Plausibility check of the output

After a successful run, the program outputs a number of information items:

For example, the block with the calculation results for the 1st pose is:

Example

example

Output of the calculation results of a sphere

% Centre of sphere [X, Y, Z] = [102.5305658, 244.3467991, 189.8131021]

% Radius R    = 14.94823794

% Variance    = 9.56062833e-006

% Residual[0] = 0.00154694034

% Residual[1] = -0.004483877805

% Residual[2] = 0.003641747138

% Residual[3] = -0.001544225933

% Residual[4] = -0.002650543572

% Residual[5] = 0.003491883191

"Centre of sphere" is the centre of the calibration sphere in XYZ coordinates for the pose specified. Radius “R” specifies the radius of the calibration sphere (more exactly: the sum of the radii of the calibration sphere and the dial gauge tip or switching point of the probe tip).

Criterion (A) – radii of the calibration spheres

The calibration sphere radii measured across all poses should be constant except for the measurement precision. If this is not the case, the calibration sphere was not correctly measured in one or several poses. Repeat the measurement for these poses.

Criterion (B) – Residuals

The residuals (Residual[i]) should be within the range of a few 1/100mm. They indicate the amount by which each of the measurement points deviate from the calibration sphere surface. If a measurement point deviates strongly from the sphere surface, reject the affected pose and repeat this measurement.

Calculation output

The 2nd part of the output contains the calculated Compensation values (P-CHAN-00438) and several additional items of information. This block looks like this, for example for kinematic 80:

Example

example

Output of calculated kinematic parameters

#timestamp                        12.12.2016 12:58:17:000

#kinematic id                     80

#kinematic id variant             0

#number of measurement records    8

#

#record 0 = [102.5305658,244.3467991,189.8131021, 0,0,0]

#record 1 = [102.545449, 258.949268, 125.6186118, 20,0,0]

#record 2 = [29.18053549,233.8711714,175.957552, -20,-90,0]

#record 3 = [176.2562104,233.7423534,176.3588772, -20.90,0]

#record 4 = [29.20839651,244.3560828,116.1031577, 0,-90,0]

#record 5 = [176.2877121,244.3573191,116.5414503, 0.90,0]

#record 6 = [176.2795252,233.8784364,56.68825135, 20.90,0]

#record 7 = [29.21147408,233.7284348,56.33260263, 20,-90,0]

#

#solution precision       0.04409219769

#probe position PCS, [mm]   [-0.2058572907, 0, 73.53215927]

#

kinematik[80].corr[1]    E39A4FA5FA49658AFD79AE9E

kinematik[80].corr[2]    E39A4FA5FA49658AFD79AE9E

kinematik[80].corr[3]    E39A4FA5FA49658AFD79AE9E

kinematik[80].corr[4]    E39A4FA5FA49658AFD79AE9E

kinematik[80].corr[5]    6E9B4FA534A809BEB3FDC8FB

kinematik[80].corr[6]    E39A4FA5FA49658AFD79AE9E

kinematik[80].corr[7]    E9954FA50E96594B756EB4F7A79567C2084AC996

kinematik[80].corr[8]    E39A4FA5FA49658AFD79AE9E

kinematik[80].corr[9]    A18A4FA512510AB45F3259BCA79567C2084AC996

kinematik[80].corr[10]   60914FA56EE1CFF3A6AA2953

kinematik[80].corr[11]   AD924FA53196E9D01A53CB3D

Besides the time stamp, the kinematic ID and the number of poses, the solution precision is also output, among other items. This value describes in mm the largest deviation between model and measurement values.

The deviation is produced by measurement errors or by errors not mapped in the model, e.g. axis misalignments, etc.

Notice

notice

The compensation parameters are output in encrypted form (P-CHAN-00438) and later decrypted on import.

Criterion (C) – solution precision

The solution precision should not exceed a certain limit. A limit for this value can be specified in the main program. If this limit is exceeded, an error is output and the measurement equipment, measurement method etc. must be checked.

Criterion (D) – kinematic parameters calculated

If the check is successful, the above mentioned block can be copied to the channel parameter list. After the latest import of the channel parameters, the compensation values are available to the kinematic.

Attention

attention

Then a careful, slow test run should be executed with the kinematic enabled to check whether the compensation values have the correct effectiveness. This also checks that the transfer of compensation values is error-free and that no unintended collisions occur.