Search

KR-20260062585-A - Floating Dock Loadout Control System Using Circular Neural Network

KR20260062585AKR 20260062585 AKR20260062585 AKR 20260062585AKR-20260062585-A

Abstract

This document is about a recurrent neural network-based floating dock loadout control system. The proposed load-out control system includes a plurality of tanks disposed on the hull of the floating dock; and a control unit configured to control the amount of water supplied to the plurality of tanks when the load-out is performed. At this time, the control unit is driven based on an automated algorithm composed of a plurality of layers based on a plurality of variables corresponding to the amount of water supplied to the plurality of tanks, and the plurality of layers are configured with a structure including a first layer that controls the value of the plurality of variables based on hogging characteristics for maintaining a step difference, and one or more remaining layers following the first layer.

Inventors

  • 김남훈
  • 한준영

Assignees

  • 에이치디한국조선해양 주식회사
  • 에이치디현대중공업 주식회사
  • 에이치디현대삼호 주식회사

Dates

Publication Date
20260507
Application Date
20241029

Claims (7)

  1. In a loadout control system using a floating dock, A plurality of tanks disposed on the hull of the above-mentioned floating dock; and When performing the above loadout, it includes a control unit configured to control the amount of water supplied to the plurality of tanks, and The above control unit is operated based on an automated algorithm composed of multiple layers based on multiple variables corresponding to the water supply amounts of the plurality of tanks, and The above plurality of layers are, A load-out control system comprising a structure including a first layer that controls the values of a plurality of variables based on a hogging feature for maintaining a step difference, and one or more remaining layers following the first layer.
  2. In Article 1, The above-mentioned first layer or the second layer following the above-mentioned first layer is, A loadout control system configured to control one or more of sudden sagging or local sagging that occurs when watering or draining is not performed within a time determined by the vessel's entry speed.
  3. In Article 1, One or more remaining layers following the above-mentioned first layer are, A second layer controlling the height difference and equal drainage volume; and Third layer controlling the entry of the main vessel and equal discharge volume Includes, A load-out control system configured to reduce the parameters of the second and third layers compared to the parameters in the case where there is no processing of the first layer, according to the processing of the first layer.
  4. In Article 1, One or more of the last predetermined number of layers among the above plurality of layers are, A load-out control system composed of layers that perform trim adjustment and equal water level control.
  5. In Article 4, The above plurality of layers are, Prior to the layer performing the above trim adjustment and equal water level control, A loadout control system including a layer that performs bow-seat synchronization and trim adjustment.
  6. In Article 1, The above plurality of layers are, The above first layer; A second layer controlling the height difference and equal drainage volume; A third layer controlling the entry of the main vessel and equal discharge volume; A fourth layer that performs verification based on the plurality of pump capacities mentioned above; Fifth layer performing bow-stern synchronization and trim adjustment; A sixth layer that performs equal water level conversion; and A load-out control system comprising a seventh layer that restores the target bending moment to its original state after outputting the values of the plurality of variables.
  7. In Article 1, The above control unit is, A load-out control system configured to apply a fixed amount of water to the central tank among the plurality of tanks and to variably control the amount of water to the bow tank and the stern tank when performing the load-out above.

Description

Floating Dock Loadout Control System Using Circular Neural Network The following description relates to a system for automatically controlling the loadout of a ship under construction using a floating dock, specifically a system configured to improve longitudinal strength and local strength when performing floating dock loadout based on a recurrent neural network. Figure 1 is a diagram illustrating the concept of a floating dock and load-out process. Floating Dock (110) is primarily used to move ships (120) built on land from the quay (130) toward the sea and launch them, or to redocking ships that require repair or repainting, instead of the existing dry dock. Normally, it is used as a quay (130) for outfitting work on launched ships. In particular, the process of moving ships (120) from land to sea is subdivided into Loadout among land-based construction methods. As shown in FIG. 1, a link beam (140) may additionally be used for the connection between the quay wall (130) and the floating dock (110). At this time, ballasting and deballasting are determined according to the increase in light weight (LWT) of the large vessel that has finished preparing for launching, based on the change in currents on the west coast and the speed of moving it to land, and it is possible to check whether they have been properly performed through land/sea level difference monitoring. As described above, the load-out process using a floating dock is defined by the following main operating principles, primarily reflecting the characteristics of the 6,500 TEU container ship and 155k LNGC LWT, which were the main vessel types about 17 years ago. 1) Selective ballasting and deballasting applied to only some tanks (no margin during manual operation) 2) Adjustment of longitudinal strength using the central tank (for the purpose of minimizing the step difference at the onshore/offshore connection) 3) Adjust trim with forward and stern tanks (to maintain Even Keel conditions) However, compared to the existing system, the main types of vessels currently applicable to the onshore construction method have become larger, such as 8,000 TEU container ships and 174k LNGCs, and accordingly, the existing operating principles have the following problems. 1) Some tanks alone cannot balance the tank ballasting corresponding to the main ship's LWT. 2) Since the pump in the central tank is used in conjunction with other longitudinal tanks, 2-1) the step difference is widened due to the slow response speed of the floating dock when the valve is opened or closed, or 2-2) the step difference is widened by reducing the efficiency of the other tank used in conjunction. 3) Even when adjusting the trim with the bow and stern tanks, 3-1) the even kill balance was not achieved while meeting the pump capacity, 3-2) the trim change amount was too sensitive compared to the ballasting change amount, or 3-3) the total watering time increased because it was not adjusted to the same level as other tanks in the transverse direction. Figure 1 is a diagram illustrating the concept of a floating dock and load-out process. FIG. 2 is a diagram illustrating the concept of a ballast sequence to which embodiments of the present invention are applied. FIG. 3 is a drawing for explaining the arrangement relationship of a plurality of tanks in a floating dock according to one embodiment of the present invention. FIG. 4 is a diagram illustrating a strategic approach for designing a neural network model for loadout automation according to an embodiment of the present invention. FIG. 5 is a diagram illustrating the post-processing concept of a recurrent neural network according to one embodiment of the present invention. FIG. 6 is a diagram illustrating the layers of a neural network structure for automatically calculating the amount of water supplied to a tank during a loadout according to an embodiment of the present invention. Figure 7 is a diagram illustrating a hierarchical structure for setting initial conditions of the neural network structure of Figure 6. FIGS. 8 and 9 are drawings for explaining the recurrent neural network layer-by-layer processing method at each stage of a loadout according to an embodiment of the present invention. Hereinafter, embodiments of the present invention are described in detail with reference to the attached drawings so that those skilled in the art can easily implement the invention. However, the present invention may be embodied in various different forms and is not limited to the embodiments described herein. Furthermore, in order to clearly explain the present invention in the drawings, parts unrelated to the explanation have been omitted, and similar parts throughout the specification are denoted by similar reference numerals. Throughout the specification, when a part is described as "including" a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional comp