SPRING 2017

Schedule
Week 4, Guoyuan
Week 6, Wei (Postponed)
Week 7, Yingguang
Week 10, Lars Ivar
Week 11, Birger
Week 12, Wei
Week 14, Robert
Week 16, Thiago
Week 18, Girts
Week 20, Hao or Di
Week 22, Hao or Di

1 – A bezier curve based ship trajectory optimization for close-range maritime operations

Speaker: Guoyuan Li, Researcher, Department of Ocean Operations and Civil Engineering, NTNU i Aalesund, Norway

Time: Thursday 26 January 2017, 13:00-14:00

Place: Sukkertoppen Auditorium

Abstract

Ship maneuvering in close-range maritime operations is challenging for pilots, since they have to not only prevent the ship from collisions and compensate environmental impacts, but also steer it close to the target towards a proper heading. This paper presents a path planner to assist the pilots to foresee the optimal trajectory in the scenario. The path planning is formatted as an optimizing problem to minimize the turning variation fluctuation and the fuel consumption of the ship through ocean current while satisfying the constraint of orientations at the start and the end positions. Taking advantages of Bézier curves’ smoothness and adjustability, feasible trajectories are divided into two categories based on the location of the intersection between the start and end directions, and are designed as a set of parameterized Bézier curves. The variables in the Bézier curves become the state space. By searching the space using an evolutionary technique, the candidate of the Bézier curve that has the best turning and the minimized fuel consumption can be obtained. Through two case studies, the feasibility and effectiveness of the proposed planner is verified.

2 – Marine crane virtual prototyping simulator– OOM modeling and co-simulation based on the FMI specification

Speaker: Yingguang Chu, PhD, Department of Ocean Operations and Civil Engineering, NTNU i Aalesund, Norway

Time: Thursday 16 Februar 2017, 13:30-14:00

Place: B433

Abstract

This paper presents an object-oriented approach to model development for engineering systems, specifically, the hydraulic systems of marine cranes. Benefited from the rapid development of computation technology, many modeling and simulation techniques and software tools have been used purposefully in various stages of the product and system development processes. However, there are still many challenges for modeling and simulation due to the increasing complexity of the physical engineering systems, such as model flexibility, simulation accuracy, stability and efficiency. The goal of introducing the object-orientated modeling approach to engineering systems is to provide flexible and effective models for different simulation applications. A virtual prototyping framework based on co-simulation using the Functional Mock-up Interface specification was proposed and presented for marine crane systems. The advantages of using co-simulation is that the simulations of complex systems, in particular, with stiff and tightly-coupled systems can be distributed. As a result, the simulation characteristics, specifically, the trade-offs between accuracy and efficiency can be improved. The essential features of object-oriented modeling and its application within engineering system domain are highlighted through a case study, e.g., model causality, model encapsulation and inheritance, which are essential for the decomposition of a complex system model for co-simulations. The simulation results based on the proposed virtual prototyping framework using FMI co-simulation showed that the simulation efficiency was improved significantly.

3 – Preliminary results from the SFI MOVE – Wind installation concept

Speaker: Lars Ivar Hatledal, Research assistant, NTNU i Aalesund, Norway

Time: Thursday 3 March 2017, 14:00-14:30

Place: Aksla

Abstract

Current methods for installation of offshore wind turbines are all sensitive to the weather conditions and the present cost level of offshore wind power is more than twice the cost of land-based units, increasing with water depth. This paper presents numerical simulations of a novel experimental gripper design to reduce the environmental effects applied to a catamaran type of vessel during wind turbine installation. In SFI MOVE project in NTNU Aalesund, our team proposed a novel wind turbine installation process. A new catamaran vessel will carry pre-assembled wind turbines to the installation location. Two new designed grippers on the deck will make a lifting operation to install the wind turbine onto the turbine foundation. Three prismatic grippers with several rolling contact points at the end are attached in an arc at the catamaran’s aft, designed to grasp the turbine foundation in order to make a connection between the two in the horizontal plane. This paper will only emphasize the contact responses between the turbine foundation and the three grippers during the wind turbine installation process. Numerical simulations are carried out using the virtual prototyping framework Vicosim which is developed by NTNU Aalesund. The simulation results show validation of a key part of the proposed new wind turbine installation idea.

4 – Crane and winch design prototyping tools

Speaker: Birger Skogeng Pedersen, Research assistant, NTNU i Aalesund, Norway

Time: Thursday 16 March 2017, 14:00-14:30

Place: Aksla

Abstract

To meet desired design criteria for maritime cranes and winches, engineers choose values for the design parameters based on simplified physical models, previous experience and other heuristic methods such as rules-of-thumb. The results are assessed and if the results do not satisfy the desired performance, the process is repeated until suitable parameters are found. This trial-and-error process can be tedious, costly and may not always succeed in meeting the design criteria. Applications to aid our industrial partners with this design process has been developed. The applications utilizes artifical intelligence methods, such as genetic algorithms, to propose new and better design. The applications are web-based for enhanced availability. Though the applications are initially reserved for our industrial partners, the computational engines are accesible through common networking techonlogies. As a result, other collagues from the NTNU in Aalesund university have utilized the computational engines for their own research on artifical intelligence.

5 – A FPGA based real-time SAR extended object simulator

Speaker: Wei Li, Researcher, NTNU i Aalesund, Norway

Time: Thursday 23 March 2017, 14:00-14:30

Place: Aksla

Abstract

An FPGA based SAR extended object simulator will be presented focused on backscattering coefficients computing. It is aimed to be used for hardware-in-the-loop SAR raw signal simulation with applications in UAV or other platforms SAR sensors test and assessment. This simulator is based on geometrical optics (GO) model and can work in real time. The GO algorithm is analyzed and implemented with multiple processing units in FPGA. Arithmetic units are optimized to decrease the FPGA resource cost. The design is implemented on a FPGA board and verified with initial test results.

6 – Data based ship motion prediction

Speaker: Robert Skulstad, PhD, NTNU i Aalesund, Norway

Time: Thursday 6 April 2017, 14:00-14:30

Place: Aksla

Abstract

An overview of my work up until now will be presented along with results from prediction run in OSC (Offshore Simulator Centre) simulator. Key topics are:

  • Motivation for ship motion prediction
  • Different strategies for constructing models for prediction
  • Neural networks in motion prediction
  • Description of simulator used to test prediction algorithms
  • Results from tests using static/non-updating neural networks in simulator
  • Future work

Comments are closed