APPARATUS FOR MACHINE LEARNING-BASED SHIP DESIGN ERROR EARLY WARNING AND METHOD THEREOF

摘要

The present invention relates to an apparatus for providing a prior warning for a ship design error based on machine learning and a method thereof. According to an embodiment of the present invention, in a method for predicting a reason for revision using the apparatus for providing a prior warning for a ship design error, the method comprises: a step of making a database by receiving inputs of drawing numbers, hull numbers, revision codes, and reasons for revision of a plurality of ships whose designs are completed; a step of creating a learning model by applying the input drawing numbers, hull numbers, revision codes, and reasons for revision to a probability-based algorithm and making the learning performed on the association between the drawing numbers and the hull numbers and the association between the revision codes and the reasons for revision; a step of receiving inputs of the drawing number and the hull number of a ship to be designed from a designer; a step of applying the input drawing number and hull number to the learning model and selecting one or more revision codes by priority; and a step of predicting and providing one or more reasons for revision falling under the selected revision code. As such, according to the present invention, a designer can identify potential design errors in advance that have frequently occurred in previous situations that were similar to the current design work, and additional demands from the ship owner, before starting the design work, thereby taking proactive measures when necessary. In addition, the present invention is able to predict design changes involving extensive work-hours, and prevent such changes, thereby shortening design time.