A method, device, equipment and medium for adjusting posture of an assembly robot

Abstract

The present invention relates to the field of robot technology, and in particular to a method, device, equipment and medium for adjusting the posture of an assembly robot. Before assembling a target accessory, the present invention can establish a contact stress estimation model to obtain multiple pose matrices corresponding to multiple preset contact points between the target accessory and the assembly robot, and input the shape and position errors corresponding to each pose matrix into the contact stress estimation model respectively to obtain a contact stress distribution map corresponding to each pose matrix, and then select the pose matrix corresponding to the contact stress distribution map with the minimum average contact stress and less than the preset average contact stress as the target pose matrix, and adjust the posture of the assembly robot according to the target pose matrix. The present invention predetermines the posture of the assembly robot to clamp the accessory before assembly, which can reduce damage caused by the clamping posture.

Description

Technical Field

[0001] The present invention relates to the field of robot technology, and in particular to a posture adjustment method, device, equipment and medium for assembling a robot. Background Art

[0002] During the assembly process of large helicopter windshields and cabin doors, which include special-shaped curved surfaces, the different postures of the assembly robot's mechanical arm to clamp the windshield will also affect the contact stress of the windshield installed on the helicopter frame. In traditional assembly, workers operate the mechanical arm to clamp the windshield and align it to the helicopter frame. The workers adjust the different postures of the mechanical arm to clamp the windshield based on their personal experience, which makes the windshield more likely to be damaged due to contact stress during installation.

[0003] Based on this, the present invention proposes a posture adjustment method and device for an assembly robot to solve the above technical problems. Summary of the invention

[0004] The present invention describes a posture adjustment method, device, equipment and medium for an assembly robot, which can predetermine the posture of the assembly robot for clamping accessories before assembly, thereby reducing damage caused by the clamping posture.

[0005] According to a first aspect, the present invention provides a posture adjustment method for an assembly robot, comprising:

[0006] A contact stress estimation model is established based on a mapping equation between the shape and position error of a defect position on a target accessory and the corresponding contact stress when the target accessory is assembled; the defect position is a point on the target accessory where contact stress exists when the target accessory is assembled with a target skeleton, and the target accessory includes a special-shaped curved surface accessory in a helicopter;

[0007] Acquire multiple pose matrices corresponding to multiple preset contact points between the target accessory and the assembly robot, and calculate the shape and position errors corresponding to each pose matrix; the pose matrix is ​​used to characterize the posture of the assembly robot gripping the target accessory;

[0008] Inputting the shape and position errors corresponding to each posture matrix into the contact stress estimation model respectively, and obtaining the contact stress distribution diagram corresponding to each posture matrix output by the contact stress estimation model;

[0009] From the contact stress distribution maps corresponding to the multiple posture matrices, select the posture matrix corresponding to the contact stress distribution map with the smallest average contact stress and less than the preset average contact stress as the target posture matrix, and adjust the posture of the assembly robot according to the target posture matrix.

[0010] According to a second aspect, the present invention provides a posture adjustment device for an assembly robot, comprising:

[0011] A modeling unit, for establishing a contact stress estimation model based on a mapping equation between a shape and position error of a defect position on a target accessory and a corresponding contact stress when the target accessory is assembled; the defect position is a point on the target accessory where contact stress exists when the target accessory is assembled with a target skeleton, and the target accessory includes a special-shaped curved surface accessory in a helicopter;

[0012] A calculation unit, used to obtain a plurality of posture matrices corresponding to a plurality of preset contact points between the target accessory and the assembly robot, and calculate the shape and position errors corresponding to each posture matrix; the posture matrix is ​​used to characterize the posture of the assembly robot gripping the target accessory;

[0013] An output unit, used to input the shape and position errors corresponding to each posture matrix into the contact stress estimation model respectively, to obtain a contact stress distribution diagram corresponding to each posture matrix output by the contact stress estimation model;

[0014] An adjustment unit is used to select, from the contact stress distribution maps corresponding to the multiple posture matrices, a posture matrix corresponding to a contact stress distribution map in which the average contact stress is the smallest and is less than a preset average contact stress as a target posture matrix, and adjust the posture of the assembly robot according to the target posture matrix.

[0015] According to a third aspect, the present invention provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method of the first aspect is implemented.

[0016] According to a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to execute the method of the first aspect.

[0017] According to the posture adjustment method, device, equipment and medium of the assembly robot provided by the present invention, before assembling the target accessory, the present invention can establish a contact stress estimation model to obtain multiple posture matrices corresponding to multiple preset contact points between the target accessory and the assembly robot, calculate the shape and position errors corresponding to each posture matrix, and input the shape and position errors corresponding to each posture matrix into the contact stress estimation model respectively to obtain a contact stress distribution map corresponding to each posture matrix output by the contact stress estimation model, and then select the posture matrix corresponding to the contact stress distribution map with the minimum average contact stress and less than the preset average contact stress as the target posture matrix, adjust the posture of the assembly robot according to the target posture matrix, and predetermine the posture of the assembly robot to clamp the accessory before assembly, so as to reduce damage caused by the clamping posture. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 A schematic flow chart of a method for adjusting the posture of an assembly robot according to an embodiment is shown;

[0020] Figure 2 A schematic block diagram of a posture adjustment device for an assembly robot according to an embodiment is shown. DETAILED DESCRIPTION

[0021] The solution provided by the present invention is described below in conjunction with the accompanying drawings.

[0022] Figure 1 A flow chart of a method for adjusting the posture of an assembly robot according to an embodiment is shown. It can be understood that the method can be executed by any device, equipment, platform, or equipment cluster with computing and processing capabilities. Figure 1 As shown, the method includes:

[0023] Step 101: establishing a contact stress estimation model based on a mapping equation between a shape and position error of a defect position on a target component and a corresponding contact stress when the target component is assembled; wherein the defect position is a point on the target component where contact stress exists when the target component is assembled with a target skeleton, and the target component includes a special-shaped curved surface component in a helicopter;

[0024] In this embodiment, the target accessory refers to an accessory including a special-shaped curved surface such as a windshield and a cabin door of a large helicopter before assembly. The target frame refers to a blank frame to which the target accessory is to be installed, such as a helicopter frame. In this embodiment, the assembly of the target accessory and the target frame is achieved by clamping the target accessory to the target frame with the mechanical arm of the assembly robot.

[0025] Step 102, obtaining a plurality of pose matrices corresponding to a plurality of preset contact points between the target component and the assembly robot, and calculating the shape and position errors corresponding to each pose matrix; wherein the pose matrix is ​​used to characterize the posture of the assembly robot gripping the target component;

[0026] In this embodiment, the pose matrix corresponding to the preset contact point includes the coordinates of a set number of positioning feature points, for example, the pose matrix corresponding to the preset contact point includes the coordinates of 6 positioning feature points in the contact point, and the 6 positioning feature points are respectively located at 1 center point of the coverage area formed by the contact point and 5 edge points of the coverage area. Optionally, the coordinates of the positioning feature points include six-dimensional feature data, and the six-dimensional feature data is related to the six-dimensional force sensor at the end of the mechanical arm of the assembly robot, for example, the six-dimensional data includes the force in the three-dimensional direction and the torque in the three-dimensional direction on the positioning feature point.

[0027] Step 103, inputting the shape and position errors corresponding to each posture matrix into the contact stress estimation model respectively, and obtaining the contact stress distribution diagram corresponding to each posture matrix output by the contact stress estimation model;

[0028] The above-mentioned contact stress distribution diagram is used to characterize the contact stress that the target accessory will bear when the target accessory and the target skeleton are assembled. The contact stress here is mainly caused by the contact stress caused by the mismatch of the shape or position of the assembly holes when the target accessory and the target skeleton are connected by screws during assembly, the contact stress generated when the edges of the two are fitted together due to abnormal protrusions, depressions, etc. on the target accessory or the target skeleton, and the contact stress generated when the edges of the two are fitted together due to deformation of the target accessory or the target skeleton.

[0029] Step 104: From the contact stress distribution maps corresponding to the multiple pose matrices, select the pose matrix corresponding to the contact stress distribution map with the smallest average contact stress and less than the preset average contact stress as the target pose matrix, and adjust the posture of the assembly robot according to the target pose matrix.

[0030] After determining the target posture matrix, professional assembly workers can adjust the posture of the robotic arm to be close to the target posture matrix by manipulating the robotic arm. The workers can adjust the posture of the robotic arm step by step based on the posture data fed back in real time by the six-dimensional force sensor at the end of the robotic arm.

[0031] In this embodiment, before assembling the target accessory, the present invention can establish a contact stress estimation model to obtain multiple pose matrices corresponding to multiple preset contact points between the target accessory and the assembly robot, and input the shape and position errors corresponding to each pose matrix into the contact stress estimation model respectively to obtain the contact stress distribution map corresponding to each pose matrix. Then, the pose matrix corresponding to the contact stress distribution map with the minimum average contact stress and less than the preset average contact stress can be selected as the target pose matrix, and the posture of the assembly robot can be adjusted according to the target pose matrix. The present invention predetermines the posture of the assembly robot to clamp the accessory before assembly, which can reduce damage caused by the clamping posture.

[0032] Described below Figure 1 How the various steps are performed.

[0033] For step 101:

[0034] In one embodiment of the present invention, the method for obtaining the mapping equation in step 101 includes:

[0035] Obtaining a first model corresponding to the target accessory and a second model corresponding to the target skeleton;

[0036] For the first model and the second model, determining a defective position where contact stress exists when the first model and the second model are fitted to simulate assembly of the target accessory and the target skeleton, and a predicted contact stress corresponding to the defective position;

[0037] Obtain the shape and position errors of each defect position, which include height error and shape error;

[0038] According to the shape and position errors of each defect position and the corresponding predicted contact stress, a mapping equation between shape and position errors and contact stress is established.

[0039] In this embodiment, the first model corresponding to the target accessory and the second model corresponding to the target skeleton are established based on the scanning results obtained by scanning the target accessory and the target skeleton based on a preset planned scanning path.

[0040] In this embodiment, the height error refers to the height difference generated when the first model and the second model are fitted. Ideally, the height difference after the first model and the second model are fitted should be 0. However, since the first model and the second model may have defects such as deformation or protrusions and depressions, there may be height differences at certain points after the first model and the second model are fitted. The height error is represented by the three-dimensional coordinates corresponding to the defective position. The shape error mainly refers to the errors of inconsistent shapes and mismatched positions that may exist in some of the assembly holes corresponding to the first model and the second model when the first model and the second model are fitted. The shape error is represented by the two-dimensional coordinates corresponding to the defective position.

[0041] As an embodiment, the predicted contact stress corresponding to each defect position can be calculated in the following manner: first, based on the known stiffness of the target accessory, a flexible body mechanics model that can characterize the relationship between the deformation and the force twist of the target accessory is established; then, when the target accessory and the target skeleton are fitted, contact stress is generated due to the height difference at the defect position. The function of the generated contact stress is to cause the target accessory to deform to adapt to the height difference. Therefore, for each defect position, a simulation test can be used to estimate the degree of deformation corresponding to each defect position on the target accessory when the target skeleton fits the defect position in different postures; then, based on the stiffness of the target accessory and the estimated degree of deformation of the defect position, the force twist (i.e., the predicted contact stress) when the defect position is deformed corresponding to the height difference is determined.

[0042] Furthermore, according to the shape and position errors of each defect position and the corresponding predicted contact stress, a mapping equation between the shape and position errors and the contact stress is established. It can be assumed that the mapping equation between the shape and position errors and the contact stress is: shape and position error*k=contact stress. The predicted contact stress and shape and position errors corresponding to each defect position have been obtained through the steps of the above embodiment, but the coefficient k is unknown. Therefore, to establish a mapping equation between the shape and position errors and the contact stress, it is necessary to calculate the value of k. For example, k can be obtained by fitting the shape and position errors and the contact stress.

[0043] Preferably, a mapping equation between the shape and position errors and the contact stress is established according to the shape and position errors of each defect position and the corresponding predicted contact stress, and it may also include: for the first model, a mapping equation between the shape and position errors and the contact stress is established according to the shape and position errors of each defect position and the corresponding predicted contact stress, combined with the stiffness of the target accessory, the rigid body coordinates of the simulated contact points between the assembly robot and the first model, and the flexible body coordinates of each assembly hole on the first model relative to the simulated contact points.

[0044] In this embodiment, the contact stress is associated with factors such as the rigidity of the target accessory, the rigid body coordinates of the simulated contact points, and the flexible body coordinates of each assembly hole relative to the simulated contact points. In addition, when calculating the predicted contact stress, the different postures of the target skeleton have a corresponding relationship with the rigid body coordinates of the simulated contact points. When the posture of the target skeleton is determined, the rigid body coordinates of the simulated contact points are also known. The flexible body coordinates of each assembly hole relative to the simulated contact points have a corresponding relationship with the rigid body coordinates. When the rigid body coordinates are known, the flexible body coordinates can be calculated based on the rigid body coordinates. Therefore, preferably, when calculating k in this embodiment, k can be set to include factors such as the rigid body coordinates, the flexible body coordinates, and the rigidity of the target accessory, and the predicted contact stress and rigid body coordinates corresponding to multiple postures of the target skeleton are comprehensively considered to calculate the final k.

[0045] The rigid body coordinates of the simulated contact point between the assembly robot and the first model refer to the coordinates of the simulated contact point between the assembly robot and the first model. The flexible body coordinates of each assembly hole on the first model relative to the simulated contact point refer to: when the target accessory is a weak rigid structure, the assembly robot may cause a slight deformation of the target accessory when gripping the target accessory, which will cause a slight coordinate change in each assembly hole. This coordinate change causes the position coordinates of each assembly hole to change relative to the change of the rigid body coordinates of the simulated contact point, so the position coordinates of the assembly hole after the change are flexible body coordinates.

[0046] As an embodiment, determining a defective position where contact stress exists when the first model and the second model are fitted to simulate the assembly of the target accessory and the target skeleton includes:

[0047] Identify and mark each defect point and assembly hole position in the first model and the second model; wherein the defect point is the position of abnormal protrusion and / or abnormal depression carried by the target accessory and the target frame when they leave the factory;

[0048] For each defect position and assembly hole position, the regional features, grayscale features and gradient features corresponding to the defect position and the assembly hole position are calculated as the defect features corresponding to the defect position and the assembly hole position;

[0049] Compare each defect point and assembly hole position in the first model and the second model, and obtain the shape and position errors corresponding to each defect point and assembly hole position based on the comparison result and the defect features corresponding to each defect point and assembly hole position;

[0050] Defective point positions and / or assembly hole positions whose shape and position errors meet the preset rules for generating contact stress are determined as defective positions.

[0051] In this embodiment, before identifying and marking each defect point and assembly hole position in the first model and the second model, the first model and the second model can be pre-processed for denoising (including removing noise from images within the contact surface range of the first model and the second model (i.e., the surface that fits together when the two are assembled)), and the first model and the second model can be segmented (the region segmentation is performed here because the target accessory is too large and is divided into small images for easy processing).

[0052] By identifying and marking each defective point and assembly hole in the first model and the second model, the position of the defective point and the height difference with the surrounding adjacent points can be determined for each segmented area according to the differences in regional characteristics, grayscale characteristics and gradient characteristics of the pixel points corresponding to the defective point and those of the surrounding pixels. The position of the assembly hole and the height difference with the surrounding adjacent points and the shape of the assembly hole can also be determined according to the differences in regional characteristics, grayscale characteristics and gradient characteristics of the pixel points corresponding to the defective point and those of the surrounding pixels.

[0053] As an embodiment, based on the assembly relationship between the first model and the second model, the assembly holes need to correspond when the two are assembled. Therefore, each defect point in the first model and the second model is compared with the assembly hole. The shapes and positions of the assembly holes with the same coordinates can be compared to determine the height errors and shape errors corresponding to the assembly holes. In addition, the defect points in one model can be compared with the corresponding points in another model to determine the height errors and shape errors corresponding to the defect points.

[0054] Preferably, in this embodiment, the comparison result obtained by comparing each defect point and assembly hole in the first model and the second model can be used as the first comparison result, the comparison result obtained by comparing each defect point and assembly hole in the first model with the first standard model can be used as the second comparison result, and the comparison result obtained by comparing each defect point and assembly hole in the second model with the second standard model can be used as the third comparison result. The first comparison result, the second comparison result and the third comparison result are combined to determine the shape and position errors corresponding to each defect point and assembly hole. Among them, the first standard model refers to the standard model corresponding to the ideal target accessory without any defects, and the second standard model refers to the standard model corresponding to the ideal target skeleton without any defects.

[0055] Finally, in this embodiment, the preset rule for generating contact stress can be that when the corresponding height error is greater than the preset height error, and / or the shape error is greater than the preset shape error, the defective points or assembly holes in the first model or the second model will generate contact stress.

[0056] Regarding step 102:

[0057] Obtain multiple pose matrices corresponding to multiple preset contact points between the target component and the assembly robot, including:

[0058] randomly generating a preset number of candidate pose matrices according to a preset genetic algorithm;

[0059] The fitness value corresponding to each candidate pose matrix is ​​calculated, and the candidate pose matrix with a fitness value greater than a preset fitness value is determined as the acquired pose matrix.

[0060] In this embodiment, the optimal posture of the assembly robot is obtained by iterative calculation of the genetic algorithm. In this embodiment, the posture matrix is ​​randomly generated by the genetic algorithm. For example, in each iteration of the posture matrix, the genetic algorithm generates 100 candidate posture matrices, and finally selects n candidate posture matrices whose fitness values ​​​​in the direction of decreasing average contact stress are greater than the preset fitness from the 100 candidate posture matrices to be determined as the acquired posture matrix.

[0061] Then, from the n pose matrices, the pose matrix corresponding to the contact stress distribution diagram with the minimum average contact stress and less than the preset average contact stress is selected as the target pose matrix.

[0062] As an embodiment, if the minimum average contact stress among the average contact stresses corresponding to the contact stress distribution diagrams corresponding to multiple posture matrices is not less than the preset average contact stress, then return to the step of obtaining multiple posture matrices corresponding to multiple preset contact points between the target accessory and the assembly robot.

[0063] For example, among n pose matrices, among the average contact stresses corresponding to the contact stress distribution diagrams corresponding to the n pose matrices, the minimum average contact stress is not less than the preset average contact stress. Then, the genetic algorithm is used to iteratively generate 100 candidate pose matrices, and then m pose matrices are determined from the 100 candidate pose matrices until the target pose matrix is ​​selected from the pose matrices.

[0064] Further, the shape and position errors corresponding to each pose matrix calculated in step 102 include:

[0065] For each pose matrix, the pose matrix is ​​input into the preset flexible body mechanics model to obtain the flexible coordinates of each assembly hole on the target accessory relative to the preset contact point, and the shape and position errors between the flexible coordinates and the actual coordinates of each assembly hole on the target accessory are calculated.

[0066] Among them, the flexible body mechanics model can be calculated based on the generalized Newton-Euler equation of the flexible rod. By substituting the system parameters such as the rigidity of the target accessory into the generalized Newton-Euler equation, the mapping relationship between the deformation of the target accessory due to the force spin and the rigid coordinates that generate the force spin is obtained.

[0067] The generalized Newton-Euler equation for a flexible rod can be expressed as:

[0068]

[0069] Among them, M rri is the inertia matrix of the rigid system, is the rigid-flexible coupling inertia matrix, M eei is the inertia matrix corresponding to the flexible mode, Bei is the nonlinear term corresponding to the modal coordinate velocity, k qi is the modal stiffness matrix. The meanings of other parameters can be found in the relevant materials of the generalized Newton-Euler equation, which will not be elaborated here.

[0070] Regarding step 103:

[0071] It should be noted that the contact stress distribution diagram corresponding to each pose matrix output by the contact stress estimation model is obtained by substituting the shape and position errors corresponding to the pose matrix into the first model and the second model, and comparing and matching the edges of the first model and the second model according to the installation direction of the target accessory to the target skeleton, so as to obtain the corresponding chromatogram when the first model and the second model are fitted, and then digitally simulate the fitting of the first model and the second model based on the chromatogram and the mapping equation, and predict the contact stress distribution diagram when the target accessory and the target skeleton are assembled.

[0072] The chromatogram is used to characterize the shape and position errors of each point on the edge of the first model and the second model.

[0073] In this embodiment, the edges of the first model and the second model are compared and matched by simulation to obtain the height error and shape error (here mainly the height difference) on the contact surface (i.e., edge) of the first model and the second model. Since the shape and position errors on the contact surface of the first model and the second model have formed a surface from points, this embodiment uses a chromatogram to characterize the shape and position errors of each point on the edge of the first model and the second model. For example, the darker the color of the chromatogram, the greater the shape and position error at that location.

[0074] Here, for the digital simulation of the fitting of the first model and the second model based on the chromatogram and the mapping equation, reference may be made to relevant digital simulation techniques, and the present application does not limit the simulation techniques.

[0075] The above describes specific embodiments of the present invention. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0076] According to another aspect of the embodiment, the present invention provides a posture adjustment device for an assembly robot. Figure 2 FIG. 1 is a schematic block diagram of a posture adjustment device for an assembly robot according to an embodiment. It is understood that the device can be implemented by any device, equipment, platform, or device cluster with computing and processing capabilities. Figure 2As shown, the device includes: a modeling unit 201, a calculation unit 202, an output unit 203 and an adjustment unit 204. The main functions of each component unit are as follows:

[0077] A modeling unit 201 is used to establish a contact stress estimation model based on a mapping equation between a shape and position error of a defect position on a target component and a corresponding contact stress when the target component is assembled; the defect position is a point on the target component where contact stress exists when the target component is assembled with the target skeleton;

[0078] The calculation unit 202 is used to obtain a plurality of posture matrices corresponding to a plurality of preset contact points between the target component and the assembly robot, and calculate the shape and position errors corresponding to each posture matrix; the posture matrix is ​​used to represent the posture of the assembly robot gripping the target component;

[0079] An output unit 203 is used to input the shape and position errors corresponding to each posture matrix into the contact stress estimation model respectively, so as to obtain a contact stress distribution diagram corresponding to each posture matrix output by the contact stress estimation model;

[0080] The adjustment unit 204 is used to select, from the contact stress distribution maps corresponding to the multiple posture matrices, a posture matrix corresponding to a contact stress distribution map with a minimum average contact stress and less than a preset average contact stress as a target posture matrix, and adjust the posture of the assembly robot according to the target posture matrix.

[0081] As a preferred implementation, the calculation unit 202 calculates the shape and position errors corresponding to each pose matrix, including:

[0082] For each pose matrix, the pose matrix is ​​input into a preset flexible body mechanics model to obtain the flexible coordinates of each assembly hole on the target accessory relative to the preset contact point, and the shape and position errors between the flexible coordinates and the actual coordinates of each assembly hole on the target accessory are calculated.

[0083] As a preferred implementation, the calculation unit 202 obtains a plurality of posture matrices corresponding to a plurality of preset contact points between the target component and the assembly robot, including:

[0084] randomly generating a preset number of candidate pose matrices according to a preset genetic algorithm;

[0085] The fitness value corresponding to each candidate pose matrix is ​​calculated, and the candidate pose matrix having a fitness value greater than a preset fitness value is determined as the acquired pose matrix.

[0086] As a preferred embodiment, the device is also used for:

[0087] If the adjustment unit 204 determines that the minimum average contact stress among the average contact stresses corresponding to the contact stress distribution diagrams corresponding to the multiple pose matrices is not less than the preset average contact stress, the calculation unit 202 is notified to obtain multiple pose matrices corresponding to multiple preset contact points between the target accessory and the assembly robot.

[0088] As a preferred implementation, the method for obtaining the mapping equation includes:

[0089] Acquire a first model corresponding to the target accessory and a second model corresponding to the target skeleton;

[0090] For the first model and the second model, determining a defective position where contact stress exists when the first model and the second model are fitted to simulate assembly of the target accessory and the target skeleton, and a predicted contact stress corresponding to the defective position;

[0091] Obtaining shape and position errors of each defect position, wherein the shape and position errors include height errors and shape errors;

[0092] According to the shape and position errors of the defect positions and the corresponding predicted contact stress, a mapping equation between the shape and position errors and the contact stress is established.

[0093] As a preferred embodiment, the determining of a defective position where contact stress exists when the first model and the second model are fitted to simulate assembly of the target accessory and the target skeleton includes:

[0094] Identify and mark each defect point and assembly hole position in the first model and the second model; the defect point is the position of abnormal protrusions and / or abnormal depressions carried by the target accessory and the target skeleton when they leave the factory;

[0095] For each defect position and assembly hole position, calculating the regional features, grayscale features and gradient features respectively corresponding to the defect position and the assembly hole position as the defect features respectively corresponding to the defect position and the assembly hole position;

[0096] Comparing each defect point and assembly hole position in the first model and the second model, and obtaining the shape and position errors corresponding to each defect point and assembly hole position based on the comparison result and the defect features corresponding to each defect point and assembly hole position;

[0097] Defective point positions and / or assembly hole positions whose shape and position errors meet the preset rules for generating contact stress are determined as defective positions.

[0098] According to another embodiment, there is also provided a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to execute a combination of Figure 1 The method described.

[0099] According to another embodiment of the present invention, an electronic device is provided, including a memory and a processor, wherein an executable code is stored in the memory, and when the processor executes the executable code, the Figure 1 method.

[0100] The various embodiments of the present invention are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0101] Those skilled in the art should be aware that in one or more of the above examples, the functions described in the present invention can be implemented by hardware, software, firmware or any combination thereof. When implemented by software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium.

[0102] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.

Claims

1. A posture adjustment method for an assembly robot, characterized in that: include: A contact stress estimation model is established based on a mapping equation between the shape and position errors of the defective position on the target accessory and the corresponding contact stress when the target accessory is assembled; wherein the defective position is a point on the target accessory where contact stress exists when the target accessory is assembled with the target skeleton, and the target accessory includes a special-shaped curved surface accessory in a helicopter; Acquire multiple pose matrices corresponding to multiple preset contact points between the target accessory and the assembly robot, and calculate the shape and position errors corresponding to each pose matrix; wherein the pose matrix is ​​used to characterize the posture of the assembly robot gripping the target accessory; Inputting the shape and position errors corresponding to each posture matrix into the contact stress estimation model respectively, and obtaining the contact stress distribution diagram corresponding to each posture matrix output by the contact stress estimation model; From the contact stress distribution maps corresponding to the multiple posture matrices, select the posture matrix corresponding to the contact stress distribution map with the smallest average contact stress and less than the preset average contact stress as the target posture matrix, and adjust the posture of the assembly robot according to the target posture matrix.

2. The method according to claim 1, characterized in that Calculate the shape and position errors corresponding to each pose matrix, including: For each pose matrix, the pose matrix is ​​input into a preset flexible body mechanics model to obtain the flexible coordinates of each assembly hole on the target accessory relative to the preset contact point, and the shape and position errors between the flexible coordinates and the actual coordinates of each assembly hole on the target accessory are calculated.

3. The method according to claim 1, characterized in that The step of obtaining a plurality of posture matrices corresponding to a plurality of preset contact points between the target component and the assembly robot includes: Randomly generate a preset number of candidate pose matrices according to a preset genetic algorithm; The fitness value corresponding to each candidate pose matrix is ​​calculated, and the candidate pose matrix having a fitness value greater than a preset fitness value is determined as the acquired pose matrix.

4. The method according to claim 1 or 3, characterized in that: The method further comprises: If the minimum average contact stress among the average contact stresses corresponding to the contact stress distribution diagrams corresponding to the multiple pose matrices is not less than the preset average contact stress, return to the step of obtaining multiple pose matrices corresponding to multiple preset contact points between the target accessory and the assembly robot.

5. The method according to claim 1, characterized in that The method for obtaining the mapping equation includes: Acquire a first model corresponding to the target accessory and a second model corresponding to the target skeleton; For the first model and the second model, determining a defective position where contact stress exists when the first model and the second model are fitted to simulate assembly of the target accessory and the target skeleton, and a predicted contact stress corresponding to the defective position; Obtaining shape and position errors of each defect position, wherein the shape and position errors include height errors and shape errors; According to the shape and position errors of the defect positions and the corresponding predicted contact stress, a mapping equation between the shape and position errors and the contact stress is established.

6. The method according to claim 5, characterized in that The step of determining a defective position where contact stress exists when the first model and the second model are fitted to simulate assembly of the target accessory and the target skeleton comprises: Identify and mark each defect point and assembly hole position in the first model and the second model; the defect point is the position of abnormal protrusions and / or abnormal depressions carried by the target accessory and the target skeleton when they leave the factory; For each defect position and assembly hole position, calculating the regional features, grayscale features and gradient features respectively corresponding to the defect position and the assembly hole position as the defect features respectively corresponding to the defect position and the assembly hole position; Comparing each defect point and assembly hole position in the first model and the second model, and obtaining the shape and position errors corresponding to each defect point and assembly hole position based on the comparison result and the defect features corresponding to each defect point and assembly hole position; Defective point positions and / or assembly hole positions whose shape and position errors meet the preset rules for generating contact stress are determined as defective positions.

7. A posture adjustment device for an assembly robot, characterized in that: include: A modeling unit, for establishing a contact stress estimation model based on a mapping equation between a shape and position error of a defect position on a target accessory and a corresponding contact stress when the target accessory is assembled; wherein the defect position is a point on the target accessory where contact stress exists when the target accessory is assembled with a target skeleton, and the target accessory includes a special-shaped curved surface accessory in a helicopter; A calculation unit, used to obtain a plurality of posture matrices corresponding to a plurality of preset contact points between the target accessory and the assembly robot, and calculate the shape and position errors corresponding to each posture matrix; wherein the posture matrix is ​​used to characterize the posture of the assembly robot gripping the target accessory; An output unit, used to input the shape and position errors corresponding to each posture matrix into the contact stress estimation model respectively, to obtain a contact stress distribution diagram corresponding to each posture matrix output by the contact stress estimation model; An adjustment unit is used to select, from the contact stress distribution maps corresponding to the multiple posture matrices, a posture matrix corresponding to a contact stress distribution map in which the average contact stress is the smallest and is less than a preset average contact stress as a target posture matrix, and adjust the posture of the assembly robot according to the target posture matrix.

8. The device according to claim 7, characterized in that The calculation unit calculates the shape and position errors corresponding to each pose matrix, including: For each pose matrix, the pose matrix is ​​input into a preset flexible body mechanics model to obtain the flexible coordinates of each assembly hole on the target accessory relative to the preset contact point, and the shape and position errors between the flexible coordinates and the actual coordinates of each assembly hole on the target accessory are calculated.

9. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 6 is implemented.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed in a computer, the computer is caused to execute the method according to any one of claims 1 to 6.

Images