Feature ArticleAn Automated Calibration System For Fisheries Acoustic Surveys
Stan D. Tomich
Lawrence C. Hufnagle Jr.
Northwest Fisheries Science Center
NOAA Fisheries Service
Current at-sea calibrations of ship-mounted echosounders are based on the standard target calibration method, which involves deploying one or more standard targets (calibration spheres) to depths that are in the farfield of the transducers. The spheres are tethered with three lines fed by three downriggers mounted on the port and starboard sides of the vessel, bracketing the centerboard, allowing an operator to control the spheres' location. The efficiency of the calibration depends heavily on the experience of the operator, especially when calibration of a hull-mounted multibeam sonar is involved.
To more effectively conduct the acoustic system calibration at sea, the Fishery Resources Analysis and Monitoring (FRAM) division of the Northwest Fisheries Science Center developed the automated acoustic calibration system (AACS).
The AACS includes three highly modified commercially available fishing downriggers, each of which contains mechanical and electrical components that have been integrated with a microcontroller, allowing precise line control and monitoring. Custom-designed software, accessed via a graphical user interface (GUI), can control each of the three downriggers independently so that the calibration spheres can be moved to a desired location in three dimensions within the acoustically insonified volume. The software calculates the desired line movement for each of the downriggers to control the sphere movement horizontally while maintaining proper depth under the transducers of the ship. Line tension is constantly monitored to prevent line fouling and excessive strain on components and to detect possible line snagging on the ship's hull.
The AACS can improve the efficiency of the calibration, especially for mapping the beam pattern of transducers mounted on the ship's centerboard. The system is capable of being operated by a less experienced staff, and it can control calibration-sphere position more precisely and with much less fluctuation than conventional manually operated at-sea calibration. This allows for more repeatable measurements and maintains sphere position at approximately the same range.
Difficulties of Manual Calibration
Sphere calibration of scientific echosounders has been recognized by fisheries acoustic scientists as being critical to obtaining accurate quantitative acoustic measurements of fish backscatter, which can then be used to determine biomass for stock assessment.
Manual calibration involves moving a calibration sphere below a ship-mounted transducer within the acoustically insonified sample volume, allowing researches to map the beam pattern of the transducer. This method has become the standard for fisheries scientific echosounder calibration. However, moving the sphere to completely map the beam pattern can be a long, tedious process requiring significant expertise.
Manual calibration systems composed of three fishing rods and reels or downriggers have traditionally been used to deploy and move the sphere during calibration. This method was slow and labor-intensive, requiring an experienced operator to direct the sphere movement by instructing other operators at the downriggers to let line out or bring line in to reposition the sphere.
Improvements in this method were made with the introduction of remotely controlled electric motors and reels or downriggers connected to a central controller. An experienced operator was still required to move the sphere in the beam pattern of the transducer during the calibration. Though a great improvement, this method was still time-consuming and variable, dependent on the experience and skill of the operator to move the sphere efficiently to cover the beam pattern.
Photo of the motorized fishing downrigger. A microcontroller is included in the electronic unit, while the line-tension and counter-sensing unit is on the end of the downrigger rod.
Automated Calibration System
As fisheries scientific multibeam, broadband and multifrequency echosounders have advanced, FRAM researchers realized that system calibrations were going to become more complicated and time-consuming.
FRAM researchers began developing an acoustic calibration system that increased efficiency by reducing the time and operator experience necessary to perform high-quality calibrations of new and existing systems.
The manually controlled system was modified by adding sensors to measure line movement with a centimeter resolution and to monitor line tension constantly. Manual operation was controlled by a computer running MATLAB software, which allows downrigger control in a manual method similar to the old controller. The MATLAB controller is capable of running several automated swing patterns or moving the sphere to a selected location within the beam pattern and maintaining the target sphere depth (more precisely, the range between the calibration sphere and the designated transducer). This automated method reduces the experience required by the calibration operator and reduces the time required to cover the entire beam pattern.
Field acoustic calibration systems commonly involve three motorized fishing downriggers that are mounted on the rails of both the port and starboard sides of a fish survey or research vessel. Calibration target spheres are attached to lines fed by the downriggers. The AACS builds upon and modifies the typical calibration system in two major categories: hardware and software.
Hardware. The AACS consists of a laptop control computer, communications cables and three electronically controlled fishing downriggers, specifically Cannon Mini-Mag downriggers manufactured by Johnson Outdoors Marine Electronics Inc. (Racine, Wisconsin). The electronics that control the fishing downriggers have four main functions: keeping track of line in/out movement, providing communications to and from the laptop computer, controlling the motor, and detecting the tension of the three downrigger lines.
The electronically controlled downriggers are commercially available fishing downriggers in which motor control has been automated. The downrigger motor is powered by an electronic circuit—a modified H-bridge design—that allows switching of power and ground to drive motors in two directions. In this application, two relays can supply positive power to either side of the motor, giving it a clockwise or counterclockwise rotational drive preference. Using electronic switching at the ground side of the motor with electronic components, variable power and speed can be accomplished. Pulse duration and frequency are used to control the speed and motor torque: A longer pulse provides more torque and a higher frequency increases speed. The motor drive is activated by an embedded microcontroller (PIC 16F877A) through software written by FRAM researchers.
In concert with driving the motor, a sensor monitors the line as it moves around an idler wheel. Line can be measured with this system in approximately 1.27-centimeter increments. As line is played out from the line reel, it passes over the idler wheel, which has a sensor that can measure shaft angle. As shaft angle increases or decreases, line length can be computed based on the degree of rotation and idler wheel diameter.
A separate microcontroller keeps track of measured line as it winds and unwinds from the line reel—the same microcontroller monitors the tension sensor. The additional tension sensor monitors the tension arm, which provides feedback to the overall system about line tension. Under conditions of no load, linear springs hold the tension arm up in a horizontal, or nontension, state. As line tension increases, the tension arm rotates and the spring stretches. The angle at which the tension arm is moved can be translated to approximate the line tension based on the load value of the spring. For this system, line tension can be measured from zero to five kilograms over a tension-arm movement of 30°.
Communication commands and the status of the downrigger motor are sent to and received from the controlling laptop computer via RS-232 serial communications. Commands to the hardware include distance to move, direction, motor speed and various other queries. Typical commands would include a motor speed (and torque) value and a prescribed distance to move in centimeters. Error conditions can result from line movements that exceed maximum allowed tension, in which case the hardware automatically stops line movement until the problem is resolved.
Software. The software is a GUI-driven program written in MATLAB 7.2 scripts. The software was designed to provide an intuitive, easy-to-use interface, allowing a variety of operation modes. The key feature of the software is to control the calibration sphere, moving it to a specified location while maintaining it at a constant range between the calibration sphere and the designated transducer.
Through the software, the AACS can achieve a number of functionalities that are challenging to accomplish through manual methods.
One of these functionalities is maintaining a constant range between the calibration sphere and the designated transducer. This is an important function that computer-controlled movement is capable of achieving, but one that is extremely hard, if not practically impossible, to achieve with the traditional, manually operated calibration method. This feature allows users to set a narrower depth range to exclude false echoes, which could interfere with the echoes from the calibration sphere. It also reduces any potential errors introduced by using a different range, such as time-varied gain and seawater absorption corrections.
Another AACS feature is that the software can move the calibration sphere to a specific location within the beam or to a selective swing pattern that allows the sphere to move around beyond the main lobe in order to map the transducer beam patterns.
Finally, the software can calibrate the range by reading real-time echogram data to ensure the range determined by the downrigger readings matches the actual range determined acoustically. In FRAM's applications involving the geometry of the ship and downrigger positions, the error introduced by equations that determine sphere position were very small (less than one millimeter).
Diagram of the motor control unit.
FRAM researchers conducted a computer simulation to explore the theoretical accuracy of the AACS without real-world variables such as current or ship motion. For the simulation, the following assumptions are made: First, that the readings from the downrigger moving counts are consistent with the actual move. This function has been tested repeatedly on land and consistently performed flawlessly. Second, that the position of the sphere is the same as the junction of the three lines. In other words, the swing of the sphere from current and motor movement is ignored. Finally, that the stretch of line is constant. This assumption will potentially introduce some error, but the error can be accounted for when the range between the sphere and the transducer is calibrated, with the range determined by the real-time echogram.
The position of the sphere can be obtained by incorporating the relative movement and is recalculated after each step to ensure that the sphere is approaching the correct final position. At each step, the tension readings provided by the microcontroller are checked. If the tension of a particular downrigger is too high, the 'IN' action that draws line in will be ignored. If the tension is too low, the 'OUT' action that allows line out will be ignored. This will cause the movement of the sphere to follow a more or less granular pattern, but will prevent the lines from tangling and the motor controller electronics from overheating.
A field test was conducted on the NOAA Ship Miller Freeman on June 29, 2009, in Monterey Bay, California. The swing operation was conducted with the sphere positioned about 36 meters below the centerboard. Instead of using a swing command, NOAA researchers moved the sphere by directly controlling the downrigger. The results were the output of Simrad (Horten, Norway) ER60 software with a Simrad EK60 echosounder system, and the coverage was quite uniform within the mapped beam region (9° in polar angle). The coordinates of the downriggers relative to the centerboard were measured using a laser measuring device when the ship was in dry dock.
Although the downrigger electronics can track line movement and tension, the actual position of the calibration spheres—and in turn, the overall accuracy of the positioning system—depends on the line stretching under tension and the stability of the sphere after each movement, both of which are a function of the spheres' speed and the current. Because of this, the actual location of the calibration sphere is determined by the echosounder (a split-beam system), which allows researchers to control the lines more accurately than the AACS electronics and software alone.
Although bench testing of the integrated hardware and software has been conducted successfully, due to the time constraints, the team was not able to test all aspects of this system during the cruise. Future work will include tests of various fully automated swing routines used to map the beam pattern.
The NOAA research team has developed an automated, easy-to-operate calibration system. The simulation and the initial field tests provided promising results. This automated calibration system will not only simplify future calibrations, but will also improve the accuracy and consistency of calibration results. Future work will focus on testing a fully integrated system, especially the fully automated sphere swing.
The team believes that the success of such a system will have a significant influence on field calibration within the fisheries acoustics community.
The authors would like to acknowledge Steve de Blois and Matt Barnhart for assistance with setup and testing of the AACS; facilities manager Jim Herkelrath, facilities team supervisor John Rheaume and the entire facilities team for testing assistance; Lt. Sean D. Cimilluca of the NOAA Diving Center for assistance conducting system tests in the dive tank; and commanding officer Mike Hopkins and the crew of the Miller Freeman for assistance conducting the field calibration. Finally, the authors thank the Northwest Fisheries Science Center, FRAM division, for providing funding and support.
Stan D. Tomich, an advanced technology engineer for the Northwest Fisheries Science Center, Fishery Resources Analysis and Monitoring division, specializes in developing unique instrumentation that supports scientific research. He received a B.S. in electronics engineering from Arizona State University, Tempe, in 1980.
Lawrence C. Hufnagle Jr., a physical scientist with the Northwest Fisheries Science Center, Fishery Resources Analysis and Monitoring division, has a research focus on multibeam sonar and is also working on developing calibration methods for multibeam sonar systems to be used in quantitative fisheries surveys. He holds a B.Sc. in biology and chemistry from Bridgewater State College.
Dezhang Chu, who leads the acoustics team at the Northwest Fisheries Science Center, Fishery Resources Analysis and Monitoring division, currently studies acoustic scattering, characterization and classification of fish. He received a bachelor's in electrical engineering from the China University of Geosciences, Wuhan, in 1982 and a Ph.D. in geophysics from the University of Wisconsin, Madison, in 1989.