ABSTRACT
In this chapter, we considered the fundamentals underlying the biomechanical optimisation of sports movements, with an emphasis on theory-driven statistical modelling, computer simulation modelling and optimisation, and the approaches of artificial intelligence, particularly that of artificial neural networks. The relationships that can exist between a performance criterion and various performance variables were explained, the cross-sectional, longitudinal and contrast approaches to statistical modelling were described and the limitations of statistical modelling in sports biomechanics were evaluated. The use of computer simulation modelling, when seeking to evaluate and improve sports movements, were covered; brief explanations of modelling, simulation, simulation evaluation and optimisation were also provided. The differences between static and dynamic optimisation and global and local optimums were covered. We considered in detail the optimisation of javelin release as an example of computer simulation modelling of sports movements, and touched on the modelling of the human performer, with a couple of examples of models of skeletal muscle. Finally, we looked at the various approaches from artificial intelligence that have been used in the context of the biomechanical improvement of sports performance, focusing mainly on artificial neural networks and the links between this approach and the non-linear dynamical systems theory of movement coordination.
