Donelan, M. | Search bioRxiv

bioRxiv2024

Effects of sensorimotor delays and muscle force capacity limits on the performance of feedforward and feedback control in animals of different sizes

Animals rely on both feedforward and feedback control for perturbation responses. When comparing animals of different sizes, we find that several features that affect perturbation responses change--larger animals have longer sensorimotor time delays, heavier body segments and proportionally weaker muscles. We used simple computational models to compare fast perturbation response times under feedforward and feedback control, as a function of animal size. We developed two tasks representing common perturbation response scenarios in animal locomotion: a distributed mass pendulum approximating swing limb repositioning (swing task), and an inverted pendulum approximating whole body posture recovery (posture task). First, we used a normalized feedback control system to show how feedback response times can either be limited by the force generation capacity of muscles (force-limited), or by sensorimotor delays which constrain the maximum feedback gains that can be used to produce stable responses (delay-limited). Next, we used more detailed scaled models which represent the full-size range of terrestrial mammals and parameterized the sensorimotor delays, maximum muscle forces, and inertial properties using published scaling relationships from literature. Across animal size and in both tasks, we found that feedback control was primarily delay-limited--the fastest responses used a fraction of the available muscle force capacity. Feedforward control, which is able to fully activate muscles and produce faster responses--was about four times faster than feedback control in the smallest animals, and around two times faster in the largest animals. For rapid perturbation responses, feedback control appears ineffective for terrestrial mammals of all sizes, as the fastest response times exceeded available movement times, while feedforward control did not. Thus, feedforward control is more effective for reacting quickly to sudden and large perturbations in animals of all sizes.

neuroscience↗

bioRxiv2018

How People Initiate Energy Optimization and Converge on Their Optimal Gaits

A central principle in motor control is that the coordination strategies learned by our nervous system are often optimal. Here we combined human experiments with computational reinforcement learning models to study how the nervous system navigates possible movements to arrive at an optimal coordination. Our experiments used robotic exoskeletons to reshape the relationship between how participants walk and how much energy they consume. We found that while some participants used their relatively high natural gait variability to explore the new energetic landscape and spontaneously initiate energy optimization, most participants preferred to exploit their originally preferred, but now suboptimal, gait. We could nevertheless reliably initiate optimization in these exploiters by providing them with the experience of lower cost gaits suggesting that the nervous system benefits from cues about the relevant dimensions along which to re-optimize its coordination. Once optimization was initiated, we found that the nervous system employed a local search process to converge on the new optimum gait over tens of seconds. Once optimization was completed, the nervous system learned to predict this new optimal gait and rapidly returned to it within a few steps if perturbed away. We model this optimization process as reinforcement learning and find behavior that closely matches these experimental observations. We conclude that the nervous system optimizes for energy using a prediction of the optimal gait, and then refines this prediction with the cost of each new walking step.

neuroscience↗