Optimization-Based Prediction of the Motion of a Soldier Performing the ‘Going Prone’ and ‘Get Up from Prone’ Military Tasks

Abstract

In this work, an optimization based 3-D motion prediction of the motion of a soldier transitioning between a “Standing” posture and a “Prone” posture is presented. Based on the time discretization strategy between these two postures, a set of motion frames are defined such that they include important moments of time during the motion (such as moments when the parts of soldier’s body that touch the ground alter). Then every frame of the motion is analyzed, predicted and certified to be both feasible and optimal considering all dynamic properties of a motion frame such as velocities, accelerations and all higher derivatives of the human’s position. The digital human model is a full-body, three dimensional model with 55° of freedom. Six degrees of freedom specify the global position and orientation of the coordinate frame attached to the pelvic point of the digital human and 49° of freedom represent the revolute joints which model the human joints and determine the kinematics of the entire digital human. Motion is generated by a multi-objective optimization approach minimizing the mechanical energy and joint discomfort simultaneously. The optimization problem is subject to constraints which represent the limitations of the environment, the digital human model and the motion task. Design variables are the joint angle profiles. All the forces, inertial, gravitational as well as external, are known, except the ground reaction forces. The feasibility of the generation of that arbitrary motion by using the given ground contact areas is ensured by using the well-known Zero Moment Point (ZMP) constraint. During the kneeling motion, different parts of the body come in contact and lose contact with the ground which is modeled using a general approach. The ground reaction force on each transient ground contact area is determined using the equations of motion. Using these ground reaction forces, the required torques at all joints are calculated by the recursive Lagrangian formulation. This simulation is able to predict feasible and optimal motions that vary when loading (such as backpack, etc.) or the equipment that the human model carries change as or when the human model’s strength, size or weight is altered.