Development of a human walking model comprising springs and positive force feedback to generate stable gait

Biomechanics is a very important field of study when one considers the importance that motion, such as walking, has in our everyday life. Although we perform tasks such as running and climbing stairs in an automatic way, without even thinking of it, these tasks are extremely complex from a mechanical and biological point of view. Nowadays, the progression of science towards the understanding of the human body and locomotion has been exciting, with huge advancements being held in fields such as prosthetics and humanoid robots. In the latter case, humanoid robots such as ASIMO by Honda and Petman by Boston Dynamics are now able to perform much of the tasks required for human locomotion. However, much of the state-of-the-art humanoids are not as robust as desirable, generating a gait pattern that is frequently dependent on predefined joint trajectories. Inspired in neuro-musculo-skeletal models, principles of spring mechanics, feedback control theory and a careful study of the human gait cycle, we present in this work a new reflex-type controller to generate a human walking pattern in a 2D simulation. Developed in a Matlab, Simulink and SimMechanics environment, this control strategy comprises the motion of the three main joints of walking (ankle, knee and hip) in the sagittal plane.