Stable, Autonomous, Unknown Terrain Locomotion for Quadrupeds Based on Visual Feedback and Mixed-Integer Convex Optimization

Ahn, Min Sung; Chae, Hosik; Hong, Dennis

doi:10.1109/iros.2018.8594015

Cited by 13 publications

(12 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…x, y, ψ represents our state variables (desired planar position and yaw or heading angle), and U = v x , v y , ψ represents our control variables (desired planar velocity and yaw rate). Matrices A and B are the same as shown in (1), except for an additional row/column for yaw and yaw rate.…”

Section: B General Mpc Formulationmentioning

confidence: 99%

“…Necessary requirements for robots to autonomously perform such complex tasks include, but are not limited to, online low-level feedback controls, localization, vision, motion planning, high-level reasoning, and reasoning under uncertainty. Currently, all individual components are welldeveloped, but integrating multiple pieces together into a single system, especially for environments that are not wellknown, has proven to be a daunting challenge because of issues related to robustness [1]. For example, simultaneous planning, localization, and mapping (SPLAM, or "Active SLAM") is an active area of research that attempts to satisfy some of these requirements.…”

Section: Introductionmentioning

confidence: 99%

“…To resolve the above issues, we propose a multifaceted approach that uses model predictive control (MPC), SLAM 1 , and recurrent neural network (RNN) algorithms to address the problem of Active SLAM and account for uncertainties in both current and future robot positions. Our architecture is based on MPC because MPC operates online, continually satisfies the dynamic state of the robot over a prediction horizon N , and naturally offsets estimation errors [12].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Risk-Averse MPC via Visual-Inertial Input and Recurrent Networks for Online Collision Avoidance

Schperberg,

Chen,

Tsuei

et al. 2020

Preprint

View full text Add to dashboard Cite

In this paper, we propose an online path planning architecture that extends the model predictive control (MPC) formulation to consider future location uncertainties for safer navigation through cluttered environments. Our algorithm combines an object detection pipeline with a recurrent neural network (RNN) which infers the covariance of state estimates through each step of our MPC's finite time horizon. The RNN model is trained on a dataset that comprises of robot and landmark poses generated from camera images and inertial measurement unit (IMU) readings via a state-of-the-art visualinertial odometry framework. To detect and extract object locations for avoidance, we use a custom-trained convolutional neural network model in conjunction with a feature extractor to retrieve 3D centroid and radii boundaries of nearby obstacles. The robustness of our methods is validated on complex quadruped robot dynamics and can be generally applied to most robotic platforms, demonstrating autonomous behaviors that can plan fast and collision-free paths towards a goal point.and Aerospace Engineering. {aschperberg28, mbjewett318, hooksj, dennishong}@ucla.edu 2 K. Chen and A. Mehta are with the Laboratory for Embedded Machines and Ubiquitous Robots, Department of Electrical and Computer Engineer-

show abstract

Section: B General Mpc Formulationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Risk-Averse MPC via Visual-Inertial Input and Recurrent Networks for Online Collision Avoidance

Schperberg,

Chen,

Tsuei

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…To fully realize this capability, they often use motion planning to autonomously choose their footholds and plan their body movements to avoid slipping and tumbling. Optimization-based techniques have been exploited by researchers to resolve the motion planning problem [1]- [4]. The easiest way to implement optimization for motion planning would be to include all the linear or nonlinear equations into a single optimization problem.…”

Section: Introductionmentioning

confidence: 99%

Designing Multi-Stage Coupled Convex Programming with Data-Driven McCormick Envelope Relaxations for Motion Planning

Lin¹,

Ahn²,

Hong³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

For multi-limbed robots, motion planning with posture and force constraints tends to be a difficult optimization problem due to nonlinearities, which also present extended solve times. We propose a multi-stage optimization framework with data-driven inter-stage coupling constraints to address the nonlinearity. Both clustering and evolutionary approaches to find the McCormick envelope relaxations are used to find the problem-specific parameters. The learned constraints are then used in the prior stages, which provides advanced knowledge of the following stages. This leads to improved solve times and interpretability of the results. The planner is validated through multiple walking and climbing tasks on a 10 kg hexapod robot.

show abstract

“…In this paper, we present another approach which utilizes optimization methods for multi-limbed climbing robots to plan trajectories when climbing. Optimization based methods, such as mixed-integer convex programming (MICP) and nonlinear programming (NLP), have been implemented in many situations to plan motions for walking robots [9] [10] [11] [12]. This paper extends these methods to wall-climbing applications.…”

Section: Introductionmentioning

confidence: 99%

Optimization Based Motion Planning for Multi-Limbed Vertical Climbing Robots

Lin

Zhang

Shen

et al. 2019

2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

View full text Add to dashboard Cite

Motion planning trajectories for a multilimbed robot to climb up walls requires a unique combination of constraints on torque, contact force, and posture. This paper focuses on motion planning for one particular setup wherein a six-legged robot braces itself between two vertical walls and climbs vertically with end effectors that only use friction. Instead of motion planning with a single nonlinear programming (NLP) solver, we decoupled the problem into two parts with distinct physical meaning: torso postures and contact forces. The first part can be formulated as either a mixed-integer convex programming (MICP) or NLP problem, while the second part is formulated as a series of standard convex optimization problems. Variants of the two wall climbing problem e.g. , obstacle avoidance, uneven surfaces, and angled walls, help verify the proposed method in simulation and experimentation.

show abstract

Stable, Autonomous, Unknown Terrain Locomotion for Quadrupeds Based on Visual Feedback and Mixed-Integer Convex Optimization

Cited by 13 publications

References 18 publications

Risk-Averse MPC via Visual-Inertial Input and Recurrent Networks for Online Collision Avoidance

Risk-Averse MPC via Visual-Inertial Input and Recurrent Networks for Online Collision Avoidance

Designing Multi-Stage Coupled Convex Programming with Data-Driven McCormick Envelope Relaxations for Motion Planning

Optimization Based Motion Planning for Multi-Limbed Vertical Climbing Robots

Contact Info

Product

Resources

About