ABSTRACT
The development of a dynamic strength model for humans would be extremely useful in predicting the forces and torques to be applied when performing a physical task. This will enhance the design of equipment, workstation and workplace in order to maximise efficiency and at the same time minimising energy and stress. In the past biomechanic strength models tended to be based on static controlled exertions. This is partly due to the mathematical complexity of dynamic biomechanical analysis. One approach in resolving this problem is to develop a model based on empirical data rather than attempting to model it as a mechanical system with rigid links, rotational springs and dampers. There have been attempts to develop polynomial equations determining joint torque as a function of joint position and velocity using the least squares regression technique. In this paper we present an alternative approach using neural networks.
