Micro gap assembly path planning method based on high-precision interference detection

Abstract

The invention relates to a micro-gap assembly path planning method based on high-precision interference detection. The method comprises the following steps: acquiring an actual surface shape of a matching surface of a workpiece to be assembled through three-dimensional scanning equipment; generating an initial assembly path according to the relative position of the point cloud and the theoretical assembly path, and initializing an assembly path optimization algorithm based on a particle swarm; according to the key position of the assembly path, an assembly intelligent path optimization algorithm based on a particle swarm is used for searching an optimal path point, and the assembly path is fitted; performing interference and collision detection on the assembly path; and outputting the feasible assembly path. According to the method, the virtual high-precision interference detection technology and the assembly path planning are combined to dynamically adjust the assembly path so as to adapt to the error and deformation of the actual workpiece surface. In the assembling process, the position and the size of the micro gap in the assembling process are calculated in a digital space through a digital twinning means, and an assembling path is adjusted according to a detection result so as to ensure that the micro gap is accurately controlled.

Description

Technical Field

[0001] The present invention relates to the field of industrial manufacturing, and the specific technical field is high-precision assembly in large-scale equipment manufacturing. Specifically, it is a small gap assembly path planning method based on high-precision interference detection, which is used in the high-precision assembly process of large-scale equipment, and is specifically aimed at the insertion path planning when the assembly datum of two parts to be assembled is deformed, the assembly gap is small, and the assembly path window is narrow. Background Art

[0002] The manufacture of large equipment usually requires the precise assembly of various components to ensure the performance and quality of the final product. However, due to various factors in the manufacturing and transportation process, such as surface manufacturing errors, deformation, thermal expansion, etc., it becomes very difficult to control the tiny gaps during the assembly process. Traditional assembly path planning methods usually only consider the assembly path under ideal conditions, without fully considering the changes in the actual state of the workpiece. Summary of the invention

[0003] In order to solve the above technical problems, the purpose of the present invention is to provide a small gap assembly path planning method based on high-precision interference detection, which is mainly used for insertion path planning when the assembly datum of two parts to be assembled is deformed, the assembly gap is small, and the assembly path window is narrow. It is used to achieve precise control of small gaps and high-quality assembly in the high-precision assembly process of large equipment.

[0004] The technical solution adopted by the present invention to achieve the above-mentioned purpose is: a small gap assembly path planning method based on high-precision interference detection, comprising the following steps:

[0005] Step S1: obtaining the actual surface shape of the mating surface of the workpieces to be assembled by a three-dimensional scanning device, and performing preliminary point cloud data processing to make the relative positions of the point cloud data of the two workpieces to be assembled consistent with the theoretical assembly positions of the workpieces;

[0006] Step S2: Generate an initial assembly path according to the relative position of the point cloud in step S1 and the theoretical assembly path, set several key positions of the assembly path based on the fixed workpiece, and initialize the assembly path optimization algorithm based on the particle swarm;

[0007] Step S3: according to the key positions of the assembly path, the assembly path optimization algorithm based on particle swarm is used to find the optimal path point, and the optimal path point is fitted as the assembly path; if the assembly path optimization based on particle swarm cannot find the optimal path point at any key assembly position, it is determined that the workpiece assembly cannot be completed;

[0008] Step S4: performing interference and collision detection on the assembly path obtained in step S3, and calculating the assembly interference and clearance between the workpieces on the assembly path;

[0009] Step S5: Analyze assembly interference and clearance values; if there is interference or the clearance value is less than the threshold, add the location where the interference occurs as the key location of the assembly path, and repeat steps S3-S4, otherwise output a feasible assembly path.

[0010] Described step S1 comprises the following steps:

[0011] Step S1.1: De-noising and smoothing the three-dimensional point cloud data of the workpiece to be assembled obtained by the three-dimensional scanning device to reduce the noise of the three-dimensional point cloud data;

[0012] Step S1.2: respectively aligning the point cloud data of the two workpieces to be assembled obtained in step S1.1 with the two assembled workpieces in the assembly CAD model, so that the relative positions of the point cloud data of the two workpieces to be assembled are consistent with the theoretical assembly positions of the workpieces.

[0013] Described step S2 comprises the following steps:

[0014] Step S2.1: generating an initial assembly path according to the relative positions of the point cloud data of the two workpieces to be assembled determined in step S1.2, the assembly direction of the workpieces during the assembly process, and the theoretical assembly path;

[0015] Step S2.2: several key positions of the assembly path are set on the initial assembly path according to the assembly features of the assembly workpiece, and the key positions of the assembly path are used as the cross-sectional cutting points of the point cloud for optimizing the assembly path points; wherein the key positions include the positions where the stepped hole stop begins to fit and the positions where the assembly clearance accuracy is higher than the threshold;

[0016] Step S2.3: Initialize the particle swarm-based assembly path optimization algorithm: each particle is the two-dimensional deviation value of the non-interference assembly position in the workpiece assembly process compared with the theoretical assembly position along the direction perpendicular to the assembly movement, the number of particles, the initial position of the particle, the particle value range, the number of iterations, and the particle acceleration factor. The constraint condition is that the matching gap between the point cloud data of the two workpieces at the point cloud cross-sectional cutting point is greater than the given threshold δ, and the fitness function is the variance of the matching gap between the point cloud data of the two workpieces at the point cloud cross-sectional cutting point and the convergence threshold of the fitness function.

[0017] Described step S3 comprises the following steps:

[0018] Step S3.1: For each key position of the assembly path, the point cloud data of the two workpieces to be assembled are cut along the vertical assembly direction. The optimization goal is to make the point cloud data gaps of the two workpieces at the key position of the assembly path point cloud cross section uniform, that is, the fitness function is minimized. The assembly path optimization algorithm based on particle swarm is used to find the optimal path point. If no particles that meet the constraint conditions can be found, step S3.2 is executed, otherwise step S3.3 is executed;

[0019] Step S3.2: If no particle that meets the constraint conditions can be found, it is determined that the workpiece cannot be assembled, and the assembly path optimization algorithm is terminated;

[0020] Step S3.3: Use the particle swarm-based intelligent path optimization algorithm to find the best key position path point that minimizes the fitness function, determine whether the optimization algorithm meets the optimization termination condition, and record the position of the optimal path point;

[0021] Step S3.4: Repeat steps S3.1-3.3 until all key path points of the assembly path are optimized, and use the least squares method to fit the optimal path points of several key assembly paths into a straight line to obtain the assembly path fitted with the optimal path points.

[0022] Described step S4 comprises the following steps:

[0023] Step S4.1: n point cloud cross-section points are set on the fixed assembly workpiece along the movement direction of the assembly, and the assembly path is divided into m segments according to the assembly path verification accuracy requirement;

[0024] Step S4.2: the point cloud data of the moving assembly workpiece is moved according to the assembly path, and each section of the assembly path is subjected to a point cloud cross-section sectioning point during the assembly movement, so as to obtain n point cloud data matching cross sections of the fixed assembly workpiece and the moving assembly workpiece;

[0025] Step S4.3: Repeat step S4.2 until the point cloud data of the moving assembly workpiece is assembled according to the assembly path, and finally obtain n*m point cloud data matching cross sections of the fixed assembly workpiece and the moving assembly workpiece;

[0026] Step S4.4: Calculate the interference and gap values ​​between the point cloud data of the fixed assembly workpiece and the moving assembly workpiece on the n*m ​​cross sections.

[0027] The step S5 includes the following steps:

[0028] Step S5.1: Analyze and count the interference and gap values ​​between the point cloud data of the fixed assembly workpiece and the moving assembly workpiece on the n*m ​​cross sections. If there is interference, that is, the gap value is less than 0, or the gap value is less than the safety threshold specified by the assembly process, execute step S5.2; otherwise, execute step S5.3:;

[0029] Step S5.2: adding the position where interference occurs, i.e., the clearance value is less than 0, or the clearance value is less than the safety threshold specified by the assembly process as the key position of the assembly path, and returning to step S3;

[0030] Step S5.3: Outputting the assembly path optimized by the particle swarm-based assembly path optimization algorithm as a feasible assembly path.

[0031] The micro-gap assembly path planning system based on high-precision interference detection includes:

[0032] The point cloud data acquisition module is used to acquire the actual surface shape of the mating surface of the workpieces to be assembled through a three-dimensional scanning device, and perform preliminary point cloud data processing to make the relative position of the point cloud data of the two workpieces to be assembled consistent with the theoretical assembly position of the workpieces;

[0033] The assembly path optimization initialization module is used to generate the initial assembly path according to the relative position of the point cloud and the theoretical assembly path, set several key positions of the assembly path based on the fixed workpiece, and initialize the assembly path optimization algorithm based on the particle swarm;

[0034] The assembly path optimization module is used to find the optimal path point according to the key positions of the assembly path using the assembly path optimization algorithm based on particle swarm, and fit the optimal path point into the assembly path; if the assembly path optimization based on particle swarm cannot find the optimal path point at any key assembly position, it is determined that the workpiece assembly cannot be completed;

[0035] The assembly path output module is used to perform interference and collision detection on the assembly path, calculate the assembly interference and clearance between the workpieces on the assembly path; analyze the assembly interference and clearance values; if there is interference or the clearance value is less than the threshold, the location where the interference occurs is added as the key location of the assembly path, otherwise a feasible assembly path is output.

[0036] A device for planning a small gap assembly path based on high-precision interference detection comprises a memory and a processor; the memory is used to store a computer program; the processor is used to implement the small gap assembly path planning method based on high-precision interference detection when executing the computer program.

[0037] A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the method for planning a small gap assembly path based on high-precision interference detection is implemented.

[0038] The present invention has the following beneficial effects and advantages:

[0039] 1. This method takes into account the manufacturing error and deformation of the workpiece surface, and aims to achieve precise control of small gaps and high-quality assembly during the assembly process. The present invention is applicable to various large-scale equipment manufacturing fields, such as aerospace, automobile manufacturing, heavy machinery, etc., especially for assembly scenarios with extremely high assembly precision requirements.

[0040] 2. The innovation of the present invention is that it combines virtual high-precision interference detection technology with assembly path planning to dynamically adjust the assembly path to adapt to the errors and deformations of the actual workpiece surface. During the assembly process, the position and size of the tiny gaps in the assembly process are calculated in the digital space by means of digital twins, and the assembly path is adjusted according to the detection results to ensure that the tiny gaps are precisely controlled. This method can improve the accuracy and repeatability of assembly, reduce the scrap rate, thereby reducing manufacturing costs and improving product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is the overall process of the small gap assembly path planning method based on high-precision interference detection of the present invention;

[0042] Figure 2 is a schematic diagram of two assembly workpieces and key positions of an assembly path in an embodiment of the present invention;

[0043] Figure 3 is a schematic diagram of the cross-sectional cutting points of the point cloud set along the movement direction of the assembly in an embodiment of the present invention;

[0044] Figure 4 It is a schematic diagram of the point cloud cross-section cutting points and the point cloud data matching cross-section in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments, but the scope of protection claimed in the present invention is not limited to the following specific embodiments.

[0046] Example

[0047] This embodiment discloses a small gap assembly path planning method based on high-precision interference detection

[0048] like Figure 1 As shown, the small gap assembly path planning method based on high-precision interference detection includes the following steps:

[0049] Step S1: obtaining the actual surface shape of the mating surface of the workpieces to be assembled by a three-dimensional scanning device, and performing preliminary point cloud data processing to make the relative positions of the point cloud data of the two workpieces to be assembled consistent with the theoretical assembly positions of the workpieces;

[0050] Step S2: Determine a theoretical assembly path as the initial assembly path based on the ideal CAD model of the workpiece, manually define several key positions of the assembly path based on the fixed workpiece, and initialize the assembly path optimization algorithm based on the particle swarm;

[0051] Step S3: According to the key positions of the assembly path, the particle swarm-based intelligent path optimization algorithm is used to find the optimal path point, and the optimal path point is fitted as the assembly path. If the particle swarm-based intelligent path optimization cannot find the optimal path point at any key assembly position, it is determined that the workpiece cannot be assembled;

[0052] Step S4: performing interference and collision detection on the assembly path obtained in S3, and calculating the assembly interference and clearance between the workpieces on the assembly path;

[0053] Step S5: Analyze assembly interference and clearance value. If there is interference or the clearance value is less than the threshold, add the location where the interference occurs as the key location of the assembly path and repeat steps S3-S4. Otherwise, output a feasible assembly path.

[0054] Specifically, the step S1 includes the following steps:

[0055] Step S1.1: De-noising and smoothing the three-dimensional point cloud data of the workpiece to be assembled obtained by the three-dimensional scanning device to reduce the noise of the three-dimensional point cloud data;

[0056] Specifically, the step S1.1 includes the following steps:

[0057] Step S1.1.1: Set the filter radius and threshold, and use radius filtering to remove isolated noise points in the point cloud data;

[0058] Step S1.1.2: Set the mean and variance of the filter and use the statistical filtering method to further remove outliers in the point cloud data;

[0059] Step S1.1.3: For the point cloud data obtained in S2.2, use the random sampling method to simplify the data; use the Gaussian filtering method to smooth the data to improve the accuracy of the three-dimensional point cloud data.

[0060] Step S1.2: aligning the point cloud data of the workpiece in S1.1 with the workpiece in the assembly CAD model respectively, so that the relative positions of the point cloud data of the two workpieces to be assembled are consistent with the theoretical assembly positions of the workpieces;

[0061] Specifically, the step S1.2 includes the following steps:

[0062] Step S1.2.1: First, perform a rough registration of the actual point cloud data and the ideal point cloud data, use the local point pair feature descriptor combined with the random consistency sampling method to determine the initial coordinate transformation matrix, and complete the rough registration of the point cloud data collected by the 3D scanning device;

[0063] Specifically, the step S1.2.1 includes the following steps:

[0064] Step S1.2.1.1: Calculate the local point pair feature descriptors of all points in the actual point cloud data in S2 and the ideal point cloud data in S1 respectively.

[0065] Step S1.2.1.2: using a random consistency sampling method to identify point pairs having a corresponding relationship between the actual point cloud data and the ideal point cloud data;

[0066] Step S1.2.1.3: Based on the point pairs obtained by the random consistency sampling method in step S3.1.2, calculate the coordinate transformation matrix from the actual point cloud data to the ideal point cloud data, and apply the matrix transformation to the actual point cloud data, which is transformed from the coordinate system of the scanning device to the same coordinate system as the ideal point cloud data.

[0067] Step S1.2.2: Perform precise registration of the point cloud data collected by three-dimensional scanning, use the iterative closest point (ICP) algorithm to calculate the coordinate transformation matrix of the point cloud precise registration, and obtain the precisely registered three-dimensional scanning point cloud data.

[0068] Specifically, the step S1.2.2 includes the following steps:

[0069] Step S1.2.2.1: Initialize the ICP algorithm and set the convergence conditions: the error is less than a certain threshold or the number of iterations reaches a predetermined value;

[0070] Step S1.2.2.2: For each point in the actual point cloud data, find the point in the ideal point cloud closest to it and establish a point pair matching relationship;

[0071] Step S1.2.2.3: Use Euclidean distance to measure the distance error between each point pair;

[0072] Step S1.2.2.4: Assign weights to each point pair using a Gaussian weight function based on the error value;

[0073] Step S1.2.2.5: Use numerical optimization methods to minimize the error between weighted point pairs and adjust the coordinate transformation matrix between the point pairs;

[0074] Step S1.2.2.6: Check whether the convergence condition is met. If not, execute step S1.2.2.7. If the convergence condition is met, execute step S1.2.2.8.

[0075] Step S1.2.2.7: Continue iterating and update the coordinate transformation matrix;

[0076] Step S1.2.2.8: Obtain the final coordinate transformation matrix, apply the matrix to the actual point cloud data, and accurately align it to the ideal point cloud data.

[0077] Described step S2 comprises the following steps:

[0078] Step S2.1: Determine the assembly direction and theoretical assembly path according to the ideal CAD model of the workpiece as the initial assembly path of the particle swarm-based assembly path optimization algorithm;

[0079] Specifically, the step S2.1 includes the following steps:

[0080] Step S2.1.1: An ideal CAD assembly model of the workpiece is used to determine the assembly direction and final assembly position between the two workpieces to be assembled;

[0081] Step S2.1.2: Define a theoretical assembly path according to the requirements of the assembly task and the geometric features of the two workpieces to be assembled;

[0082] Step S2.2: several key positions of the assembly path are defined on the initial assembly path according to the assembly features of the assembly workpiece, such as (the position where the stopper begins to engage with the joint, the position where the engagement requirement is high), and the key positions of the assembly path are used as the cross-sectional cutting points of the point cloud for optimizing the assembly path points; Figure 2 shown.

[0083] Specifically, the step S2.2 includes the following steps:

[0084] Step S2.2.1: Analyze the geometric features and assembly requirements of the workpiece to determine which positions are more important in the assembly process (including the positions where the stopper begins to engage, positions with higher engagement requirements, etc.).

[0085] Step S2.2.2: defining the key positions in step S2.2.1 on the initial assembly path as a series of key points along the assembly path, i.e., the key positions of the assembly path;

[0086] Step S2.3: Initialize the particle swarm-based assembly path optimization algorithm. Define each particle as the two-dimensional deviation value of the workpiece assembly process along the direction perpendicular to the assembly movement compared to the theoretical assembly position, define the number of particles, the initial position of the particles, the particle value range, the number of iterations, and the particle acceleration factor, define the constraint condition as the matching gap between the point cloud data of the two workpieces at the point cloud cross-section cutting point is greater than the given threshold δ, and define the fitness function as the variance of the matching gap between the point cloud data of the two workpieces at the point cloud cross-section cutting point and the convergence threshold of the fitness function;

[0087] Step S2.3.1: Define each particle to represent the deviation value of the workpiece assembly process along the direction perpendicular to the assembly movement relative to the theoretical assembly position. Define the particle dimensions to be six-dimensional, which respectively represent the deviation value of the axis of the moving assembly workpiece from the theoretical assembly position in six degrees of freedom in space when the assembly reaches a critical position of a certain assembly path, where the deviation value along the assembly direction should be zero;

[0088] Step S2.3.2: Define the algorithm parameters such as the number of particles, the initial position of the particles, the particle value range (the value allowed to deviate from the theoretical assembly position), the number of iterations and the convergence threshold of the fitness function (the termination condition of the particle swarm-based assembly path optimization algorithm) and the particle acceleration factor.

[0089] Step S2.3.3: Define a fitness function for evaluating the performance of each particle. The fitness function is defined as the variance of the fit gap between the point cloud data of the two workpieces on the point cloud cross section at all section points. The optimization goal is defined as the uniform gap between the point cloud data of the two workpieces on the point cloud cross section at the key position of the assembly path (i.e., the fitness function value is the smallest) to obtain the best assembly path.

[0090] Step S2.3.4: randomly generate the initial particle swarm and ensure that they are within the defined value range;

[0091] Based on the path optimization algorithm, the optimal planning of the assembly path is performed, and the step S3 includes the following steps:

[0092] Step S3.1: For each key position of the assembly path, cut the point cloud data of the two workpieces to be assembled along the vertical assembly direction, define the optimization goal as the uniform gap between the point cloud data of the two workpieces on the cross section of the point cloud at the key position of the assembly path (that is, the fitness function is minimized), use the particle swarm-based intelligent path optimization algorithm to find the optimal path point, if no particles that meet the constraint conditions can be found, execute step S3.2, otherwise execute step S3.3;

[0093] Step S3.2: If no particle that meets the constraint conditions can be found, it is determined that the workpiece cannot be assembled and the path optimization algorithm is terminated;

[0094] Step S3.3: Use the particle swarm-based intelligent path optimization algorithm to find the best key position path point that minimizes the fitness function, determine whether the optimization algorithm meets the optimization termination condition, and record the position of the optimal path point;

[0095] Specifically, the step S3.3 includes the following steps:

[0096] Step S3.3.1: Determine whether the optimization algorithm has reached the maximum number of iterations or the fitness function meets the threshold requirement. If the termination condition is reached, execute step S3.3.2, otherwise execute step S3.3.3;

[0097] Step S3.3.2: Record the positions of the global optimal path points during multiple iterations, and complete the optimization of the assembly path points at the key positions of the assembly path;

[0098] Step S3.3.3: Update the position and velocity of the particles, calculate the fitness of each particle, and repeat step S3.3.1;

[0099] Step S3.4: Repeat steps S3.1-3.3 until all key path points of the assembly path are optimized, and the optimal path points of the key locations of the assembly path are fitted into a straight line using the least squares method to obtain an assembly path fitted with the optimal path points;

[0100] Specifically, the step S3.4 includes the following steps:

[0101] Step S3.4.1: Repeat steps S3.1-3.3 until the optimization of all key path points of the assembly path is completed.

[0102] Step S3.4.2: Use the least square method to fit all key path points of the assembly path into a straight line to generate an assembly path with optimal path point fitting;

[0103] Verify the feasibility of the assembly path of the optimal path point fitting, the step S4 includes the following steps:

[0104] Step S4.1: Figure 3 As shown, n point cloud cross-section points are set on the fixed assembly workpiece along the movement direction of the assembly, and the assembly path is divided into m segments according to the assembly path verification accuracy requirements. The larger the m, the higher the accuracy of the path verification calculation.

[0105] Specifically, the step S4.1 includes the following steps:

[0106] Step S4.1.1: Set n point cloud cross-section points on the fixed assembly workpiece along the direction of assembly movement. These points will be used to capture point cloud data at different positions to verify assembly accuracy and calculate interference and clearance.

[0107] Step S4.1.2: According to the verification accuracy requirements of the assembly path, the assembly path is divided into m segments. This is determined according to the specific requirements of the application. The larger the m, the higher the accuracy of the path verification calculation. Each segment of the path represents a step or position of the assembly.

[0108] Step S4.2: the point cloud data of the moving assembly workpiece is moved according to the assembly path, and each section of the assembly path is subjected to a point cloud cross-section sectioning point during the assembly movement, so as to obtain n point cloud data matching cross sections of the fixed assembly workpiece and the moving assembly workpiece;

[0109] Specifically, the step S4.2 includes the following steps:

[0110] Step S4.2.1: making the point cloud data of the moving assembly workpiece move according to the assembly path;

[0111] Step S4.2.2: On each assembly path, perform a point cloud cross-section cut according to the pre-set point cloud cross-section cut points to obtain n point cloud data matching cross-sections of the fixed assembly workpiece and the moving assembly workpiece;

[0112] Step S4.3: Repeat step S4.2 until the point cloud data of the moving assembly workpiece is assembled according to the assembly path, and finally obtain n*m point cloud data matching cross sections of the fixed assembly workpiece and the moving assembly workpiece;

[0113] Step S4.4: Calculate the interference and gap values ​​between the point cloud data of the fixed assembly workpiece and the moving assembly workpiece on the n*m ​​cross sections;

[0114] Judging whether the assembly path is feasible according to the interference and clearance values, the step S5 includes the following steps:

[0115] Step S5.1: Analyze and count the interference and gap values ​​between the point cloud data of the fixed assembly workpiece and the moving assembly workpiece on n*m cross sections. If there is interference (the gap value is less than 0) or the gap value is less than the safety threshold specified by the assembly process, execute step S5.2, otherwise execute step S5.3;

[0116] Specifically, the step S5.1 includes the following steps:

[0117] Step S5.1.1: For each cross section, calculating the interference or clearance value between the fixed assembly workpiece and the moving assembly workpiece;

[0118] Step S5.1.2: Repeat step S5.1.1 until the calculation of the interference or clearance value between the fixed assembly workpiece and the moving assembly workpiece on each cross section is completed;

[0119] Step S5.1.3: Figure 4 As shown, the interference and gap values ​​between the point cloud data of the fixed assembly workpiece and the moving assembly workpiece on the n*m ​​cross sections are counted and analyzed. If there is interference (the gap value is less than 0) or the gap value is less than the safety threshold specified by the assembly process, the assembly path position corresponding to the cross section is recorded and step S5.2 is executed; otherwise, step S5.3 is executed;

[0120] Step S5.2: adding the position where interference occurs (clearance value is less than 0) or the position where the clearance value is less than the safety threshold specified by the assembly process as the key position of the assembly path, and repeating steps S3-S4;

[0121] Step S5.3: Outputting the assembly path optimized by the particle swarm-based assembly path optimization algorithm as a feasible assembly path.

Claims

1. A small gap assembly path planning method based on high-precision interference detection, characterized in that: The following steps are involved: Step S1: obtaining the actual surface shape of the mating surface of the workpieces to be assembled by a three-dimensional scanning device, and performing preliminary point cloud data processing to make the relative positions of the point cloud data of the two workpieces to be assembled consistent with the theoretical assembly positions of the workpieces; Step S2: Generate an initial assembly path according to the relative position of the point cloud in step S1 and the theoretical assembly path, set several key positions of the assembly path based on the fixed workpiece, and initialize the assembly path optimization algorithm based on the particle swarm; Step S3: according to the key positions of the assembly path, the assembly path optimization algorithm based on particle swarm is used to find the optimal path point, and the optimal path point is fitted as the assembly path; if the assembly path optimization based on particle swarm cannot find the optimal path point at any key assembly position, it is determined that the workpiece assembly cannot be completed; Step S4: performing interference and collision detection on the assembly path obtained in step S3, and calculating the assembly interference and clearance between the workpieces on the assembly path; Step S5: Analyze assembly interference and clearance values; if there is interference or the clearance value is less than the threshold, add the location where the interference occurs as the key location of the assembly path, and repeat steps S3-S4, otherwise output a feasible assembly path.

2. The method for small gap assembly path planning based on high-precision interference detection according to claim 1 is characterized in that: Described step S1 comprises the following steps: Step S1.1: De-noising and smoothing the three-dimensional point cloud data of the workpiece to be assembled obtained by the three-dimensional scanning device to reduce the noise of the three-dimensional point cloud data; Step S1.2: respectively aligning the point cloud data of the two workpieces to be assembled obtained in step S1.1 with the two assembled workpieces in the assembly CAD model, so that the relative positions of the point cloud data of the two workpieces to be assembled are consistent with the theoretical assembly positions of the workpieces.

3. The method for small gap assembly path planning based on high-precision interference detection according to claim 1 is characterized in that: Described step S2 comprises the following steps: Step S2.1: generating an initial assembly path according to the relative positions of the point cloud data of the two workpieces to be assembled determined in step S1.2, the assembly direction of the workpieces during the assembly process, and the theoretical assembly path; Step S2.2: several key positions of the assembly path are set on the initial assembly path according to the assembly features of the assembly workpiece, and the key positions of the assembly path are used as the cross-sectional cutting points of the point cloud for optimizing the assembly path points; wherein the key positions include the positions where the stepped hole stop begins to fit and the positions where the assembly clearance accuracy is higher than the threshold; Step S2.3: Initialize the particle swarm-based assembly path optimization algorithm: each particle is the two-dimensional deviation value of the non-interference assembly position in the workpiece assembly process compared with the theoretical assembly position along the direction perpendicular to the assembly movement, the number of particles, the initial position of the particle, the particle value range, the number of iterations, and the particle acceleration factor. The constraint condition is that the matching gap between the point cloud data of the two workpieces at the point cloud cross-sectional cutting point is greater than the given threshold δ, and the fitness function is the variance of the matching gap between the point cloud data of the two workpieces at the point cloud cross-sectional cutting point and the convergence threshold of the fitness function.

4. The method for small gap assembly path planning based on high-precision interference detection according to claim 1 is characterized in that: Described step S3 comprises the following steps: Step S3.1: For each key position of the assembly path, the point cloud data of the two workpieces to be assembled are cut along the vertical assembly direction. The optimization goal is to make the point cloud data gaps of the two workpieces at the key position of the assembly path point cloud cross section uniform, that is, the fitness function is minimized. The assembly path optimization algorithm based on particle swarm is used to find the optimal path point. If no particles that meet the constraint conditions can be found, step S3.2 is executed, otherwise step S3.3 is executed; Step S3.2: If no particle that meets the constraint conditions can be found, it is determined that the workpiece cannot be assembled, and the assembly path optimization algorithm is terminated; Step S3.3: Use the particle swarm-based intelligent path optimization algorithm to find the best key position path point that minimizes the fitness function, determine whether the optimization algorithm meets the optimization termination condition, and record the position of the optimal path point; Step S3.4: Repeat steps S3.1-3.3 until all key path points of the assembly path are optimized, and use the least squares method to fit the optimal path points of several key assembly paths into a straight line to obtain the assembly path fitted with the optimal path points.

5. The method for small gap assembly path planning based on high-precision interference detection according to claim 1 is characterized in that: Described step S4 comprises the following steps: Step S4.1: n point cloud cross-section points are set on the fixed assembly workpiece along the movement direction of the assembly, and the assembly path is divided into m segments according to the assembly path verification accuracy requirement; Step S4.2: the point cloud data of the moving assembly workpiece is moved according to the assembly path, and each section of the assembly path is subjected to a point cloud cross-section sectioning point during the assembly movement, so as to obtain n point cloud data matching cross sections of the fixed assembly workpiece and the moving assembly workpiece; Step S4.3: Repeat step S4.2 until the point cloud data of the moving assembly workpiece is assembled according to the assembly path, and finally obtain n*m point cloud data matching cross sections of the fixed assembly workpiece and the moving assembly workpiece; Step S4.4: Calculate the interference and gap values ​​between the point cloud data of the fixed assembly workpiece and the moving assembly workpiece on the n*m ​​cross sections.

6. The method for small gap assembly path planning based on high-precision interference detection according to claim 1 is characterized in that: The step S5 includes the following steps: Step S5.1: Analyze and count the interference and gap values ​​between the point cloud data of the fixed assembly workpiece and the moving assembly workpiece on the n*m ​​cross sections. If there is interference, that is, the gap value is less than 0, or the gap value is less than the safety threshold specified by the assembly process, execute step S5.2; otherwise, execute step S5.3:; Step S5.2: adding the position where interference occurs, i.e., the clearance value is less than 0, or the clearance value is less than the safety threshold specified by the assembly process as the key position of the assembly path, and returning to step S3; Step S5.3: Outputting the assembly path optimized by the particle swarm-based assembly path optimization algorithm as a feasible assembly path.

7. A small gap assembly path planning system based on high-precision interference detection, characterized in that: include: The point cloud data acquisition module is used to acquire the actual surface shape of the mating surface of the workpieces to be assembled through a three-dimensional scanning device, and perform preliminary point cloud data processing to make the relative position of the point cloud data of the two workpieces to be assembled consistent with the theoretical assembly position of the workpieces; The assembly path optimization initialization module is used to generate the initial assembly path according to the relative position of the point cloud and the theoretical assembly path, set several key positions of the assembly path based on the fixed workpiece, and initialize the assembly path optimization algorithm based on the particle swarm; The assembly path optimization module is used to find the optimal path point according to the key positions of the assembly path using the assembly path optimization algorithm based on particle swarm, and fit the optimal path point into the assembly path; if the assembly path optimization based on particle swarm cannot find the optimal path point at any key assembly position, it is determined that the workpiece assembly cannot be completed; The assembly path output module is used to perform interference and collision detection on the assembly path, calculate the assembly interference and clearance between the workpieces on the assembly path; analyze the assembly interference and clearance values; if there is interference or the clearance value is less than the threshold, the location where the interference occurs is added as the key location of the assembly path, otherwise a feasible assembly path is output.

8. A small gap assembly path planning device based on high-precision interference detection, characterized in that: It comprises a memory and a processor; the memory is used to store a computer program; the processor is used to implement the small gap assembly path planning method based on high-precision interference detection as described in any one of claims 1-6 when executing the computer program.

9. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the processor, the method for small gap assembly path planning based on high-precision interference detection as described in any one of claims 1 to 6 is implemented.

Images