Simulation Of Human Jumping - Task Alteration

Abstract

Similar human motions are often grouped together into a single movement class. In jumping, the question of similarity in control and coordination of different tasks within a movement class has been addressed by altering the direction of maximal effort jumps between vertical and horizontal (Jensen & Phillips, 1991). Other researchers have studied vertical jumping using computer simulations (Pandy & Zajac, 1991; van Soest et al., 1993), but have not addressed the issue of modifying the jumping task. The purpose of the present study was to investigate the effect of task alteration on forward .dynamic simulations of jumping. The simulation model was comprised of 4 linked rigid segments (Fig. 1). Segmental motion was controlled by 3 torque generators defined with pre-set magnitudes and activation time constants. All jumps began from a static posture with all torques set to zero. Each jumping performance was dictated by the onset times of each torque generator. The choice of task was specified by one of two objective functions: 1) vertical height, or 2) horizontal distance. The optimization algorithm searched for the pattern of activation onset times for the 3 torque generators which maximized each objective function. A variety of tests were performed to compare the model's optimal vertical and horizontal jumping performance, and the underlying coordination of torque generation. With the set of initial conditions shown in Fig. la (00 = (1.0,-0.7,1.1,-0.8}), the model's maximum height was 1.814 m, with a forward displacement of 0.39 m. For the optimal forward jump the displacement increased to only 0.73 m. However, by changing the initial posture (00 = (0.8,-1.0,0.8,-1.0}) the optimal forward jump was improved to 1.87 m (Fig. lb), but the optimal vertical jump was reduced to 1.465 m. The relative magnitude of onset times varied substantially between the two optimal jumps. For the conditions studied, the order of onset times does not follow the proximo-distal sequence often .proported for humans, and the relative timing of joint torques changes with task. The results illustrate the dependence of the optimal solution on initial conditions. Further work will examine the relations between initial conditions, coordination and perfommance.