Based upon this work, this shared library provides an easy-to-use interface for the analytical inverse kinematics solver for the Franka Emika Panda robot. The available functionalities are callable from both C++ and Python. A Python example is provided here.
git clone [email protected]:Anatr1/panda-IK.git
cd panda-IK
make configure
make build
make install
To obtain an array of doubles corresponding to the joint angles of the robot for a given end-effector pose, use the following function:
#include <panda_ik.hpp>
// ...
std::array<double, 7> result = compute_inverse_kinematics(array<double, 6> xyzrpy, array<double, 7> q_actual);
xyzrpy is an array of doubles containing the desired end-effector pose in the format [x, y, z, roll, pitch, yaw]. q_actual is an array of doubles containing the current joint angles of the robot. result will contain the solution, if any, for the given input. If no solution is found, the function will return the current joint angles of the robot.
To obtain an array of doubles corresponding to the joint angles of the robot for a given end-effector pose, use the following function:
import ctypes
# Load the shared library
pandaik_lib = ctypes.CDLL('/usr/local/diana/lib/libPanda-IK.so')
# Define the argument types and return type
pandaik_lib.compute_inverse_kinematics_void.argtypes = [ctypes.POINTER(ctypes.c_double), ctypes.POINTER(ctypes.c_double), ctypes.POINTER(ctypes.c_double)]
pandaik_lib.compute_inverse_kinematics_void.restype = None
xyzrpy = (ctypes.c_double * 6)(X, Y, Z, ROLL, PITCH, YAW)
q_actual = (ctypes.c_double * 7)(Q1, Q2, Q3, Q4, Q5, Q6, Q7)
output = (ctypes.c_double * 7)()
# Call a function from the library
pandaik_lib.compute_inverse_kinematics_void(xyzrpy, q_actual, output)X, Y, Z, ROLL, PITCH, YAW are the desired end-effector pose. Q1, Q2, Q3, Q4, Q5, Q6, Q7 are the current joint angles of the robot. output will contain the solution, if any, for the given input. If no solution is found, the function will return the current joint angles of the robot.