Til NTNUs hovedside

Tightly Integrated GPS/INS Systems for Dynamic Positioning


Department of Engineering Cybernetics
NTNU, Trondheim
Professor Thor I. Fossen (head of project)
Post. Doc. Bjørnar Vik
ABB Industri AS
Dr.Ing. Alf Kåre Ådnanes
Dr.Ing. Jann Peter Strand
Norsk Hydro Produksjon

Dr.Ing. Svein Ivar Sagatun


The primary objective of this project is to develop and test a reference system which may act as a third independent position reference system for consequence class III DP operations [1] in deep water. DnV [1] paragraph C100 states:

C100 General: "When more than one position reference system are required, they are not all to be of the same type, but based on different principlec and suitable for the operating condition".

The requirement for three independed systems is strengthened in the revised Dnv [1] which will be released primo 2001. The third reference system will be an integrated INS/GPS (Inertial Navigation System/Global Positioning System) system which will produce reliable position references for a period of time after the GPS system has failed due to total or partly GPS failure. A secondary objective is to improve the performance and reliability of the DGPS system, and thus the operability of the marine operation, by aiding the DGPS system with inertial data (from the INS) when the GPS system has lock on to less than four good satellites.


There are requirements (governmental requirements or recommended practice from the International Marine Organization, IMO) on a number of independent position reference systems available in order to perform specific marine operations, such as drilling, diving etc. Similar system requirements are defined by the classification societies. As an example, the IMO DP Class 3 requirement (for a dynamic positioning system) is that 3 redundant sensors systems must be available during a critical operation - in particular 3 independent position reference systems. This requirement is only partially met when using, say two redundant GPS receivers in combination with a hydro-acoustic system. In this case only the hardware redundancy requirement is fulfilled - only two different principles are used. Another problem is the availability of reliable measurements from these systems. In several areas of the world the satellite-based systems may be distorted for several hours a day, whereas in deep-waters the communication link in the hydro-acoustic systems are strongly deteriorated. The DGPS (Differential GPS) and HPR (Hydroacoustic Position Reference) systems are the most common utilized position reference systems and at present the only suitable systems for most deep water applications, e.g. drilling vessels, floating production units, supply and support vessels. For numerous deepwater oil fields such as outside Brazil and West-Africa, there are massive problems with HPR due to varying acoustic properties in different layers of the water and with DGPS due to ionospheric interference phenomenon. These problems imply reduced operational availability, reliability and safety, and hence enormous additional cost and risk. For instance, the operational unavailability could be reduced with more than 50% by introducing position reference systems based on different concepts than the HPR and the DGPS. Another trend for GPS position reference systems, often integrated with electronic map systems, is that the operators tend to overestimate the reliability of the measurements. Hence, there is a strong marked requirement for position reference systems with high reliability and availability that are based on other principles and signal measurement media. This applies particularly for oil and gas related vessels on deep water and for vessels with extreme requirements regarding operational availability.

Integrated DGPS and IMU Solutions

Because of possible disastrous consequences when relying on GPS for navigating marine vessels it is mostly desired with a reliable and accurate maintaining positioning system. One approach to the solution is combination of the IMU/INS (Integrated Motion Unit) and GPS application. Either of these systems has some disadvantages which can be reduced by integration.

The following advantages can be achieved through IMU-DGPS integration:

  • Alignment of the INS can be performed in Marine environment (GPS provides needed attitude information for alignment)
  • Improved GPS acquisition/reacquisition times (INS provides data for initialization of GPS acquisition/reacquisition)
  • Improved accuracy, integrity, continuity and availability than stand-alone INS or GPS solutions (INS smoothing of noise in GPS attitude and position outputs, bounded INS attitude, velocity and position errors, INS solution available during GPS outage, GPS navigation data, including attitude information, is available during INS outage).
  • GPS multi-path error effects are reduced through blended INS-GPS solution.

A few years ago the price and the size of an optical based IMU were beyond the acceptable level of commercial marine deliveries. Today, the size and price have been reduced dramatically and next year a new type of optical gyros is released see prognosis. The reduction in IMU pricing is one of the driving forces for this project. The current advances in GPS technology suggests a price reduction of approximately 20% within two years.

Project Description

For the dynamic positioning case it is important to be able to avoid shutting down operation in case of loss of the DGPS solution. Just as important is the ability to predict when this happen, so that the captain of the vessel has time to make a decision whether to carry out a costly shutdown of operation or not. In most GPS receivers, no position or velocity measurements are given if the number of satellites fall below four. The instance of this event is difficult to predict, and a sudden loss of measurements is the result. This causes problems both for the control system and to the decision making of the captain. A tightly integrated GPS/INS system can utilize GPS measurements even if fewer than four satellites are available. With a good model of the internal clock of the GPS receivers (or a stable external clock reference) it is possible to maintain sufficiently accurate position and velocity measurements for some period of time. The length of this period is dependent on the quality of the INS. Another advantage is that the accuracy degrades gracefully, and estimates can be given for how long the measurements will be within acceptable limits with the given number of satellites available. It is also possible to utilize the information from the integrated system to reducing the time it takes for the receiver to re-acquire lost satellite signals. More advanced architectures can even prevent the loss of satellite signals in the first place. The aim of this project is to determine the performance of tightly integrated GPS/INS systems for DP purposes as the number of satellites available is reduced from four to zero. The project will be divided into two phases:

Phase I: Van Testing

The GPS/INS Laboratory at Departments of Engineering Cybernetics and Telecommunication, NTNU has the capability to carry out van testing with realistic wave excitations.

The laboratory consists of the following key components:

  • A dedicated van with a platform on the roof which is controllable in pitch and yaw. The platform can be programmed with actual pitch and yaw measurements from sea trials or by using computer generated wave disturbances.
  • Two NovAtel RT-2 GPS L1/L2 receivers.
  • One IMU of tactical quality, Litton LN-250
  • One IMU of navigation quality, Litton LN-200

Litton LN-250

Litton LN-200

NovAtel RT-2 GPS L1/L2 Receivers

Phase II: Sea trials

Testing the drift of the integrated GPS/INS system on a DP vessel as the number of satellites goes below four.


  • [1] DnV Dynamic Posistining Systems Rules for Classifications of Mobile Offshore Units, Part 6. Ch. 7, new revision expected January 2001.
  • [2] Henk Nijmeijer and Thor I. Fossen (1999) (eds.). New Directions in Nonlinear Observer Design, Springer-Verlag London.
  • [3] Bjørnar Vik (1999). Nonlinear Design and Analysis of Integrated GPS and Inertial Navigation Systems, Dr.Ing. dissertation, Department of Engineering Cybernetics. NTNU
  • [4] BjørnarVik and T. I. Fossen (2001). Nonlinear Observer Design for Integration of GPS and Inertial Navigation Systems, Proc. of the IEEE CDC'2001, Florida, USA, pp. 2956-2961


Editor: Head of department, Professor Rolf Henriksen, Contact address: Webmaster, Updated: March 30, 2004 .