A "force-vision-position" coupled compliant assembly method for industrial robots

Abstract

The present invention discloses a "force-vision-position" coupled compliant assembly method for industrial robots. Based on multiple binocular vision sensors, the conversion matrix of the assembled workpiece relative to the main robot, that is, the theoretical assembly position, is obtained; based on the six-dimensional force sensor, the gravity of the assembled workpiece is identified, and real-time gravity compensation is performed according to the robot posture to obtain the actual external contact force of the assembled workpiece; the actual external contact force is monitored in real time and the displacement correction is performed on the current direction. Based on the binocular vision sensor, the actual position of the current assembled workpiece is obtained, and the relative distance between the actual position and the theoretical assembly position is combined to move the assembled workpiece until the relative distance is less than the error threshold, and the assembly is completed. The present invention realizes the precision compensation method of online detection of assembly errors, which can meet assembly tasks under different postures.

Description

Technical Field

[0001] The invention belongs to the technical field of industrial robot assembly, and in particular relates to a "force-vision-position" coupled compliant assembly method for an industrial robot. Background Art

[0002] At present, the application of assembly robots in my country's manufacturing industry has achieved certain development, but there is still much room for improvement. First, there are a large number of assembly tasks with low repeatability in the manufacturing industry. These tasks often have unfixed spatial positions and assembly positions. They need to be actually planned according to specific assembly tasks. If a single task is planned once, it will often affect the assembly efficiency and assembly time. Secondly, with the progress and development of the manufacturing industry, the precision requirements for workpiece assembly are also increasing. The manual positioning assembly method and the repetitive and single robot assembly in the past are often only suitable for work with low assembly precision requirements and repetitive and single processes. It is difficult to complete assembly tasks with more precision requirements and complicated assembly processes. Finally, there are a large number of large-load workpiece assembly tasks in the manufacturing industry. Such assembly tasks are often carried out by tower cranes, rope cranes or robots clamping workpieces and multiple workers manually positioning and assembling. In such assembly methods, the assemblers, assembly workpieces, and auxiliary assembly tools often need to be in a fixed assembly operation area, and the spatial freedom is poor. At the same time, the traditional assembly method not only requires a high level of proficiency for the assemblers, but also has problems such as cumbersome processes and high labor intensity. This not only makes the assembly efficiency low, but also brings great safety hazards to the entire assembly process. The application of intelligent manufacturing technology in the assembly process is an urgent need of my country's manufacturing industry. Especially in the field of aerospace, there are extremely high requirements for the assembly accuracy, assembly efficiency, and assembly safety of large load parts. In the past, the manual installation method had fixed processes and cumbersome steps. There were many potential dangers for long-term assembly of large workpiece loads. On the other hand, manual installation could not meet the increasingly high requirements for spatial positioning and assembly accuracy.

[0003] With the continuous optimization and upgrading of visual sensors, force sensors and robot hardware and control technology. On the one hand, the recognition and positioning accuracy of visual sensors is getting higher and higher. On the other hand, robots have the characteristics of dexterity, high efficiency, high flexibility, strong versatility, low cost and easy integration. With the continuous optimization and upgrading of hardware and control technology, the absolute positioning accuracy and stability of industrial robots have been greatly improved, which can meet the precision assembly of assembly workpieces. This paper aims to combine the high flexibility, high precision and easy integration of visual sensors, force sensors and industrial robots to achieve the smooth assembly of "force-vision-position" coupling.

[0004] Document CN104625676A discloses an industrial robot system for shaft hole assembly. This method requires a fixed and precise relative position between the robot and the assembly platform, and then uses a wrist force sensor to control the contact force to complete the assembly operation. This method has great limitations on the assembly position. Document CN104690551A discloses a robot automated assembly system, which obtains assembly position and posture data through multiple visual sensors and laser rangefinders to perform motion planning. This method uses non-contact measurement, which has great limitations for assembly tasks that are sensitive to external contact forces.

[0005] Document CN106272416A discloses a robot slender shaft precision assembly system and method based on force perception and vision. The accuracy of the robot end is ensured by tracking a known trajectory in the position control direction, while active compliance is achieved in the force control direction, and small contact force and torque are maintained over a long stroke to achieve precision assembly of the slender shaft. However, this method does not perform gravity compensation on the workpiece clamped at the end to obtain the gravity magnitude and center of gravity position of the workpiece in the force sensor coordinate system. As a result, there are greater constraints on the assembly posture. The robot end clamps the workpiece in a fixed posture to perform the assembly task to achieve a better force control effect, which has certain limitations.

[0006] In summary, how to better combine the high flexibility, high precision, and easy integration characteristics of force sensors, visual sensors, and industrial robots to achieve assembly tasks in different positions and postures and those that are sensitive to external forces is an important research issue. Summary of the invention

[0007] The technical problem to be solved by the present invention is to provide a "force-vision-position" coupled compliant assembly method for an industrial robot in view of the deficiencies of the above-mentioned prior art.

[0008] In order to achieve the above technical objectives, the technical solution adopted by the present invention is:

[0009] A "force-vision-position" coupled compliant assembly method for industrial robots realizes industrial robot assembly tasks in different positions by combining force sensors and vision sensors, including:

[0010] Step S1, obtaining a transformation matrix of the workpiece under assembly relative to the main robot based on multiple binocular vision sensors, that is, a theoretical assembly position;

[0011] Step S2, performing gravity identification on the assembled workpiece based on the six-dimensional force sensor, and performing real-time gravity compensation according to the robot posture to obtain the actual external contact force of the assembled workpiece;

[0012] Step S3, monitor the actual external contact force in real time and perform displacement correction on the current direction, obtain the actual position of the workpiece currently assembled based on the binocular vision sensor, and move the workpiece on the assembly based on the relative distance between the actual position and the theoretical assembly position obtained in step S1 until the relative distance is less than the error threshold and the assembly is completed.

[0013] To optimize the above technical solutions, the specific measures taken also include:

[0014] The above step S1 includes:

[0015] Step S11: calibrate and measure the transformation matrix of the main robot base coordinate system under the large field of view binocular vision sensor

[0016] Step S12: Use the first small field binocular vision sensor to measure the target point calibration plate to obtain a conversion matrix of the target point calibration plate relative to the first small field binocular vision sensor.

[0017] Step S13: Use a large field of view binocular vision sensor to measure the target point calibration plate to obtain a conversion matrix of the target point calibration plate relative to the large field of view binocular vision sensor.

[0018] Step S14: Obtain the transformation matrix of the first small field of view binocular vision sensor relative to the main robot base coordinate system according to the transformation matrix obtained in steps S11, S12 and S13.

[0019]

[0020] Step S15: Use the second small field binocular vision sensor to measure the target point calibration plate to obtain the conversion matrix of the target point calibration plate relative to the second small field binocular vision sensor. Then the transformation matrix of the second small field of view binocular vision sensor relative to the main robot base coordinate system is obtained:

[0021]

[0022] Step S16: Obtain the conversion matrix of the second small field of view binocular vision sensor relative to the first small field of view binocular vision sensor according to the conversion matrix obtained in step S14 and step S15. Used for coordinate system conversion after feature recognition of workpiece;

[0023]

[0024] Step S17, using the first small field of view binocular vision sensor and the second small field of view binocular vision sensor to measure the assembled workpiece coordinate system, converting the features measured by the second small field of view binocular vision sensor to the first small field of view binocular vision sensor coordinate system through the conversion matrix in step S16, and obtaining the conversion matrix of the assembled workpiece coordinate system relative to the first small field of view binocular vision sensor through the feature fitting method

[0025] Step S18: Read the current X, Y, Z, A, B, and C of the robot to obtain the transformation matrix of the current main robot flange relative to the base coordinate system. Then the transformation matrix of the workpiece coordinate system on the assembly relative to the flange of the main robot is obtained

[0026]

[0027] The X, Y, Z, A, B, and C represent the displacement and Euler angle of the flange of the main robot in the base coordinate system respectively;

[0028] Step S19, using the first small field of view binocular vision sensor and the second small field of view binocular vision sensor to measure the coordinate system of the assembled workpiece, converting the features measured by the second small field of view binocular vision sensor to the coordinate system of the first small field of view binocular vision sensor through the conversion matrix in step S16, and obtaining the conversion matrix of the assembled workpiece coordinate system relative to the first small field of view binocular vision sensor through the feature fitting method Then the theoretical transformation matrix of the flange relative to the main robot during assembly is obtained That is, the theoretical assembly position [XYZABC] theory :

[0029]

[0030] The above step S2 includes:

[0031] Step S21, gravity identification is performed on the assembled workpiece. The robot grabs the workpiece and runs a certain trajectory, changes the robot posture, and reads the force sensor data and inertial force compensation during the operation of the robot to obtain N groups of force sensor data. Then, the center of gravity position (x, y, z) and gravity magnitude G of the assembled workpiece in the six-dimensional force sensor coordinate system and the initial zero drift value [F x0 F y0 F z0 T x0 T y0 T z0 ] T ;

[0032] Step S22, read the six-dimensional force data of the six-dimensional force sensor, and obtain the X, Y, Z, A, B, C of the robot in real time, perform real-time gravity compensation according to the obtained robot posture, and obtain the component of the gravity of the assembled workpiece in the six-dimensional force sensor coordinate system [G x0 G y0 G z0 G x0 G y0 G z0 ] T And the zero drift value of the six-axis force sensor [F x0 F y0 F z0 T x0 T y0 T z0 ] T up-part , and then the actual contact force of the workpiece on the assembly [F x F y F z T x T y T z ];

[0033] Step S23: The industrial robot moves the assembled workpiece to the initial position and calculates the current external actual contact force [F x F y F z T x T y T z ] T up-part , and add it to the calculated zero drift value of the six-axis force sensor to reset the zero drift of the six-axis force sensor and obtain the current zero drift value of the sensor [F x0 F y0 F z0 T x0 T y0 T z0 ] T up-part’ , return to step S22.

[0034] The above step S3 includes:

[0035] Step S31: Assume that the six-axis single-step movement of the industrial robot is [△x△y△z△a△b△c] robot , by interpolating the distance that the workpiece needs to move for assembly with the single-step movement of the six-axis, the number of steps required for interpolation is obtained;

[0036] Step S32: Setting the actual contact force threshold of the workpiece [F x F y Fz T x T y T z ] Limited-Lower And the acceptable error threshold between the actual position of the workpiece and the theoretical assembly position [△x△y△z△a△b△c] Limited-Dis ;

[0037] Real-time monitoring of the actual external contact force. If the actual external contact force is greater than the contact force threshold, the current direction is corrected.

[0038] Step S33, according to the interpolation steps obtained in step S31, control the industrial robot to perform interpolation. After the interpolation of the industrial robot is completed, use the first small-field binocular vision sensor and the second small-field binocular vision sensor to capture the actual position of the assembled workpiece, and calculate the relative distance that the assembled workpiece needs to move. If it is less than the error threshold, the assembly is completed; if it is greater than the threshold, return to step S14 until it is less than the error threshold, and the assembly is completed.

[0039] In the above step S33, the first small field binocular vision sensor and the second small field binocular vision sensor are used to photograph the characteristic hole of the currently assembled workpiece to obtain the actual position [X 0 Y 0 Z 0 A 0 B 0 C 0 ] up-part , and calculate the relative distance between the theoretical assembly position and the actual position of the assembled workpiece That is the relative distance the workpiece needs to be moved to assemble.

[0040] The present invention has the following beneficial effects:

[0041] 1) The "force-vision-position" coupled compliant assembly method for industrial robots proposed in the present invention combines the high-precision and high-flexibility characteristics of a six-dimensional force sensor and multiple binocular vision sensors to achieve an accuracy compensation method for online detection of assembly errors.

[0042] 2) The present invention performs gravity identification on the assembled workpiece, thereby being able to obtain the gravity size of the assembled workpiece and the position of the center of gravity in the force sensor coordinate system, and then to achieve gravity compensation for the assembled workpiece itself. Combined with the data from the six-dimensional force sensor, the actual external contact force of the assembled workpiece in different postures is obtained, ultimately satisfying the assembly tasks in different postures.

[0043] 3) The present invention obtains the displacement of the robot in the assembly task through visual detection, and at the same time has the function of external force detection. On the one hand, it realizes real-time compensation of the visually guided displacement, thereby correcting its offset error; on the other hand, it can also minimize the risk of collision of the assembled workpieces during the entire assembly process. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a schematic diagram of the process of the present invention;

[0045] Figure 2 It is a schematic diagram of the positions of various components in the assembly task of the present invention;

[0046] Figure 3 Schematic diagram of the coordinate transformation relationship among the robot, force sensor and load in the present invention;

[0047] Figure 4 Schematic diagram of the force sensor coordinate system and load gravity center in the present invention. DETAILED DESCRIPTION

[0048] The embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings.

[0049] like Figure 1-4 As shown, a "force-vision-position" coupled compliant assembly method for industrial robots realizes industrial robot assembly tasks in different positions by combining force sensors and vision sensors, and is based on the following hardware: industrial robot, assembly upper workpiece, assembly lower workpiece, six-dimensional force sensor, large field of view binocular vision sensor, first small field of view binocular vision sensor, second small field of view binocular vision sensor. The robot clamps the assembly upper workpiece with a quick-change fixture, and the robot is within its own working range during the entire assembly process. The method includes:

[0050] Step S1, obtaining a transformation matrix of the workpiece under assembly relative to the main robot based on multiple binocular vision sensors, that is, a theoretical assembly position;

[0051] Step S2, performing gravity identification on the assembled workpiece based on the six-dimensional force sensor, and performing real-time gravity compensation according to the robot posture to obtain the actual external contact force of the assembled workpiece;

[0052] Step S3, monitor the actual external contact force in real time and perform displacement correction on the current direction, obtain the actual position of the workpiece currently assembled based on the binocular vision sensor, and move the workpiece on the assembly based on the relative distance between the actual position and the theoretical assembly position obtained in step S1 until the relative distance is less than the error threshold and the assembly is completed.

[0053] In the embodiment, the step S1 comprises:

[0054] Step S11: calibrate and measure the transformation matrix of the main robot base coordinate system under the large field of view binocular vision sensor

[0055] Step S12: Use the first small field binocular vision sensor to measure the target point calibration plate to obtain a conversion matrix of the target point calibration plate relative to the first small field binocular vision sensor.

[0056] Step S13: Use a large field of view binocular vision sensor to measure the target point calibration plate to obtain a conversion matrix of the target point calibration plate relative to the large field of view binocular vision sensor.

[0057] Step S14: Obtain the transformation matrix of the first small field of view binocular vision sensor relative to the main robot base coordinate system according to the transformation matrix obtained in steps S11, S12 and S13.

[0058]

[0059] Step S15: Use the second small field binocular vision sensor to measure the target point calibration plate to obtain the conversion matrix of the target point calibration plate relative to the second small field binocular vision sensor. Then the transformation matrix of the second small field of view binocular vision sensor relative to the main robot base coordinate system is obtained:

[0060]

[0061] Step S16: Obtain the conversion matrix of the second small field of view binocular vision sensor relative to the first small field of view binocular vision sensor according to the conversion matrix obtained in step S14 and step S15. Used for coordinate system conversion after feature recognition of workpiece;

[0062]

[0063] Step S17, using the first small field of view binocular vision sensor and the second small field of view binocular vision sensor to measure the assembled workpiece coordinate system, converting the features measured by the second small field of view binocular vision sensor to the first small field of view binocular vision sensor coordinate system through the conversion matrix in step S16, and obtaining the conversion matrix of the assembled workpiece coordinate system relative to the first small field of view binocular vision sensor through the feature fitting method

[0064] Step S18: Read the current X, Y, Z, A, B, and C of the robot to obtain the transformation matrix of the current main robot flange relative to the base coordinate system. Then the transformation matrix of the workpiece coordinate system on the assembly relative to the flange of the main robot is obtained

[0065]

[0066] The X, Y, Z, A, B, and C represent the displacement and Euler angle of the flange of the main robot in the base coordinate system respectively;

[0067] Step S19, using the first small field of view binocular vision sensor and the second small field of view binocular vision sensor to measure the coordinate system of the assembled workpiece, converting the features measured by the second small field of view binocular vision sensor to the coordinate system of the first small field of view binocular vision sensor through the conversion matrix in step S16, and obtaining the conversion matrix of the assembled workpiece coordinate system relative to the first small field of view binocular vision sensor through the feature fitting method Then the theoretical transformation matrix of the flange relative to the main robot during assembly is obtained That is, the theoretical assembly position [XYZABC] theory :

[0068]

[0069] In the embodiment, step S2 comprises:

[0070] Step S21, gravity identification is performed on the assembled workpiece. The robot grabs the workpiece and runs a certain trajectory, changes the robot posture, and reads the force sensor data and inertial force compensation during the operation of the robot to obtain N groups of force sensor data. Then, the center of gravity position (x, y, z) and gravity magnitude G of the assembled workpiece in the six-dimensional force sensor coordinate system and the initial zero drift value [F x0 F y0 F z0 T x0 T y0 T z0 ] T ;

[0071] Step S22, read the six-dimensional force data of the six-dimensional force sensor, and obtain the X, Y, Z, A, B, C of the robot in real time, perform real-time gravity compensation according to the obtained robot posture, and obtain the component of the gravity of the assembled workpiece in the six-dimensional force sensor coordinate system [G x0 G y0 G z0 G x0 G y0 G z0 ] T And the zero drift value of the six-axis force sensor [F x0 F y0 F z0 T x0 Ty0 T z0 ] T up-part Then, the actual external contact force of the assembled workpiece [F x F y F z T x T y T z ];

[0072] Step S23: The industrial robot moves the assembled workpiece to the initial position and calculates the current external actual contact force [F x F y F z T x T y T z ] T up-part , and add it to the calculated zero drift value of the six-axis force sensor to reset the zero drift of the six-axis force sensor and obtain the current zero drift value of the sensor [F x0 F y0 F z0 T x0 T y0 T z0 ] T up-part’ , return to step S22.

[0073] In the embodiment, step S3 includes:

[0074] Step S31: Assume that the six-axis single-step movement of the industrial robot is [△x△y△z△a△b△c] robot , by interpolating the distance that the workpiece needs to move for assembly with the single-step movement of the six-axis, the number of steps required for interpolation is obtained;

[0075] Step S32: Setting the actual contact force threshold of the workpiece [F x F y F z T x T y T z ] Limited-Lower And the acceptable error threshold between the actual position of the workpiece and the theoretical assembly position [△x△y△z△a△b△c] Limited-Dis ;

[0076] Real-time monitoring of the actual external contact force. If the actual external contact force is greater than the contact force threshold, the current direction is corrected.

[0077] Step S33, according to the interpolation steps obtained in step S31, control the industrial robot to perform interpolation. After the interpolation of the industrial robot is completed, use the first small-field binocular vision sensor and the second small-field binocular vision sensor to capture the actual position of the assembled workpiece, and calculate the relative distance that the assembled workpiece needs to move. If it is less than the error threshold, the assembly is completed; if it is greater than the threshold, return to step S14 until it is less than the error threshold, and the assembly is completed.

[0078] In the embodiment, in step S33, the first small field binocular vision sensor and the second small field binocular vision sensor are used to photograph the characteristic hole of the currently assembled workpiece to obtain the actual position [X 0 Y 0 Z 0 A 0 B 0 C 0 ] up-part , and calculate the relative distance between the theoretical assembly position and the actual position of the assembled workpiece That is the relative distance the workpiece needs to be moved to assemble.

[0079] The above are only preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should be regarded as the protection scope of the present invention.

Claims

1. A "force-vision-position" coupled compliant assembly method for industrial robots, which realizes industrial robot assembly tasks in different positions by combining force sensors and visual sensors, and is characterized in that: include: Step S1, obtaining a transformation matrix of the workpiece under assembly relative to the main robot based on multiple binocular vision sensors, that is, a theoretical assembly position; Step S2, performing gravity identification on the assembled workpiece based on the six-dimensional force sensor, and performing real-time gravity compensation according to the robot posture to obtain the actual external contact force of the assembled workpiece; Step S3, real-time monitoring of the actual external contact force and displacement correction of the current direction, obtaining the actual position of the currently assembled workpiece based on the binocular vision sensor, and moving the assembled workpiece in combination with the relative distance between the actual position and the theoretical assembly position obtained in step S1, until the relative distance is less than the error threshold, and the assembly is completed; The step S1 comprises: Step S11: calibrate and measure the transformation matrix of the main robot base coordinate system under the large field of view binocular vision sensor Step S12: Use the first small field binocular vision sensor to measure the target point calibration plate to obtain a conversion matrix of the target point calibration plate relative to the first small field binocular vision sensor. Step S13: Use a large field of view binocular vision sensor to measure the target point calibration plate to obtain a conversion matrix of the target point calibration plate relative to the large field of view binocular vision sensor. Step S14: Obtain the transformation matrix of the first small field of view binocular vision sensor relative to the main robot base coordinate system according to the transformation matrix obtained in steps S11, S12 and S13. Step S15: Use the second small field binocular vision sensor to measure the target point calibration plate to obtain the conversion matrix of the target point calibration plate relative to the second small field binocular vision sensor. Then the transformation matrix of the second small field of view binocular vision sensor relative to the main robot base coordinate system is obtained: Step S16: Obtain the conversion matrix of the second small field of view binocular vision sensor relative to the first small field of view binocular vision sensor according to the conversion matrix obtained in step S14 and step S15. Used for coordinate system conversion after feature recognition of workpiece; Step S17, using the first small field of view binocular vision sensor and the second small field of view binocular vision sensor to measure the assembled workpiece coordinate system, converting the features measured by the second small field of view binocular vision sensor to the first small field of view binocular vision sensor coordinate system through the conversion matrix in step S16, and obtaining the conversion matrix of the assembled workpiece coordinate system relative to the first small field of view binocular vision sensor through the feature fitting method Step S18: Read the current X, Y, Z, A, B, and C of the robot to obtain the transformation matrix of the current main robot flange relative to the base coordinate system. Then the transformation matrix of the workpiece coordinate system on the assembly relative to the flange of the main robot is obtained The X, Y, Z, A, B, and C represent the displacement and Euler angle of the flange of the main robot in the base coordinate system respectively; Step S19, using the first small field of view binocular vision sensor and the second small field of view binocular vision sensor to measure the coordinate system of the assembled workpiece, converting the features measured by the second small field of view binocular vision sensor to the coordinate system of the first small field of view binocular vision sensor through the conversion matrix in step S16, and obtaining the conversion matrix of the assembled workpiece coordinate system relative to the first small field of view binocular vision sensor through the feature fitting method Then the theoretical transformation matrix of the flange relative to the main robot during assembly is obtained That is, the theoretical assembly position [XYZ AB C] theory : The step S3 comprises: Step S31: Assume that the six-axis single-step movement of the industrial robot is [△x△y△z△a△b△c] robot , by interpolating the distance that the workpiece needs to move for assembly with the single-step movement of the six-axis, the number of steps required for interpolation is obtained; Step S32: Setting the actual contact force threshold of the workpiece [F x F y F z T x T y T z ] Limited-Lower And the acceptable error threshold between the actual position of the workpiece and the theoretical assembly position [△x△y△z△a△b△c] Limited-Dis ; Monitor the actual external contact force in real time. If the actual external contact force is greater than the contact force threshold, perform displacement correction on the current direction. Step S33, according to the interpolation step number obtained in step S31, control the industrial robot to perform interpolation, after the interpolation of the industrial robot is completed, use the first small field of view binocular vision sensor and the second small field of view binocular vision sensor to capture the actual position of the assembled workpiece, and calculate the relative distance that the assembled workpiece needs to move, if it is less than the error threshold, the assembly is completed; if it is greater than the threshold, return to step S14, until it is less than the error threshold, the assembly is completed; In the step S33, the first small field binocular vision sensor and the second small field binocular vision sensor are used to photograph the characteristic hole of the currently assembled workpiece to obtain the actual position [X0Y0Z0A0B0C0] of the currently assembled workpiece. up-part , and calculate the relative distance between the theoretical assembly position and the actual position of the assembled workpiece That is the relative distance the workpiece needs to be moved to assemble.

2. The "force-vision-position" coupled compliant assembly method for an industrial robot according to claim 1, characterized in that: The step S2 comprises: Step S21, gravity identification is performed on the assembled workpiece. The robot grabs the workpiece and runs a certain trajectory, changes the robot posture, reads the force sensor data during the operation of the robot to obtain N groups of force sensor data, and then obtains the center of gravity position (x, y, z) of the assembled workpiece in the six-dimensional force sensor coordinate system and the gravity magnitude G and the initial zero drift value [F x0 F y0 F z0 T x0 T y0 T z0 ] T ; Step S22, read the six-dimensional force data of the six-dimensional force sensor, and obtain the X, Y, Z, A, B, C of the robot in real time, perform real-time gravity compensation according to the obtained robot posture, and obtain the component of the gravity of the assembled workpiece in the six-dimensional force sensor coordinate system [G x0 G y0 G z0 G x0 G y0 G z0 ] T And the zero drift value of the six-axis force sensor [F x0 F y0 F z0 T x0 T y0 T z0 ] T up-part , and then the actual contact force of the workpiece on the assembly [F x F y F z T x T y T z ]; Step S23: The industrial robot moves the assembled workpiece to the initial position and calculates the current external actual contact force [F x F y F z T x T y T z ] T up-part , and add it to the calculated zero drift value of the six-axis force sensor to reset the zero drift of the six-axis force sensor and obtain the current zero drift value of the sensor [F x0 F y0 F z0 T x0 T y0 T z0 ] T up-part’ , return to step S22.

Images