Fault-Tolerant Inertial Sensor Fusion for Marine Vessels

The project has developed new methods for inertial navigation systems onboard ships. This includes a new software architecture for integration of sensor data from various position reference systems and inertial sensors. The overall goal was to estimate the position, velocity and attitude (PVA) of a floating vessel with better accuracy and reliability using nonlinear observer theory. This goal has been achieved and the proposed solutions have been demonstrated to improve redundancy and failure handling. The results have been validated by installing a strapdown inertial navigation system (INS) and datalogger onboard an offshore anchor handling vessel and the Rolls-Royce Marine dynamic positioning system has been used as reference system. The main results have been published in top ranked international journals.

This project is driven by a need to improve fault-tolerance, safety and performance of vessel control, monitoring and advisory systems. Taking advantage of recent developments in low-cost Micro-electromechanical system (MEMS) based sensors, the approach i s sensor fusion that combines redundant inertial sensors with other redundant vessel state, motion and position sensors in order to achieve more accurate and reliable motion and position estimates. Meeting these objectives are essential for reaching perf ormance and availability targets, and is challenging on existing technologies, in particular at a cost level that can enable a wider usage of high quality position reference systems and motion sensors. Our vision is to develop novel architecture and algo rithms for integrating sensor data from various position reference systems and inertial sensors, to be able to provide motion and position measurements with better accuracy and reliability with less stringent requirements to each of the sensors. The solut ions will improve redundancy and failure handling, and use software algorithms to gain high quality information from data from lower grade sensors. These objectives will be met through research centered around the following 3 work packages: WP1: Optimal sensor fusion for marine vessels using redundant inertial and position sensors WP2: Fault-tolerant sensor fusion by exploiting redundant inertial measurements WP3: Full-scale validation, pilot installations, technology qualification, and demonstrations The project targets mainly MAROFF theme 2: Challenging safe maritime operations, although the technology also has benefits towards fuel efficiency, emissions and cost.