Multiobjective Codesign Optimization of a Planar Pneumatic Artificial Muscle‐Based Snake‐Like Robot for Enhanced Agility and Energy Efficiency

In this work, a codesign multiobjective optimization framework was developed to concurrently optimize the morphology and controller of a pneumatic artificial muscle (PAM)‐driven snake‐like robot during planar locomotion with enhanced agility and power consumption efficiency. The framework includes a parametric dynamic model of the PAM‐driven snake robot and a joint space computed muscle‐force controller designed to track the desired trajectories. Motion parameters were analyzed as inputs to the controller, assessing their impact on the robot's steady‐state forward velocity and power consumption. A global Sobol sensitivity analysis identified the design parameters with dominant contributions to performance, guiding the optimization design space. The controller's performance was examined under different friction conditions and gait's motion parameters. A codesign approach was used to optimize both the robot's link geometry and controller's gains, aiming for the best balance between the robot's forward velocity and energy efficiency. Additional constraints were incorporated to ensure physical feasibility of PAM forces, leading to optimal designs with improved actuator range utilization and preserved velocity‐power tradeoff. Results show that adjusting gait characteristics maximizes velocity while minimizing power consumption. Optimization further suggested that shorter muscles, positioned closer to the joints, combined with longer linkages, yield peak velocity with minimal power usage.