Abhilash Sharma Somayajula

A majority of marine accidents that occur can be attributed to errors in human decisions. Through automation, the occurrence of such incidents can be minimized. Therefore, automation in the marine industry has been receiving increased attention in the recent years. This paper investigates the automation of the path following action of a ship. A deep Q-learning approach is proposed to solve the path-following problem of a ship. This method comes under the broader area of deep reinforcement learning (DRL) and is well suited for such tasks, as it can learn to take optimal decisions through sufficient experience. This algorithm also balances the exploration and the exploitation schemes of an agent operating in an environment. A three-degree-of-freedom (3-DOF) dynamic model is adopted to describe the ship's motion. The Krisco container ship (KCS) is chosen for this study as it is a benchmark hull that is used in several studies and its hydrodynamic coefficients are readily available for numerical modeling. Numerical simulations for the turning circle and zig-zag maneuver tests are performed to verify the accuracy of the proposed dynamic model. A reinforcement learning (RL) agent is trained to interact with this numerical model to achieve waypoint tracking. Finally, the proposed approach is investigated not only by numerical simulations but also by model experiments using 1:75.5 scaled model.

Unlike the traditional displacement vessels, the modern roll-on roll-off (Ro–Ro), container and cruise vessels designed over the past two decades are seen to be prone to dynamic instabilities, which in some cases may lead to capsizing. Although the vulnerability of a design to dynamic instabilities can be assessed through simulations, this approach is time consuming and unsuitable for analyzing several interim designs during the design spiral iterations. Recent global efforts by the International Maritime Organization (IMO) towards a second generation level 2 criterion attempt to adopt a first principles approach without resorting to time consuming numerical simulations or expensive physical model tests. This work provides such a tool for one of the identified capsizing mechanisms known as parametric rolling in a realistic random seaway. The technique of stochastic averaging is applied to a previously developed realistic model for parametric excitation in random waves. A semi-analytic design criterion for the comparative assessment of different hull forms to parametric roll in random seas is formulated in terms first passage statistics of the system. A sensitivity analysis is performed on the C11 container ship hull form to quantify and gain a deeper understanding of the relative importance of both physical parameters (restoring arm and damping) and environmental parameters (wave spectra intensity and characteristic frequency) on the instability.

Heave compensation is a vital part of various marine and offshore operations. It is used in various applications, including the transfer of cargo between two vessels in the open ocean, installation of topsides of an offshore structure, offshore drilling and for surveillance, reconnaissance and monitoring. These applications typically involve a load suspended from a hydraulically powered winch that is connected to a vessel that is undergoing dynamic motion in the ocean environment. The goal in these applications is to design a winch controller to keep the load at a regulated height by rejecting the net heave motion of the winch arising from ship motions at sea. In this study, we analyze and compare the performance of various control algorithms in stabilizing a suspended load while the vessel is subjected to changing sea conditions. The KCS container ship is chosen as the vessel undergoing dynamic motion in the ocean. The negative of the net heave motion at the winch is provided as a reference signal to track. Various control strategies like Proportional-Derivative (PD) Control, Model Predictive Control (MPC), Linear Quadratic Integral Control (LQI), and Sliding Mode Control (SMC) are implemented and tuned for effective heave compensation. The performance of the controllers is compared with respect to heave compensation, disturbance rejection and noise attenuation

Both the time domain simulations based on potential flow theory and wave tank model tests are used to evaluate the motion responses of a wave energy device. In this work, system identification is applied to obtain the frequency dependent transfer functions (between motions and excitations) from a series of model test runs of a wave energy platform exposed to irregular (random) waves. This is the first application of Reverse-Multiple Input Single Output (R-MISO) to a realistically (catenary) moored system (typical characteristics of wave energy devices) comparing physical model tests with our in-house time domain simulation program with addition of a mooring model. Based on the comparisons between the transfer functions from the time domain simulations and those from the model tests, reasonable frequency dependent dampings have been directly pulled out from the test cases under random sea states. System identification derived corrections to the linear or quadratic damping in pitch significantly improved the accuracy of motion responses. In this sense, this methodology can be a powerful tool in assisting the accurate simulation and design of wave energy devices under random sea states.

Heave compensation is an essential part in various offshore operations. It is used in various applications, which include on-loading or off-loading systems, offshore drilling, landing helicopter on oscillating structures, and deploying and retrieving manned submersibles. In this paper, a reinforcement learning (RL) based controller is proposed for active heave compensation using a deep deterministic policy gradient (DDPG) algorithm. A DDPG algorithm which is a model-free, online reinforcement learning method, is adopted to capture the experience of the agent during the training trials. The simulation results demonstrate up to 10 % better heave compensation performance of RL controller as compared to a tuned Proportional-Derivative Control. The performance of the proposed method is compared with respect to heave compensation, offset tracking, disturbance rejection, and noise attenuation.

The designs of modern container ships, roll-on–roll-off vessels and cruise vessels have evolved over the years, and in recent times, some of them have been observed to experience dynamic instabilities during operation in the open ocean. These catastrophic events demonstrate that satisfying prescriptive stability rules set forth by International Maritime Organization (IMO), national authorities (e.g., Coast Guard) and other classification societies are not sufficient to ensure dynamic stability of ships at sea. In light of these events, IMO is organizing efforts to make way toward a second generation of intact stability criteria that are better equipped to deal with these dynamic instabilities. This paper discusses the development of such a tool for parametric rolling in a realistic random seaway, which is one of the critical phenomena identified by IMO. In this study, a previously developed analytical model for roll restoring moment, which was found to be effective in modeling the problem of parametric roll, is analyzed using the Melnikov approach. The stability of the system is quantified in terms of rate of phase space flux of the system. This approach is further compared with another technique known as the Markov approach that is based on stochastic averaging and quantifies stability in terms of mean first passage time. The sensitivity of both of these metrics to environmental parameters is investigated. Finally, the nature of random response is analyzed using Lyapunov exponents to determine whether the vessel exhibits any chaotic dynamics.

Traditional linear time-domain analysis is used widely for predicting the motions of floating structures. When it comes to a wave energy structure, which usually is subjected to larger relative (to their geometric dimensions) wave and motion amplitudes, the nonlinear effects become significant. This paper presents the development of an in-house blended time-domain program (SIMDYN). SIMDYN's "blend" option improves the linear option by accounting for the nonlinearity of important external forces (e.g., Froude-Krylov). In addition, nonlinearity due to large body rotations (i.e., inertia forces) is addressed in motion predictions of wave energy structures. Forced motion analysis reveals the significance of these nonlinear effects. Finally, the model test correlations examine the simulation results from SIMDYN under the blended option, which has seldom been done for a wave energy structure. It turns out that the blended time-domain method has significant potential to improve the accuracy of motion predictions for a wave energy structure.