No abstract

Over the past decade, there has been a rapid increase in applications of multibody system dynamics to the predictive simulation of human movement. Using predictive "what-if" human dynamic simulations that do not rely on experimental testing or prototypes, new medical interventions and devices can be developed more quickly, cheaply, and safely. In this paper, we provide a comprehensive review of research into the predictive multibody dynamic simulation of human movements, with applications in clinical practice, medical and assistive device design, sports, and industrial ergonomics. Multibody models of human neuromusculoskeletal systems are reviewed, including models of joints, contacts, and muscle forces or torques, followed by a review of simulation approaches that use optimal control methods and a cost function to predict human movements. Modelling and optimal control software are also reviewed, and directions for future research are suggested.

This paper suggests an approach to develop dynamics model and proposes a novel method to solve inverse dynamics of a biped robot with active toe joints. In order to obtain closed-form dynamics model, first, equations of motion are developed for the robot with no interaction. Then, consistent with constraints, unknown forces and moments are considered, using constraints relaxation method. These forces and moments are different during various phases of motion for the robot which is equipped with active toe joints. In our dynamics modeling method, the double support phase is divided into two subphases, depending on the action of toe joint. Then, to compute joints actuating torques and ground reaction forces and moments, inverse dynamics problem for each subphase is solved. The simulation results for walking on flat ground surface illustrate that the proposed approach is effectively used for dynamics modeling of biped robots with redundant legs. Finally, the actual ZMP trajectory is computed using obtained reaction forces and moments to guarantee the feasibility of the generated walking pattern.

No abstract

No abstract