Abstract:Contact detection is the key part in the unified event-based motion control framework of hydraulic quadruped robots. Without the use of an external force sensor, this process usually requires estimating the contact force, which is then compared with a certain set threshold to judge whether contact occurs. But there are some challenges when applied to hydraulic robots, due to the inherent complexity of the hydraulic system, such as high-order, non-linearity, strong interference, etc. The typical threshold method uses only single force information, which is easy to make incorrect inferences when the feedback hydraulic signal has a large unknown perturbation. What’s more, the dynamic model of a hydraulic robot is more complex and difficult to establish accurately, making some ways of estimating force based on complete dynamics inapplicable. To tackle such problems, this paper introduces the method of probabilistic model fusion for contact estimation, and applies it to our hydraulic single-leg platform. As far as we know, it may be the first time that this probability method is applied to a hydraulic legged robot. The method fuses multiple available information and makes optimal estimation, which can reduce the misjudgment caused by the interference of single information. Moreover, a simplified method for estimating contact force through impedance relationship is proposed. Since the method does not require complex dynamic modeling and torque feedback, it is computationally simple and effective and suitable for hydraulic robots. The above theory is tested in a general framework combined with an event-based state machine, and the experimental results verify the effectiveness. The work of this paper can provide a basis for the quadruped robot to cross the unstructured terrain without external perception in the follow-up work.

Terrain Estimation for Quadruped Robots Based on Kalman Filter

Invariant EKF based State Estimator for Quadruped Robots

Learning Accurate and Robust Velocity Tracking for Quadrupedal Robots

Real-time Localization, Mapping and Navigation for Quadruped Robots

Efficient Motion Planning Based on Kinodynamic Model for Quadruped Robots Following Persons in Confined Spaces

Adaptive Robust Invariant Extended Kalman filtering for Biped Robot.

Contact Detection by Probabilistic Model Fusion for Hydraulic Quadruped Robots

State Estimation for Quadruped Robots on Non-Stationary Terrain via Invariant Extended Kalman Filter and Disturbance Observer

Nonlinear MPC-Based Control Framework for Quadruped Robots: Touch-Down in Complex Terrain

State estimation and traversability map construction method of a quadruped robot on soft uneven terrain

SVM-based Extended Kalman Filter State Estimation of Biped Robot.

Simultaneous State Estimation and Contact Detection for Legged Robots by Multiple-Model Kalman Filtering

State Estimation for Hybrid Locomotion of Driving-Stepping Quadrupeds

Contact detection with multi-information fusion for quadruped robot locomotion under unstructured terrain

State Estimation for a Humanoid Robot

Vision-Assisted Localization and Terrain Reconstruction with Quadruped Robots

Proprioceptive State Estimation for Quadruped Robots using Invariant Kalman Filtering and Scale-Variant Robust Cost Functions

Design and Dynamic Locomotion Control of Quadruped Robot with Perception-Less Terrain Adaptation

State Estimation for Legged Robots Using Contact-Centric Leg Odometry

Probabilistic Contact State Estimation for Legged Robots using Inertial Information

OptiState: State Estimation of Legged Robots using Gated Networks with Transformer-based Vision and Kalman Filtering