Integrated inertial navigation system structure

An INS calculates position, velocity and attitude using high frequency data from an Inertial Measurement Unit (IMU). An IMU consists of three accelerometers measuring specific force and three gyros measuring angular rate. If left unaided the INS will, after a short period of time, have unacceptable position errors. The position error growth is determined by the class of the IMU, see Table 1. Currently the best available IMU with a feasible size for an AUV gives a position error growth in the order of 1 nmi/h (1 ct), when integrated in an INS. To reduce this error the INS needs to be aided by redundant sensor measurements. These sensors are usually integrated with the INS through a Kalman filter, which performs this integration in a mathematically optimal manner.

Fig 1 shows a schematic view of the HUGIN integrated INS, where the Kalman filter is based on an error-state model and provides a much higher total navigation performance than is obtained from the individual sensors alone.

IMU class

Gyro bias

Accelerometer bias

>10 nmi/h

1°/h

1 mg

1 nmi/h

0.005°/h

30 ^g

Table 1. Classification of feasible IMUs for AUVs 2.2 DVL aided INS

The solution for most modern AUVs is a low drift Doppler Velocity Log (DVL) aided INS that can integrate various forms of position measurement updates. DVL accuracy is dependent on acoustic frequency. Higher frequency yields better accuracy at the cost of decreased range, as illustrated in Table 2. Prioritisation between range and accuracy is dependent on the application.

Table 1. Classification of feasible IMUs for AUVs 2.2 DVL aided INS

The solution for most modern AUVs is a low drift Doppler Velocity Log (DVL) aided INS that can integrate various forms of position measurement updates. DVL accuracy is dependent on acoustic frequency. Higher frequency yields better accuracy at the cost of decreased range, as illustrated in Table 2. Prioritisation between range and accuracy is dependent on the application.

Frequency

Long term accuracy

Range

150 kHz

±0.5% o.s.

± 2 mm/s

425 - 500 m

300 kHz

±0.4% o.s.

± 2 mm/s

200 m

600 kHz

±0.2% o.s.

± 1 mm/s

90 m

1200 kHz

±0.2% o.s.

± 1 mm/s

30 m

Table 2. DVL range and accuracy depends on its acoustic frequency (o.s is of speed).

Table 2. DVL range and accuracy depends on its acoustic frequency (o.s is of speed).

Inertial Navigation System

Underwater transponder positioning

Fig. 1. Block diagram of the HUGIN integrated inertial navigation system.

Underwater transponder positioning

Fig. 1. Block diagram of the HUGIN integrated inertial navigation system.

The horizontal position drift in a DVL aided INS is determined by the error in the estimated Earth-fixed velocity (i.e. North and East velocity). The main contributors to this error are body-fixed velocity error, and heading error.

The error in estimated body-fixed velocity is mainly determined by the low-frequency error in the DVL itself. This error is not observable if the vehicle is travelling along a straight line and without position aiding. High frequency velocity errors on the other hand, are estimated by means of the IMU. Without aiding, even the most accurate INS will have a velocity uncertainty larger than the DVL accuracy after a short period of time. This means that the AUV's along track position error is quickly dominated by the low-frequency error of its DVL.

Consider an AUV travelling at 2 m/ s. According to Table 2, the along track error drift of a 300 kHz DVL is 28.8 m/hour, while a 1200 kHz DVL has a drift of 14.4 m/hour. The error in heading is determined by the gyrocompassing capability of the integrated system. The heading estimation error will typically be of low frequency, corresponding to non-observable gyro bias dynamics. Referring to Table 1, a 1 nmi/h navigation class IMU typically gyrocompasses to an accuracy of <r(8y) = 0.02 deg sec(lat). This corresponds to an error drift of ct(8^)-100 % of traveled distance in radians). At 45° latitude, this equals

0.05% of traveled distance, or 3.4 m/hour at 2 m/s AUV speed.

The Kalman filter can compensate for part of the scale factor error of the DVL when running more complex missions than a straight line, or when position updates are available. The expected error drift of a state-of-the-art DVL-aided INS is summarised in Table 3, both for a typical straight-line transit phase and for a typical lawnmower survey pattern.

Position error drift (% of traveled distance)

Lawnmower Straight line pattern with 1 km lines

Along track 0.11% 0.01% _Across track_0.03%_0.001%_

Table 3. Typical position error drift for a high quality DVL-aided INS 2.3 Model and DVL water-track aided INS

As mentioned in Section 2.2, a DVL with bottom-track is an efficient aiding tool for limiting the position error drift of an INS. Even when including this capability however, situations will occur where it fails to work or measurements are discarded due to decreased quality. This will for instance occur when operating above the sensor range or over very rough bathymetry due to loss of bottom lock. In either case, in the absence of DVL bottom-track data, alternative velocity information is required to achieve an acceptable low drift INS solution between position updates. Two especially efficient alternatives are to utilize a kinetic vehicle model describing the vehicle dynamics or by including measurements from a DVL with water-track. The latter measures the velocity of the vehicle relative to the surrounding fluid. Similarly, the kinetic vehicle model predicts the same water-relative velocity by solving a set of ordinary differential equations (with measured actuation signals, e.g. propeller RPM and fin deflections as inputs). The effect of including the kinetic model in the navigation system can be seen in Fig 2, where USBL position measurements were lost for 30 minutes. The system without any velocity aiding quickly becomes useless, while the navigation accuracy in the system with model aiding remains good throughout the USBL outage.

Logged from 0lï-0et-2006 04:20:29 to 06-0ct-200l3 07:20:28 Logged from 06-0ct-2006 04:29:29 to 06-0et>20015 07:20:28

Logged from 0lï-0et-2006 04:20:29 to 06-0ct-200l3 07:20:28 Logged from 06-0ct-2006 04:29:29 to 06-0et>20015 07:20:28

Underwater Scuba Diving Vechicles

-500 0 500 1000 1500 -500 0 500 1000 1500

Fig. 2. Navigation during a USBL outage: The left plot shows the system without modelaiding and right when it is included. The red (solid) trajectory is considered ground truth. The blue (o) data show the USBL measurements. The segment without measurements corresponds to 30 minutes. The green (dashed) is the real-time navigation solution.

-500 0 500 1000 1500 -500 0 500 1000 1500

Fig. 2. Navigation during a USBL outage: The left plot shows the system without modelaiding and right when it is included. The red (solid) trajectory is considered ground truth. The blue (o) data show the USBL measurements. The segment without measurements corresponds to 30 minutes. The green (dashed) is the real-time navigation solution.

Note that when aiding the INS with both DVL water-track measurements and data from the kinetic model, it is important to also include real-time sea current estimation in the navigation system. The sea current will often constitute the dominant error source when aiding the INS with water-relative velocity data, in particular when applying DVL water track data. For additional information on model-aiding and sea current estimation, the reader may refer to (Hegren^s et al., 2007) and (Hegren^s et al., 2008).

+1 0

Responses

  • pasqualina li fonti
    HOW INERTIAL NAVIGATION SYSTEM WORKS , BLOCK DIAGRAM?
    6 years ago
  • Mentha Sandyman
    What is long term accuracy dvl?
    5 years ago

Post a comment