new_gripper_design #2
Conversation
There was a problem hiding this comment.
Thanks for your work! I have a few points I suggest to change:
There is no need for 3 python scripts to update each transform/axis. You can just update the axis positions in the action server and remove a lot of duplicated code and make it easier to understand. You also don't need the cmd_vel topic then to send the information to the other scripts. Same for the static_tf_gripper_origin.py
To Gripper.action in gripper_msgs:
- remove the outcommented variables
- feedback is not needed since the transform is given all necessary feedback already
To the README.md:
- reformat it so the code is copy-able and well structured like the README in gripper_tf_simulator
To gripper_tf_simulator/resource/gripper_tf_simulator
- remove this empty file/folder
| self.transf.transform.rotation.y = 0.0 | ||
| self.transf.transform.rotation.z = 0.0 | ||
| self.transf.transform.rotation.w = 1.0 | ||
|
|
||
| self.tf_broadcaster = TransformBroadcaster(self) | ||
| self.prev_time = self.get_clock().now() | ||
| self.timer = self.create_timer(0.1, self.tf_timer) |
There was a problem hiding this comment.
Call the topic something different than cmd_vel. You don't want to send velocity commands and it is already in use by the robotino driver. Call it cmd_gripper, gripper_commands, gripper_target or something like that.
| @@ -51,18 +69,24 @@ def tf_timer(self): | |||
| self.position_x += self.linear_x * dt | |||
| self.position_y += self.linear_y * dt | |||
| self.position_z += self.linear_z * dt | |||
|
|
|||
| self.yaw += self.angular_yaw * dt | |||
There was a problem hiding this comment.
What is the reason for this for the x axis?
| # Update the previous time | ||
| self.prev_time = current_time | ||
|
|
||
|
|
||
|
|
||
| def main(): | ||
| rclpy.init() |
There was a problem hiding this comment.
remove the try catch block below and just do
rclpy.spin(node)
rclpy.shutdown()
| # self.transf.transform.rotation.z = quat[2] | ||
| # self.transf.transform.rotation.w = quat[3] | ||
|
|
||
| quat = euler2quat(self.angular_roll, self.angular_pitch, self.angular_yaw) |
There was a problem hiding this comment.
this code is doubled and it is unclear for what you use it except for setting self.angular_yaw
| @@ -34,14 +47,19 @@ def arm_callback(self, msg: Twist): | |||
| self.linear_x = msg.linear.x | |||
There was a problem hiding this comment.
It is unclear why you do more than setting self.linear_x in this function.
| # feedback_msg.y_angle = goal_handle.request.pitch_target - orientation.y | ||
| # feedback_msg.z_angle = goal_handle.request.yaw_target - orientation.z | ||
| # feedback_msg.w_angle = goal_handle.request.qua_target - orientation.w | ||
| current_yaw = math.atan2( |
There was a problem hiding this comment.
Use euler2quat from transforms3d.euler for that instead. Else nobody can follow if this is correct.
| 1.0 - 2.0 * (orientation.y ** 2 + orientation.z ** 2) | ||
| ) | ||
| target_yaw = math.atan2(goal_handle.request.y_target, goal_handle.request.x_target) | ||
| z_angle = target_yaw - current_yaw | ||
| goal_handle.publish_feedback(feedback_msg) |
There was a problem hiding this comment.
Feedback is not needed since you update the TFs of the axis as feedback for the action clients.
| if (abs(feedback_msg.position_x) <= tolerance and | ||
| abs(feedback_msg.position_y) <= tolerance and | ||
| abs(feedback_msg.position_z) <= tolerance): | ||
| abs(z_angle) <= tolerance and |
There was a problem hiding this comment.
No need to check it like this. If you have the execution in this script as well, you know when the position is reached.
|
|
||
| # Check if the target position is reached (within a small tolerance) | ||
| tolerance = 0.01 # Define your tolerance | ||
| tolerance = 0.1 # Define your tolerance |
There was a problem hiding this comment.
10cm is a very large threshold. Try 0.005 instead.
| cmd_vel.linear.x = 3 * feedback_msg.position_x | ||
| cmd_vel.linear.y = 3 * feedback_msg.position_y | ||
| cmd_vel.linear.z = 3 * feedback_msg.position_z | ||
| if goal_handle.request.x_target <0 : |
There was a problem hiding this comment.
I will change this mathematical part to what we discussed today.
updates on the new gripper design