Optimal setpoint learning of a thruster-assisted position mooring system using a deep deterministic policy gradient appr
- PDF / 2,332,043 Bytes
- 12 Pages / 595.276 x 790.866 pts Page_size
- 4 Downloads / 201 Views
ORIGINAL ARTICLE
Optimal setpoint learning of a thruster‑assisted position mooring system using a deep deterministic policy gradient approach Shangyu Yu1,2,3 · Lei Wang1,2,3 · Bo Li1,2,3 · Huacheng He1,2,3 Received: 21 May 2019 / Accepted: 17 September 2019 © The Japan Society of Naval Architects and Ocean Engineers (JASNAOE) 2019
Abstract Thruster-assisted position mooring (PM) systems use both mooring lines and thrusters for station keeping of marine structures in ocean environments. To operate in an energy-efficient manner in moderate sea conditions, setpoints need to be appropriately chosen for the setpoint controller, so that the mooring system counteracts main environmental loads, while the thrusters reduce oscillatory motions of the marine structure. In this paper, reinforcement learning is used to design a decision-making agent for setpoint selection. In particular, a deep deterministic policy gradient (DDPG) approach is adopted with the powerful actor–critic architecture to continuously modify the setpoint setting at an optimal position. Extensive numerical experiments demonstrated that with the DDPG-based PM system, the intelligent agent is able to successfully identify the optimal positioning region in an unknown and stochastic environment, and the power consumption of the thrusters is maintained at a considerably low level. Keywords Thruster-assisted position mooring · Optimal setpoint · Reinforcement learning · DDPG · Neural network
1 Introduction A thruster-assisted position mooring system, usually referred to as PM system, consists of mooring lines and thrusters to maintain the position and heading of a marine structure. The main contribution of the mooring lines is to compensate the mean environmental loads, while the thrusters provide forces to reduce structure offset and keep the mooring line tensions within a safety limit to prevent line breakage. Compared with the dynamic positioning (DP) system, the PM system provides a more cost-effective solution for station keeping in many deep-water operations. Over the years, extensive studies on control strategies have been proposed for PM systems. A fundamental model test for a PM FPSO can be found in the work of Aalbers and Merchant [1]. In their tests, a PID controller * Lei Wang [email protected] 1
State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, Shanghai 200240, China
2
Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration, Shanghai 200240, China
3
School of Naval Architecture, Ocean and Civil Engineering, Shanghai Jiao Tong University, Shanghai 200240, China
was implemented to maintain the center of the vessel turret at a reference position. Later, Strand et al. [2] presented a mathematical model of a PM vessel. Four control modes of operation including manual, damping, setpoint and tracking control are described in detail. Simulation results showed that heading setpoint control was considered to be the most important function for turret-anchored ships, and damping control of surge
Data Loading...
