Biped walking pattern generation by using preview control of zero-moment point

By Shuuji Kajita, Fumio Kanehiro, K. Kaneko, K. Fujiwara, K. Harada, K. Yokoi, Hirohisa Hirukawa. IEEE International Conference on Robotics and Automation, 2003

2,383 citations recorded by the source corpus.

Subfields: Biomedical Engineering, Control and Systems Engineering

View Paper: https://www.semanticscholar.org/paper/3b979cca666dcb0d7f2cfb49223227772458c38b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

In Control and Systems Engineering

Citations in the corpus

Viability and predictive control for safe locomotion (2008, positive): “The fact that minimizing the jerk of the CoM generates stable walking motions has already been acknowledged in [8], but through an optimal Linear Quadratic Regulator scheme only related to Lyapunov stability and therefore inducing viability only locally.”
Learning, planning, and control for quadruped locomotion over challenging terrain (2011, methodology): “In order to formulate stability constraints, we consider the cart-table ZMP model (Kajita et al. 2003) as an approximation of the LittleDog robot, which has a heavy body relative to its legs.”
Dynamic Walking with Compliance on a Cassie Bipedal Robot (2019, methodology): “A significant subset of the work lies in viewing walking dynamics as a reduction problem, wherein the complex real world dynamics are assumed to be governed by the evolution of some reduced system, such as a LIP models (Linear Inverted Pendulum [13]), SLIP models (Spring Loaded Inverted Pendulum [22”
CHOMP: Gradient optimization techniques for efficient motion planning (2009, methodology): “The initial trunk trajectory is given by a Zero Moment Point (ZMP) preview controller [13], and the initial swing leg trajectory is generated by interpolation through a collection of knot points intended to guide the swing foot a specified distance above the convex hull of the terrain.”
Three-Dimensional Bipedal Walking Control Based on Divergent Component of Motion (2015, methodology): “D force acting on the CoM can be expressed as F LIP = ⎡ ⎢⎣ 0 0 −mg ⎤ ⎥⎦ ︸︷︷︸ F g + mg z − pz (x − p) ︸︷︷︸ F e x t = mg z − pz ⎡ ⎢⎣ x − px y − py 0 ⎤ ⎥⎦ (2) where x = [x y z]T and p = [px py pz ]T denote CoM position and LIP base joint coordinates.”
Hierarchical quadratic programming: Fast online humanoid-robot motion generation (2014, background): “…to satisfy are • ejl: avoid the joint limits • epg: track the center-of-mass and flying foot references issued from a walking pattern generator [Kajita et al., 2003, Stasse et al., 2008b] • eθ: keep the angle of the umbrella stick below a fixed threshold • eumb: keep the companion below…”
Humanoid robots learning to walk faster: from the real world to simulation and back (2013, background): “Such works include [7], [13], [27], and [32].”
Real time motion generation and control for biped robot -1st report: Walking gait pattern generation- (2009, methodology): “An approach based on the control theory proposed by Kajita et al. [9] designs the position of the center of gravity (CoG) to converge above the position of desired ZMP at the end of the previewing period.”
Whole-body motion planning with centroidal dynamics and full kinematics (2014, background): “On the other hand, there exists a large arsenal of planning algorithms that use a simple dynamics model like the linear inverted pendulum [12], [5].”
From human to humanoid locomotion—an inverse optimal control approach (2010, methodology): “We use the walking pattern generator by Kajita et al. (2003), which is based on preview control of zero moment point (ZMP) using the table-cart inverted pendulum model, and which produces appropriate footprints and generates a desired ZMP trajectory.”
Gait and Trajectory Optimization for Legged Systems Through Phase-Based End-Effector Parameterization (2018, methodology): “This fast and effective approach has been successfully applied in bipedal and quadrupedal locomotion on real hardware [4]–[7].”
Online Walking Motion Generation with Automatic Footstep Placement (2010, background): “[4] to generate walking motions proposes to focus on the motion of the CoM of the walking robot.”
Trajectory Free Linear Model Predictive Control for Stable Walking in the Presence of Strong Perturbations (2006, methodology): “Introduced for the first time in the reference [11], the ZMP Preview Control scheme proposes to generate a trajectory of the CoM of a humanoid robot under the constraint that the footsteps are fixed and impossible to change.”
Control of Instantaneously Coupled Systems applied to humanoid walking (2010, methodology): “Preview control of the CoP can be used to generate CoM trajectories [2].”
Capturability-based analysis and control of legged locomotion, Part 2: Application to M2V2, a lower-body humanoid (2012, positive): “ZMP-based trajectory tracking control usually encompasses choosing a desired ZMP trajectory based on available footholds and desired gait properties, and calculating the center of mass (CoM) motion that results in that ZMP trajectory [13,14].”
Feedback Control of a Cassie Bipedal Robot: Walking, Standing, and Riding a Segway (2018, background): “The majority of walking gaits in 3D humanoid robots are stabilized by regulating the Zero-Moment Point (ZMP) [12], [19], [20], [22] (roughly the center of pressure of the stance foot).”