Warning: Undefined array key "DOKU_PREFS" in /is/htdocs/wp1019470_OPI92FFHXV/www/_lauflabor_v3/inc/common.php on line 2082

Warning: Cannot modify header information - headers already sent by (output started at /is/htdocs/wp1019470_OPI92FFHXV/www/_lauflabor_v3/inc/common.php:2082) in /is/htdocs/wp1019470_OPI92FFHXV/www/_lauflabor_v3/inc/common.php on line 2090

Warning: Cannot modify header information - headers already sent by (output started at /is/htdocs/wp1019470_OPI92FFHXV/www/_lauflabor_v3/inc/common.php:2082) in /is/htdocs/wp1019470_OPI92FFHXV/www/_lauflabor_v3/inc/actions.php on line 38
Locomotion Laboratory
Stylesheet style.css not found, please contact the developer of "django_v3" template.

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision 		Previous revision
Next revision Both sides next revision
projects:projects [2019/09/27 13:42]
Martin Grimmer [Improving walking assistance techniques for lower limb exoskeletons (2019 - 2023)] 		projects:projects [2020/06/05 12:12]
Guoping Zhao [Generating human walking gait with a neuromuscular model based on deep reinforcement learning (2018 - 2019)]
Line 32: Line 32:
 ===== Generating human walking gait with a neuromuscular model based on deep reinforcement learning (2018 - 2019) ===== ===== Generating human walking gait with a neuromuscular model based on deep reinforcement learning (2018 - 2019) =====
  
 +{{ youtube>ig8yQA80kgE?medium|Complex neuromuscular gait model with DRL}}
 A gait model capable of generating human-like walking behavior at both the kinematic and the muscular level can be a very useful framework for developing control schemes for humanoids and wearable robots such as exoskeletons and prostheses. In this work we are aiming at demonstrating the feasibility of using deep reinforcement learning based approach for neuromuscular gait modelling. A lower limb gait model consists of seven segments, fourteen degrees of freedom, and twenty two Hill-type muscles was built to capture human leg dynamics and the characteristics of muscle properties. We implemented the proximal policy optimization algorithm to learn the sensory-motor mappings (control policy) and generate human-like walking behavior for the model. Human motion capture data, muscle activation patterns and metabolic cost estimation were included in the reward function for training. A gait model capable of generating human-like walking behavior at both the kinematic and the muscular level can be a very useful framework for developing control schemes for humanoids and wearable robots such as exoskeletons and prostheses. In this work we are aiming at demonstrating the feasibility of using deep reinforcement learning based approach for neuromuscular gait modelling. A lower limb gait model consists of seven segments, fourteen degrees of freedom, and twenty two Hill-type muscles was built to capture human leg dynamics and the characteristics of muscle properties. We implemented the proximal policy optimization algorithm to learn the sensory-motor mappings (control policy) and generate human-like walking behavior for the model. Human motion capture data, muscle activation patterns and metabolic cost estimation were included in the reward function for training.
 +
  
 [[projects:projects_LearnWalk|Read more...]] [[projects:projects_LearnWalk|Read more...]]
  
-Contact: [[zhao@sport.tu-darmstadt.de|M.Sc. Guoping Zhao]] +Contact: [[zhao@sport.tu-darmstadt.de|Dr. Guoping Zhao]] 
  
  
Line 43: Line 45:
  
 Funded by DFG Funded by DFG
-{{ :projects:epa.jpg?200|Schematics of PAM arrangement}}+{{ :projects:epahopper.mp4?400|EPA-hopper hopping}}
 A better understanding of how actuator design supports locomotor function may 
help design and develop novel and more functional powered assistive or robotic legged
 systems. Legged locomotion can be described as a composition of locomotor
 sub-functions, namely axial leg function, leg swinging and balancing. In this 
project, we focus on the axial leg function (e.g., spring-like hopping) based on a novel concept of a hybrid electric-pneumatic actuator (EPA). This principal locomotor sub-function determines 
the movement of the body center of mass. We will design and manufacture EPA prototypes 
as enhanced variable impedance actuators (VIA). In contrast to other VIAs, the EPA provides not only adaptable compliance (e.g. an adjustable spring) 
but with the pneumatic artificial muscle (PAM) also 
an additional powerful actuator with muscle-like properties, which can be
arranged in different configurations (e.g., in series or parallel) to the electric motor (EM). This novel hybrid actuator
 shares the advantages of EM and PAM combining precise control with compliant
 energy storage required for efficient, robust and versatile human-like leg motions via simple control 
laws.  A better understanding of how actuator design supports locomotor function may 
help design and develop novel and more functional powered assistive or robotic legged
 systems. Legged locomotion can be described as a composition of locomotor
 sub-functions, namely axial leg function, leg swinging and balancing. In this 
project, we focus on the axial leg function (e.g., spring-like hopping) based on a novel concept of a hybrid electric-pneumatic actuator (EPA). This principal locomotor sub-function determines 
the movement of the body center of mass. We will design and manufacture EPA prototypes 
as enhanced variable impedance actuators (VIA). In contrast to other VIAs, the EPA provides not only adaptable compliance (e.g. an adjustable spring) 
but with the pneumatic artificial muscle (PAM) also 
an additional powerful actuator with muscle-like properties, which can be
arranged in different configurations (e.g., in series or parallel) to the electric motor (EM). This novel hybrid actuator
 shares the advantages of EM and PAM combining precise control with compliant
 energy storage required for efficient, robust and versatile human-like leg motions via simple control 
laws.