|
|
Explore!
 |
 |
Field Trials of the
THESEUS AUV
Theseus is a large, long-range AUV designed for operating in ice-covered waters. The vehicle's first dive was in September of 1994, and it has since undergone extensive field trials in BC waters. Arctic trials of the vehicle were completed in April of 1995. Following a description of the vehicle, its performance and capabilities, this paper describes some of the results obtained during both local and arctic sea trials.
Theseus Theseus consists of a segmented aluminum pressure hull with free-flooding fiberglass hull sections fore and aft. A general arrangement of the vehicle is shown in Figure 1.
Figure 1 The critical design parameters for Theseus are:
The nose contains a forward-looking obstacle avoidance sonar, a computer-controlled variable ballast tank, and transducers for acoustic telemetry and acoustic homing. The 6-section pressure hull houses the batteries, electrical power distribution, computer, and most of the "dry" electronics. The sensor bay immediately aft of the pressure hull contains the doppler sonar transducer and electronics, a medium frequency tracking transponder and an rf telemetry antenna. The payload bay contains fiber optic cable packs and buoyancy compensation tanks. The tail section houses a variable ballast tank, a low-frequency emergency transponder, thruster motor and gearbox. For flight control, Theseus uses six independent hydroplanes: two located forward at the nose, with four located aft. Eight different aftplane configurations were examined: '+' (cruciform), 'X', 'Y', inverted-'Y', 'T', inverted-'T', 'V' and inverted-'V'. Each of these configurations was evaluated using the following performance criteria: manoeuvring, susceptibility to planes damage when parking on the seabed or up against an ice cover, drag, redundancy, weight and previous experience with the configuration. An 'X' configuration appeared to provide the best manoeuvrability, parking capability and redundancy at the cost of some drag and weight.
The vehicle's payload section is currently configured to carry a number of buoyancy- compensated fiber optic cable packs. Many other payload configurations are possible, as long as the payload meets the following approximate requirements:
Vehicle Performance To take advantage of the redundancy provided by the hydroplanes configuration, four different manoeuvring control modes were designed, each using the planes differently to provide depth, pitch, roll and yaw control. These are summarized in Table 1.
Table 1 Manoeuvring Control Mode Hydroplane Assignments Manoeuvring mode 1 is used during most trials where water depths are less than 100m, as it is more conservative than modes 2, 3 and 4. These latter modes provide a greater amount of vertical performance due to their pitching action. Mode 4 provides the greatest redundancy. Table 2 summarizes the vehicle's manoeuvring performance for control mode 1. These results are all for non-combined manoeuvres, at the vehicle's normal operating speed of 4 knots, with 200 lbs positive buoyancy. Following control system tuning, steady-state oscillations of less than ñ1ø in each axis were attained. Vertical performance using manoeuvring modes 2 and 3 is typically twice that obtained with mode 1.
Table 2 Theseus Manoeuvring Performance (Mode 1) Measured power consumption for Theseus is shown in Table 3. Propulsive efficiency has been measured at approximately 80%. During both local and arctic trials, Theseus was outfitted with a NiCd battery bank for onboard power. These batteries provide approximately 20 kWhrs of energy, at a nominal DC bus voltage of 117V, which allows for approximately 5 hours of running time. For long missions, Theseus will be outfitted with 3 banks of Yardney AgZn cells, providing a nominal 280 kWhrs, enough for a mission of approximately 70 hours.
Table 3 Theseus Power Consumption
Communications By definition, an AUV does not require any communication with its human operators in order to function. However, it has been our experience that having one or more communication links with an AUV for monitoring (and possibly control) during sea trials is a necessity. These links are used to provide real-time or near real-time communications during initial checkout, control system tuning and testing of higher level control functions such as guidance and obstacle avoidance. Relying solely on an onboard "flight recorder" inevitably introduces significant delays in trials. Five different communication links have been used at various phases of vehicle trials:
A electrical umbilical, 150 meters in length, is used to provide 19.2 kbps serial communications with the vehicle during shop tests, on-deck tests and static in-water tests. The umbilical also provides a spare RS-422 serial channel for diagnostic purposes, coaxes for video and ethernet, and conductors which allow vehicle power to be supplied externally. During local sea trials, when the vehicle is first being operated in-water and when the vehicle control system is being tuned, real-time communications is preferred. For these situations, the vehicle is outfitted with a Dataradio asynchronous packet radio, which provides a full duplex, 9600 bps real-time communications link with the operator console on the support vessel. A faired, waterproofed rf coax, 25 meters in length, connects the vehicle radio to an antenna float (a modified surfboard), as shown in Figure 2. This system provides real-time communications with the vehicle operating at depths up to 10 meters. When the antenna float is no longer required, a smaller stub antenna is mounted directly on top of the vehicle, amidships. This provides the same real-time communications whenever the vehicle is floating on the surface. Between dives, this system can be used to upload vehicle logfiles or download new configuration software.
Figure 2: Radio telemetry using Antenna Float Acoustic telemetry is used when the vehicle is operating underwater where no other communications links are available. Theseus uses the Datasonics ATM-800 series acoustic modems, which provide a half-duplex link between the vehicle and operator console. To provide good communication paths when the vehicle is parked on the seabed or up against the ice, two acoustic telemetry transducers are mounted in the nose of the vehicle, one pointing up and the other pointing down. The operator or the vehicle computer can select which transducer is active depending on the current situation. When real-time communications are required and the vehicle is operated below the depth
range of the antenna float, two different methods of fiber optic telemetry are used. The
first method is shown in Figure 3. A Rochester Steel-Light fiber optic tether (0.060"
dia) runs from the vehicle up to the surface float, then to a support vessel following
closely behind the float. This fiber optic tether has a working load of 100 lbs, and is
strong enough to pull the float at 4 knots in a light chop. Using this system, the vehicle
has been successfully operated at depths of 20m with a 4:1 scope - deeper operations are
possible. For longer, deeper dives, fiber optic cable can be dispensed directly onto the
seabed from one or more cable packs located in the payload bay. Each cable pack is
surrounded by a buoyancy compensation tank which fills with water as the cable is
dispensed - this ensures that the vehicle's net buoyancy does not change significantly.
Many kilometers of cable, built by Laser Armor Tech, have been successfully dispensed. The
cable consists of a Corning single-mode fiber inside a stainless steel tube, surrounded by
an E-glass strength member and covered by a Surlin jacket. Figure 3: Fiber Optic Tether Telemerty
Navigation Performing long range, multi-day missions requires an accurate navigation system. Operating under a permanent ice cover places additional constraints on navigation. A hybrid solution to navigation was chosen: a medium-accuracy, doppler-aided inertial navigation system, and an acoustic positioning system for terminal guidance. Theseus uses a Honeywell MAPS Inertial Navigation Unit (INU) combined with an EDO 3050 doppler velocity sonar. Together, these two units provide a dead-reckoned position estimate which has an accuracy of approximately 1% of distance travelled. In order for the INU to provide an accurate true heading, it must be allowed to complete an internal gyro-compass alignment process. During this process, the INU uses its known initial position (supplied externally) along with its 3-axis rate gyros to sample the earth rate and determine the direction of true North. At relatively low latitudes (less than ± 55þ), the INU can align to its specified accuracy of 0.1ø within 8 minutes. At very high latitudes (> 80þ), field trials have concluded that a minimum of 120 minutes is required to gyro-compass align to specification. This long period is required due to the low earth rate at high latitudes. For terminal guidance, a Datasonics ACU-206 Tracking system is used. The transducer is mounted on the nose of the vehicle, and provides range and bearing to one or more ORE low frequency acoustic transponders deployed from the surface. When enabled by the vehicle's autonomous mission executor, the homing control system uses the range and bearing data to steer the vehicle to the transponder.
Transportation Deploying a large AUV for sea trials requires careful planning. Even during the early stages of the design process, one of the vehicle's design requirements was that it could be broken down into packages easily manageable for transportation. To mobilize the vehicle from ISER facilities to its support vessel Researcher, the vehicle is completely assembled and placed on a flatbed truck which transports it to the marina where Researcher is waiting. A Travel-Lift, commonly used to lift boats in and out of the water, lifts the vehicle off the truck and lowers it into the water. Theseus is then floated over to the nearby Researcher and lifted onboard using Researcher's crane. To mobilize the vehicle to the Arctic, the modularity of the vehicle was taken advantage of. The vehicle was broken down into five manageable groups: nose and hull section 1; hull sections 2, 3 and 4; hull sections 5, 6 and sensor bay; payload bay; and tail section. These were each strapped down onto custom-built, shock absorbent pallets with protective covers. Each pallet has special features designed to make transportation easier: skids for sliding across ice/snow, slots for a fork-lift, and lifting points for helicopter slinging. Flatbed trucks transported the pallets to the airport, where they were loaded onboard a C-130 Hercules aircraft. Once the C-130 arrived at its destination, a Bell 212 helicopter slung the pallets to the remote ice camp. At the ice camp, a Posi-Trak (a tracked fork-lift) was used to move the pallets on site.
Launch and Recovery Theseus has two different requirements for launch and recovery: deployment from a support vessel for local trials, and deployment through an ice hole. The vehicle normally rests on a set of cradles which roll along a pair of parallel tracks - these tracks can be installed on the deck of a ship or anchored into an ice surface. The cradles allow the vehicle hull to be separated at any joint, providing easy access. To launch the vehicle from the support vessel, a wire lifting harness is attached to lifting lugs on the vehicle fore and aft. The vessel's crane simply lifts the vehicle off its cradles, rotates it around and lowers the vehicle into the water. Two crew in a small zodiac release the harness and move the vehicle away from the ship. This procedure is reversed for recovery. Launch and recovery in the Arctic requires more preparation, but is in many cases easier than from a ship, since the ice provides a flat, stable platform. Preparation begins with locating a relatively flat area of ice and clearing it of snow. A large 36' x 65' WeatherHaven tent is then erected, followed by a pair of gantries, each with a 6 ton hoist. The ice hole, 5' x 42', is then made by cutting 3' x 5' ice blocks using a hot water slot cutter. The vehicle's cradle tracks are then assembled and anchored into the ice beside the ice hole. Finally, a wooden floor is put down over the inside of the tent to provide insulation and a skid-free work area. See Figure 4. To launch the vehicle:
Fugure 4: Under Ice Launch The recovery procedure begins when the vehicle returns to the ice hole using its acoustic homing system. Once within a pre-defined range of the homing transponder, the vehicle shuts off its thruster, trims itself to the required buoyancy, and parks on the seabed under the ice hole. To recover the vehicle:
Arctic Trials The Theseus vehicle, along with all its support equipment and personnel, was deployed in March of 1995 to an ice camp on the Lincoln Sea, at the northern end of Ellesmere Island, Canada. The large operations tent, lifting gantries, ice hole, vehicle tracks and cradles were already in place at the ice camp when the vehicle arrived. The ice camp was situated on ice 2 meters thick, on water 60 meters deep. Transportation of the vehicle to the ice camp went smoothly. One week was required to completely assemble and thoroughly checkout all systems. Following checkout, a series of static in-water tests were performed to verify that all subsystem operated correctly in the -1.5þ water. Field trials of the acoustic telemetry and acoustic homing systems were performed to verify acceptable acoustics in the trials range. Homing transponders were deployed at the launch site and at the outer-most extent of the range. Temporary ice camps were setup at several locations to provide deployment sites for vehicle operator console and acoustic telemetry systems. Five dives were performed during April of 1995, three autonomously. Each dive was "out and back", travelling up to 5.7 km at depths up to 65 meters. During each autonomous dive, the vehicle dispensed fiber optic cable, providing a real-time telemetry link for monitoring. Immediately following vehicle release, the vehicle's autonomous mission executor executed a series of mission verbs, which specified the planned autonomous behaviours. Each verb defined the position, speed and vertical control setpoints. Since the vehicle's autonomous acoustic homing capabilities were not fully ready in time for the spring arctic trials, human intervention was required for terminal manoeuvring. As the vehicle autonomously approached the final waypoint (the launch site), the operator took control over the fiber optic cable and steered the vehicle towards the transponder, using the range and bearing feedbacks from the homing system. Using heading setpoints to keep the transponder bearing at 0ø (relative), it was possible to bring the vehicle to a stop directly under the launch ice hole. The vehicle was then commanded to park itself on the seabed. During these dives, a negative buoyancy of approximately 50 lbs was found to provide a gentle landing with a minimum of descent time. Figure 5 [not available]shows the track plot of the vehicle during its longest under-ice dive, as it travelled through six geographic waypoints, starting and ending at the launch site ice hole. The total distance travelled was approximately 5700 meters. The farthest distance from the launch ice hole was 2720 meters. The arctic water, exceptionally clear of suspended matter, provided excellent visibility. However, it appeared to present a problem for the doppler sonar. At altitudes less than 10 meters, the doppler would often report erratic bottom-tracked velocities and incorrect altitudes. This was never observed in local, murkier waters. The manufacturer indicated that this was possibly a result of the minimum transmitter power setting being too high when the doppler was in automatic tracking mode - a higher amount of backscatter is being expected. This will be investigated further. Acoustic telemetry coverage was available throughout the trials range (1-2 km) with a field console located roughly in the middle of the trials range. The Datasonics acoustic modems appeared to operate best when set to 150 or 300 baud, with MFSK. Field trials have indicated that ranges of 2-4 km can be expected in deeper waters.
Summary Theseus was designed to meet a number of specific requirements, many of which were specified to permit autonomous operation in ice-covered waters. Following a series of field trials in BC waters, Theseus was transported to the high arctic for further trials. These trials proved to be very successful. The vehicle's mechanical and electrical design permitted disassembly, transportation and re-assembly at an ice camp without any significant problems. Vehicle launch and recovery procedures went better than planned. It was successfully demonstrated that under-ice operations of a large AUV are feasible. The remainder of 1995 will see full automation of acoustic homing and obstacle detection/avoidance, calibration of the navigation system, integration of AgZn batteries and enhanced fault management.
Acknowledgements The development of the Theseus AUV was performed under contract to the Department of National Defence (Canada). The entire Theseus team would like to acknowledge the sponsorship, support and input from the scientists and engineers from the Esquimalt Defence Research Detachment (formerly Defence Research Establishment Pacific). |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright International Submarine Engineering Ltd., 1734 Broadway
Street, Port Coquitlam, B.C. Canada, V3C 2M8
Phone: (604) 942-5223 Fax: (604) 942-7577 Email: info@ise.bc.ca
1999 All Rights Reserved
