Module-Based Static Structural Design of a Modular Reconfigurable Robot

Abstract

In this paper, the structural design of modular reconfigurable robots (MRRs) is studied. This problem is defined as the determination of proper module sizes according to the robot’s payload and end-effector deflection specifications. Because an MRR has multiple configurations, a simple design process is proposed in order to avoid performing the structural design stage at each configuration. The final structural design is only carried out at a single configuration that can guarantee the robot’s satisfactory performance for all remaining feasible configurations. It is shown that the module structural design stage can be performed at the local coordinate frame of each module. While the module local force requirement can be fully determined, the determination of the module local deformation requirement is redundant. Thus, there can exist multiple design solutions. To overcome this problem, a nonlinear approach using a genetic algorithm is used to search for an optimal solution. Finally, a design simulation is performed on an example MRR, and the results show the effectiveness of the proposed design method.