pybullet学习笔记(二）——Dynamics, Kinematics

控制一个机器人，参见 PyBullet快速上手教程(二).

Kinematics，Dynamics

目录

Kinematics，Dynamics

calculateInverseKinematics

calculateInverseDynamics

calculateJacobian

---

calculateInverseKinematics

You can compute the joint angles that makes the end-effector reach a given target position in Cartesian world space. Internally, Bullet uses an improved version of Samuel Buss Inverse Kinematics library. At the moment only the Damped Least Squares method with or without Null Space control is exposed, with a single end-effector target. Optionally you can also specify the target orientation of the end effector. In addition, there is an option to use the null-space to specify joint limits and rest poses. This optional null-space support requires all 4 lists (lowerLimits, upperLimits, jointRanges, restPoses), otherwise regular IK will be used. See also inverse_kinematics.py example in Bullet/examples/pybullet/examples folder for details.

  逆向运动学计算：给定对象id，当前关节位置和末端执行器的目标位置，计算逆向运动学并返回新的关节状态。输入参数如下

required	bodyUniqueId	int	机器人的唯一ID，就是使用loadURDF...时返回的数值
required	endEffectorLinkIndex	int	末端执行器link索引
required	targetPosition	vec3, list of 3 floats	目标位置（笛卡尔世界坐标系）
optional	targetOrientation	vec3, list of 3 floats	目标姿态（笛卡尔世界坐标系），四元数[x，y，z，w]。 如果未指定，将计算纯位置IK。
optional	lowerLimits	list of floats [0..nDof]	可选的null-space IK,需要所有4个列表（lowerLimits，upperLimits， jointRanges，restPoses）。否则将使用常规IK
optional	upperLimits	list of floats [0..nDof]	关节限位
optional	jointRanges	list of floats [0..nDof]
optional	restPoses	list of floats [0..nDof]	Favor an IK solution closer to a given rest pose
optional	jointDamping	list of floats [0..nDof]	jointDamping allow to tune the IK solution using joint damping factors
optional	solver	int	p.IK_DLS or p.IK_SDLS, Damped Least Squares or Selective Damped Least Squares, as described in the paper by Samuel Buss "Selectively Damped Least Squares for Inverse Kinematics".
optional	physicsClientId	int	if you are connected to multiple servers, you can pick one.

calculateInverseKinematics返回关节位置列表。 有关示例，请参阅 Bullet / examples / pybullet / inverse_kinematics.py。
请注意，calculateInverseKinematics的计算结果只是一个近似值，并不精确。 使用“ccurateCalculateInverseKinematics”可以获得更准确的结果。

calculateInverseDynamics

calculateInverseDynamics将从指定的关节位置和速度开始计算达到给定关节加速度所需的力。输入参数列表:

required	bodyUniqueId	int	机器人的唯一ID，就是使用loadURDF...时返回的数值
required	objPositions	list of float	每个自由度（DoF）的关节位置（角度）。Note that fixed joints have 0 degrees of freedom. The base is skipped/ignored in all cases (floating base and fixed base).
required	objVelocities	list of float	每个关节的速度
required	objAccelerations	list of float	每个关节的期望加速度
optional	physicsClientId	int	if you are connected to multiple servers, you can pick one

calculateInverseDynamics返回每个自由度的关节力列表。 

calculateJacobian