Mitsubishi Electric CR800-D Series, User Manual

Page: 38 / 194

Startup and initial settings 36

Detection level setting method
(1) Reduction gear
The detection level of the reduction gear abnormality detection for each joint axis is "-1" (detection
disabled) as a factory default. Since the score varies depending on the operating speed and operating
conditions such as the hand load by the hand and workpiece, set the detection level according to the
actual operating conditions. When the operating conditions such as the workpiece, hand, and operation
program are changed, review the setting of the detection level. If the setting of the detection level is not
appropriate, abnormalities may not be detected or mis-detection may occur.
The score is retained while the robot is stopped, and is gradually changed when the robot is operated.
The score is converged when the cycle operation is continued under the actual operating conditions.
After the score is sufficiently converged, set a larger value than the maximum value considering
variations. (The score may fluctuate when the robot is restarted.)
・
<Procedure>
1. Operate the robot under the actual operating conditions, and collect score data for about 7 to 30
days.
2. Check the maximum score of the log data.
365 days worth of the maximum score of the day can be saved as log data.
(Log data can be checked by going to [Comprehensive] - [Abnormality detection] - [Gear]. (Refer to
"7 Abnormality detection function ".)
3.Set the detection level higher than the maximum value (+10 as a guide) based on the log data.
* The score may fluctuate depending on whether there are changes in the ambient temperature or in
how much the robot vibrates. Review the detection level if there are changes in the environment
surrounding the robot.
* Review the detection level when changing operating conditions (such as program, hand, and
speed).
Setting example of detection level for reduction gears
The following figures show an example of log data for the past 30 days after starting operation. In the
following example, the detection level is set 10 points higher than the maximum value of the past 30 days.
The log data can be checked in detail by exporting it as a CSV file.