Nonparametric Inverse Dynamic Models for Multimodal Interactive Robots

TL;DR

This work addresses inverse-dynamics learning for interactive robots operating with multimodal load dynamics and unmeasured external inputs. It introduces a nonparametric framework based on multimodal Gaussian process regression, with an EM/SEM-based scheme to cluster data into dynamic modes and learn mode-specific inverse dynamics, while also identifying intermittent disturbances via disturbance covariance. The approach provides a passivity-preserving feedforward plus impedance compensation and is validated on a test actuator, showing improved impedance rendering and reliable mode classification, even in the presence of Coulomb friction. Overall, it enables autonomous, safe handling of changing payloads and external perturbations in unstructured environments, advancing the capability of semi-autonomous robots to adapt their interaction behavior without heavily relying on predefined models.

Abstract

Direct design of a robot's rendered dynamics, such as in impedance control, is now a well-established control mode in uncertain environments. When the physical interaction port variables are not measured directly, dynamic and kinematic models are required to relate the measured variables to the interaction port variables. A typical example is serial manipulators with joint torque sensors, where the interaction occurs at the end-effector. As interactive robots perform increasingly complex tasks, they will be intermittently coupled with additional dynamic elements such as tools, grippers, or workpieces, some of which should be compensated and brought to the robot side of the interaction port, making the inverse dynamics multimodal. Furthermore, there may also be unavoidable and unmeasured external input when the desired system cannot be totally isolated. Towards semi-autonomous robots, capable of handling such applications, a multimodal Gaussian process regression approach to manipulator dynamic modelling is developed. A sampling-based approach clusters different dynamic modes from unlabelled data, also allowing the seperation of perturbed data with significant, irregular external input. The passivity of the overall approach is shown analytically, and experiments examine the performance and safety of this approach on a test actuator.

Figures (10)

• Figure 1: Interactive System Model
• Figure 2: Upper-Limb Exoskeleton
• Figure 3: Friction Model with Coulomb Friction and Stribek Effect
• Figure 4: Experimental Setup
• Figure 5: Rendering of Zero Impedance