...
Geometric Properties: These are offsets and distances specific to the mechanical setup. They are likely used in the inverse kinematics calculations to account for the physical dimensions and positions of the components.
Test Positions: These arrays define the different positions and orientations that will be tested. Each loop combination represents a different configuration of the system.
Kinematic Solving: The function inverse_kin_5 is called to solve the inverse kinematics for each set of positions. This function is expected to return joint positions required to achieve the specified x, y, pitch, roll, and yaw.
Simulink Model Execution: The Simulink model SCU_FULL.slx is run for each configuration. The results of the simulation, including positions and orientations, are captured and processed.
Data Storage and Export: Results are stored in a preallocated table and finally saved as a CSV file for further analysis. The default location is project_root/Results
After running the space explorer for the following positions (it takes around 500 s to simulate that range):
| X | Y | PITCH | ROLL | YAW |
|---|---|---|---|---|
| -1 | -1 | -0.00127 | -0.00127 | -0.00127 |
| 0 | 0 | 0 | 0 | 0 |
| 1 | 1 | 0.00127 | 0.00127 | 0.00127 |
We obtain the following results:
By looking at the error for x and y it is clear that we are quite far from the accuracy requirement of 1 um. In particular Y it's very close (1.13 um of max error) while X it's completely out with an error of almost +- 100 um. However, even the Y error is quite far from the target since a good number is around one order of magnitude better than the overall accuracy requirement. That means that we must aim for an error of less than 10^-4.
We need to perform the optimization and compute the error function.
