng Daxue Xuebao/Journal of Harbin Engineering University

卷号：42

期号：1

页面范围：82-88

ISSN号：1006-7043

摘要：To solve the influence of the six-degree-of-freedom (DOF) movement of a floating platform on the comfort of platform personnel, this paper examines the comfort platform of floating platform personnel based on a convolutional neural network (CNN). Based on fractal theory and statistical analysis methods, a dimensionality reduction analysis of measured load information is performed in this study to select mixed feature parameters. Meanwhile, the motion response model of the semi-submersible platform is simplified to the six-DOF motion of a rigid body. The central difference and vector superposition method are used to derive the correspondence between the acceleration and the six DOF at any point of the platform. A personnel comfort evaluation method based on the ISO 6897-1984(E) specification is proposed for the vertigo problem caused by the platform movement. The relationship model between the load characteristic parameters and personnel comfort level is established using the fully connected CNN method, and the personnel comfort assessment method based on the environmental load parameters is proposed. The train accuracy reaches 99.77%, and the test accuracy achieves 100%. The prediction results can provide some guidance for platform operations and services. Copyright ©2021 Journal of Harbin Engineering University.

备注：新增回溯数据