89
A School of Robotic Fish for Pollution Detection in Port
for example, the kinematics of fish. It should be noticed that both steady and
unsteady swimming behaviors of fish can be realized by BCF and MPF propulsions or their combination. This reflects the complexity of fish swimming
movements (Drucker and Lauder 2002).
To mimic fish swimming abilities, current robotic fish projects have been
focused on three aspects: (1) fish locomotion and hydrodynamics, (2) artificial muscle technologies, and (3) sensor-based control mechanisms. Most
researchers have worked on one of these aspects. Man-made ships and
underwater vehicles are based on steady-state hydrodynamics for the high
stability and high loading capability and are unable to match the turning and
maneuvering capability of real fish. Therefore, the construction of the robotic
fish relies on a full understanding of undulatory or oscillatory movements of
real fish, corresponding hydrodynamics, new materials, and advanced control mechanisms.
Different from most previous research, we focus on two levels of complexity of fish locomotion systems; that is, swimming patterns for propulsion
and multiple behaviors for temperate features. Layered control architecture
has been developed at Essex by Liu, Hu, and Gu (2006) to realize fish-like
swimming behaviors, especially unsteady behaviors such as sharp turning
and fast starts. Our research has focused on undulatory BCF propulsion,
that is, carangiform, in order to realize it in our robotic fish and study its
advantages in the engineering field. Other types of BCF and MPF will be
investigated gradually in our future research.
5.3 Biologically Inspired Design
We have built a number of generations of robotic fishes at Essex. Figure 5.3
shows six generations. In general, these robotic fish are about 50 cm long and
have three or four powerful R/C servo motors and 2 DC motors. Servo motors
are concatenated together in the tail to act as joints, as shown in Figure 5.4.
Additionally, one DC motor is fixed in the head to change center of gravity
of the fish for diving and another DC motor controls the micropump. On the
back of the fish body, a dorsal fin is fixed vertically to keep the fish from swaying. The high quality of the servo motors and the very soft structure of the tail
make it possible for the robotic fish to bend its body at a large angle in a short
time, which is novel, and nobody, as far as we know, has done this before us.
Each robotic fish has over ten embedded sensors: one gyroscope, one pressure sensor, two position sensors, two current sensors, one voltmeter, four
infrared sensors, and one inclinometer. These embedded sensors enable the
fish to detect its depth, the yaw/roll/pitch angle of its body, and the distance
to the obstacle in front of it. Additionally, the servo position and current
consumption information can be obtained. Bluetooth and RS232 serial ports
A School of Robotic Fish for Pollution Detection in Port
for example, the kinematics of fish. It should be noticed that both steady and
unsteady swimming behaviors of fish can be realized by BCF and MPF propulsions or their combination. This reflects the complexity of fish swimming
movements (Drucker and Lauder 2002).
To mimic fish swimming abilities, current robotic fish projects have been
focused on three aspects: (1) fish locomotion and hydrodynamics, (2) artificial muscle technologies, and (3) sensor-based control mechanisms. Most
researchers have worked on one of these aspects. Man-made ships and
underwater vehicles are based on steady-state hydrodynamics for the high
stability and high loading capability and are unable to match the turning and
maneuvering capability of real fish. Therefore, the construction of the robotic
fish relies on a full understanding of undulatory or oscillatory movements of
real fish, corresponding hydrodynamics, new materials, and advanced control mechanisms.
Different from most previous research, we focus on two levels of complexity of fish locomotion systems; that is, swimming patterns for propulsion
and multiple behaviors for temperate features. Layered control architecture
has been developed at Essex by Liu, Hu, and Gu (2006) to realize fish-like
swimming behaviors, especially unsteady behaviors such as sharp turning
and fast starts. Our research has focused on undulatory BCF propulsion,
that is, carangiform, in order to realize it in our robotic fish and study its
advantages in the engineering field. Other types of BCF and MPF will be
investigated gradually in our future research.
5.3 Biologically Inspired Design
We have built a number of generations of robotic fishes at Essex. Figure 5.3
shows six generations. In general, these robotic fish are about 50 cm long and
have three or four powerful R/C servo motors and 2 DC motors. Servo motors
are concatenated together in the tail to act as joints, as shown in Figure 5.4.
Additionally, one DC motor is fixed in the head to change center of gravity
of the fish for diving and another DC motor controls the micropump. On the
back of the fish body, a dorsal fin is fixed vertically to keep the fish from swaying. The high quality of the servo motors and the very soft structure of the tail
make it possible for the robotic fish to bend its body at a large angle in a short
time, which is novel, and nobody, as far as we know, has done this before us.
Each robotic fish has over ten embedded sensors: one gyroscope, one pressure sensor, two position sensors, two current sensors, one voltmeter, four
infrared sensors, and one inclinometer. These embedded sensors enable the
fish to detect its depth, the yaw/roll/pitch angle of its body, and the distance
to the obstacle in front of it. Additionally, the servo position and current
consumption information can be obtained. Bluetooth and RS232 serial ports