Page 101 - Handbook of Biomechatronics
P. 101

Model-Based Control of Biomechatronic Systems                 97


                 The biomechanics of human movement can be simulated in computers
              through inverse and forward dynamics simulations. The natural flow of
              human motion starts from the motor-neuron spikes in the CNS (i.e., includ-
              ing the brain and spinal cord) leading to the production of muscle twitches
              and a force pulling the bones to reach the desired position. A forward
              dynamic simulation can properly capture these neuromuscular dynamics
              since it follows the same natural flow. Equations of motion are integrated
              forward in time to obtain motion trajectories from neuromuscular inputs.
              In contrast, an inverse dynamics approach processes information in the
              opposite direction: the measured joint trajectories and limb motion and
              external loads from a motion capture system and force sensors are the sim-
              ulation inputs, and the muscle twitches are the simulation outputs. While an
              inverse dynamics approach is useful for clinical decision making, it cannot
              explain the underlying cause-and-effect relationships between motor
              neuron-spikes and system kinematics. The forward dynamic simulation
              can also be used to simulate what-if scenarios such as what happens if the
              stiffness of a foot-ankle orthoses increases? The biomechanical model
              parameters can be adjusted to represent different individuals with various
              physical abilities and disorders.



              1.2.1 Inverse Dynamic Simulation
              To study the biomechanics of a task, one can measure the kinematics
              (motion) and perhaps a portion of kinetics (e.g., external loads) of that par-
              ticular task in the laboratory. The kinematics can be measured using optical
              movement-monitoringsystemswithactiveorpassivemarkers(e.g.,Optotrak
              and Vicon motion capture systems, respectively) or with a markerless system
              (e.g., Microsoft Kinect), or using other movement assessment tools such as
              electro-goniometers and inertial measurement units (e.g., MVN suit). Force
              sensors can measure external loads applied to the body (e.g., foot-ground
              reaction forces during walking). Knowing the kinematics and external forces
              acting on the system, one can compute the required generalized forces
              (e.g., net joint torques and forces) to perform the given task by means of
              an inverse dynamic simulation. Before an inverse dynamic simulation can
              be performed, the equations of motion representing the task should be
              extracted using a dynamic modeling method:


                                          ð
                                   _ xtðÞ ¼ fx tðÞ, TtðÞ, FtðÞÞ            (1a)
                                    gx tðÞ, TtðÞ, FtðÞÞ ¼ 0                (1b)
                                     ð
   96   97   98   99   100   101   102   103   104   105   106