As a researcher in the field of robotics, I have always been fascinated by the challenges posed by complex amphibious environments. These terrains, which combine land and water elements, are notoriously difficult for conventional robots to navigate due to their limited adaptability and simplistic motion capabilities. In response to this, my team and I embarked on a project to develop a novel amphibious bionic robot, inspired by the remarkable adaptability of the salamander. This bionic robot aims to overcome the shortcomings of existing designs by incorporating a sophisticated pitch attitude control system, enabling it to perform a wide range of maneuvers such as surfacing, diving, pitching up, and pitching down. The ultimate goal is to create a bionic robot with high fidelity to biological motion, excellent environmental adaptability, and the ability to execute complex tasks in diverse amphibious settings. This article details the structural design, finite element analysis, and kinematics simulation of this innovative amphibious bionic robot.
The amphibious environment is a multi-media complex terrain, often characterized by muddy land, shallow water, and uneven surfaces. Traditional amphibious bionic robots typically suffer from simple structural designs and limited motion repertoires, making them inadequate for special and complex missions. Our design philosophy centers on biomimicry, specifically emulating the salamander, a creature known for its exceptional agility both on land and in water. This bionic robot is not merely a mechanical replica; it is an engineering solution that integrates advanced control systems, robust驱动 mechanisms, and adaptive buoyancy control to achieve seamless transition and operation in amphibious domains. The development of such a bionic robot holds significant promise for applications in resource exploration, disaster relief, reconnaissance, and environmental monitoring, where human presence is risky or impractical.
The overall design of our amphibious bionic robot is meticulously crafted to balance weight, strength, and functionality. The robot has a total length of 1.3 meters, with a height of 0.2 meters when the legs are retracted and 0.28 meters when extended for terrestrial locomotion. It can carry a payload of up to 8 kg, and its total mass without payload is approximately 24.47 kg. The weight distribution across various components is critical for stability and performance, as summarized in the table below.
| Component | Mass (kg) |
|---|---|
| Drive System | 13 |
| Control System | 0.3 |
| Buoyancy Control Mechanism | 5.35 |
| Battery | 4 |
| Data Acquisition System | 0.13 |
| Payload | 8 |
| Shell | 1.69 |
| Total Mass (with payload) | 32.47 |
This weight distribution ensures that the bionic robot maintains a low center of gravity and optimal buoyancy, essential for both terrestrial and aquatic mobility. The drive system is the core of the bionic robot’s movement capabilities. It consists of two main subsystems: the body joint drive system and the limb joint drive system. Each joint is powered by servo motors selected based on the required torque and range of motion. The摆动 ranges for all joints are designed to mimic the salamander’s natural kinematics, as detailed in the following table.
| Joint Position | Swing Range |
|---|---|
| Head Joint | (-15°, 15°) |
| Trunk Joint | (-25°, 25°) |
| Tail Joint | (-25°, 25°) |
| Thigh Lift Joint | (-30°, 30°) |
| Thigh Horizontal Swing Joint | (-90°, 60°) |
| Calf Extension Joint | (-30°, 60°) |
The body joint drive system comprises six horizontal swing joints, each equipped with a speed sensor for real-time monitoring. The motors are connected via couplings to adjacent fixed base plates, forming revolute pairs that allow relative rotation between segments. The limb system, which includes both forelimbs and hindlimbs, is primarily responsible for terrestrial support and propulsion. Each limb has three joints: one for thigh lifting, one for thigh horizontal swinging, and one for calf extension. The coordination between limb and body joints is crucial for achieving lifelike motion in this bionic robot. The specifications of the drive system components are summarized below.
| Component | Model/Material | Rated Torque (N·m) | Max Stress (MPa) | Max Strain (mm) | Mass (kg) |
|---|---|---|---|---|---|
| Fixed Base Plate | Carbon Fiber | – | – | – | 0.22 |
| Thigh Assembly | Aluminum Alloy | – | 4.8 | 4.64×10⁻⁵ | 0.54 |
| Calf Flat Tube | Carbon Fiber | – | – | – | 0.04 |
| Fixed Base | Aluminum Alloy | – | 22.7 | 2.27×10⁻⁴ | 0.27 |
| Leg Motor | RMD-X6-s2 | 18 | – | – | 0.59 |
| Body Motor | MG6010-v3 | 5 | – | – | 0.34 |
One of the most innovative features of this amphibious bionic robot is its buoyancy and pitch attitude control system. This system enables the bionic robot to perform precise maneuvers in water, such as surfacing, diving, pitching up, and pitching down. It consists of five fixed buoyancy blocks and two flexible airbags located at the shoulder and hip units. The buoyancy blocks are made of plastic and filled with high-pressure gas at 0.15 MPa, providing inherent buoyancy. The flexible airbags are connected to high-pressure air storage tanks and air pumps via solenoid valves, allowing for dynamic adjustment of buoyancy and center of buoyancy.
The control logic for buoyancy adjustment can be described using two primary modes. In the first mode, for surfacing, solenoid valves A and D are opened, allowing gas to flow from the storage tanks into the airbags until pressure equilibrium is reached. This maximizes buoyancy, causing the bionic robot to float. For diving, valves A and D are closed, and valves B and E are opened, activating the air pumps to transfer gas from the airbags back to the storage tanks, reducing buoyancy until the robot submerges. The second mode involves transferring gas between the shoulder and hip airbags to shift the center of buoyancy, thereby controlling the pitch angle. For instance, to pitch down, gas is pumped from the shoulder airbag to the hip airbag using the shoulder pump, moving the center of buoyancy toward the tail. This induces a downward pitch, facilitating diving under propulsion. Conversely, pumping gas from the hip to the shoulder airbag causes an upward pitch for surfacing. The pitch angle is monitored by attitude sensors and controlled via closed-loop feedback, ensuring precise maneuverability for this bionic robot.
The shell of the bionic robot is constructed from plastic, weighing approximately 1.69 kg. It serves as a protective cover for the joints and internal components, shielding them from damage during operation. Physical limit blocks are integrated into the shell to prevent joints from exceeding their designed swing ranges, adding an extra layer of safety. For example, the leg joint limit modules prevent over-rotation of the thigh or calf that could harm motors or sensors, while body joint limits protect adjacent modules from collisions during extreme movements.
Power supply is a critical aspect of any autonomous bionic robot. We selected a battery with a capacity of 0.7 kWh, weighing about 4 kg, to meet the energy demands of all subsystems. The battery is housed within a dedicated compartment in the torso buoyancy block, ensuring waterproofing and safety. The power consumption was calculated based on the operational profiles of the motors and air pumps. On land, all joints are active, with motors operating intermittently. The average torque outputs are: thigh lift joint at 15 N·m, thigh horizontal swing and calf extension joints at 7 N·m, and body joints at 5 N·m. The static torque during standing is about 3.4 N·m per leg joint. The total energy consumption for terrestrial locomotion is estimated using the formula for mechanical work: $$W = \sum_{i=1}^{N} \tau_i \cdot \theta_i \cdot f_i \cdot t_i$$ where $\tau_i$ is the torque, $\theta_i$ is the angular displacement, $f_i$ is the frequency, and $t_i$ is the duration for each joint. For our bionic robot, the approximate terrestrial energy consumption is 0.236 kWh. In water, only the body joints are active for swimming, with a higher oscillation frequency. The air pumps consume about 50 W during buoyancy adjustments. The aquatic energy consumption is around 0.108 kWh. Thus, the battery provides sufficient endurance: approximately 135 minutes on land and 240 minutes underwater without payload, ensuring reliable operation for extended missions.
The data acquisition system of this bionic robot incorporates various sensors to monitor internal states and external conditions. It includes temperature sensors, speed sensors, pressure sensors, and water tank level sensors. These sensors feed data to the central processing unit, which then relays information to the upper computer for real-time decision-making. This system enables the bionic robot to adapt to environmental changes, such as detecting water depth or terrain irregularities, enhancing its autonomy and effectiveness.
The control system is built around an STM32 microcontroller serving as the central processor. It handles data processing, communication, and execution of control algorithms. The hardware peripherals include wireless communication modules, CAN bus modules, MPU6050 gyroscope modules, GPS modules, and camera modules. The GPS provides real-time localization, while the camera captures environmental imagery for navigation and task execution. The software system, programmed in C, comprises multiple subroutines for different motion modes: terrestrial walking, aquatic swimming, and buoyancy control. The main program flowchart orchestrates these subroutines based on sensor inputs and mission requirements. For instance, upon startup, the bionic robot initializes all systems, then enters a main loop that calls specific motion subroutines depending on whether it is on land or in water. This modular software architecture allows for flexibility and robustness in controlling the bionic robot.
To validate the structural integrity of critical components, we performed finite element analysis (FEA) using SolidWorks. A key component analyzed was the thigh assembly, made of aluminum alloy with dimensions 68 mm × 228 mm × 21 mm and a mass of about 0.5 kg. This part supports the leg motor and sensors, and it must withstand the motor’s torque, especially in scenarios where the foot gets stuck or trapped. The FEA simulated a static load equal to the motor’s rated torque of 18 N·m. The results showed a maximum stress of 11.67 MPa, well below the material’s yield strength of 55.15 MPa. The maximum strain was negligible, at 4.64×10⁻⁵ mm. This confirms that the thigh assembly can endure unexpected loads without failure, ensuring the bionic robot’s durability in challenging conditions. The stress-strain relationship can be expressed using Hooke’s Law for linear elasticity: $$\sigma = E \cdot \epsilon$$ where $\sigma$ is stress, $E$ is Young’s modulus, and $\epsilon$ is strain. For aluminum alloy, $E \approx 69$ GPa, so the calculated strain aligns with the FEA results, verifying the design’s safety margin.
Kinematics simulation is essential for evaluating the motion performance of the bionic robot before physical prototyping. We used Webots software to create a virtual prototype and simulate its behavior in both aquatic and terrestrial environments. The 3D model from SolidWorks was exported in VRML format and imported into Webots, where joints and constraints were defined. Control programs were written to replicate the salamander’s gait patterns.
On land, the bionic robot extends its limbs to support the body, with coordinated oscillations between body and limb joints producing stable walking. The joint angles over time can be described by periodic functions. For example, the body joint angles might follow a sinusoidal pattern: $$\theta_i(t) = A_i \sin(2\pi f t + \phi_i)$$ where $A_i$ is the amplitude, $f$ is the frequency, and $\phi_i$ is the phase shift for the $i$-th joint. Simulation results showed smooth leg joint movements, with angles varying within the designed ranges, as illustrated in the table below for a sample gait cycle.
| Time (s) | Thigh Lift Angle (°) | Thigh Swing Angle (°) | Calf Extension Angle (°) |
|---|---|---|---|
| 0.0 | 0 | -30 | 0 |
| 0.3 | 15 | -15 | 15 |
| 0.6 | 30 | 0 | 30 |
| 0.9 | 15 | 15 | 15 |
| 1.2 | 0 | 30 | 0 |
In water, the limbs are retracted, and propulsion is achieved solely through body undulation. The swimming motion was simulated with high-frequency body joint oscillations. The joint angles exhibited coupled摆动, similar to a traveling wave along the body. The kinematics can be modeled using the serpenoid curve, common in snake-like robots: $$\theta(s,t) = A \sin(ks – \omega t)$$ where $\theta$ is the joint angle, $s$ is the arc length along the body, $k$ is the wave number, $\omega$ is the angular frequency, and $t$ is time. For our bionic robot with discrete joints, this translates to: $$\theta_i(t) = A \sin(2\pi i / N – \omega t)$$ for $i = 1$ to $N$ joints. The simulation confirmed that this bionic robot could achieve efficient swimming with realistic undulatory motions. The body joint angles over time are plotted below, showing the phase-lagged oscillations essential for generating thrust.
The effectiveness of the pitch attitude control was also simulated. By varying the volume of the flexible airbags, the bionic robot successfully demonstrated surfacing, diving, and pitch adjustments in virtual water environments. The pitch angle $\phi$ is controlled by the differential buoyancy force between the shoulder and hip. The equation of motion for pitch can be simplified as: $$I \ddot{\phi} = \Delta F \cdot d – C \dot{\phi}$$ where $I$ is the moment of inertia, $\Delta F$ is the differential buoyancy force, $d$ is the lever arm distance between the airbags, and $C$ is a damping coefficient. The simulation showed that the bionic robot could achieve precise pitch angles up to ±30° within a few seconds, validating the control system’s responsiveness.
In conclusion, this novel amphibious bionic robot design represents a significant advancement in biomimetic robotics. By emulating the salamander, we have created a bionic robot with exceptional adaptability to both terrestrial and aquatic environments. The integration of a sophisticated pitch attitude control system allows for versatile maneuvers, surpassing the capabilities of conventional amphibious bionic robots. Finite element analysis confirmed the structural robustness of key components, while kinematics simulations in Webots demonstrated smooth and efficient motion in various scenarios. This bionic robot exhibits high fidelity to biological motion, with coordinated joint movements enabling stable walking and swimming. The design’s modularity and use of lightweight materials contribute to its overall performance and potential for real-world applications. Future work will focus on physical prototyping, hydrodynamic testing, and enhancing autonomy through advanced AI algorithms. This bionic robot lays a strong foundation for the development of next-generation amphibious machines capable of tackling complex tasks in unpredictable environments.
Throughout this project, the term “bionic robot” has been central to our discourse, emphasizing the biomimetic approach that underpins the design. This bionic robot not only mimics the form of a salamander but also its functional principles, resulting in a machine that moves with lifelike grace and efficiency. The journey from concept to simulation has been rigorous, involving detailed engineering analyses and iterative refinements. As we move forward, we are confident that this amphibious bionic robot will inspire further innovations in the field, pushing the boundaries of what bionic robots can achieve in challenging multi-media terrains. The potential applications for such a bionic robot are vast, from scientific exploration in wetlands to disaster response in flooded areas, underscoring the importance of continued research and development in amphibious bionic robotics.