Towards robust online inverse dynamics learning

Meier, Franziska; Kappler, Daniel; Ratliff, Nathan; Schaal, Stefan

doi:10.1109/iros.2016.7759594

Cited by 31 publications

(30 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We support the proposed algorithm with an experimental validation on a compliant 6DoF robotic arm, shown in Figure 1, which makes use of low-power series elastic actuators [10]. The results demonstrate that the controller improves the tracking performance in terms of root mean square tracking error compared to a PID controller, the offset-free MPC scheme presented in [11], the nonlinear MPC scheme [12], and the three learning-based schemes presented in [13], [14]. In particular, the proposed controller reduces the root mean square tracking error by up to 53% compared with the bestpractice PID controller.…”

Section: Introductionsupporting

confidence: 52%

Data-Driven Model Predictive Control for Trajectory Tracking With a Robotic Arm

Carron

Arcari

Wermelinger

et al. 2019

IEEE Robot. Autom. Lett.

133

View full text Add to dashboard Cite

High-precision trajectory tracking is fundamental in robotic manipulation. While industrial robots address this through stiffness and high-performance hardware, compliant and cost-effective robots require advanced control to achieve accurate position tracking. In this paper, we present a model-based control approach, which makes use of data gathered during operation to improve the model of the robotic arm and thereby the tracking performance. The proposed scheme is based on an inverse dynamics feedback linearization and a data-driven error model, which are integrated into a model predictive control formulation. In particular, we show how offset-free tracking can be achieved by augmenting a nominal model with both a Gaussian Process, which makes use of offline data, and an additive disturbance model suitable for efficient online estimation of the residual disturbance via an extended Kalman filter. The performance of the proposed offset-free GPMPC scheme is demonstrated on a compliant 6 degrees of freedom (DoF) robotic arm, showing significant performance improvements compared to other robot control algorithms.

show abstract

Section: Introductionsupporting

confidence: 52%

Data-Driven Model Predictive Control for Trajectory Tracking With a Robotic Arm

Carron

Arcari

Wermelinger

et al. 2019

IEEE Robot. Autom. Lett.

133

View full text Add to dashboard Cite

show abstract

“…This allows for exponential smoothing of the GP hyperparameters, which increases the robustness of the GP at the cost of having slower reactiveness. Nevertheless, [17] does not provide a proof of the robust stability of the closed-loop system. In [18], the variance of the GP prediction is utilized to adapt the parameters of an outer-loop PD controller online, and the uniform ultimate boundedness of the tracking error is proved under some assumptions on the structure of the PD controller (e.g., the gain matrix was assumed to be diagonal, which imposes a decentralized gain control scheme).…”

Section: Related Workmentioning

confidence: 99%

“…Since the Lyapunov function V is strictly decreasing outsidē B δ , the tracking error e(t) eventually reaches and remains in S δ ⊂B c , and so the tracking error e(t) is uniformly ultimately bounded, and its ultimate bound is the radius of B c . Note that from (17), B δ ⊂ {e ∈ R 2N : e T P e < e(0) T P e(0)} ⊂ D, and ρ is a correct upper bound on η . One can see that δ and hence the radius of B c depend on the choice of the design parameter .…”

Section: Theoretical Guaranteesmentioning

confidence: 99%

Provably Robust Learning-Based Approach for High-Accuracy Tracking Control of Lagrangian Systems

Helwa

Heins

Schoellig

2019

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

Lagrangian systems represent a wide range of robotic systems, including manipulators, wheeled and legged robots, and quadrotors. Inverse dynamics control and feedforward linearization techniques are typically used to convert the complex nonlinear dynamics of Lagrangian systems to a set of decoupled double integrators, and then a standard, outer-loop controller can be used to calculate the commanded acceleration for the linearized system. However, these methods typically depend on having a very accurate system model, which is often not available in practice. While this challenge has been addressed in the literature using different learning approaches, most of these approaches do not provide safety guarantees in terms of stability of the learning-based control system. In this paper, we provide a novel, learning-based control approach based on Gaussian processes (GPs) that ensures both stability of the closed-loop system and high-accuracy tracking. We use GPs to approximate the error between the commanded acceleration and the actual acceleration of the system, and then use the predicted mean and variance of the GP to calculate an upper bound on the uncertainty of the linearized model. This uncertainty bound is then used in a robust, outer-loop controller to ensure stability of the overall system. Moreover, we show that the tracking error converges to a ball with a radius that can be made arbitrarily small. Furthermore, we verify the effectiveness of our approach via simulations on a 2 degree-offreedom (DOF) planar manipulator and experimentally on a 6 DOF industrial manipulator.

show abstract

“…An improvement of accuracy in the modeling then directly converts into a reduction of gains for these controllers. This reduction consequently is a major objective and performance criterion for the success of such modeling approaches [14,16,18]. …”

Section: Hybrid Modelingmentioning

confidence: 99%

Hybrid Analytical and Data-Driven Modeling for Feed-Forward Robot Control †

Reinhart

Shareef

Steil

2017

Sensors

View full text Add to dashboard Cite

Feed-forward model-based control relies on models of the controlled plant, e.g., in robotics on accurate knowledge of manipulator kinematics or dynamics. However, mechanical and analytical models do not capture all aspects of a plant's intrinsic properties and there remain unmodeled dynamics due to varying parameters, unmodeled friction or soft materials. In this context, machine learning is an alternative suitable technique to extract non-linear plant models from data. However, fully data-based models suffer from inaccuracies as well and are inefficient if they include learning of well known analytical models. This paper thus argues that feed-forward control based on hybrid models comprising an analytical model and a learned error model can significantly improve modeling accuracy. Hybrid modeling here serves the purpose to combine the best of the two modeling worlds. The hybrid modeling methodology is described and the approach is demonstrated for two typical problems in robotics, i.e., inverse kinematics control and computed torque control. The former is performed for a redundant soft robot and the latter for a rigid industrial robot with redundant degrees of freedom, where a complete analytical model is not available for any of the platforms.

show abstract

Towards robust online inverse dynamics learning

Cited by 31 publications

References 6 publications

Data-Driven Model Predictive Control for Trajectory Tracking With a Robotic Arm

Data-Driven Model Predictive Control for Trajectory Tracking With a Robotic Arm

Provably Robust Learning-Based Approach for High-Accuracy Tracking Control of Lagrangian Systems

Hybrid Analytical and Data-Driven Modeling for Feed-Forward Robot Control †

Contact Info

Product

Resources

About