Path planning methods and related devices for in-flight measurement of aerospace parts

Abstract

This application discloses a path planning method and related apparatus for in-flight measurement of aerospace parts, relating to the field of path planning technology. The method includes: after acquiring the measurement point set of each measurement feature of the aerospace part's 3D model, using a nearest-neighbor algorithm to determine the starting and ending measurement points of each feature; for each feature, using an ant colony algorithm to perform path planning to obtain a collision-free local path; simultaneously, using the RRT* algorithm to perform path planning to obtain a collision-free connecting path; and connecting all local paths and all connecting paths according to the connection order to obtain a complete measurement path for in-flight measurement of the aerospace part's 3D model. This application combines three algorithms, optimizing the path planning process without manual intervention, thereby efficiently and automatically completing path planning for in-flight measurement of aerospace parts and improving measurement efficiency.

Description

Technical Field

[0001] This application relates to the field of path planning technology, and in particular to a path planning method and related apparatus for in-flight measurement of aerospace parts. Background Technology

[0002] With the widespread adoption of digitalization in the aviation industry, the inspection methods for aerospace parts have also undergone a revolution. Consequently, using advanced algorithms to empower the inspection process and achieve a higher degree of automation has become a popular research direction. Coordinate Measuring Machines (CMMs) are one of the most effective methods for measuring and obtaining dimensional data. For the key inspection characteristics of aerospace parts, in-machine measurement using CMMs is a commonly used technology. In this process, the object to be measured is placed in a three-coordinate measurement space, allowing the acquisition of the coordinate positions of various measuring points on the object. Based on these spatial coordinate values, the geometric dimensions, shape, and position of the object are calculated.

[0003] Currently, the common method for connecting measurement points to form a measurement path is as follows: First, measurement point planning is performed. Points are manually selected or automatically placed according to rules to form a measurement point set. The measurement points in the set are arranged in the order of generation to obtain a measurement point sequence. Connecting the measurement points according to the sequence forms the original measurement path. Then, obstacle avoidance is performed. After forming the original measurement path, a safe plane approach is used for obstacle avoidance to form a collision-free measurement path. However, the above path planning process often does not pay enough attention to the order of measurement points, resulting in an excessively long and poor-quality collision-free measurement path, leading to a significant waste of measurement time. Furthermore, the safe plane approach involves manual intervention in processes such as reference plane selection and offset determination, which cannot achieve a greater degree of automation and is inefficient. Summary of the Invention

[0004] The purpose of this application is to provide a path planning method and related apparatus for on-board measurement of aerospace parts, which can efficiently and automatically complete path planning for on-board measurement of aerospace parts and improve measurement efficiency.

[0005] To achieve the above objectives, this application provides the following solution:

[0006] Firstly, this application provides a path planning method for on-flight measurement of aerospace parts, the path planning method for on-flight measurement of aerospace parts comprising:

[0007] Obtain the measurement point set for each measurement feature of the 3D model of the aerospace part; the measurement feature is a line and / or surface of the 3D model of the aerospace part; the measurement point set includes multiple measurement points located on the measurement feature;

[0008] Sort all measurement features according to their connection order. Using the measurement point set of the i-th measurement feature and the measurement point set of the (i+1)-th measurement feature as input, determine the last measurement point of the i-th measurement feature and the starting measurement point of the (i+1)-th measurement feature using the nearest point pair algorithm. Randomly select a measurement point other than the last measurement point of the i-th measurement feature as the starting measurement point of the i-th measurement feature. Randomly select a measurement point other than the starting measurement point of the i-th measurement feature as the last measurement point of the i-th measurement feature. This yields the starting and ending measurement points for each measurement feature. Here, i = 1, 2, ..., I-1, where I is the total number of measurement features.

[0009] For each measurement feature, the ant colony algorithm is used for path planning to obtain a collision-free local path that traverses all measurement points in the measurement point set of the measurement feature and runs from the starting measurement point of the measurement feature to the ending measurement point of the measurement feature.

[0010] For the i-th and i+1-th measurement features, the RRT* algorithm is used for path planning to obtain a collision-free connection path from the end measurement point of the i-th measurement feature to the start measurement point of the i+1-th measurement feature.

[0011] By connecting all local paths and all connecting paths in the order of connection, a complete measurement path for in-flight measurement of the 3D model of the aerospace part is obtained.

[0012] Secondly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described path planning method for in-flight measurement of aerospace parts.

[0013] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned path planning method for in-flight measurement of aerospace parts.

[0014] Fourthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the aforementioned path planning method for in-flight measurement of aerospace parts.

[0015] According to the specific embodiments provided in this application, this application has the following technical effects:

[0016] This application provides a path planning method and related apparatus for in-flight measurement of aerospace parts. After obtaining the measurement point set of each measurement feature of the 3D model of the aerospace part, the nearest point pair algorithm is used to determine the starting and ending measurement points of each measurement feature. For each measurement feature, the ant colony algorithm is used for path planning to obtain a collision-free local path that traverses all measurement points in the measurement point set and runs from the starting to the ending measurement point of the measurement feature. Simultaneously, for the i-th and (i+1)-th measurement features, the RRT* algorithm is used for path planning to obtain a collision-free connecting path from the ending to the starting point of the (i+1)-th measurement feature. All local paths and all connecting paths are connected in the order of connection to obtain the complete measurement path for in-flight measurement of the 3D model of the aerospace part. This application combines three algorithms, completing the path planning process without manual intervention. Furthermore, the algorithm can optimize the path planning process, thus efficiently and automatically completing path planning for in-flight measurement of aerospace parts and improving measurement efficiency. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is an application environment diagram for a path planning method for in-flight measurement of aerospace parts provided in Embodiment 1 of this application.

[0019] Figure 2 This is a flowchart illustrating a path planning method for in-flight measurement of aerospace parts, as provided in Embodiment 1 of this application.

[0020] Figure 3 This is a schematic diagram of the improved RRT* algorithm provided in Embodiment 1 of this application.

[0021] Figure 4 This is a schematic diagram of local smoothing processing provided in Embodiment 1 of this application; wherein, Figure 4 (a) in the text represents the original path. Figure 4 (b) in the diagram represents the path after removing redundant points. Figure 4 (c) in the diagram represents the path after local smoothing.

[0022] Figure 5 This is a schematic diagram of the measuring point orientation and probe collision volume provided in Embodiment 1 of this application; wherein, Figure 5 (a) in the image represents the orientation after the measurement point has been processed. Figure 5(b) in the figure represents the orientation of the measurement point without processing.

[0023] Figure 6 This is a flowchart of the collision detection algorithm provided in Embodiment 1 of this application.

[0024] Figure 7 This is a schematic diagram of the structure of a computer device provided in Embodiment 2 of this application. Detailed Implementation

[0025] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0026] Example 1

[0027] The path planning method for in-flight measurement of aerospace parts provided in this application embodiment can be applied to, for example... Figure 1In the application environment shown, the terminal communicates with the server via a network. The data storage system stores the data the server needs to process. The data storage system can be set up independently, integrated into the server, or placed in the cloud or on another server. The terminal can send the measurement point set of each measurement feature of the 3D model of the aircraft part to be processed to the server. After receiving the measurement point set of each measurement feature of the 3D model of the aircraft part to be processed, the server sorts all measurement features according to the connection order. Using the measurement point set of the i-th measurement feature and the measurement point set of the (i+1)-th measurement feature as input, the server uses the nearest point pair algorithm to determine the end measurement point of the i-th measurement feature and the start measurement point of the (i+1)-th measurement feature. A measurement point other than the end measurement point of the i-th measurement feature is randomly selected from the measurement point set of the i-th measurement feature as the start measurement point of the i-th measurement feature. A measurement point other than the end measurement point of the i-th measurement feature is randomly selected from the measurement point set of the i-th measurement feature. All measurement points except the starting point of the i-th measurement feature are used as the ending point of the i-th measurement feature, resulting in the starting and ending points of each measurement feature. For each measurement feature, ant colony optimization is used for path planning to obtain a collision-free local path that traverses all measurement points in the measurement feature's set and runs from the starting point to the ending point. For the i-th and (i+1)-th measurement features, RRT* algorithm is used for path planning to obtain a collision-free connecting path from the ending point of the i-th measurement feature to the starting point of the (i+1)-th measurement feature. All local paths and all connecting paths are connected in the order of connection to obtain the complete measurement path for in-flight measurement of the 3D model of the aerospace part. The server can then feed back the obtained complete measurement path to the terminal.

[0028] Furthermore, in some embodiments, the path planning method for the on-board measurement process of aerospace parts can also be implemented by a server or a terminal. For example, the terminal can directly process the measurement point set of each measurement feature of the three-dimensional model of the aerospace part to be processed, or the server can obtain the measurement point set of each measurement feature of the three-dimensional model of the aerospace part to be processed from the data storage system and process the measurement point set of each measurement feature of the three-dimensional model of the aerospace part to be processed.

[0029] The terminals can be, but are not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, and smart in-vehicle devices, while portable wearable devices can include smartwatches, smart bracelets, and head-mounted devices. Servers can be implemented using independent servers, server clusters composed of multiple servers, or cloud servers.

[0030] In one exemplary embodiment, such as Figure 2 As shown, a path planning method for in-flight measurement of aerospace parts is provided. This method is executed by computer equipment, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1 The following steps are used as an example of a server in the example.

[0031] Step S1: Obtain the measurement point set of each measurement feature of the three-dimensional model of the aerospace part; the measurement feature is a line and / or surface of the three-dimensional model of the aerospace part; the measurement point set includes multiple measurement points located on the measurement feature.

[0032] Step S2: Sort all measurement features according to the connection order. Using the measurement point set of the i-th measurement feature and the measurement point set of the (i+1)-th measurement feature as input, use the nearest point pair algorithm to determine the end measurement point of the i-th measurement feature and the start measurement point of the (i+1)-th measurement feature. Randomly select a measurement point other than the end measurement point of the i-th measurement feature from the measurement point set of the i-th measurement feature as the start measurement point of the i-th measurement feature. Randomly select a measurement point other than the start measurement point of the i-th measurement feature from the measurement point set of the i-th measurement feature as the end measurement point of the i-th measurement feature. This yields the start and end measurement points for each measurement feature. Where i = 1, 2, ..., I-1, and I is the total number of measurement features.

[0033] Step S3: For each measurement feature, use the ant colony algorithm to perform path planning to obtain a collision-free local path that traverses all measurement points in the measurement point set of the measurement feature and runs from the starting measurement point of the measurement feature to the ending measurement point of the measurement feature.

[0034] Step S4: For the i-th measurement feature and the (i+1)-th measurement feature, use the RRT* algorithm to perform path planning to obtain a collision-free connection path from the end measurement point of the i-th measurement feature to the starting measurement point of the (i+1)-th measurement feature.

[0035] Step S5: Connect all local paths and all connection paths in the connection order to obtain the complete measurement path for in-flight measurement of the 3D model of the aerospace part.

[0036] By implementing steps S1 to S5 above, this embodiment uses the nearest point pair algorithm to determine the starting and ending measurement points of each measurement feature, uses the ant colony algorithm for path planning to obtain a collision-free local path from the starting measurement point to the ending measurement point of each measurement feature, and uses the RRT* algorithm for path planning to obtain a collision-free connecting path between two adjacent measurement features from the ending measurement point of the i-th measurement feature to the starting measurement point of the (i+1)-th measurement feature. Finally, all local paths and all connecting paths are connected to obtain a complete measurement path. The entire path planning process does not require manual intervention and can achieve efficient and automatic path planning for on-board measurement of aerospace parts, realizing the generation of globally efficient collision-free paths.

[0037] Before performing on-machine measurement using a coordinate measuring machine (CMM), a measurement path file needs to be generated through path planning within the CATIA environment. This path planning process is implemented based on the MBD (Model Based Definition) model. In this embodiment, the MBD model refers to the 3D model of the aerospace part. To enable customized development of the MBD model, the Component Application Architecture (CAA) provided by CATIA offers interfaces for operating on elements at various levels, including document level (such as CATIDocument), product level (such as CATIProduct), topology level (such as CATCell), and geometry level (such as CATPoint). In addition, it provides a large number of interfaces for operating on tree structure nodes. The path planning method in this embodiment is based on the interfaces provided by CAA and combined with advanced algorithms.

[0038] The path planning process is performed on the MBD model. The generation of a measurement path may involve multiple features of the MBD model, which may include various lines, surfaces, etc. Measurement points are located on these measurement features. At this time, there is a set of measurement points on each measurement feature. The idea of ​​the path planning method in this embodiment is to first determine the nearest point pair between each adjacent measurement feature and use it as the first and last measurement points of the cross-measurement feature path. Then, local path planning is performed on each measurement feature's respective set of measurement points (ensuring that the first and last measurement points remain unchanged) to form multiple local paths. Next, cross-measurement feature connection path planning is performed to form multiple connection paths. By combining these two types of paths, the final complete measurement path can be obtained.

[0039] In this embodiment, the measurement features, measurement point sets, and connection order are all pre-designed according to user requirements. It is only necessary to obtain the measurement point set of each measurement feature of the three-dimensional model of the aerospace part. The measurement features are the lines and / or surfaces of the three-dimensional model of the aerospace part. The measurement point set includes multiple measurement points located on the measurement features. The connection order refers to the order in which the various measurement features are connected. For example, if the measurement features include A, B, and C, the connection order may be A connected to B first, then B connected to C, or it may be another order.

[0040] In this embodiment, all measurement features are sorted according to the connection order. For example, if the measurement features include A, B, and C, and the connection order is A connects to B first, then B connects to C, then the sorting result is A as the first measurement feature, B as the second measurement feature, and C as the third measurement feature. Using the measurement point set of the i-th measurement feature and the measurement point set of the (i+1)-th measurement feature as input, the nearest point pair algorithm is used to determine the last measurement point of the i-th measurement feature and the starting measurement point of the (i+1)-th measurement feature. A measurement point other than the last measurement point of the i-th measurement feature is randomly selected from the measurement point set of the i-th measurement feature as the starting measurement point of the i-th measurement feature. Similarly, a measurement point other than the starting measurement point of the i-th measurement feature is randomly selected from the measurement point set of the i-th measurement feature as the last measurement point of the i-th measurement feature. This yields the starting and ending measurement points for each measurement feature, where i = 1, 2, ..., I-1, and I is the total number of measurement features.

[0041] The reason why this embodiment first performs path preprocessing between measurement features based on the nearest point pair algorithm is that if path planning is performed directly, the path planning algorithm will take too long to run due to the large number and wide distribution of global measurement points. In order to optimize the running time of the path planning algorithm, the nearest point pair algorithm is introduced to predetermine the first and last measurement points of each measurement feature, and then local path planning is performed within each measurement feature. Finally, the optimized local path is combined with the connection path across measurement features to obtain the complete measurement path.

[0042] The traditional nearest-point-pair algorithm first decomposes the input point set, sorts all points in the set according to their x-coordinates, and then divides the set into two subsets, left and right, by the midpoint. Each subset contains half of the points. The divide-and-conquer method is then recursively applied to the left and right subsets to find the nearest point pair and its corresponding distance in each subset. Point pairs that span the left and right subsets require special handling. This involves a brute-force search of points within a certain width (twice the current minimum distance) on the midline, because these points may form closer point pairs with points on the other side. Then, the results of the left and right subsets are merged. The nearest point pair may be entirely on the left, entirely on the right, or one point may be on the left and the other on the right.

[0043] Since the input in this embodiment is two point sets instead of a single point set, the nearest point pair algorithm is improved. First, the two point sets are merged into a single point set and sorted according to their x-coordinates to facilitate efficient divide-and-conquer operations in subsequent steps. During the divide-and-conquer process, the sorted point set is recursively divided into left and right parts, and the shortest distance point pair is found in each part. To ensure that the shortest distance point pairs are derived from two different point sets, a cross-group constraint is introduced each time the results of the left and right parts are recursively merged. To implement this constraint, the points in the input point set are marked, so that in the calculations near the left and right parts and the midline, only the distance between cross-group point pairs is calculated. Furthermore, during the check of the region near the midline, point pairs whose distance to the midline is less than the current minimum distance are filtered out, ensuring that the shortest distance point pair that meets the condition is found only among cross-group point pairs, thus ensuring the minimization of the cross-group distance. Through the above process, the nearest point pairs that cross the point sets and satisfy the conditions can be obtained. In subsequent steps, each pair of closest points serves as the last measurement point of the previous measurement feature and the starting measurement point of the next measurement feature, respectively.

[0044] In this embodiment, the measurement point set of the i-th measurement feature and the measurement point set of the (i+1)-th measurement feature are used as input. The nearest point pair algorithm is used to determine the last measurement point of the i-th measurement feature and the starting measurement point of the (i+1)-th measurement feature. Specifically, this includes:

[0045] (1) Merge the measurement point set of the i-th measurement feature and the measurement point set of the (i+1)-th measurement feature to obtain the overall point set.

[0046] (2) Under preset constraints, using the entire point set as input, the nearest point pair algorithm is used to determine the nearest point pair. The measurement point in the nearest point pair belonging to the measurement point set of the i-th measurement feature is taken as the last measurement point of the i-th measurement feature, and the measurement point in the nearest point pair belonging to the measurement point set of the (i+1)-th measurement feature is taken as the starting measurement point of the (i+1)-th measurement feature. The preset constraint is that only the distance between cross-group point pairs is calculated when determining the nearest point pair. The cross-group point pair with the closest distance is the nearest point pair. The cross-group point pair consists of one measurement point in the measurement point set of the i-th measurement feature and one measurement point in the measurement point set of the (i+1)-th measurement feature.

[0047] After determining the starting and ending measurement points of each measurement feature, this embodiment uses the ant colony algorithm to perform path planning for each measurement feature, resulting in a collision-free local path that traverses all measurement points in the measurement point set of the measurement feature and runs from the starting measurement point to the ending measurement point of the measurement feature.

[0048] This embodiment generates local paths based on an improved ant colony algorithm, which can achieve the shortest path optimization within the same measurement feature. To achieve this goal, the traditional ant colony algorithm has been improved (mainly in the path selection part). The specific algorithm flow after the improvement is as follows.

[0049] (1) Initialize the ant colony

[0050] First, the object characteristic management interface (CATISpecObject) is called to obtain the measurement features of different measurement points. Ant colonies are independently deployed on each measurement feature for computation, and the number of deployed ants is calculated based on the number of measurement points on that feature, following the principle that the more measurement points, the more ants. After this, the pheromone concentration is initialized to a constant.

[0051] (2) Path selection

[0052] Ants will start from the initial test point and select the next test point according to heuristic rules until they reach the final test point. Specifically, they will combine heuristic factors such as pheromone concentration, distance, and whether a collision occurs to make the decision on the next test point. The formula for calculating the transition probability is as follows:

[0053]

[0054] Where, p ij τ is the probability of moving from measurement point i to measurement point j; ij η is the pheromone concentration from measurement point i to measurement point j; α is the first parameter that adjusts the importance of pheromone concentration to path selection; η is the pheromone concentration from measurement point i to measurement point j. ij τ represents the heuristic information from measurement point i to measurement point j; β is a second parameter that adjusts the importance of the heuristic information for path selection; allowed is the set of the next measurement points that do not collide, meaning that the line connecting measurement point i and the next measurement point does not collide with the 3D model of the aerospace part; ik η represents the pheromone concentration from measurement point i to measurement point k. ik This represents the heuristic information from measurement point i to measurement point k; K is the total number of the next measurement point within the allowed range; R i,j Let be the probability coefficient from measurement point i to measurement point j.

[0055] η ik Defined using the reciprocal of the distance between two measuring points, as follows:

[0056]

[0057] Where, d ik Let be the distance between measuring point i and measuring point k.

[0058] R i,j Defined as:

[0059]

[0060] Here, "if a collision happens between i and j" means that the line connecting measurement points i and j collides with the 3D model of the aerospace part.

[0061] After all ants in an ant colony have completed their search, each ant generates a planned path. To improve the diversity of the ant colony and avoid getting trapped in local optima, the idea of ​​a selection function in a genetic algorithm is used. The probability of a planned path being selected is calculated by weighting the quality (also known as fitness, represented by the reciprocal of the path length) of each planned path, as follows:

[0062]

[0063] Where P(i) is the probability of path i being selected, representing the probability that a certain path is selected as the next generation; fitness(i) is the fitness of path i; N is the size of the ant colony, that is, the total number of ants, which is also the total number of paths; and fitness(j) is the fitness of path j.

[0064] Based on the selection probability, a certain number of planned paths are selected as selected paths using roulette wheel betting. Then, an even number of selected paths are selected as paths to be transformed, and the remaining selected paths are recorded as unselected paths. Finally, the selected paths to be transformed are intersected. Specifically, a crossover point is selected using roulette wheel betting. Then, the parts of the two paths to be transformed that are located after this crossover point are swapped to obtain two mutated paths. All unselected paths and all mutated paths are used as update paths to update pheromone concentration.

[0065] (3) Pheromone update

[0066] The pheromone concentration on the updated path is updated based on the quality of the updated path. More pheromones will be retained on better paths. The pheromone update is performed using the following formula:

[0067] τ ij =(1-ρ)·τ ij +Δτ ij ;

[0068] Where ρ is the pheromone evaporation coefficient, used to control the pheromone evaporation rate; Δτ ij The pheromone increment between measurement point i and measurement point j is calculated using the following formula:

[0069]

[0070] Where Q is the pheromone enhancement coefficient; L α and L β These are the total lengths of path α and path β, respectively. α represents the unselected path, and β represents the mutated path obtained through crossover and mutation operations.

[0071] (4) Iterate through the loop and update the path.

[0072] The path on the measured features is iterated repeatedly to eventually generate the optimal local path.

[0073] In this embodiment, ant colony optimization is used for path planning to obtain a collision-free local path that traverses all measurement points in the measurement feature set and runs from the starting measurement point to the ending measurement point of the measurement feature. Specifically, this includes:

[0074] (1) Initialize the ant colony and pheromone concentration. The ant colony consists of several ants.

[0075] (2) Place each ant at the starting point of the measurement feature. For each ant, continuously select the next measurement point according to the transition probability until all measurement points in the measurement feature's measurement point set are traversed and the last measurement point of the measurement feature is reached, generating a collision-free planning path.

[0076] (3) Determine whether the first maximum number of iterations has been reached.

[0077] (4) If so, stop the iteration and select the shortest planned path in the current iteration as the collision-free local path that traverses all measurement points in the measurement point set of the measurement feature and from the starting measurement point of the measurement feature to the ending measurement point of the measurement feature.

[0078] (5) If not, continue iterating, update the pheromone concentration based on all planned paths, and return to the step of "place each ant at the starting point of the measurement feature".

[0079] This includes updating pheromone concentrations based on all planned paths, specifically:

[0080] (1) Calculate the fitness of each planning path. For each planning path, calculate the probability of the planning path being selected based on the fitness of the planning path. The fitness is the reciprocal of the length of the planning path.

[0081] (2) Using the probability of all planned paths being selected as input, a portion of the planned paths are selected as the selected paths using roulette wheel betting.

[0082] (3) Randomly select an even number of selected paths as paths to be transformed, and record the remaining selected paths as unselected paths.

[0083] (4) Combine all the paths to be transformed in pairs to obtain several combinations.

[0084] (5) For each combination, take the two paths to be transformed in the combination as input, use roulette to select the intersection point, and swap the part of the two paths to be transformed in the combination that is located after the intersection point to obtain two variant paths.

[0085] (6) Update pheromone concentration based on all unselected paths and all mutated paths.

[0086] In this embodiment, for the i-th measurement feature and the (i+1)-th measurement feature, the RRT* algorithm is used for path planning to obtain a collision-free connection path from the end measurement point of the i-th measurement feature to the starting measurement point of the (i+1)-th measurement feature.

[0087] This embodiment uses an improved RRT* algorithm to generate obstacle avoidance paths across measurement features. For most measurement tasks, path collisions typically occur when crossing measurement features, such as... Figure 3 As shown, the improved RRT* algorithm first defines the state space, and forms nodes across measurement features by randomly sampling in the state space. Connecting these nodes forms the initial connection path across measurement features. Then, the initial connection path is locally smoothed to obtain a reasonable connection path across measurement features. The specific process is as follows.

[0088] (1) Initially form a collision-free path using the improved RRT* algorithm.

[0089] In three-dimensional space, a tree T is first constructed with the starting node as the root node, containing the set of new nodes generated during each iteration. The second maximum number of iterations N and the target node q are set. goal .

[0090] Next, within the specified feasible sampling region X = {min rand ,max rand Random sampling nodes within}, min rand ,max rand The minimum and maximum values ​​of the sampled nodes are shown below:

[0091]

[0092] Where, q rand p is the sampling node; p is the target node sampling rate, which is the probability of directly selecting the target node as the next node; RandomSample(X) represents random sampling within the feasible sampling area.

[0093] Determine whether the connection between the current node and the sampling node (including the case of directly sampling the target node) collides with the 3D model of the aerospace part. If yes, proceed to the next iteration; otherwise, further find the distance q from the sampling node in tree T. rand Nearest neighbor node q nearest ,as follows:

[0094]

[0095] Where q is a node in tree T.

[0096] Based on this, starting from the neighboring node q with a preset step size Δq nearest Along the direction of sampling node q rand Expand the tree T in a directional direction and generate a new node q. new ,as follows:

[0097]

[0098] Determine the new node q new With neighboring node q nearest If the connection q collides with the 3D model of the aerospace part, proceed to the next iteration; otherwise, change the new node q. new Add to tree T.

[0099] After this, examine the set of adjacent nodes N = {q ∈ T |||qq} within the search radius r. new Find the new node q among the nodes in ≤r||}. new The parent node, further finds the new node q new The optimal path for connection is determined, and it is then assessed whether the reselected optimal path is better, thus deciding whether to replace it. A better path requires that it be shorter and collision-free, and the parent nodes of the relevant nodes are updated. c is a constant used to adjust the growth rate of the search radius, n is the number of nodes in the tree, and d is the spatial dimension.

[0100] New node q new The parent node is determined as follows:

[0101]

[0102] Where, q min Let be the parent node; Cost(q) is the cost of node q. To accelerate path convergence, Euclidean distance is used as a heuristic function to constrain the reconnection process, and the cost function is constructed as follows:

[0103] Cost(q) = g(n) + h(n);

[0104] Where g(n) is the actual cost from the starting node to node q, and h(n) is the actual cost from node q to the target node. In this embodiment, Euclidean distance is used, as follows:

[0105]

[0106] h(n) = ||q k -q goal ||;

[0107] Where k is the index of node q, and q i Let q be the i-th node in the path. i-1 q is the (i-1)th node in the path; k That is, node q.

[0108] Repeat the iterations until the second maximum number of iterations N is reached, or until a node is found that is close enough to the target node without colliding with it.

[0109] (2) Preliminary local smoothing of collision-free paths

[0110] The initial collision-free path generated above is optimized. To shorten the probe's travel distance, a smoothing function is designed, such as... Figure 4 As shown, redundant points are first removed. The system iteratively checks whether the path after removing intermediate nodes still avoids collisions. If the requirement is met, the intermediate nodes are removed, and this process is repeated until no such intermediate nodes exist. Furthermore, local smoothing is used, attempting to replace two adjacent nodes with their midpoints. If no collision occurs, the midpoint is used to replace the two adjacent nodes.

[0111] In this embodiment, the RRT* algorithm is used for path planning to obtain a collision-free connection path from the end point of the i-th measurement feature to the starting point of the (i+1)-th measurement feature, specifically including:

[0112] (1) Use the RRT* algorithm to perform path planning and obtain a collision-free initial connection path from the end point of the i-th measurement feature to the starting point of the (i+1)-th measurement feature.

[0113] (2) Perform local smoothing on the initial connection path to obtain a collision-free connection path from the end point of the i-th measurement feature to the starting point of the (i+1)-th measurement feature.

[0114] Specifically, the RRT* algorithm is used for path planning to obtain a collision-free initial connection path from the last measurement point of the i-th measurement feature to the starting measurement point of the (i+1)-th measurement feature, which includes:

[0115] (1) Randomly sample nodes within the preset sampling area to obtain sample nodes, and determine whether the connection between the current node of the current iteration and the sample node collides with the 3D model of the aerospace parts; if so, increment the iteration count by 1 and return to the step of "randomly sample nodes within the preset sampling area" until the second maximum iteration count is reached; if not, find the nearest neighbor node to the sample node in the tree of the current iteration, and generate a new node based on the sample node and the neighbor node; where the current node of the current iteration is the last node in the partial generation path of the previous iteration.

[0116] (2) Determine whether the connection between the new node and its neighboring nodes collides with the 3D model of the aerospace parts; if so, increment the iteration count by 1 and return to the step of "randomly sampling nodes within the preset sampling area" until the second maximum iteration count is reached; if not, add the new node to the tree of the current iteration, obtain the tree of the next iteration, and determine the parent node of the new node in the tree of the current iteration.

[0117] (3) According to the parent-child node relationship, a new partial generation path from the starting node to the new node is obtained, and it is determined whether the new partial generation path is better. If so, the new partial generation path is used as the partial generation path of the current iteration, and the new node is used as the last node in the partial generation path of the current iteration. If not, the partial generation path of the previous iteration is used as the partial generation path of the current iteration, which is equivalent to retaining the path of the previous iteration and using the last node in the partial generation path of the previous iteration as the last node in the partial generation path of the current iteration. Wherein, the starting node is the last measurement point of the i-th measurement feature.

[0118] (4) Determine whether the connection between the last node in the partial generation path of the current iteration and the target node collides with the 3D model of the aerospace part; if yes, increment the iteration count by 1 and return to the step of "randomly sampling nodes in the preset sampling area" until the second maximum iteration count is reached; if no, end the iteration and connect the last node in the partial generation path of the current iteration with the target node to obtain a collision-free initial connection path from the end measurement point of the i-th measurement feature to the starting measurement point of the i+1-th measurement feature; where the target node is the starting measurement point of the i+1-th measurement feature.

[0119] Specifically, the initial connection path undergoes local smoothing to obtain a collision-free connection path from the end point of the i-th measurement feature to the starting point of the (i+1)-th measurement feature, which includes:

[0120] (1) Select any intermediate node in the initial connection path as the node to be judged; where the intermediate node is the node in the initial connection path other than the starting node and the target node.

[0121] (2) Determine whether the line connecting the previous node and the next node of the node to be judged collides with the 3D model of the aerospace parts; if not, remove the node to be judged, connect the previous node and the next node of the node to be judged to obtain a new initial connection path, and use the new initial connection path as the initial connection path for the next iteration, and return to the step of "selecting any intermediate node in the initial connection path as the node to be judged"; if yes, select the unselected intermediate node in the initial connection path as the node to be judged, and return to the step of "determining whether the line connecting the previous node and the next node of the node to be judged collides with the 3D model of the aerospace parts", until all intermediate nodes have been selected, and the connection path after preliminary processing is obtained.

[0122] (3) Select any two adjacent nodes in the connection path after preliminary processing as the node group to be judged.

[0123] (4) Determine the midpoint of the node group to be judged, and determine whether the line connecting the previous node of the node group to be judged and the midpoint collides with the 3D model of the aerospace part, or whether the line connecting the midpoint and the next node of the node group to be judged collides with the 3D model of the aerospace part; if not, remove the node group to be judged, connect the previous node of the node group to be judged and the midpoint, and connect the midpoint and the next node of the node group to be judged to obtain a new preliminary processing connection path, and use the new preliminary processing connection path as the preliminary processing connection path for the next iteration, and return to the step of "selecting any two adjacent nodes in the preliminary processing connection path as the node group to be judged"; if yes, select any two unselected adjacent nodes in the preliminary processing connection path as the node group to be judged, and return to the step of "determining the midpoint of the node group to be judged", until any two adjacent nodes have been selected, and obtain a collision-free connection path from the end measurement point of the i-th measurement feature to the start measurement point of the i+1-th measurement feature.

[0124] It should be noted that "previous" and "next" are defined according to the connection order from the starting node to the target node.

[0125] After obtaining the local paths and connection paths, this embodiment connects all local paths and all connection paths in the connection order to obtain a complete measurement path for in-flight measurement of the three-dimensional model of the aerospace parts.

[0126] This embodiment further adjusts the orientation of the measuring point, such as... Figure 5As shown, since the collision volume of the probe cannot be ignored, the spatial orientation of the original probe points was already defined during the probe point planning period, which is the direction of the probe entry. These directions were not affected during the path planning period and do not need to be adjusted. For newly generated probe points (i.e., intermediate nodes on the connecting path), based on the model design algorithm, the probe point orientation is set to the normal direction of the surface of the obstacle it crosses.

[0127] In this embodiment, after obtaining the complete measurement path, the path planning method for the on-board measurement process of aerospace parts further includes: setting the orientation of the nodes in the complete measurement path other than the measuring points to the normal direction of the surface of the obstacle being crossed.

[0128] After the above steps, a collision-free optimized path can be generated.

[0129] This embodiment uses a collision detection function to determine whether a collision has occurred. Specifically, it implements a collision detection algorithm based on the MBD model, such as... Figure 6 As shown, firstly, the MBD model (i.e., the 3D model of the aerospace part) is parsed to obtain the surface information, and all surface information is stored. Then, using the start and end points of the path to be detected (i.e., the line connecting two points), a line is drawn in 3D space. Next, all stored surfaces are traversed, and attempts are made to intersect them using the lines to find the intersection point. If an intersection point is successfully obtained, to eliminate the influence of other factors, it is necessary to reverse-check whether the intersection point is located on the model surface. If this condition is met, it is determined that the path to be detected has collided with the model; otherwise, it is determined that no collision has occurred. It should be noted that since the position coordinates of the model and the position coordinates of the path to be detected are known, the process of finding the intersection point and determining whether the intersection point is located on the model surface can be completed directly using geometric relationships.

[0130] Among them, the path to be detected is the line connecting the aforementioned measurement point o to the next measurement point, the line connecting the current node of the current iteration to the sampling node, the line connecting the new node to the neighboring node, the line connecting the last node in the partially generated path of the current iteration to the target node, the line connecting the previous node of the node to be judged to the next node of the node to be judged, the line connecting the previous node of the node group to be judged to the midpoint, and the line connecting the midpoint to the next node of the node group to be judged.

[0131] This embodiment proposes a path planning method for in-flight measurement of aerospace parts. Under the CATIA V5 environment, secondary development is performed using the Component Application Architecture (CAA) provided by CATIA. The Ant Colony Optimization (ACO) algorithm is combined with and improved upon the Rapid Expanding Random Tree (RRT*) algorithm to achieve globally efficient collision-free path generation. Its usability has been verified in a real-world project. After initial measurement point planning, measurement points crossing measurement features are preprocessed using the nearest point pair algorithm. Then, ants are placed on different measurement features for tracking, and the improved ACO algorithm is used for path planning to obtain local paths. Further, for connecting paths crossing measurement features, the improved RRT* algorithm is used for random sampling in the legal space, followed by local smoothing to form collision-free connecting paths across measurement features. Finally, the orientation of the newly generated measurement points is adjusted based on the model to form the final collision-free optimized path.

[0132] This embodiment enables automatic generation of collision-free optimized paths for coordinate measuring machines (CMMs), significantly improving generation efficiency and reducing operational complexity compared to traditional path generation and obstacle avoidance methods. Traditional methods for path generation directly sort global measurement points, resulting in low efficiency and poor optimization. This embodiment preprocesses measurement points across measurement features using a nearest-neighbor algorithm before path optimization, then places ants on each feature for tracking, greatly improving the algorithm's speed. For obstacle avoidance, collision detection is first implemented based on the MBD model. Furthermore, the traditional RRT* algorithm is extended to 3D space, and redundant points are removed, along with local smoothing to optimize the initially generated collision-free path. Finally, the orientation of newly generated measurement points is adjusted based on the MBD model to avoid collisions caused by neglecting the probe's collision volume.

[0133] This embodiment, within the CATIA V5 environment, utilizes the CAA secondary development tool to achieve the generation and optimization of collision-free measurement paths. Addressing the issues of excessively long detection times and the need for significant manual intervention in traditional CMM path planning methods, an advanced algorithm is employed for improvement. Previously, the order of measurement points was often insufficiently considered when forming measurement paths, directly using the planned order or employing ineffective traditional algorithms for optimization, resulting in substantial wasted measurement time. This embodiment uses an improved ACO (Automatic Optimization Coding) for measurement point sequence planning, generating shorter optimized paths. In handling path collisions, the traditional approach often employs a safety plane, involving manual intervention in processes such as reference plane selection and offset determination, hindering greater automation. This embodiment improves the RRT* algorithm, achieving automatic generation of obstacle-free paths through random sampling within legal space. Finally, considering probe collision volumes, the orientation of newly generated measurement points is adjusted to avoid volume collisions outside of path interference. This embodiment significantly improves the efficiency of the CMM inspection process, reduces manual intervention for greater automation, and contributes to the digitalization of aerial inspection.

[0134] This application also provides an application scenario in which the above-described path planning method for in-flight measurement of aerospace parts is applied. Specifically, the path planning method for in-flight measurement of aerospace parts provided in this embodiment can be applied to in-flight measurement scenarios for aerospace parts. The in-flight measurement scenario for aerospace parts includes a data acquisition stage, a path planning stage, and an in-flight measurement stage. The data acquisition stage is used to acquire the measurement point set of each measurement feature of the aerospace part's 3D model. The path planning stage is used to plan a path based on the measurement point set of each measurement feature to obtain a complete measurement path. The in-flight measurement stage is used to perform in-flight measurement of the aerospace part using a coordinate measuring machine based on the complete measurement path. The path planning method for in-flight measurement of aerospace parts provided in this embodiment belongs to the path planning stage.

[0135] Example 2

[0136] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 7As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When executed by the processor, the computer program implements a path planning method for in-flight measurement of aerospace parts.

[0137] Those skilled in the art will understand that Figure 7 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0138] In one exemplary embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program that, when executed by the processor, implements the path planning method for in-flight measurement of aerospace parts as described in Embodiment 1.

[0139] Example 3

[0140] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the path planning method for in-flight measurement processes of aerospace parts in Embodiment 1.

[0141] Example 4

[0142] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the path planning method for in-flight measurement processes of aerospace parts in Embodiment 1.

[0143] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0144] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0145] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A path planning method for in-flight measurement of aerospace parts, characterized in that, The path planning method for on-board measurement of aerospace parts includes: Obtain the measurement point set for each measurement feature of the 3D model of the aerospace part; the measurement feature is a line and / or surface of the 3D model of the aerospace part; the measurement point set includes multiple measurement points located on the measurement feature; Sort all measurement features according to their connection order. Using the measurement point set of the i-th measurement feature and the measurement point set of the (i+1)-th measurement feature as input, determine the last measurement point of the i-th measurement feature and the starting measurement point of the (i+1)-th measurement feature using the nearest point pair algorithm. Randomly select a measurement point other than the last measurement point of the i-th measurement feature as the starting measurement point of the i-th measurement feature. Randomly select a measurement point other than the starting measurement point of the i-th measurement feature as the last measurement point of the i-th measurement feature. This yields the starting and ending measurement points for each measurement feature. Here, i = 1, 2, ..., I-1, where I is the total number of measurement features. For each measurement feature, the ant colony algorithm is used for path planning to obtain a collision-free local path that traverses all measurement points in the measurement point set of the measurement feature and runs from the starting measurement point of the measurement feature to the ending measurement point of the measurement feature. For the i-th measurement feature and the (i+1)-th measurement feature, using The algorithm performs path planning to obtain a collision-free connection path from the last measurement point of the i-th measurement feature to the starting measurement point of the (i+1)-th measurement feature. By connecting all local paths and all connecting paths in the order of connection, a complete measurement path for in-flight measurement of the 3D model of the aerospace part is obtained.

2. The path planning method for on-board measurement of aerospace parts according to claim 1, characterized in that, Using the measurement point set of the i-th measurement feature and the measurement point set of the (i+1)-th measurement feature as input, the nearest point pair algorithm is used to determine the last measurement point of the i-th measurement feature and the starting measurement point of the (i+1)-th measurement feature, specifically including: Merge the measurement point set of the i-th measurement feature and the measurement point set of the (i+1)-th measurement feature to obtain the overall point set; Under preset constraints, the entire point set is used as input, and the nearest point pair algorithm is used to determine the nearest point pair. The measurement point in the nearest point pair belonging to the measurement point set of the i-th measurement feature is taken as the last measurement point of the i-th measurement feature, and the measurement point in the nearest point pair belonging to the measurement point set of the (i+1)-th measurement feature is taken as the starting measurement point of the (i+1)-th measurement feature. The preset constraint is that only the distance between cross-group point pairs is calculated when determining the nearest point pair. The cross-group point pair with the closest distance is the nearest point pair. The cross-group point pair consists of a measurement point in the measurement point set of the i-th measurement feature and a measurement point in the measurement point set of the (i+1)-th measurement feature.

3. The path planning method for on-board measurement of aerospace parts according to claim 1, characterized in that, Using the ant colony algorithm for path planning, a collision-free local path is obtained that traverses all measurement points in the measurement feature set and runs from the starting measurement point to the ending measurement point of the feature. Specifically, this path includes: Initialize the ant colony and pheromone concentration; the ant colony includes several ants; Each ant is placed at the starting point of the measurement feature. For each ant, the next measurement point is continuously selected according to the transition probability until all measurement points in the measurement feature's measurement point set are traversed and the last measurement point of the measurement feature is reached, generating a collision-free planning path. Determine whether the first maximum number of iterations has been reached; If so, stop the iteration and select the shortest planned path in the current iteration as a collision-free local path that traverses all measurement points in the measurement point set of the measurement feature and runs from the starting measurement point of the measurement feature to the ending measurement point of the measurement feature. If not, continue iterating, update the pheromone concentration based on all planned paths, and return to the step of "place each ant at the starting point of the measured feature".

4. The path planning method for on-board measurement of aerospace parts according to claim 3, characterized in that, The formula for calculating the transition probability is: ； in, To measure the point Transfer to measuring point The probability of; To measure the point to the measuring point pheromone concentration; The first parameter is used to adjust the importance of pheromone concentration to path selection; To measure the point to the measuring point Heuristic information; The second parameter is used to adjust the importance of heuristic information in path selection; The set of the next measurement points that will not collide; "not colliding" means that the measurement points... The line connecting to the next measurement point does not collide with the 3D model of the aerospace part; To measure the point to the measuring point pheromone concentration; To measure the point to the measuring point Heuristic information; for The total number of the next measurement points within; To measure the point to the measuring point The probability coefficient; The pheromone concentration is updated based on all planned paths, specifically including: Calculate the fitness of each planning path separately, and for each planning path, calculate the probability of the planning path being selected based on the fitness of the planning path; the fitness is the reciprocal of the length of the planning path. Using the probability of all planned paths being selected as input, a subset of planned paths are selected as the chosen paths using a roulette wheel. Randomly select an even number of the selected paths as the paths to be transformed, and record the remaining selected paths as unselected paths; Combine all the paths to be transformed in pairs to obtain several combinations; For each combination, the two paths to be transformed within the combination are used as input. The intersection point is selected using a roulette wheel, and the parts of the two paths to be transformed within the combination that are located after the intersection point are swapped to obtain two mutated paths. Update pheromone concentrations based on all unselected paths and all mutated paths.

5. The path planning method for on-board measurement of aerospace parts according to claim 1, characterized in that, use The algorithm performs path planning to obtain a collision-free connection path from the last measurement point of the i-th feature to the starting measurement point of the (i+1)-th feature, specifically including: use The algorithm performs path planning to obtain a collision-free initial connection path from the last measurement point of the i-th measurement feature to the starting measurement point of the (i+1)-th measurement feature. The initial connection path is locally smoothed to obtain a collision-free connection path from the end point of the i-th measurement feature to the starting point of the (i+1)-th measurement feature.

6. The path planning method for on-board measurement of aerospace parts according to claim 5, characterized in that, use The algorithm performs path planning to obtain a collision-free initial connection path from the last measurement point of the i-th feature to the starting measurement point of the (i+1)-th feature, specifically including: Randomly sample nodes within a preset sampling area to obtain sampled nodes, and determine whether the connection between the current node of the current iteration and the sampled node collides with the 3D model of the aerospace part; if so, increment the iteration count by 1 and return to the step of "randomly sample nodes within the preset sampling area" until the second maximum iteration count is reached; if not, find the nearest neighbor node to the sampled node in the tree of the current iteration, and generate a new node based on the sampled node and the neighbor node; where the current node of the current iteration is the last node in the partial generation path of the previous iteration; Determine whether the connection between the new node and its neighboring nodes collides with the 3D model of the aerospace part; if so, increment the iteration count by 1 and return to the step of "randomly sampling nodes within the preset sampling area" until the second maximum iteration count is reached; if not, add the new node to the tree of the current iteration, obtain the tree of the next iteration, and determine the parent node of the new node in the tree of the current iteration. Based on the parent-child node relationship, a new partial generation path is obtained from the starting node to the new node, and it is determined whether the new partial generation path is better. If so, the new partial generation path is used as the partial generation path of the current iteration, and the new node is used as the last node in the partial generation path of the current iteration. If not, the partial generation path of the previous iteration is used as the partial generation path of the current iteration, and the last node in the partial generation path of the previous iteration is used as the last node in the partial generation path of the current iteration. Here, the starting node is the last measurement point of the i-th measurement feature. Determine whether the connection between the last node in the current iteration's partially generated path and the target node collides with the 3D model of the aerospace part. If yes, increment the iteration count by 1 and return to the step of "randomly sampling nodes within the preset sampling area" until the second maximum iteration count is reached. If no, end the iteration and connect the last node in the current iteration's partially generated path with the target node to obtain a collision-free initial connection path from the end point of the i-th measurement feature to the starting point of the (i+1)-th measurement feature. The target node is the starting point of the (i+1)-th measurement feature. The initial connection path is locally smoothed to obtain a collision-free connection path from the end point of the i-th measurement feature to the starting point of the (i+1)-th measurement feature, specifically including: Select any intermediate node in the initial connection path as the node to be judged; where the intermediate node is any node in the initial connection path other than the starting node and the target node; Determine whether the line connecting the previous node and the next node of the node to be judged collides with the 3D model of the aerospace part. If not, remove the node to be judged, connect the previous node and the next node of the node to be judged to obtain a new initial connection path, and use the new initial connection path as the initial connection path for the next iteration, returning to the step of "selecting any intermediate node in the initial connection path as the node to be judged". If yes, select the unselected intermediate node in the initial connection path as the node to be judged, and return to the step of "determining whether the line connecting the previous node and the next node of the node to be judged collides with the 3D model of the aerospace part", until all intermediate nodes have been selected, and the preliminary processed connection path is obtained. Select any two adjacent nodes in the connection path after preliminary processing as the node group to be judged; Determine the midpoint of the node group to be judged. Determine whether the line connecting the previous node of the node group to be judged and the midpoint collides with the 3D model of the aerospace part, or whether the line connecting the midpoint and the next node of the node group to be judged collides with the 3D model of the aerospace part. If not, remove the node group to be judged, connect the previous node of the node group to the midpoint, and connect the midpoint to the next node of the node group to be judged to obtain a new preliminary processing connection path. Use the new preliminary processing connection path as the preliminary processing connection path for the next iteration, and return to the step of "selecting any two adjacent nodes in the preliminary processing connection path as the node group to be judged". If yes, select any two unselected adjacent nodes in the preliminary processing connection path as the node group to be judged, and return to the step of "determining the midpoint of the node group to be judged", until any two adjacent nodes have been selected, resulting in a collision-free connection path from the end point of the i-th measurement feature to the start point of the (i+1)-th measurement feature.

7. The path planning method for on-board measurement of aerospace parts according to claim 1, characterized in that, After obtaining the complete measurement path, the path planning method for the on-board measurement process of aerospace parts further includes: setting the orientation of the nodes in the complete measurement path, excluding the measurement points, to the normal direction of the surface of the obstacle being crossed.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the path planning method for in-flight measurement of aerospace parts according to any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the path planning method for in-flight measurement of aerospace parts as described in any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the path planning method for in-flight measurement of aerospace parts as described in any one of claims 1-7.

Images