The application of high-strength steel plates can reduce ship weight, and the saddle plate is one of the most common types of double-curved hull plates. To fill the research gap regarding high-strength steel saddle plates, two prediction models are established here to predict deformation in saddle plate forming. Deflection is a key parameter reflecting the overall deformation of a curved plate. Therefore, first of all, the influencing factors of the line heating of high-strength steel saddle plates were analyzed. The influence of plate geometric parameters and forming parameters on deflection was researched. Second, a multiple linear regression model between deflection and the geometric parameters and forming parameters of high-strength steel saddle plates was established. Finally, to solve the problem of a large error in the multivariate regression model for extrapolation, an intelligent prediction program for deflection based on a support vector machine (SVM) was developed using the Python language. The results show that the error of the multiple regression model was less than 5% for data interpolation. The error of the intelligent prediction model for deflection was less than 5% for data extrapolation. This research can provide data support for the automatic forming of marine saddle plates.
Keywords: deflection prediction; high-strength steel; line heating; plate forming; ship lightweight; support vector machine.