One of the fascinating questions facing neuroscience is how the nervous system controls, with such sophistication, the high dimensional, highly redundant and noisy system that is our skeletomuscular system. In recent years optimal feedback control has increasingly been used as a normative model to explain human motor behaviour. However models so far have failed to consider and account for variable impedance which is an important feature of observed human movement, both under normal conditions and in adaptation to new environments. The goal of the here proposed research is to investigate if variable impedance can be understood in the framework of optimal feedback control. In particular we intend to test models on novel artificial systems which are more realistic than previously used toy systems and to verify predictions in human experiments.
Related Publications and Presentations
- Jun Nakanishi, Konrad Rawlik, and Sethu Vijayakumar, “Stiffness and Temporal Optimization in Periodic Movements: An Optimal Control Approach”, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2011.
- Konrad Rawlik, Marc Toussaint, and Sethu Vijayakumar, “Path Integral Control by Reproducing Kernel Hilbert Space Embedding”, International Joint Conference in Artificial Intelligence (IJCAI), 2013.
- Konrad Rawlik, Marc Toussaint, and Sethu Vijayakumar, “On Stochastic Optimal Control and Reinforcement Learning by Approximate Inference”, Robotics: Science and Systems, 2012.