Controlling exoskeletons with EMG signals and a biomechanical body model

This work presents a control system for exoskeletons that ut ilizes electrical signals from the muscles as the main means of information transporta ti n between the human operator and the exoskeleton. Those signals are picked up from the skin on top of selected mu scles and reflect the activation of the observed muscle. They are evaluated by a so phi ticated but simplified biomechanical model of the human body to derive the desired a ction of the operator. A support action is computed in accordance to the desired act ion and is executed by the exoskeleton. The biomechanical model fuses results from different biome chanical and biomedical research groups and performs a sensible simplification considering the intended application. It contains parameters which reflect properti es of the human operator and his or her current body state. A calibration algorithm for th ose parameters is presented which relies exclusively on sensors mounted on the exoskele ton, and provides deep inside into the mechanisms of the model. An exoskeleton for the knee joint support was designed and co structed to verify the model and investigate the interaction between the human operator and the machine in experiments with force support during everyday movement s. Those results are also presented here.