Movement reproduction and obstacle avoidance with dynamic movement primitives and potential fields

Park, Dae-Hyung; Hoffmann, Herbert; Pástor, Peter; Schaal, Stefan

doi:10.1109/ichr.2008.4755937

Cited by 101 publications

(20 citation statements)

References 12 publications

(13 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…where p(x, v) is the negative gradient of the potential field [27] and 𝜆 is a constant that indicates the strength of the entire field [10,28]. The potential field depends on the relative position and velocity of the end effector with respect to the obstacle.…”

Section: Obstacle Avoidance Based On the Dmp Methodsmentioning

confidence: 99%

“…Different types of APFs are constructed to apply in various obstacle avoidance scenarios. For point-like obstacles, static and dynamic potential fields are used to address static and moving obstacles, respectively [9,10]. Superquadric potential functions have also been adopted to model obstacles and achieve volumetric obstacle avoidance [11].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A model predictive obstacle avoidance method based on dynamic motion primitives and a Kalman filter

Wang

et al. 2022

Asian Journal of Control

View full text Add to dashboard Cite

A dynamic motion primitive (DMP) is a robust framework that generates obstacle avoidance trajectories by introducing perturbative terms. The perturbative term is usually constructed with an artificial potential field (APF) method.Dynamic obstacle avoidance is rarely considered with this approach; furthermore, even when dynamic obstacles are considered, only the velocity and position information of the current state are incorporated into the obstacle avoidance framework. However, if the position of an obstacle changes suddenly, a robot may be placed in a dangerous position close to the obstacle, resulting in large obstacle avoidance accelerations, sharp trajectories, or even obstacle avoidance failure. Therefore, we present a model predictive obstacle avoidance method based on dynamic motion primitives and a Kalman filter. This method has three main components: Dynamic motion primitives are used to generate the desired trajectory and introduce perturbations to achieve obstacle avoidance; the Kalman filter method is adopted to estimate the future positions of the obstacles; and model predictive control is employed to optimize the repulsive force generated by the APF while minimizing the defined cost function, thus guaranteeing the safety and flexibility of the method. We validate the presented method with 2D and 3D obstacle avoidance simulations. The method is also verified with a real robot: the-Kinova MOVO. The simulation and experimental results show that the proposed method not only avoids dynamic obstacles but also tracks the desired trajectory more smoothly and precisely. KEYWORDS artificial potential field (APF), dynamic motion primitive (DMP), dynamic obstacle avoidance, kalman filter, model predictive control (MPC)

show abstract

Section: Obstacle Avoidance Based On the Dmp Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A model predictive obstacle avoidance method based on dynamic motion primitives and a Kalman filter

Wang

et al. 2022

Asian Journal of Control

View full text Add to dashboard Cite

show abstract

“…Following the modified formulation of positional DMP introduced by ( Park et al, 2008 ), the differential equation of a one-dimensional positional DMP has three components. The first component is the transformation system that creates the trajectory plan:

where

and

are the position and velocity of a prescribed point of the system, respectively.…”

Section: Preliminariesmentioning

confidence: 99%

“…DMPs can generate a trajectory or control signal that can be flexibly adjusted to guide the real system without manual parameter tuning or affecting the overall convergence and stability. They can also be modulated to meet different requirements, such as obstacle avoidance ( Park et al, 2008 ; Hoffmann et al, 2009 ), by adding feedback terms. Therefore, we consider using DMPs to adapt an existing trajectory profile to new tasks so that we can learn new control policies based on the generalized trajectory profiles.…”

Section: Introductionmentioning

confidence: 99%

An Adaptive Imitation Learning Framework for Robotic Complex Contact-Rich Insertion Tasks

et al. 2022

View full text Add to dashboard Cite

Complex contact-rich insertion is a ubiquitous robotic manipulation skill and usually involves nonlinear and low-clearance insertion trajectories as well as varying force requirements. A hybrid trajectory and force learning framework can be utilized to generate high-quality trajectories by imitation learning and find suitable force control policies efficiently by reinforcement learning. However, with the mentioned approach, many human demonstrations are necessary to learn several tasks even when those tasks require topologically similar trajectories. Therefore, to reduce human repetitive teaching efforts for new tasks, we present an adaptive imitation framework for robot manipulation. The main contribution of this work is the development of a framework that introduces dynamic movement primitives into a hybrid trajectory and force learning framework to learn a specific class of complex contact-rich insertion tasks based on the trajectory profile of a single task instance belonging to the task class. Through experimental evaluations, we validate that the proposed framework is sample efficient, safer, and generalizes better at learning complex contact-rich insertion tasks on both simulation environments and on real hardware.

show abstract

“…When inferring the trajectory for an unseen task, the phase as well as the task-specific parameters will be passed. Some other extensions for DMPs focused on online adaptation, where one of the typical tasks could be collision avoidance (Park et al, 2008;Hoffmann et al, 2009;Tan et al, 2011).…”

Section: Movement Primitivesmentioning

confidence: 99%

Using Probabilistic Movement Primitives in Analyzing Human Motion Differences Under Transcranial Current Stimulation

Xue¹,

Herzog²,

Berger³

et al. 2021

Front. Robot. AI

View full text Add to dashboard Cite

In medical tasks such as human motion analysis, computer-aided auxiliary systems have become the preferred choice for human experts for their high efficiency. However, conventional approaches are typically based on user-defined features such as movement onset times, peak velocities, motion vectors, or frequency domain analyses. Such approaches entail careful data post-processing or specific domain knowledge to achieve a meaningful feature extraction. Besides, they are prone to noise and the manual-defined features could hardly be re-used for other analyses. In this paper, we proposed probabilistic movement primitives (ProMPs), a widely-used approach in robot skill learning, to model human motions. The benefit of ProMPs is that the features are directly learned from the data and ProMPs can capture important features describing the trajectory shape, which can easily be extended to other tasks. Distinct from previous research, where classification tasks are mostly investigated, we applied ProMPs together with a variant of Kullback-Leibler (KL) divergence to quantify the effect of different transcranial current stimulation methods on human motions. We presented an initial result with 10 participants. The results validate ProMPs as a robust and effective feature extractor for human motions.

show abstract

Movement reproduction and obstacle avoidance with dynamic movement primitives and potential fields

Cited by 101 publications

References 12 publications

A model predictive obstacle avoidance method based on dynamic motion primitives and a Kalman filter

A model predictive obstacle avoidance method based on dynamic motion primitives and a Kalman filter

An Adaptive Imitation Learning Framework for Robotic Complex Contact-Rich Insertion Tasks

Using Probabilistic Movement Primitives in Analyzing Human Motion Differences Under Transcranial Current Stimulation

Contact Info

Product

Resources

About