Simulation based design optimization framework for a gait pattern generation of a small biped robot with tiptoe mechanism

The research of biped robot is still far from proposing a solution which generates a level of flexibility and reliability gait pattern that would enable practical walking on the variety of rough ground human's negotiation with ease on a regular basis. To solve its issue, we learn from the mammal, belong to the plantigrade, on the natural world. The plantigrade, i.e., human, monkey or bear walk by adapting to a ground from a toe up to heel. From this adaptive walk, its flexibility and stability are excellent although this gait is not suitable for a quick walk. In this study, to get flexibility and reliability of gait pattern, we introduce a tiptoe mechanism to a small biped robot through inspiration from its adaptive walk, and develop the Simulation Based Design Optimization framework (SBDO). In this paper, the SBDO framework by approximated optimization process using Adaptive Plan system with Genetic Algorithm (APGA) is proposed. Moreover, to validate an ability of this framework, a small biped robot with tiptoe mechanism is simulated on flat ground at first trial. We discuss this validation result for estimating the SBDO framework.