Huxley-type muscle models in largish-scale musculoskeletal models; a feasibility study

Using maximum height squat jumping as an example task, it was shown that optimization of neural inputs to a largish-scale musculoskeletal model in which the actuators are represented by Huxley-type muscles, results in macroscopic behaviour similar to that of an otherwise identical model with Hill-type actuators. The computational cost for the Huxley-type model was approximately 20000 times that of the Hill-type model. For the Huxley-type model, the computation time for one optimization was about 30 hours. We conclude that task optimization of largish-scale musculoskeletal models with Huxley-type muscle models is now a feasible option.