RFFMIDT: An Approach toward Optimal Control of Ship anoeuvring in Offshore Operations (2016-2017)

Project partners

Rolls Royce Marine AS
Offshore Simulation Center
NTNU i Ålesund, Mechtronics Lab

Project leaders

Dr. Guoyuan Li, Prof. Houxiang Zhang

Project period


Project type

RFFMIDT pre-project

Short project description

With the growth of emerging demands from offshore applications, such as seabed survey, pipeline maintenance and offshore oil installations, the complexity of ship manoeuvring during offshore operations increases as more constraints from position accuracy, limited working space, and collision avoidance between vessels and floating structures need to be taken into consideration. Meanwhile, ship manoeuvring performance such as power consumption is of great concern by ship owners and operators for economically and safety beneficial reasons, as well by the International Maritime Organization (IMO) that considers good manoeuvring have potential to contribute to a reduction of greenhouse gas emissions. To address the complexity and guarantee the performance, new knowledge and technology for such constrained ship manoeuvring are urgently demanded.


This project focuses on developing the control scheme for constrained ship manoeuvring. We will develop a new control scheme that will provide operators with a good understanding and prediction of the ship’s manoeuvring behaviour together with an iterative optimisation method in order to accomplish safe and efficient manoeuvring during demanding operations.

 Objectives / aims for the project

In this project, we will make a theoretical and experimental study on combining neural network structures and optimisation algorithms to establish a complete control scheme for optimised manoeuvring in offshore operations.

  1. Investigate and analyse on-board ship data for optimised manoeuvring in demanding offshore operations
    • Investigate possible optimisation demands and constrains in offshore manoeuvring applications
    • Study the manoeuvring characteristics of a ship through analysing recorded on-board ship data
  2. Develop an optimised control system for constrained ship manoeuvring
    • A robust learning module for prediction of ship behaviours
    • Real-time optimised control strategy based on neural network model and optimisation methods
  3. Make a demonstrative case study of trajectory tracking within partners’ framework

Research content

  1. For different offshore operations, there would be different optimising demands. We will investigate the related optimising matters and summarise them as mathematical models.
  2. We will also obtain the data from our partners and start the analysis for both physical constrains and manoeuvrability.
  3. Considering the challenge of obtaining the precise dynamic model of marine vehicles, we plan to approximate it by combining regression technique with neural network structure.
  4. We will develop a robust optimising solver to respond to different offshore operations considering the physical constraints from the ship.

Level of Innovation

Even though many technologies for ship manoeuvring have been under development, the necessity of optimal control is less used in practice today. This project will be devoted to solving constrained manoeuvring problems encountered in offshore operations. It will promote offshore operation applications, thereby enhancing the local economy and improving the technological level of the maritime industry:

  1. On-board decision support systems
  2. Real-time suggestions
  3. Evaluation

Project impact

Constrained ship manoeuvring is imperative that it enables operators to manage the vessel in ways that are more economical and ecologically beneficial, in compliance with safety and physical constraints. Therefore, the project will be strategically important for sustaining and improving the manoeuvring performance of the Norwegian offshore industry for advanced marine operations.

Comments are closed