无需额外电机，机械臂自重构+自适应抓取，轻松应对多变任务。

Fixed degree-of-freedom (DoF) loading mechanisms often suffer from excessive actuators, complex control, and limited adaptability to dynamic tasks. This study proposes an innovative mechanism of underactuated metamorphic loading manipulators (UMLM), integrating a metamorphic arm with a passively adaptive gripper. The metamorphic arm exploits geometric constraints, enabling the topology reconfiguration and flexible motion trajectories without additional actuators. The adaptive gripper, driven entirely by the arm, conforms to diverse objects through passive compliance. A structural model is developed, and a kinetostatics analysis is conducted to investigate isomorphic grasping configurations. To optimize performance, Particle-Swarm Optimization (PSO) is utilized to refine the gripper's dimensional parameters, ensuring robust adaptability across various applications. Simulation results validate the UMLM's easily implemented control strategy, operational versatility, and effectiveness in grasping diverse objects in dynamic environments. This work underscores the practical potential of underactuated metamorphic mechanisms in applications requiring efficient and adaptable loading solutions. Beyond the specific design, this generalized modeling and optimization framework extends to a broader class of manipulators, offering a scalable approach to the development of robotic systems that require efficiency, flexibility, and robust performance.