Assembly optimization method based on big data and intelligent manufacturing

Abstract

The invention relates to an assembly optimization method based on big data and intelligent manufacturing. The method comprises the steps that a feature analysis module obtains a first assembly angle feature of a target to-be-assembled surface and a second assembly angle feature of a target combined surface according to a plurality of assembly feature points and a plurality of combined feature points; a point location association module establishes a space point location mapping relation between each datum point to be assembled and each combined datum point according to the orientation similarity; the action decomposition module analyzes the assembly action of the assembly robot in real time to obtain the corresponding assembly movement direction and assembly movement distance; the error analysis module performs error analysis according to the assembly movement direction, the assembly movement distance, the second assembly angle feature and the first assembly angle feature to obtain an assembly direction deviation error; and the orientation correction module corrects the motion trail and the assembly angle of the assembly robot in real time to obtain orientation correction parameters on the corresponding space motion points, and sends the orientation correction parameters to the corresponding assembly robot.

Description

Technical Field

[0001] The present invention relates to the field of big data and intelligent manufacturing, and in particular to an assembly optimization method based on big data and intelligent manufacturing. Background Art

[0002] The core goal of intelligent manufacturing is to continuously connect the supply, manufacturing, sales and other data in production through information-physical systems, the Internet of Things, end-to-end technologies and various application software (enterprise resource planning systems, supply chain management, product lifecycle management, etc.), and establish a highly flexible and flexible product production and service model, thereby realizing the vertical integration of workshop production processes and the horizontal integration of the manufacturing value network. Through the Internet of Things technology, any objects or processes that need to be monitored, connected, and interacted with are collected in real time, realizing ubiquitous links between objects and objects, and between objects and people, and then realizing active perception of important data in the manufacturing process, thereby providing timely, accurate, and comprehensive manufacturing process operation information for intelligent decision-making of the manufacturing system.

[0003] In the field of intelligent manufacturing, robot assembly technology, as a new generation of service and production methods, has been identified as one of the cutting-edge technologies with a great impact on human development. It can be applied not only in national fields such as industrial manufacturing, bioengineering, medical treatment, and education, but also in fields such as marine engineering and aerospace. Therefore, robot assembly technology is widely used in various industrial production. In practical applications, since the assembly parts are fixed at the specified position in the processing area, even a small error may lead to assembly failure. Therefore, it is difficult for assembly robots in traditional modes to maintain high efficiency and high quality assembly in a complex and changeable working environment, which seriously limits the actual application scope of assembly robots. Summary of the invention

[0004] In view of this, the present invention provides an assembly optimization method based on big data and intelligent manufacturing, which includes:

[0005] The management terminal sends automatic assembly instructions to the corresponding assembly robot;

[0006] The assembly robot assembles the corresponding components to be assembled in response to the received automatic assembly instructions and sends the multi-dimensional sensor data collected in real time during the assembly process to the intelligent manufacturing cloud platform, wherein the multi-dimensional sensor data includes the assembly action measurement data of the assembly robot, the three-dimensional image of the components to be assembled of the components to be assembled, and the three-dimensional image of the assembled components of the assembled components;

[0007] The feature extraction module extracts feature points from the three-dimensional image of the component to be assembled and the three-dimensional image of the assembled component to obtain a number of assembly feature points of the target surface to be assembled in the component to be assembled and a number of combination feature points of the target combination surface in the assembled component;

[0008] The vector acquisition unit of the feature analysis module performs depth analysis on each assembly feature point according to the grayscale features of all assembly feature points to obtain depth information of each assembly feature point, and obtains a first distance feature vector and a first angle feature vector between each assembly feature point according to the depth information, grayscale features and position information of all assembly feature points;

[0009] The shape analysis unit of the feature analysis module performs feature analysis according to the first distance feature vector and the first angle feature vector between each assembly feature point and the second distance feature vector and the second angle feature vector between each combination feature point to obtain the first geometric shape feature of the target to-be-assembled surface in the three-dimensional image of the to-be-assembled component and the second geometric shape feature of the target combination surface in the three-dimensional image of the assembled component;

[0010] The angle analysis unit of the feature analysis module performs angle analysis on the target to-be-assembled surface and the target combined surface according to the first geometric shape feature and the second geometric shape feature to obtain a first assembly angle feature and a second assembly angle feature;

[0011] The point association module obtains the orientation similarity between each reference point to be assembled and each combined reference point according to the first orientation feature of each reference point to be assembled on the target surface to be assembled and the assembly center key point and the second orientation feature between each combined reference point on the target combined surface and the combined center key point, and establishes the spatial point mapping relationship between each reference point to be assembled and each combined reference point according to the orientation similarity;

[0012] The motion decomposition module performs real-time analysis on the assembly motion of the assembly robot according to the assembly motion measurement data and the spatial point mapping relationship to obtain the assembly motion direction and assembly motion distance of the end effector of the assembly robot at the corresponding spatial motion point;

[0013] The error analysis module performs error analysis according to the assembly movement direction, the assembly movement distance, the second assembly angle feature and the first assembly angle feature to obtain an assembly direction deviation error between the target assembly surface and the target combination surface;

[0014] The orientation correction module corrects the motion trajectory and assembly angle of the assembly robot in real time according to the assembly direction deviation error, the first assembly angle feature and the second assembly angle feature to obtain an orientation correction parameter at the corresponding spatial motion point, and sends the orientation correction parameter to the corresponding assembly robot;

[0015] The assembly robot performs real-time motion corrections on the motion trajectory and assembly angle during assembly according to the received orientation correction parameters until the assembly work on the target assembly surface of the assembly component is completed.

[0016] According to a preferred embodiment, the multi-dimensional sensor data is a multi-dimensional fusion of all sensor data collected by multiple internal state sensors and multiple external state sensors of the assembly robot.

[0017] According to a preferred embodiment, the vector acquisition unit of the feature analysis module obtains the first distance feature vector and the first angle feature vector between each assembly feature point according to the depth information, grayscale features and position information of all assembly feature points, and further includes:

[0018] The vector acquisition unit performs depth analysis on each combined feature point according to the grayscale features of all combined feature points to obtain depth information of each combined feature point;

[0019] The vector acquisition unit obtains a second distance feature vector and a second angle feature vector between each combined feature point according to the depth information, grayscale features and position information of all combined feature points.

[0020] According to a preferred embodiment, the point association module obtains the orientation similarity between each reference point to be assembled and each combined reference point according to the first orientation feature between each reference point to be assembled on the target surface to be assembled and the assembly center key point and the second orientation feature between each combined reference point on the target combined surface and the combined center key point, including:

[0021] The point association module obtains the relative distance and relative angle between each reference point to be assembled and the key point of the assembly center according to the position information of each reference point to be assembled and the position information of the key point of the assembly center, and obtains the relative distance and relative angle between each combined reference point and the key point of the combined center according to the position information of each combined reference point and the position information of the key point of the assembly center;

[0022] The point association module obtains the first orientation feature between the corresponding reference point to be assembled and the assembly center key point according to the relative distance and relative angle between each reference point to be assembled and the assembly center key point, and obtains the second orientation feature between the corresponding combined reference point and the combined center key point according to the relative distance and relative angle between each combined reference point and the combined center key point.

[0023] According to a preferred embodiment, the first orientation feature is the orientation feature between the corresponding assembly reference point and the assembly center key point in the target assembly surface; the second orientation feature is the orientation feature between the corresponding combination reference point and the combination center key point in the target combination surface.

[0024] According to a preferred embodiment, the point association module obtains the orientation similarity between each reference point to be assembled and each combined reference point according to the first orientation feature between each reference point to be assembled on the target surface to be assembled and the assembly center key point and the second orientation feature between each combined reference point on the target combined surface and the combined center key point, and further includes:

[0025] The point association module performs centroid recognition on the target assembly surface and the target combination surface according to the first geometric shape feature and the second geometric shape feature to obtain the assembly center key point of the target assembly surface and the combination center key point of the target combination surface.

[0026] The first geometric shape feature is a geometric shape feature of a target surface to be assembled; the second geometric shape feature is a geometric shape feature of a target assembly surface.

[0027] The point association module obtains the orientation similarity between each reference point to be assembled and each combined reference point according to the first orientation feature of each reference point to be assembled on the target surface to be assembled and the assembly center key point and the second orientation feature between each combined reference point on the target combined surface and the combined center key point, and further includes:

[0028] The point association module performs key point detection on the target surface to be assembled according to the direction feature and grayscale feature of each assembly feature point to obtain multiple assembly reference points of the target surface to be assembled;

[0029] The point association module performs key point detection on the target composite surface according to the directional features and grayscale features of each composite feature point to obtain multiple composite reference points of the target composite surface.

[0030] According to a preferred embodiment, the error analysis module performs error analysis based on the assembly movement direction, the assembly movement distance, the second assembly angle feature, and the first assembly angle feature to obtain the assembly direction deviation error between the target assembly surface and the target assembly surface, including:

[0031] The error analysis module performs a difference analysis based on the first assembly angle feature of the component to be assembled and the second assembly angle feature of the assembled component to obtain an angle difference between the target surface to be assembled and the target assembly surface;

[0032] The error analysis module performs error analysis based on the assembly movement direction, assembly movement distance and angle difference to obtain the assembly direction deviation error between the target assembly surface and the target combination surface.

[0033] The present invention has the following beneficial effects:

[0034] The present invention can assist an assembly robot to complete corresponding assembly tasks more accurately and with higher quality, by performing error analysis on the assembly movement direction, assembly movement distance and assembly angle characteristics between the target assembly surface of the assembly component and the target combined surface of the assembled component during the assembly process of the assembly robot, so as to obtain the assembly direction deviation error of the end effector of the assembly robot at the corresponding spatial movement point, so as to make real-time corrections to the assembly action and assembly direction of the assembly robot, thereby improving the assembly accuracy of the assembly robot and reducing the assembly error caused by the deviation of the assembly angle and movement direction of the assembly robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a flow chart of the assembly optimization method based on big data and intelligent manufacturing of the present invention;

[0036] Figure 2 It is a structural block diagram of the assembly optimization system for running the present invention. DETAILED DESCRIPTION

[0037] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0038] The terms used in the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The singular forms "a", "the" and "the" used in the present invention and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0039] See also Figure 1 In one embodiment, the assembly optimization method based on big data and intelligent manufacturing may include:

[0040] S1. The management terminal sends an automatic assembly instruction to the corresponding assembly robot.

[0041] Optionally, the assembly robot includes a manipulator, a controller and a robot sensor; the manipulator includes an arm, a rod, a joint, an end effector connection device and an end effector; the controller includes a teaching box, which is a handheld unit connected to the control system to move the robot; the robot sensor includes an internal state sensor and an external state sensor.

[0042] Optionally, the internal state sensor is a robot sensor for measuring the internal state of the robot, which includes a code disk, a potentiometer, a tachometer generator and a gyroscope;

[0043] The external state sensor is a robot sensor used to measure the state of the robot's environment and the state of its interaction with the environment, and includes a global positioning system, a visual sensor, a distance sensor, a force sensor, a tactile sensor and an acoustic sensor.

[0044] S2. The assembly robot assembles the corresponding parts to be assembled in response to the received automatic assembly instructions and sends the multi-dimensional sensor data collected in real time during the assembly process to the intelligent manufacturing cloud platform, wherein the multi-dimensional sensor data includes the assembly action measurement data of the assembly robot, the three-dimensional image of the parts to be assembled and the three-dimensional image of the assembled parts.

[0045] Optionally, the component to be assembled is an original component waiting to be assembled with one or more assembly components adapted to it; and the assembly component is a component waiting to be assembled with the component to be assembled with it.

[0046] Optionally, the multi-dimensional sensing data is a multi-dimensional fusion of all sensing data collected by multiple internal state sensors and multiple external state sensors of the assembly robot.

[0047] S3. The feature extraction module extracts feature points from the three-dimensional image of the component to be assembled and the three-dimensional image of the assembled component respectively to obtain a number of assembly feature points of the target surface to be assembled in the component to be assembled and a number of combination feature points of the target combination surface in the assembled component.

[0048] Optionally, the target surface to be assembled is an assembly joint surface of a component to be assembled; and the target combined surface is an assembly joint surface of an assembled component.

[0049] S4. The vector acquisition unit of the feature analysis module performs depth analysis on each assembly feature point according to the grayscale features of all assembly feature points to obtain the depth information of each assembly feature point, and obtains the first distance feature vector and the first angle feature vector between each assembly feature point according to the depth information, grayscale features and position information of all assembly feature points.

[0050] Optionally, the vector acquisition unit obtains a first distance feature vector and a first angle feature vector between each assembly feature point based on the depth information, grayscale features and position information of all assembly feature points, wherein the position information is the coordinate value of the assembly feature point in the three-dimensional image of the corresponding assembly component, and the grayscale feature is the grayscale value of the assembly feature point in the three-dimensional image of the corresponding assembly component.

[0051] Specifically, the vector acquisition unit of the feature analysis module obtains the first distance feature vector and the first angle feature vector between each assembly feature point according to the depth information, grayscale features and position information of all assembly feature points, and further includes:

[0052] The vector acquisition unit performs depth analysis on each combined feature point according to the grayscale features of all combined feature points to obtain depth information of each combined feature point;

[0053] The vector acquisition unit obtains a second distance feature vector and a second angle feature vector between each combined feature point according to the depth information, grayscale features and position information of all combined feature points.

[0054] Optionally, the first distance feature vector is a vector obtained by transforming relative distance features between different assembly feature points, and the first angle feature vector is a vector obtained by transforming relative angle features between different assembly feature points;

[0055] The second distance feature vector is a vector obtained by transforming the relative distance features between different combination feature points, and the second angle feature vector is a vector obtained by transforming the relative angle features between different combination feature points.

[0056] Optionally, the vector acquisition unit obtains a second distance feature vector and a second angle feature vector between each combined feature point based on the depth information, grayscale features and position information of all combined feature points, wherein the position information is the coordinate value of the combined feature point in the three-dimensional image of the corresponding assembly component, and the grayscale feature is the grayscale value of the combined feature point in the three-dimensional image of the corresponding assembly component.

[0057] S5. The shape analysis unit of the feature analysis module performs feature analysis based on the first distance feature vector and the first angle feature vector between each assembly feature point and the second distance feature vector and the second angle feature vector between each combination feature point to obtain the first geometric shape feature of the target surface to be assembled in the three-dimensional image of the component to be assembled and the second geometric shape feature of the target combination surface in the three-dimensional image of the assembled component.

[0058] Optionally, the first geometric shape feature is a geometric shape feature of a target surface to be assembled; and the second geometric shape feature is a geometric shape feature of a target combined surface.

[0059] S6. The angle analysis unit of the feature analysis module performs angle analysis on the target to-be-assembled surface and the target combined surface according to the first geometric shape feature and the second geometric shape feature to obtain a first assembly angle feature and a second assembly angle feature.

[0060] Optionally, the first assembly angle feature is an angle feature of a target surface to be assembled; and the second assembly angle feature is an angle feature of a target combined surface.

[0061] S7, the point association module obtains the orientation similarity between each reference point to be assembled and each combined reference point according to the first orientation feature of each reference point to be assembled on the target surface to be assembled and the assembly center key point and the second orientation feature between each combined reference point on the target combined surface and the combined center key point, and establishes a spatial point mapping relationship between each reference point to be assembled and each combined reference point according to the orientation similarity.

[0062] Optionally, the point association module obtains the orientation similarity between each to-be-assembled reference point and each combined reference point according to the first orientation feature between each to-be-assembled reference point of the target to-be-assembled surface and the assembly center key point and the second orientation feature between each combined reference point of the target combined surface and the combined center key point, including:

[0063] The point association module obtains the relative distance and relative angle between each reference point to be assembled and the key point of the assembly center according to the position information of each reference point to be assembled and the position information of the key point of the assembly center, and obtains the relative distance and relative angle between each combined reference point and the key point of the combined center according to the position information of each combined reference point and the position information of the key point of the assembly center;

[0064] The point association module obtains the first orientation feature between the corresponding reference point to be assembled and the assembly center key point according to the relative distance and relative angle between each reference point to be assembled and the assembly center key point, and obtains the second orientation feature between the corresponding combined reference point and the combined center key point according to the relative distance and relative angle between each combined reference point and the combined center key point.

[0065] Optionally, the first orientation feature is an orientation feature between a corresponding assembly reference point and an assembly center key point in the target assembly surface; the second orientation feature is an orientation feature between a corresponding combination reference point and a combination center key point in the target combination surface.

[0066] Specifically, the point association module obtains the orientation similarity between each to-be-assembled reference point and each combined reference point according to the first orientation feature of each to-be-assembled reference point and the assembly center key point of the target to-be-assembled surface and the second orientation feature between each combined reference point and the combined center key point of the target combined surface, and further includes:

[0067] The point association module performs centroid recognition on the target assembly surface and the target combination surface according to the first geometric shape feature and the second geometric shape feature to obtain the assembly center key point of the target assembly surface and the combination center key point of the target combination surface.

[0068] Specifically, the point association module obtains the orientation similarity between each to-be-assembled reference point and each combined reference point according to the first orientation feature of each to-be-assembled reference point and the assembly center key point of the target to-be-assembled surface and the second orientation feature between each combined reference point and the combined center key point of the target combined surface, and further includes:

[0069] The point association module performs key point detection on the target surface to be assembled according to the direction feature and grayscale feature of each assembly feature point to obtain multiple assembly reference points of the target surface to be assembled;

[0070] The point association module performs key point detection on the target composite surface according to the directional features and grayscale features of each composite feature point to obtain multiple composite reference points of the target composite surface.

[0071] Optionally, the reference points to be assembled include corner points and local key points of the target surface to be assembled; and the combined reference points include corner points and local key points of the target combined surface.

[0072] S8. The motion decomposition module performs real-time analysis on the assembly motion of the assembly robot according to the assembly motion measurement data and the spatial point mapping relationship to obtain the assembly motion direction and assembly motion distance of the end effector of the assembly robot at the corresponding spatial motion point.

[0073] S9. The error analysis module performs error analysis according to the assembly movement direction, the assembly movement distance, the second assembly angle feature and the first assembly angle feature to obtain the assembly direction deviation error between the target assembly surface and the target combination surface.

[0074] Specifically, the error analysis module performs error analysis according to the assembly movement direction, the assembly movement distance, the second assembly angle feature, and the first assembly angle feature to obtain the assembly direction deviation error between the target assembly surface and the target assembly surface, including:

[0075] The error analysis module performs a difference analysis based on the first assembly angle feature of the component to be assembled and the second assembly angle feature of the assembled component to obtain an angle difference between the target surface to be assembled and the target assembly surface;

[0076] The error analysis module performs error analysis based on the assembly movement direction, assembly movement distance and angle difference to obtain the assembly direction deviation error between the target assembly surface and the target combination surface.

[0077] S10, the orientation correction module corrects the motion trajectory and assembly angle of the assembly robot in real time according to the assembly direction deviation error, the first assembly angle feature and the second assembly angle feature to obtain the orientation correction parameters at the corresponding spatial motion point, and sends the orientation correction parameters to the corresponding assembly robot.

[0078] S11. The assembly robot performs real-time motion correction on the motion trajectory and assembly angle during assembly according to the received orientation correction parameters until the assembly work on the target assembly surface of the assembly component is completed.

[0079] The present invention can assist an assembly robot to complete corresponding assembly tasks more accurately and with higher quality, by performing error analysis on the assembly movement direction, assembly movement distance and assembly angle characteristics between the target assembly surface of the assembly component and the target combined surface of the assembled component during the assembly process of the assembly robot, so as to obtain the assembly direction deviation error of the end effector of the assembly robot at the corresponding spatial movement point, so as to make real-time corrections to the assembly action and assembly direction of the assembly robot, thereby improving the assembly accuracy of the assembly robot and reducing the assembly error caused by the deviation of the assembly angle and movement direction of the assembly robot.

[0080] like Figure 2 As shown, the assembly optimization system in the present invention includes a management terminal, an assembly robot and an intelligent manufacturing cloud platform. The intelligent manufacturing cloud platform is connected to the management terminal and the assembly robot respectively. The management terminal is a device with computing, storage and communication functions used by management personnel, including: a smart phone, a desktop computer and a laptop computer.

[0081] The intelligent manufacturing cloud platform includes feature extraction module, feature analysis module, point association module, action decomposition module, error analysis module and orientation correction module.

[0082] The feature extraction module is used to extract feature points from the three-dimensional image of the to-be-assembled component and the three-dimensional image of the assembled component to obtain several assembly feature points of the target to-be-assembled surface in the to-be-assembled component and several combination feature points of the target combination surface in the assembled component.

[0083] The feature analysis module includes a vector acquisition unit, a shape analysis unit and an angle analysis unit.

[0084] The vector acquisition unit is used to perform depth analysis on each assembly feature point according to the grayscale features of all assembly feature points to obtain depth information of each assembly feature point, and to obtain a first distance feature vector and a first angle feature vector between each assembly feature point according to the depth information, grayscale features and position information of all assembly feature points.

[0085] The shape analysis unit is used to perform feature analysis based on the first distance feature vector and the first angle feature vector between each assembly feature point and the second distance feature vector and the second angle feature vector between each combination feature point to obtain the first geometric shape feature of the target to-be-assembled surface in the three-dimensional image of the to-be-assembled component and the second geometric shape feature of the target combination surface in the three-dimensional image of the assembled component.

[0086] The angle analysis unit is used to perform angle analysis on the target to-be-assembled surface and the target combined surface according to the first geometric shape feature and the second geometric shape feature to obtain a first assembly angle feature and a second assembly angle feature.

[0087] The point association module is used to obtain the orientation similarity between each reference point to be assembled and each combined reference point based on the first orientation characteristics of each reference point to be assembled on the target surface to be assembled and the assembly center key point and the second orientation characteristics between each combined reference point on the target combined surface and the combined center key point, and to establish a spatial point mapping relationship between each reference point to be assembled and each combined reference point based on the orientation similarity.

[0088] The motion decomposition module is used to perform real-time analysis on the assembly motion of the assembly robot according to the assembly motion measurement data and the spatial point mapping relationship to obtain the assembly motion direction and assembly motion distance of the end effector of the assembly robot at the corresponding spatial motion point.

[0089] The error analysis module is used to perform error analysis on the assembly movement direction, assembly movement distance, second assembly angle feature and first assembly angle feature to obtain the assembly direction deviation error between the target assembly surface and the target combination surface.

[0090] The orientation correction module is used to assemble the direction deviation error, the first assembly angle feature and the second assembly angle feature to perform real-time correction on the motion trajectory and assembly angle of the assembly robot to obtain the orientation correction parameters at the corresponding spatial motion point, and send the orientation correction parameters to the corresponding assembly robot.

[0091] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An assembly optimization method based on big data and intelligent manufacturing, characterized in that: include: The management terminal sends automatic assembly instructions to the corresponding assembly robot; The assembly robot assembles the corresponding components to be assembled in response to the received automatic assembly instructions and sends the multi-dimensional sensor data collected in real time during the assembly process to the intelligent manufacturing cloud platform, wherein the multi-dimensional sensor data includes the assembly action measurement data of the assembly robot, the three-dimensional image of the components to be assembled of the components to be assembled, and the three-dimensional image of the assembled components of the assembled components; The feature extraction module extracts feature points from the three-dimensional image of the component to be assembled and the three-dimensional image of the assembled component to obtain a number of assembly feature points of the target surface to be assembled in the component to be assembled and a number of combination feature points of the target combination surface in the assembled component; The vector acquisition unit of the feature analysis module performs depth analysis on each assembly feature point according to the grayscale features of all assembly feature points to obtain depth information of each assembly feature point, and obtains a first distance feature vector and a first angle feature vector between each assembly feature point according to the depth information, grayscale features and position information of all assembly feature points; The shape analysis unit of the feature analysis module performs feature analysis according to the first distance feature vector and the first angle feature vector between each assembly feature point and the second distance feature vector and the second angle feature vector between each combination feature point to obtain the first geometric shape feature of the target to-be-assembled surface in the three-dimensional image of the to-be-assembled component and the second geometric shape feature of the target combination surface in the three-dimensional image of the assembled component; The angle analysis unit of the feature analysis module performs angle analysis on the target to-be-assembled surface and the target combined surface according to the first geometric shape feature and the second geometric shape feature to obtain a first assembly angle feature and a second assembly angle feature; The point association module obtains the orientation similarity between each reference point to be assembled and each combined reference point according to the first orientation feature of each reference point to be assembled on the target surface to be assembled and the assembly center key point and the second orientation feature between each combined reference point on the target combined surface and the combined center key point, and establishes the spatial point mapping relationship between each reference point to be assembled and each combined reference point according to the orientation similarity; The motion decomposition module performs real-time analysis on the assembly motion of the assembly robot according to the assembly motion measurement data and the spatial point mapping relationship to obtain the assembly motion direction and assembly motion distance of the end effector of the assembly robot at the corresponding spatial motion point; The error analysis module performs error analysis according to the assembly movement direction, the assembly movement distance, the second assembly angle feature and the first assembly angle feature to obtain an assembly direction deviation error between the target assembly surface and the target combination surface; The orientation correction module corrects the motion trajectory and assembly angle of the assembly robot in real time according to the assembly direction deviation error, the first assembly angle feature and the second assembly angle feature to obtain an orientation correction parameter at the corresponding spatial motion point, and sends the orientation correction parameter to the corresponding assembly robot; The assembly robot performs real-time motion corrections on the motion trajectory and assembly angle during assembly according to the received orientation correction parameters until the assembly work on the target assembly surface of the assembly component is completed.

2. The method according to claim 1, characterized in that The multi-dimensional sensing data is a multi-dimensional fusion of all sensing data collected by multiple internal state sensors and multiple external state sensors of the assembly robot.

3. The method according to claim 2, characterized in that The vector acquisition unit of the feature analysis module obtains the first distance feature vector and the first angle feature vector between each assembly feature point according to the depth information, grayscale feature and position information of all assembly feature points, and further includes: The vector acquisition unit performs depth analysis on each combined feature point according to the grayscale features of all combined feature points to obtain depth information of each combined feature point; The vector acquisition unit obtains a second distance feature vector and a second angle feature vector between each combined feature point according to the depth information, grayscale features and position information of all combined feature points.

4. The method according to claim 3, characterized in that: The point association module obtains the orientation similarity between each to-be-assembled reference point and each combined reference point according to the first orientation feature of each to-be-assembled reference point and the assembly center key point of the target to-be-assembled surface and the second orientation feature between each combined reference point and the combined center key point of the target combined surface, including: The point association module obtains the relative distance and relative angle between each reference point to be assembled and the key point of the assembly center according to the position information of each reference point to be assembled and the position information of the key point of the assembly center, and obtains the relative distance and relative angle between each combined reference point and the key point of the combined center according to the position information of each combined reference point and the position information of the key point of the assembly center; The point association module obtains the first orientation feature between the corresponding reference point to be assembled and the assembly center key point according to the relative distance and relative angle between each reference point to be assembled and the assembly center key point, and obtains the second orientation feature between the corresponding combined reference point and the combined center key point according to the relative distance and relative angle between each combined reference point and the combined center key point.

5. The method according to claim 4, characterized in that The first orientation feature is the orientation feature between the corresponding assembly reference point and the assembly center key point in the target assembly surface; the second orientation feature is the orientation feature between the corresponding combination reference point and the combination center key point in the target combination surface.

6. The method according to claim 5, characterized in that The point association module obtains the orientation similarity between each reference point to be assembled and each combined reference point according to the first orientation feature of each reference point to be assembled on the target surface to be assembled and the assembly center key point and the second orientation feature between each combined reference point on the target combined surface and the combined center key point, and further includes: The point association module performs centroid recognition on the target assembly surface and the target combination surface according to the first geometric shape feature and the second geometric shape feature to obtain the assembly center key point of the target assembly surface and the combination center key point of the target combination surface.

7. The method according to claim 6, characterized in that The first geometric shape feature is a geometric shape feature of a target surface to be assembled; the second geometric shape feature is a geometric shape feature of a target assembly surface.

8. The method according to claim 7, characterized in that The point association module obtains the orientation similarity between each reference point to be assembled and each combined reference point according to the first orientation feature of each reference point to be assembled on the target surface to be assembled and the assembly center key point and the second orientation feature between each combined reference point on the target combined surface and the combined center key point, and further includes: The point association module performs key point detection on the target surface to be assembled according to the direction feature and grayscale feature of each assembly feature point to obtain multiple assembly reference points of the target surface to be assembled; The point association module performs key point detection on the target composite surface according to the directional features and grayscale features of each composite feature point to obtain multiple composite reference points of the target composite surface.

9. The method according to claim 8, characterized in that The reference points to be assembled include corner points and local key points of the target surface to be assembled; the combined reference points include corner points and local key points of the target combined surface.

10. The method according to claim 9, characterized in that The error analysis module performs error analysis according to the assembly movement direction, the assembly movement distance, the second assembly angle feature and the first assembly angle feature to obtain the assembly direction deviation error between the target assembly surface and the target assembly surface, including: The error analysis module performs a difference analysis based on the first assembly angle feature of the component to be assembled and the second assembly angle feature of the assembled component to obtain an angle difference between the target surface to be assembled and the target assembly surface; The error analysis module performs error analysis based on the assembly movement direction, assembly movement distance and angle difference to obtain the assembly direction deviation error between the target assembly surface and the target combination surface.

Images