Modeling a Predictive Control of Human Locomotion Based on the Dynamic Behavior

Introduction

The growing number of legged impaired in the world is a reality that should be confronted. The available passive mechanic orthosis presents locomotion difficulties when walking on inclined terrain with obstacles, such as slopes and stairs (Floriano-Batista & Rosario 2013).

Modeling of a system is a fundamental resource on the study, comprehension and simulation of its dynamic behavior. Real systems are rich in complexity, making it a difficult task reducing the dynamic behavior for a mathematic approach (de Melo, Rodrigues, & Rosário, 2015). By approaching robotics, with the objective of optimizing the system, there is an attempt to emulate the biomechanical behavior present in the human locomotion in which a trajectory assumed by the lower limbs is denominated gait (Gutiérrez-Carvajal, de Melo, Rosário, & Machado, 2016). This gait presents discontinuities and mechanical impacts that make modeling a complex and laborious task by means of the conventional mechanics formalism, and frequently limits the extension of the studies. This becomes critical when the interest is mimicking, i.e. the perfect assimilation of dynamic behavior on this system, intended objective when approaching the design of exoskeletons (mechatronics devices integrated with the human body), or parts of the body, intended to map and/or amplify its motor functions, and that interact in a physical and cognitive way with its user (Pons, Ceres, & Calderón, 2008).

The development of a study platform that simulates the locomotion of human gait and that permits the dynamic integration of the continuous behavior with the event dynamic, in order to create a model that can be applied to a gait, aiding exoskeleton, is the purpose of this paper. The main objectives of this are the development of a physical model of the human locomotion system, taking the gait in the sagittal plane as a first step and, then, connecting the control system, previously developed for the human leg, integrated by the human gait modeled system.

This chapter presents the study and the simulation of an exoskeletics orthosis for reproducing the human gait movement and integrating the patient within today society and a proposal for development of a lower limb orthosis based on the continuous dynamic behavior and on the events presented on the human locomotion, when the legs alternate between different functions. Computational model was developed to approach the different functioning models related to the bipedal anthropomorphic gait. One of the investigated factors is the gait bilateralism, that provides balance to the center of mass of the human body during locomotion. The authors combines the comparison of the use of the predictive control based on dynamical study and the decoupling of the dynamical model, with auxiliary parallelograms, for locating the center of mass of the mechanism using springs in order to achieve the balancing of each leg. Virtual model is implemented and its kinematic and dynamic motion analyzed through simulation of an exoskeleton, designed to lower limbs, for training and rehabilitation of the human gait. For this, a strategy was developed for dynamic model of anthropomorphic mechanism and predictive control architecture with robust control.