In this article, a formal mechatronic design of a biped robot is addressed. It is considered a model-based system engineering methodology since the continuous updating of information, from analysis and evolution of conceptual designs, demands a large volume of data. The definition of a biped robot comes from the need of a system to perform human-like walking as the problem to be solved. A specific robot configuration results from the analysis of conceptual solutions throughout SysML as the language for modeling the synergistic and automatic integration among engineering disciplines. The general design process is developed according to the well-known V-model for mechatronic systems design; however, a three-dimensional focus is proposed in order to address a variety of domains and their interaction along the design process. The detailed study of the solution is evaluated in order to optimize the joint torques and limbs shape from an anthropometric robot to achieve effective human-like motion. Although the mechatronic design is done for the overall biped robot system, this work is particularly focused on mechanical features as the most representative subsystem that incorporates genetic algorithm optimization based on a numerical Newton–Euler model merged with topology optimization tools to define final geometry of limbs with stiffness maximization.

中文翻译：

---

基于新型机电一体化方法的仿人双足运动系统设计

在本文中，讨论了双足机器人的正式机电一体化设计。它被认为是一种基于模型的系统工程方法，因为信息的不断更新，从概念设计的分析和演变，需要大量的数据。双足机器人的定义来自于需要一个系统来执行类人行走作为要解决的问题。特定的机器人配置源自对整个 SysML 的概念解决方案的分析，SysML 作为建模工程学科之间的协同和自动集成的语言。通用设计流程是根据众所周知的机电系统设计V模型开发的；然而，为了解决设计过程中的各种领域及其相互作用，提出了一个三维焦点。对该解决方案的详细研究进行了评估，以优化人体测量机器人的关节扭矩和四肢形状，以实现有效的类人运动。虽然机电一体化设计是为整个双足机器人系统完成的，但这项工作特别侧重于机械特征作为最具代表性的子系统，它结合了基于数值牛顿-欧拉模型的遗传算法优化与拓扑优化工具相结合来定义肢体的最终几何形状与刚度最大化。