Navigation

The design of the navigation system of an AUV has to take into account several issues such as desired positioning and attitude accuracy, size, weight, and power consumption of available sensors and systems, and also overall cost. Available technologies include inertial navigation systems, digital compasses, tilt sensors, pressure cells, acoustic positioning systems, and Doppler based velocity meters. The MARES navigation package was selected taking into account the above mentioned issues but also the envisaged missions and application scenarios.

3.1 Sensors and systems

To estimate its position in real time, the vehicle carries a pressure sensor, a digital compass with a set of tilt sensors, and an acoustic system for long baseline (LBL) positioning (Vaganay et al., 1996). Vehicle depth is directly given by the pressure cell, while roll and pitch angles are obtained from the set of tilt sensors. The estimation of the horizontal position and velocity also employs dead-reckoning data.

The pressure cell has a centimetre level accuracy and outputs new data at a rate exceeding 100 Hz. This allows for a quite accurate vertical positioning and also for a software based estimation of heave velocity. The digital compass and tilt sensors unit provides data at 20 Hz with accuracies better than 0.5°.

The acoustic system is a second generation of multi frequency boards, developed by the Ocean Systems Group, following some excellent results with previous versions (Cruz et al., 2001). These boards are installed on-board the vehicle but also on the acoustic beacons that are deployed in the operation area. This system is completely reconfigurable (pre mission or on line programming of detection/reply frequency pairs, channel sensitivity, etc.), allows complete control of signal transmission times, and provides access to low level signal detection data.

3.2 Horizontal position estimation

The real time estimation of the horizontal position of MARES is computed by a Kalman filter based algorithm (Matos et al., 1999) that combines dead-reckoning data (vehicle surge velocity and heading) with range measurements based on times of flight of acoustic signals. For a typical operation, two acoustic beacons are deployed in the operation area in a way such that the AUV, during the execution of its mission, does not cross the line connecting them. In this way, the range measurements between the AUV and both beacons unambiguously determine its horizontal position. Usually, these beacons are attached to surface buoys and signal detection and transmission time are transmitted to a shore station by a wireless link. This information allows for the real time external tracking of the AUV according to the algorithm proposed in (Cruz et al., 2001).

Since velocity data is obtained with respect to water, and no direct measurement of water currents is available, the horizontal components of such current are also estimated in real time. The navigation algorithm periodically updates the real time estimates of the horizontal

Fig. 4. Surface buoy attached to an acoustic beacon.

position of the vehicle (x, y) and of the water current (cx, Cy) at a 20 Hz rate, according to the following dynamic system, x = v cos ^ + cx y = v sin y + cy (1)

At the same time, the associated error covariance matrix is updated. Whenever a new range measurement is obtained by the vehicle, the state variables and the error covariance matrix are corrected accordingly. This is accomplished by an iterative procedure based on the extended Kalman filter algorithm (Gelb, 1994), since range measurements are related to the state variables of the filter by the nonlinear relationship r, =V (x - xt )2 + (y - y, )2 + (z - zt )2 (2)

where (x, y, z) is the 3D position of the AUV, (xi, yi, Zi) is the 3D position of beacon i, and Ti is the predicted range measurement. Fig. 5 presents the range measurements between the AUV and two navigation beacons during an autonomous operation. It shows erroneous measurements caused multipath propagation of acoustic waves, as well as intervals of time when there are no measurements from at least one of the beacons, typically due to adverse propagation conditions. In order to increase the accuracy of the horizontal positioning, a validation mechanism is used to maximize the probability of rejecting those erroneous measurements.

Even in ideal propagation conditions, and neglecting error sources associated with the electronics of the transmission and reception acoustic boards, acoustic range measurements are directly affected by the sound speed in water. This, is turn, depends on the characteristics of the medium (temperature, salinity, pressure), and variations as high as 2%

are not unusual. It is therefore mandatory to determine the sound speed in the operation area and adjust range measurements accordingly. This can be done either by comparing range measurements close to the surface with DPGS based distance measurements, or by measuring the most relevant water characteristics with a CTD (conductivity, temperature, and depth) sensor. These calibration procedures allow horizontal position accuracies about 1 to 2 meters with respect to the positions of the beacons.

Was this article helpful?

0 0
Learn Photoshop Now

Learn Photoshop Now

This first volume will guide you through the basics of Photoshop. Well start at the beginning and slowly be working our way through to the more advanced stuff but dont worry its all aimed at the total newbie.

Get My Free Ebook


Post a comment