One-Step Ahead Prediction of Angular Momentum about the Contact Point for Control of Bipedal Locomotion: Validation in a LIP-inspired Controller

Gong, Yukai; Grizzle, Jessy W.

doi:10.1109/icra48506.2021.9560821

“…5 illustrates the performance of the two policies in a wide range of terrain elevations. 3 An interesting observation was that the NN Framework failed to generalize towards declines and opted to take very small steps to counter-act the incline. On the other hand linear policies show consistent and superior performance throughout the evaluated elevation range of −25 • to 25 • .…”

Section: ) Comparison With Baselinesmentioning

confidence: 99%

“…C LASSICAL works like spring loaded inverted pendulum (SLIP) with Raibert's heuristic controller [1], Zero Moment Point (ZMP) [2], the linear inverted pendulum [3] and Hybrid Zero Dynamics (HZD) [4] are designed to achieve robust walking behaviors on rough terrains. Despite the benefits of these works, like interpretability, existence of formal guarantees, scaling them for more complex tasks is not straightforward, involving a series of optimizations and tuning.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Linear Policies are Sufficient to Realize Robust Bipedal Walking on Challenging Terrains

Krishna

¹

,

Castillo

²

,

Mishra

³

et al. 2022

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

In this work, we demonstrate robust walking in the bipedal robot Digit on uneven terrains by just learning a single linear policy. In particular, we propose a new control pipeline, wherein the high-level trajectory modulator shapes the end-foot ellipsoidal trajectories, and the low-level gait controller regulates the torso and ankle orientation. The foot-trajectory modulator uses a linear policy and the regulator uses a linear PD control law. As opposed to neural network based policies, the proposed linear policy has only 13 learnable parameters, thereby not only guaranteeing sample efficient learning but also enabling simplicity and interpretability of the policy. This is achieved with no loss of performance on challenging terrains like slopes, stairs and outdoor landscapes. We first demonstrate robust walking in the custom simulation environment, MuJoCo, and then directly transfer to hardware with no modification of the control pipeline. We subject the biped to a series of pushes and terrain height changes, both indoors and outdoors, thereby validating the presented work.

show abstract

“…Classical works like spring loaded inverted pendulum (SLIP) with Raibert's heuristic controller [1], Zero Moment Point (ZMP) [2], the linear inverted pendulum [3] and Hybrid Zero Dynamics (HZD) [4] are designed to achieve robust walking behaviors on rough terrains. Despite the benefits of these works, like interpretability, existence of formal guarantees, scaling them for more complex tasks is not straightforward, involving a series of optimizations and tuning.…”

Section: Introductionmentioning

confidence: 99%

Linear Policies are Sufficient to Realize Robust Bipedal Walking on Challenging Terrains

Krishna¹,

Castillo²,

Mishra³

et al. 2021

Preprint

0

View full text Add to dashboard Cite

In this work, we demonstrate robust walking in the bipedal robot Digit on uneven terrains by just learning a single linear policy. In particular, we propose a new control pipeline, wherein the high-level trajectory modulator shapes the end-foot ellipsoidal trajectories, and the low-level gait controller regulates the torso and ankle orientation. The foot-trajectory modulator uses a linear policy and the regulator uses a linear PD control law. As opposed to neural network based policies, the proposed linear policy has only 13 learnable parameters, thereby not only guaranteeing sample efficient learning but also enabling simplicity and interpretability of the policy. This is achieved with no loss of performance on challenging terrains like slopes, stairs and outdoor landscapes. We first demonstrate robust walking in the custom simulation environment, MuJoCo, and then directly transfer to hardware with no modification of the control pipeline. We subject the biped to a series of pushes and terrain height changes, both indoors and outdoors, thereby validating the presented work.

show abstract

“…For example, SLIP models in conjunction with Raibert's controller are used to achieve stable walking [6]. Other methods include the centroidal models, the Zero-Moment Point (ZMP) [7], the linear inverted pendulum [8], and Hybrid Zero Dynamics (HZD) [9]. They are usually combined with either offline or online trajectory optimization to obtain a library of behaviors [10], [11].…”

Section: Introductionmentioning

confidence: 99%

Learning Linear Policies for Robust Bipedal Locomotion on Terrains with Varying Slopes

Krishna¹,

Mishra²,

Castillo³

et al. 2021

Preprint

0

View full text Add to dashboard Cite

In this paper, with a view toward deployment of light-weight control frameworks for bipedal walking robots, we realize end-foot trajectories that are shaped by a single linear feedback policy. We learn this policy via a model-free and a gradient free learning algorithm, Augmented Random Search (ARS), in the two robot platforms Rabbit and Digit. Our contributions are two-fold: a) By using torso and support plane orientation as inputs, we achieve robust walking on slopes of upto 20 • in simulation. b) We demonstrate additional behaviors like walking backwards, stepping-in-place, and recovery from external pushes of upto 120 N. The end-result is a robust and a fast feedback control law for bipedal walking on terrains with varying slopes. Towards the end, we also provide preliminary results of hardware transfer to Digit.

show abstract

One-Step Ahead Prediction of Angular Momentum about the Contact Point for Control of Bipedal Locomotion: Validation in a LIP-inspired Controller

Cited by 36 publications

References 22 publications

Linear Policies are Sufficient to Realize Robust Bipedal Walking on Challenging Terrains

Linear Policies are Sufficient to Realize Robust Bipedal Walking on Challenging Terrains

Linear Policies are Sufficient to Realize Robust Bipedal Walking on Challenging Terrains

Learning Linear Policies for Robust Bipedal Locomotion on Terrains with Varying Slopes

Contact Info

Product

Resources

About