REAL-TIME NEUROMUSCULOSKELETAL MODELING TO ESTIMATE MUSCLE EXCITATIONS AND JOINT MOMENTS AT DIFFERENT WALKING SPEEDS: A PILOT STUDY

Powered prosthetic legs are promising solutions to assist amputees locomotion in a large variety of conditions. However, passive devices are still the most common commercial solution. To fully benefit from the active device, the user has to rely on the dynamic control of it and, on its visual observations or incidental signals, to walk in a natural and safe manner. The absence of communication from the user to the prosthetic that can translate body states into control commands results in the adoption of non-natural compensatory movements. An alternative solution to commercial finite state machines is given by musculoskeletal models driven by electromyographic signals (EMGs) that mimic the dynamics and motor control strategies in humans. However, EMGs are not always available as sometimes muscles are removed. Instead, our control system includes a model to mimic the role of the central nervous system in giving movement commands in real-time. It may improve user-device interaction providing an intuitive and robust system