Constructing predictive models of human running

Maus, Horst-Moritz; Revzen, Shai; Ludwig, Christian; Reger, Johann; Seyfarth, André

doi:10.1098/rsif.2014.0899

Cited by 43 publications

(71 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As noted in the introduction, our approach differs from others in constraining force, kinematics and stance duration: In some studies such as in the virtual pivot models [31], only GRFs are considered leading to model parameters that are in clear conflict with the animal's COM motion as no restriction is placed on the height of the COM. Other studies normalize time to stance duration [10], allow stance time to vary [30], or only focus on fitting the COM trajectory (vertical height as a function of the horizontal displacement) [32].…”

Section: Theoretical Fits To Experimental Datamentioning

confidence: 99%

“…That is, a successful model must produce realistic GRFs within the constraints of experimentally observed COM kinematics over experimentally observed stance duration. These three constraints are rarely satisfied [31,30,32] simultaneously in most studies of locomotion. Without satisfying the constraints on force, kinematics and duration simultaneously, the problem is under constrained: The COM height determines the natural timescale of the system (∼ R nat /g); and, in understanding locomotion, one important consideration is how does the stance duration compare to the natural time constant.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Spring-loaded inverted pendulum goes through two contraction-extension cycles during the single stance phase of walking

Antoniak

Biswas

Cortés

et al. 2019

Preprint

View full text Add to dashboard Cite

Despite the overall complexity of legged locomotion, the motion of the center of mass (COM) itself is relatively simple, and can be qualitatively described by simple mechanical models. The spring-loaded inverted pendulum (SLIP) is one such model, and describes both the COM motion and the ground reaction forces (GRFs) during running. Similarly, walking can be modeled by two SLIPlike legs (double SLIP or DSLIP). However, DSLIP has many limitations and is unlikely to serve as a quantitative model for walking. As a first step to obtaining a quantitative model for walking, we explored the ability of SLIP to model the single stance phase of walking across the entire range of walking speeds. We show that SLIP can be employed to quantitatively model the single stance phase except for two exceptions: first, it predicts larger horizontal GRFs than empirically observed. A new model -angular and radial spring-loaded inverted pendulum (ARSLIP) can overcome this deficit. Second, even the single stance phase has active elements, and therefore a quantitative model of locomotion would require active elements. Surprisingly, the leg spring undergoes a contraction-extension-contraction-extension (CECE) during walking; this cycling is partly responsible for the M-shaped GRFs produced during walking. The CECE cycle also lengthens the stance duration allowing the COM to travel passively for a longer time, and decreases the velocity redirection between the beginning and end of a step. A combination of ARSLIP along with active mechanisms during transition from one step to the next is necessary to describe walking.

show abstract

Section: Theoretical Fits To Experimental Datamentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Spring-loaded inverted pendulum goes through two contraction-extension cycles during the single stance phase of walking

Antoniak

Biswas

Cortés

et al. 2019

Preprint

View full text Add to dashboard Cite

show abstract

“…Several authors have noted that done naively, the observed eigenvalues change with phase 14, 35-37 -a result contrary to theory (see §2). Estimating the return map by using multiple Poincaré sections simultaneously 35,37 improves the reliability of the results.…”

Section: The Toolbox Of Data Driven Floquet Analysismentioning

confidence: 96%

“…We recently used DDFA to provide evidence in support of a 2009 prediction 38 that control of human running allows complete ("dead-beat") recovery within 2 steps. 37 We also used the DDFA to produce a linearized controller model for the human, 37 which could then be reduced into low dimensional "factors", allowing us to identify that by adding the state of the swing-leg ankle to the existing model's state-space the predictive ability of the model could be greatly improved.…”

Section: The Toolbox Of Data Driven Floquet Analysismentioning

confidence: 99%

Data driven models of legged locomotion

Revzen

Kvalheim

2015

SPIE Proceedings

Self Cite

View full text Add to dashboard Cite

Legged locomotion is a challenging regime both for experimental analysis and for robot design. From biology, we know that legged animals can perform spectacular feats which our machines can only surpass on some specially controlled surfaces such as roads. We present a concise review of the theoretical underpinnings of Data Driven Floquet Analysis (DDFA), an approach for empirical modeling of rhythmic dynamical systems. We provide a review of recent and classical results which justify its use in the analysis of legged systems. LOCOMOTION AS AN OSCILLATORLocomotion is a process whereby the body moves through space. Up to changes of shape the body is related to its history by a continuous trajectory in the space of body frame position and orientation -the Lie group SE(3). The mathematical structure representing such a system 1 is that of a "principal fiber bundle" -a product-like construction pairing the shape-space of the body S, with SE(3). Together these are the body's "configuration space" Q. When locomoting, bodies exert forces on a medium in their environment to produce the reaction forces that move them. Mechanics dictates a relationship -the "connection" -between shape velocity and configuration velocity, i.e. between TS and TQ. This highly abstracted view is sometimes used in the undulatory locomotion literature, 1 where the arbitrariness of body frame choice makes its explicit treatment necessary. However, it remains a valid representation for virtually all self-propelling bodies. In an ideal world, we would be able to predict robust and reliable estimates of the connection for any body of interest, in any medium. Such an estimate would be a complete and self-contained representation of how that body could move through the medium.Due to the daunting complexity of body-medium interactions most work on locomotion is focused on simplified models. Low dimensional, sprung mass models have captured salient features of the dynamics of many legged locomotion systems. 2 These models can be organized using the "templates and anchors" hypotheses (TAH): 3 "Animals have many DOF, but move 'as if ' they have only a few. Animals limit pose to a behaviorally relevant family of postures". Template-and-Anchor is a relationship between models of locomotion -one being more elaborate, the other more parsimonious. Both are assumed to be good long-term predictors of the motion, making the template a dimensionally reduced model of the anchor.At this point most treatments of locomotion in biomechanics and robotics would proceed to discuss the structure of various templates, and the predictions obtained from simulation of high-dimensional anchored models. Such approaches are model-driven -the discussion is focused primarily on models, their justification from assumptions about the physics and biology, and the derivaiton of governing parameters for the models. In our data-driven approach the focus is on large information-rich datasets. We used first principles to define a broad class of models, sidestepping the issue of assumptions ...

show abstract

“…Our work revolves around the spring-loaded inverted pendulum (SLIP) model, a hybrid dynamic model ubiquitous in both biomechanics and the robotic legged-locomotion community for modeling running or hopping gaits. This model can be fit very accurately to experimental data of many different running animals [17] [18], allows accurate prediction [19], and also has been used to design controllers for simulations [20][21] [22][23] as well as actual robots [24][25] [26]. Indeed, there has been a lot of effort to give legged robots SLIP-like behavior, either through mechanical design [27] [28] or control [24] [29].…”

Section: Viability Kernel Of a Running Model A Modelmentioning

confidence: 99%

Learning from outside the viability kernel: Why we should build robots that can fall with grace

Heim¹,

Spröwitz²

2018

2018 IEEE International Conference on Simulation, Modeling, and Programming for Autonomous Robots (SIMPAR)

View full text Add to dashboard Cite

Despite impressive results using reinforcement learning to solve complex problems from scratch, in robotics this has still been largely limited to model-based learning with very informative reward functions. One of the major challenges is that the reward landscape often has large patches with no gradient, making it difficult to sample gradients effectively. We show here that the robot state-initialization can have a more important effect on the reward landscape than is generally expected. In particular, we show the counter-intuitive benefit of including initializations that are unviable, in other words initializing in states that are doomed to fail.

show abstract

Constructing predictive models of human running

Cited by 43 publications

References 33 publications

Spring-loaded inverted pendulum goes through two contraction-extension cycles during the single stance phase of walking

Spring-loaded inverted pendulum goes through two contraction-extension cycles during the single stance phase of walking

Data driven models of legged locomotion

Learning from outside the viability kernel: Why we should build robots that can fall with grace

Contact Info

Product

Resources

About