Crocoddyl is an optimal control library for robot control under contact sequence. Its solver is based on various efficient Differential Dynamic Programming (DDP)-like algorithms
I created a planar monoped robot model using Pinocchio.robotwrapper, i.e., robot = RobotWrapper.BuildFromURDF( start_path+URDF_PATH, model_path, pin.JointModelPlanar()). The planar floating base needs three variables to specify its configuration, i.e., two translations (x,y) and one rotation (theta), whereas in the robot model, robot.q0 needs four variables, i.e., x,y,sin(theta),cos(theta). However, when we create action models using 'robot', the state dimension should be 6 rather than 8. I see this is dealt with using state =crocoddyl.StateMultibody(robot.model) to get the correct state. Could you please point me to where I could find the implementation details of multibody? Currently I could only find the header file multibody.hpp but not the
Thank you for being interested in our project.
If you have questions about Pinocchio, please post it there.
If you need to find any source file of our project, you can use the search engine of github.
Hello,
I created a planar monoped robot model using Pinocchio.robotwrapper, i.e.,
robot = RobotWrapper.BuildFromURDF( start_path+URDF_PATH, model_path, pin.JointModelPlanar())
. The planar floating base needs three variables to specify its configuration, i.e., two translations (x,y) and one rotation (theta), whereas in therobot model
,robot.q0
needs four variables, i.e., x,y,sin(theta),cos(theta). However, when we create action models using 'robot', the state dimension should be 6 rather than 8. I see this is dealt with usingstate =crocoddyl.StateMultibody(robot.model)
to get the correct state. Could you please point me to where I could find the implementation details of multibody? Currently I could only find the header file multibody.hpp but not the