Jianhong Liang

Hydrodynamic investigation of a self-propelled robotic fish based on a force-feedback control method.

We implement a mackerel (Scomber scombrus) body-shaped robot, programmed to display the three most typical body/caudal fin undulatory kinematics (i.e. anguilliform, carangiform and thunniform), in order to biomimetically investigate hydrodynamic issues not easily tackled experimentally with live fish. The robotic mackerel, mounted on a servo towing system and initially at rest, can determine its self-propelled speed by measuring the external force acting upon it and allowing for the simultaneous measurement of power, flow field and self-propelled speed. Experimental results showed that the robotic swimmer with thunniform kinematics achieved a faster final swimming speed (St = 0.424) relative to those with carangiform (Stxa0=xa00.43) and anguilliform kinematics (St = 0.55). The thrust efficiency, estimated from a digital particle image velocimetry (DPIV) flow field, showed that the robotic swimmer with thunniform kinematics is more efficient (47.3%) than those with carangiform (31.4%) and anguilliform kinematics (26.6%). Furthermore, the DPIV measurements illustrate that the large-scale characteristics of the flow pattern generated by the robotic swimmer with both anguilliform and carangiform kinematics were wedge-like, double-row wake structures. Additionally, a typical single-row reverse Karman vortex was produced by the robotic swimmer using thunniform kinematics. Finally, we discuss this novel force-feedback-controlled experimental method, and review the relative self-propelled hydrodynamic results of the robot when utilizing the three types of undulatory kinematics.

Parameter Synthesis of Coupled Nonlinear Oscillators for CPG-Based Robotic Locomotion

This paper presents a numerical method for parameter synthesis of a central pattern generator (CPG) network to acquire desired locomotor patterns. The CPG network is modeled as a chain of unidirectionally or bidirectionally coupled Hopf oscillators with a novel coupling scheme that eliminates the influence of afferent signals on amplitude of the oscillator. The method converts the related CPG parameters into dynamic systems that evolve as part of the CPG network dynamics. The frequency, amplitude, and phase relations of teaching signals can be encoded by the CPG network with the proposed learning rules. The ability of the method to learn instructed locomotor pattern is proven with simulations. Application of the proposed method to online gait synthesis of a robotic fish is also presented.

Learning fish-like swimming with A CPG-based locomotion controller

This paper presents a learning method to acquire fish-liking swimming with a CPG-based locomotor controller. The proposed method converts the related CPG parameters into dynamical systems that evolve as part of the CPG network dynamics. The teaching signals are derived from the kinematic model of carangiform swimming with trajectory approximation method. A novel coupling scheme for the CPG network, which are modeled as a chain of coupled Hopf oscillators is proposed to eliminate the influence of afferent signals on amplitude of the oscillator. The learning rules of intrinsic frequency, coupling weight and amplitude are formulated with phase space representation of the oscillators. The frequency, amplitudes and phase relations of the teaching signals can be encoded by the CPG network with the adaptation mechanisms. Numerical experiments are carried out to validate the effectiveness of the proposed learning rules.

CFD based investigation on the impact acceleration when a gannet impacts with water during plunge diving

Plunge diving is the most commonly used feeding method of a gannet, which can make the gannet transit from air to water rapidly and successfully. A large impact acceleration can be generated due to the air-to-water transition. However, the impact acceleration experienced by the gannet during plunge diving has not been studied. In this paper, this issue is investigated by using the CFD method. The effect of the dropping height and the water-entry inclination angle on the impact acceleration is considered. The results reveal that the impact acceleration along the longitudinal body axis increases with either of the two parameters. The peak time decreases with the dropping height. A quadratic relation is found between the peak impact acceleration and thexa0initial water-entry velocity. According to the computation, when the dropping height is 30xa0m (most of gannets plunge from about this height), the peak impact acceleration can reach about 23xa0times the gravitational acceleration, which will exert a considerable force on the gannet body. Furthermore, the pressure distribution of different water-entry inclination angles indicates that the large pressure asymmetry caused by a small oblique angle may lead to a large impact acceleration in the direction perpendicular to the longitudinal body axis and cause damage to the neck of the gannet, which partly explains the reason why a gannet performing a high plunge diving in nature enters water with a large oblique angle from the perspective of impact mechanics. The investigation on the plunge-diving behavior in this paper will inspire and promote the development of a biomimetic amphibious robot that transits from air to water with the plunge-diving mode.

Stabilization based design and experimental research of a fish robot

To reduce the yawing, rolling and pitching of the fish robot and to improve the efficiency of tail fin propulsion, the strength environment of the tail fin thruster was analyzed, the fish robots shape is optimized under a novel design principle, which was abbreviated as SPC (stability first, propulsion second, control third), of bionic fish robot, the result was analyzed by computer fluid dynamics simulation using the lift-drag ratio as the key parameter. The experiments of three fish robots of different scales were done. The biggest one, which was 2m in length, reached the maximal speed of 1.5m/s at the frequency of 1.6Hz. And the yawing of all three fish robots were below 5/spl deg/. The experiments have shown that owing to the improvement of the stabilization the velocity and propulsion efficiency were also improved, and the Strouhal number got into the best region of 0.2-0.3.

Learning to swim: a dynamical systems approach to mimicking fish swimming with CPG

Central Pattern Generators (CPGs) can generate robust, smooth and coordinated oscillatory signals for locomotion control of robots with multiple degrees of freedom, but the tuning of CPG parameters for a desired locomotor pattern constitutes a tremendously difficult task. This paper addresses this problem for the generation of fish-like swimming gaits with an adaptive CPG network on a multi-joint robotic fish. Our approach converts the related CPG parameters into dynamical systems that evolve as part of the CPG network dynamics. To reproduce the bodily motion of swimming fish, we use the joint angles calculated with the trajectory approximation method as teaching signals for the CPG network, which are modeled as a chain of coupled Hopf oscillators. A novel coupling scheme is proposed to eliminate the influence of afferent signals on the amplitude of the oscillator. The learning rules of intrinsic frequency, coupling weight and amplitude are formulated with phase space representation of the oscillators. The frequency, amplitudes and phase relations of the teaching signals can be encoded by the CPG network with adaptation mechanisms. Since the Hopf oscillator exhibits limit cycle behavior, the learned locomotor pattern is stable against perturbations. Moreover, due to nonlinear characteristics of the CPG model, modification of the target travelling body wave can be carried out in a smooth way. Numerical experiments are conducted to validate the effectiveness of the proposed learning rules.

Underwater box-pushing with multiple vision-based autonomous robotic fish

This paper presents an underwater cooperative box-pushing scenario in which three autonomous robotic fish that sense, plan and act on their own move an elongated box from some initial location to a goal location. With the onboard monocular camera, the robotic fish can estimate the pose of the object in the swimming tank. Considering the complexity of the underwater environment and the limited capability of a single robotic fish, we address the task by decomposing it into three subtasks and assigning them to capable robotic fish. With one robotic fish observing the box at the goal location and two robotic fish pushing the left and right ends of the box, the box can be moved gradually towards the goal location. The subtask consists a series of behaviors, each designed to fulfill one step of the subtask. The robotic fish coordinate through explicit communications and distribute the subtasks with a market-based dynamic task allocation method. Task reallocation mechanism that permits robotic fish to auction its assigned task to capable ones is used to cope with unexpected changes in the environment and the limited sensing range of the robotic fish. Experiments are conducted to verify the feasibility of the proposed methods.

Mechatronic design and locomotion control of a robotic thunniform swimmer for fast cruising

This paper presents mechatronic design and locomotion control of a biomimetic robotic fish that swims using thunniform kinematics for fast cruising. Propulsion of the robotic fish is realized with a parallel four-bar propulsive mechanism that delivers combined translational and rotational motion to a lunate caudal fin. A central pattern generator controller, composed of two unidirectionally coupled Hopf oscillators, is employed to generate robust, smooth and coordinated oscillatory control signals for the tail joints. In order to maintain correct phase relation between joints during fast tail beating, a novel phase adjusting mechanism is proposed and incorporated into the controller. The attitude of the robotic fish in fast swimming is stabilized using an attitude and heading reference system unit and a pair of pitching pectoral fins. The maximum speed of the robotic fish can reach 2.0 m s(-1), which is the fastest speed that robotic fishes have achieved. Its outstanding swimming performance presents possibilities for deployment to real-world exploration, probe and survey missions.

Development and CPG-based control of a biomimetic robotic fish with advanced underwater mobility.

This paper presents a biomimetic robotic fish that swims using thunniform kinematics for advanced underwater mobility. Propulsion and maneuvering of the robotic fish are achieved with a lunate caudal fin that undergoes combined translational and rotational motion. A parallel four-bar propulsive mechanism attached to the rear of the rigid torpedo-shaped body is used to deliver motor rotation to the caudal fin. Oscillatory control signals for the tail joints are generated with a CPG controller composed of two unidirectionally coupled Hopf oscillators. Coupling terms that allow direct specification of phase relation between oscillators are formulated. The maximum speed of the robotic fish can reach 2.0 m/s and excellent maneuverability has been exhibited. The outstanding swimming performances present exciting possibilities for real-world deployment of the robotic fish.

Numerical and analytical methods for synthesis of central pattern generators

This paper presents numerical and analytical methods for synthesis of a CPG network to acquire desired locomotor patterns. The CPG network is modeled as a chain of coupled Hopf oscillators with a coupling scheme that eliminates the influence of afferent signals on amplitude of the oscillator. The numerical method converts the related CPG parameters into dynamical systems that evolve as part of the CPG network dynamics. The frequency, amplitude and phase relations of teaching signals can be encoded by the CPG network with the proposed learning rules. For direct specification of the phase relations, the expression that defines the dependence of phase difference on coupling weights is analytically derived. The ability of the numerical methods to learn instructed locomotor pattern is proved with simulations. The effectiveness of the analytical method is also validated by the numerical results.