Reference System and Sensor Data Processing

Measurements of a vessel's position and heading at any point in time are essential for dynamic positioning.

Several different position-reference systems are normally used with the SDP system. The first position-reference system selected and accepted for use with the SDP system becomes the reference origin (the origin of the internal coordinate system).

As several different position-reference systems may be used, the SDP system carries out a series of tests on each of these systems to check whether their position measurements are accurate enough for use.

If the tests show that some measurements are outside acceptable accuracy limits then these measurements are disregarded by the SDP system. In addition, the SDP system calculates a variance for each of the position-reference systems in use. The variance of a position-reference system is a filtered value based on the squared difference between the measured position and the estimated position from the mathematical model (mean square value). This variance is used in the variance test (described below) as well as for calculating a weighting factor.

The SDP system uses this weighting factor to assign weightings to the position-reference systems. The higher the system's variance, the lower its weighting factor. Most emphasis is placed by the SDP system on the reference system with the highest weighting factor while least emphasis is placed on the one with the lowest weighting. In order to calculate whether or not the position measurements from a position-reference system are accurate, the following tests are carried out by the SDP system:

• Prediction Test (short-term accuracy assessment);
• Variance Test (long-term accuracy assessment);
• Median Test.

• Prediction Test (short-term accuracy assessment). The prediction test detects sudden jumps in the measured values and immediately rejects those that lie outside the limits:

  The prediction reject limit is established as a function of the estimated position in the mathematical model and the current accuracy of this estimate. The latter depends on the overall variance of the reference system readings.

• Variance Test (long-term accuracy assessment). The variance test detects if the variance in the measured values exceeds the reject limit. The variance reject limit is continuously calculated and is based on the variance of the position-reference system with the lowest variance.

• Median Test. When three or more reference systems are in use, the median test detects when measurements from one reference system differ from the others. In this test, measurements are used independent of the mathematical model. Therefore, slow drift in systems, even in systems with low variance, will be detected by this test.

For all three tests, alarms are reported to the operator when the defined limits are exceeded. Before the tests are carried out, a received position-reference signal is checked for validity.