14.5.1.3. Parameters Calibration of Joint Torque Sensor on the Whole Machine
14.5.1.3.1. Overview
Joint torque sensor sensitivity refers to the sensor’s responsiveness to torque changes, describing the proportional relationship between the sensor’s output voltage and the actual joint torque being measured. Linearity measures how well a regression model fits the observed data. Hysteresis error is the maximum difference between measurements during forward stroke (from small to large) and reverse stroke (from large to small) under the same test conditions for the joint torque sensor’s raw data. Repeatability is the ratio of the current test result to the previous test result, used to determine the repeatability accuracy of the joint torque sensor.
The parameter calibration method involves running the robot through a predetermined trajectory, and calculating the joint torque sensor’s sensitivity, linearity, hysteresis error, and repeatability accuracy by acquiring joint gravity torque and joint torque sensor raw data under different postures.
14.5.1.3.2. Parameter Calibration
Step1: Set the tool coordinate system to “Tool0”. Click “Auxiliary Applications” -> “Tool Applications” -> “Drag Lock”. In the joint torque sensor whole-machine drag module, click “Function Enable”.
Figure 14.5‑4 Function Enable
Step2: After clicking “Function Enable”, proceed with sensitivity calibration. Click “Generate Program” to deploy the internal controller Lua script. Switch the robot to automatic mode and set the run speed to “10”. Click “Run” and wait for the robot to move.
Figure 14.5‑5 Sensitivity Calibration
Note
If the joint torque sensor sensitivity calibration is already completed, you can proceed directly to drag function parameter settings.
Step3: After the robot completes running the predetermined trajectory, the sensitivity, linearity, hysteresis error, and repeatability calibration results are automatically displayed on the web interface. Click “Set” to apply.
Figure 14.5‑6 Parameter Calibration Results
14.5.1.4. External Force Estimation and Torque Compensation Based on Momentum Observer
14.5.1.4.1. Overview
After enabling the torque compensation function, the robot reduces the dragging torque during current loop dragging, improving the dragging experience.
14.5.1.4.2. Operation Process
Step1: Set the dynamics configuration to “Dynamics 2.0”. Click “Auxiliary Applications ” -> “Tool Applications” -> “Drag Lock”. In the dual encoder torque compensation module, click the function switch to enable it.
Chart 14.5‑7 Enable Function
Step2: Set the “Function Switch” to “ON”, and set the drag gain for each axis to 0.5. Click “Set” to apply, as shown in the figure.
Chart 14.5‑8 Gain Setting
Note
Drag gain setting range: 0-1. The larger the gain, the greater the compensation torque, and the easier the dragging under the current loop.
14.5.1.5. Assisted Drag Optimization Function Based on Joint Torque Sensor
14.5.1.5.1. Overview
This user manual describes the usage of the assisted drag optimization function based on the joint torque sensor. It involves three drag modes and, compared to traditional drag teaching methods, can improve drag compliance and reduce the drag force required by each joint.
14.5.1.5.2. Assisted Drag Optimization Function Based on Joint Torque Sensor
14.5.1.5.2.1. Zero Point Calibration and Sensitivity Calibration
Step1: If the zero point calibration and sensitivity calibration have already been completed (the indicator light before calibration is green), there is no need to perform them again. Recalibrate the zero point only when there is a floating sensation during dragging. The calibration process is described below.
Figure 14.5‑9 Indicator Light Status After Zero Point and Sensitivity Calibration of Joint Torque Sensor
Step2: Zero point calibration. Click “Auxiliary Applications” → “Tool Applications” → “Drag Lock” to enter the “Joint Torque Sensor Whole Body Drag” module. Click the “Calibrate” button for zero point calibration to calibrate the zero point data of the joint torque sensor. When the calibration is complete, a “√” will appear, and the zero point calibration result will be updated.
Figure 14.5‑10 Zero Point Calibration of Joint Torque Sensor
Step3: Sensitivity calibration (Note: It is recommended to use only the robot body without any load during calibration). Switch the robot motion mode to “Automatic Mode” and set the operation speed to “10%”. Click the “Calibrate” button for sensitivity calibration and wait for the robot to complete its motion. After the robot completes the predetermined trajectory, the sensitivity coefficient, linearity, hysteresis error, and repeatability calibration results are automatically displayed on the web interface.
Figure 14.5‑11 Sensitivity Calibration of Joint Torque Sensor
Step4: Set the assisted drag function. There are three drag modes, which can be set after completing “Zero Point Calibration” and “Sensitivity Calibration”. If not set, the default is “Mode Three”, meaning that drag teaching can be performed directly in drag mode after calibration.
14.5.1.5.3. Assisted Drag Function - Mode One
Step1: Select the drag mode as “Mode One”. When the robot motion mode is “Manual Mode”, set the sliding window size, gain coefficient, and joint speed, then click “Apply”. At this point, holding down the end “Drag Button” or in drag mode enables drag teaching.
Figure 14.5‑12 Mode One: Parameter Setting
Note
The recommended setting for the sliding window size is 30, with a maximum value of 100;
The gain coefficient affects the feel during dragging. A larger coefficient makes dragging more sensitive but can easily cause instability. The recommended setting for J1-J6 is 0.5;
The recommended joint speed is 6°/s, which can alleviate overshoot during point alignment.
14.5.1.5.4. Assisted Drag Function - Mode Two
Step1: Select the drag mode as “Mode Two”. When the robot motion mode is “Manual Mode”, set the mass coefficient, damping coefficient, stiffness coefficient, and force control threshold, then click “Apply”. At this point, drag teaching can be performed in position mode.
Figure 14.5‑13 Mode Two: Parameter Setting
Note
The mass coefficient affects the joint inertial force during dragging. The recommended settings are: J1-J3: 1.0, J4-J5: 0.5, J6: 0.1;
The damping coefficient affects the feel during dragging. Higher damping results in a heavier feel. The recommended settings are: J1-J3: 10.0, J4-J5: 5.0, J6: 1.0;
The stiffness coefficient should be set to 0 for all;
The force control threshold is the activation torque during dragging. The recommended settings are: J1-J3: 0.3, J4-J5: 0.2, J6: 0.1.
14.5.1.5.5. Assisted Drag Function - Mode Three
Step1: Select the drag mode as “Mode Three”. When the robot motion mode is “Manual Mode”, set the gain coefficient for each joint, then click “Apply”. At this point, holding down the end “Drag Button” or in drag mode enables drag teaching.
Figure 14.5‑14 Mode Three: Parameter Setting
Note
The gain coefficient affects the joint drag force during low-speed dragging. When the coefficient ranges from 0.1 to 1.0, the resistance during low-speed dragging increases as the coefficient increases. For precise point alignment tasks, it is recommended to set the J1-J6 gain coefficient to 1.0. When considering overall ease of dragging and compliance, it is recommended to set the J1-J6 gain coefficient to 0.3.
14.6. Intersection Point Generation (Laser Point Capture Motion)
During welding, the laser point capture motion can be configured with posture, enabling the robot to achieve the expected posture when reaching the position point. This easily handles special scenarios such as fillet welds and groove welds.
14.6.1. Laser Point Capture Motion Function Operation Process
Step1: Before using the laser sensor, first apply the “Welding Torch” tool coordinate system to the current tool coordinate system. Open the Teach Page, click “Initial Setup”, “Basic”, “Tool Coordinates” in sequence, select “Welding Torch” as the coordinate system name and apply it. The tool coordinate system in the system status bar will display as Tool1.
Figure 14.6‑1 Apply Welding Torch Coordinate System
Step2: Write the laser point capture motion Lua program. Click “Teach Program” -> “Program Programming” -> “New Button” in sequence, and create a new user program “testPointRecord.lua”.
Figure 14.6‑2 New Laser Point Capture Motion Program
Step3: Configure the reference posture teach point (Optional). In manual mode, drag the robot to the desired welding posture. On the Teach Page, click “Teach Point Record”, “Name Point” in sequence, and save the posture teach point as “referencePoint”.
Figure 14.6‑3 Save Posture Reference Teach Point
Step4: Generate the laser point capture motion program. Click “Teach Program” -> “Program Programming” -> “Welding Instructions”, “Laser Tracking” in sequence, scroll down to the Sensor Point Capture Motion section, select the required “Motion Mode”, “Debug Speed”, and posture reference point to generate the corresponding laser point capture LUA program.
If no posture reference point is selected, the posture at the point of capture is maintained by default during motion. If a posture reference point is selected, the robot moves to the laser-captured point using the reference posture.
Figure 14.6‑4 Select Posture Reference Teach Point
Execute the laser point capture motion. Drag the robot so that the laser sensor beam points to the desired weld seam point. The laser sensor will acquire the weld seam position and capture the point. After executing the laser point capture motion, the welding torch will move to the point scanned by the laser sensor using the reference posture.
Figure 14.6‑5 Laser Acquires Weld Seam Position
Figure 14.6‑6 Welding Torch Points to Weld Seam Position with Reference Posture