Predicting construction quality of a marine drilling platform based on GRNN
Liang Jing-guo · Harbin Gongcheng Daxue Xuebao/Journal of Harbin Engineering University · 2009
Effective prediction of potential construction quality failures for marine drilling platforms can result in improved construction quality and construction techniques.Based on descriptions of equipment and detailed requirements of quality inspection,a general regression neural network(GRNN) was used to provide predictions.Taking a particular jack up rig as an example,construction errors were predicted with 1.5 years' quality inspection data.In order to show the effectiveness of this method for construction quality prediction,a comparison was made with predictions by a back-propagation neural network(BPNN).The result shows that GRNN is a useful prediction method which is more precise than BPNN.On the basis of the predictions,countermeasures were put forward to increase construction quality.