Paper Reading AI Learner

Low-Level Force-Control of MR-Hydrostatic Actuators

2022-07-06 13:37:51
Jeff Denis, Jean-Sebastien Plante, Alexandre Girard

Abstract

Precise and high-fidelity force control is critical for new generations of robots that interact with humans and unknown environments. Mobile robots, such as wearable devices and legged robots, must also be lightweight to accomplish their function. Hydrostatic transmissions have been proposed as a promising strategy for meeting these two challenging requirements. In previous publications, it was shown that using magnetorheological (MR) actuators coupled with hydrostatic transmissions provides high power density and great open-loop human-robot interactions. Still, the open-loop force fidelity at low and high frequencies are decreased by the transmission's dynamics and by nonlinear friction. This letter compares control strategies for MR-hydrostatic actuator systems to increase its torque fidelity, defined as the bandwidth (measured vs desired torque reference) and transparency (minimizing the undesired forces reflected to the end effector when backdriving the robot). Four control approaches are developed and compared experimentally: (1) Open-loop control with friction compensation; (2) non-collocated pressure feedback; (3) collocated pressure feedback; (4) LQGI state feedback. A dither strategy is also implemented to smoothen ball screw friction. Results show that approaches (1), (2) and (3) can increase the performances but are facing compromises, while approach (4) can simultaneously improve all metrics. These results show the potential of using control schemes for improving the force control performance of robots using tethered architectures, addressing issues such as transmission dynamics and friction.

Abstract (translated)

URL

https://arxiv.org/abs/2207.02671

PDF

https://arxiv.org/pdf/2207.02671.pdf


Tags
3D Action Action_Localization Action_Recognition Activity Adversarial Agent Attention Autonomous Bert Boundary_Detection Caption Chat Classification CNN Compressive_Sensing Contour Contrastive_Learning Deep_Learning Denoising Detection Dialog Diffusion Drone Dynamic_Memory_Network Edge_Detection Embedding Embodied Emotion Enhancement Face Face_Detection Face_Recognition Facial_Landmark Few-Shot Gait_Recognition GAN Gaze_Estimation Gesture Gradient_Descent Handwriting Human_Parsing Image_Caption Image_Classification Image_Compression Image_Enhancement Image_Generation Image_Matting Image_Retrieval Inference Inpainting Intelligent_Chip Knowledge Knowledge_Graph Language_Model Matching Medical Memory_Networks Multi_Modal Multi_Task NAS NMT Object_Detection Object_Tracking OCR Ontology Optical_Character Optical_Flow Optimization Person_Re-identification Point_Cloud Portrait_Generation Pose Pose_Estimation Prediction QA Quantitative Quantitative_Finance Quantization Re-identification Recognition Recommendation Reconstruction Regularization Reinforcement_Learning Relation Relation_Extraction Represenation Represenation_Learning Restoration Review RNN Salient Scene_Classification Scene_Generation Scene_Parsing Scene_Text Segmentation Self-Supervised Semantic_Instance_Segmentation Semantic_Segmentation Semi_Global Semi_Supervised Sence_graph Sentiment Sentiment_Classification Sketch SLAM Sparse Speech Speech_Recognition Style_Transfer Summarization Super_Resolution Surveillance Survey Text_Classification Text_Generation Tracking Transfer_Learning Transformer Unsupervised Video_Caption Video_Classification Video_Indexing Video_Prediction Video_Retrieval Visual_Relation VQA Weakly_Supervised Zero-Shot