Multi-element wire harness path optimization method in three-dimensional space

Abstract

The invention provides a multi-element wire harness path optimization method in a three-dimensional space, and aims at solving the problems that in the prior art, a loop is likely to be generated in path planning, and path interference is unreasonable. The optimization process comprises the steps that firstly, an elastic Astar path planning algorithm based on a union-check set is adopted, and it is ensured that no loop is generated in the path planning process; then, a BVH-tree (Bounding Volume Hierarchy Based On Tree, a hierarchical bounding box tree) is applied to implement efficient collision detection, and a cubic B-spline curve is constructed to simulate the state of the wire harness; and finally, optimizing a path penalty function by introducing an adaptive elastic energy mechanism. Through algorithm iteration processing, loop-free, interference-free and optimal automatic planning of a wire harness path in a three-dimensional space is realized; a numerical test verifies the effectiveness of the method in the field of three-dimensional wire harness path optimization.

Description

Technical Field

[0001] The present invention belongs to the technical field of path optimization, and in particular relates to a multi-element wire harness path optimization method in three-dimensional space. Background Art

[0002] The wiring harness is an important channel for transmitting signals and energy in electronic systems. The quality of the wiring harness path layout will directly affect the performance of the overall system. The wiring harness layout requires a wiring harness topology and routing path that meet engineering constraints. According to the connection relationship between the wiring harness and the port, determine the starting point and end point of the path, and then determine the wiring space. The wiring space is usually divided into feasible space and infeasible space. The feasible space refers to the area where the wiring harness is allowed to pass; the infeasible space refers to the area where the wiring harness is not allowed to pass due to obstacles, high temperature, inability to install, etc.

[0003] Wire harness path planning can be viewed as an optimization problem, which comprehensively weighs the process constraints of each path to determine the wiring optimization target. Under the wiring specifications such as collision interference constraints, minimum turning radius constraints, and wire harness laying interval requirements, the wire harness path with the lowest total process cost and the shortest total length is found to complete the wire harness path planning from a single starting point to a single end point.

[0004] In addition, the problem can be viewed as a minimum cost flow problem, with the starting point of the harness as the source point of the minimum cost flow network, the end point of the harness as the sink point, and the clamp as the middle point. The surface distance between any two points in three-dimensional space is abstracted into the distance between two vertices in the plane graph to construct a minimum cost flow network. Each clamp point is limited by the maximum clamp capacity, and the harness usage from the source point to the sink point must be equal.

[0005] The Edmonds–Karp (EK) algorithm and the Dinic algorithm are often used in the wiring harness path planning problem based on the minimum cost flow. The above methods can obtain more accurate results when the wiring harness path planning task is relatively simple, but the complex wiring harness laying planning problem has high requirements on data accuracy, especially when planning constraints for complex problems, there are problems such as large amount of calculation and difficulty in solving. Summary of the invention

[0006] Purpose of the invention: In view of the above-mentioned harness path optimization problem, the present invention provides a multivariate harness path optimization method in three-dimensional space. Under the wiring specifications such as collision interference constraints, minimum turning radius constraints, harness laying interval requirements, etc., find the harness path with the lowest total process cost and the shortest total length. The present invention adopts the elastic Astar path planning algorithm based on union-find set, and checks whether each point on the path belongs to the generated tree structure through union-find set to avoid the formation of loops in the harness path; then, the harness is collided with using BVH-tree (Bounding Volume Hierarchy Based On Tree, hierarchical bounding box tree), and a cubic B-spline curve is established to simulate the harness state; finally, adaptive elastic energy is set to improve the harness path penalty function. Through algorithm iteration, the present invention realizes automatic planning of harness paths without loops, interference, and optimal paths in three-dimensional space.

[0007] The present application provides a multi-element wire harness path optimization method in three-dimensional space, the method comprising the following steps:

[0008] Step 1: Based on the existing clamp positions in three-dimensional space as the geometric constraints of the harness at this location, read the harness wiring relationship file and wiring parameters, use the priority queue to process the clamp position information, use the Astar path planning algorithm to search for the shortest path, and check the wiring path data through the union-find set to avoid loops in the path structure;

[0009] Step 2 uses a complex harness collision detection algorithm based on the BVH-tree of the harness flexible annular surface and establishes a cubic spline curve to more accurately simulate the cable state, avoid path interference, and ensure that the bending radius meets the requirements;

[0010] Step 3 uses an angle limitation algorithm based on adaptive elastic energy to perform flexible angle penalties, reduce the number of bad turns in step 2, automatically select the method with the least turns, and determine the optimal path.

[0011] Preferably, in step 1, the Astar algorithm is used to search for the shortest path of the harness, starting from the starting end of the harness, taking the distance from the current clamp to the end point as a heuristic function, and giving priority to the point closest to the end point of the path, so as to improve the efficiency of the algorithm.

[0012] Preferably, when searching with the Astar algorithm, the wiring harness path is checked with the tree structure planning of the existing wiring harnesses by union-find, so as to avoid loops.

[0013] Preferably, in step 2, a complex cable collision detection algorithm based on a flexible annular surface BVH-tree is used.

[0014] Preferably, BVH-tree is a data structure used for collision detection.

[0015] Preferably, the BVH-tree divides the triangular facets on the aircraft model into two halves according to the x, y, and z axis coordinates in turn to form a binary tree structure. When performing collision detection, the triangular facets that may collide can be found in logarithmic time.

[0016] Preferably, when detecting whether the cable collides with the model, a cubic B-spline curve is established to more accurately simulate the cable, estimate whether the bending radius is within the limit, and use it as a basis for collision detection.

[0017] Preferably, in step 3, an adaptive elastic energy angle limitation algorithm is used for judgment and adjustment, a flexible angle penalty is proposed, and the path planning adaptive elastic energy is set. When routing, by setting and adjusting the value of the adaptive elastic energy M, the algorithm will automatically select the method with the least turns.

[0018] The beneficial effects of the present invention are:

[0019] The present invention adopts a union-find-based elastic Astar path planning algorithm to find the shortest path of the harness and avoid loops in the harness path. Then, BVH-tree (Bounding Volume Hierarchy Based On Tree) is used to perform collision detection on the harness and simulate the harness state. Finally, adaptive elastic energy is set to improve the harness path penalty function, thereby achieving automatic planning of the harness path without loops, interference, and optimal path in three-dimensional space.

[0020] The priority queue data structure is used to reduce the time complexity of planning the path using the Astar algorithm; the present invention checks the wiring path data by a union-find set to avoid loops in the path structure; the present invention establishes a cubic spline curve by using a complex harness collision detection algorithm of the BVH-tree of the harness flexible annular surface, more accurately simulates the cable state, and improves the quality compliance of harness path planning; the present invention uses an adaptive elastic energy angle restriction algorithm to perform flexible angle penalties, automatically select the method with the least turns, and determine the optimal path. The present invention can automatically and quickly find the optimal path for the harness when the harness planning constraints are complex through algorithm iteration. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 The flowchart of the method of the present invention is: a process adopted by a multi-element wire beam path optimization method in a three-dimensional space.

[0022] Figure 2 It is a test diagram of the method of the present invention: a cable path diagram obtained by using a multi-element wire harness path optimization method in a three-dimensional space. DETAILED DESCRIPTION

[0023] The present invention provides a multivariate wire harness path optimization method in three-dimensional space, aiming to solve the problem of loops and unreasonable path interference in path planning in the prior art; the optimization process includes: first, adopting the elastic Astar path planning algorithm based on union-find set to ensure that no loops are generated during the path planning process; then, using BVH-tree (Bounding Volume Hierarchy Based On Tree, hierarchical bounding box tree) to implement efficient collision detection, and constructing a cubic B-spline curve to simulate the wire harness state; finally, by introducing an adaptive elastic energy mechanism, the path penalty function is optimized. After algorithm iterative processing, the present invention realizes automatic planning of the wire harness path in three-dimensional space without loops, interference and optimal path; numerical tests verify the effectiveness of this method in the field of three-dimensional wire harness path optimization.

[0024] The present application provides a multi-element wire harness path optimization method in three-dimensional space, including the following technical means:

[0025] Step 1: Based on the existing clamp positions in three-dimensional space as the geometric constraints of the harness at this location, read the harness wiring relationship file and wiring parameters, use the priority queue to process the clamp position information, use the Astar path planning algorithm to search for the shortest path, and check the wiring path data through the union-find set to avoid loops in the path structure;

[0026] Step 2 uses a complex harness collision detection algorithm based on the harness flexible annular surface BVH-tree and establishes a cubic spline curve to more accurately simulate the cable state, avoid path interference, and ensure that the bending radius meets the requirements;

[0027] Step 3 uses an angle limitation algorithm based on adaptive elastic energy to perform flexible angle penalties, reduce the number of bad turns in step 2, automatically select the method with the least turns, and determine the optimal path.

[0028] Through the above steps, an optimal wiring harness path with no loops, no interference, and the minimum number of turns can be quickly found to complete the wiring of the wiring harness in three-dimensional space.

[0029] In the above technical solution, step 1 is specifically as follows:

[0030] According to the existing clamp positions in three-dimensional space as the geometric constraints of the harness here, the clamps are regarded as nodes in graph theory, and the node information is formed into a priority queue data structure; after reading the harness wiring relationship file, according to the clamp position data, the Astar algorithm is used to search for the shortest path of the harness, and the search is carried out from the starting end of the harness. The distance from the current clamp to the end point is used as a heuristic function, and the point closest to the end point of the path is processed first, so as to improve the efficiency of the algorithm. Since loops are not allowed when the harness is bundled, in order to avoid loops in the same harness path, when the Astar algorithm is used for search, each harness in the wiring relationship file is planned in turn. Assuming that the previous harnesses have formed a tree structure, when a new harness is added, each point on the harness path is checked by a union-find set to see if it belongs to the generated tree structure. If there are two or more intersections between this harness and the previous tree structure, the path that intersects with the tree structure is planned according to the original tree structure to avoid loops.

[0031] Furthermore, step 2 is as follows:

[0032] The cable has a flexible annular surface. A complex cable collision detection algorithm based on the flexible annular surface BVH-tree is proposed. BVH-tree is used for collision detection. For the harness path that does not meet the collision detection requirements, the clamp path is abandoned.

[0033] BVH-tree is a data structure used for collision detection. It divides the triangular facets on the aircraft model into two halves according to the x, y, and z axis coordinates in turn to form a binary tree structure. When performing collision detection, it can find the triangular facets that may collide in logarithmic time.

[0034] When detecting whether the harness collides with the physical model, four control points are constructed through information such as the starting point clamp, the end point clamp and its direction, and the lines between the control points are used as the basis for collision detection.

[0035] Read in the wiring parameter information to establish a cubic B-spline curve. The cubic B-spline curve can roughly estimate whether the bending radius is within the limit, simulate the cable status more accurately, estimate whether the bending radius is within the limit, and ensure that the bending radius of the harness path meets the wiring requirements.

[0036] Furthermore, the step 3 is specifically as follows:

[0037] According to the aircraft wiring specifications, the wiring harness path planning should meet the direction requirements, and the adaptive elastic energy angle limit algorithm is used for judgment and adjustment. First, the algorithm determines whether the wiring harness path has a turn, and then determines whether its turn angle is along the x-axis direction or perpendicular to the x-axis direction. Secondly, by taking the average of the y and z coordinates of all clamps, an approximate centerline of the aircraft can be obtained. For wiring near the center of the cabin, it is required to be along the y and z directions. If it is not in this direction, the turn should be changed, and the path planning gives priority to the path with fewer bad turns.

[0038] Since each path in the harness path optimization process involves multiple dimensions of data information such as distance, number of turns, number of bad turns, penalty value for excessive turns, number of consecutive turns, and the shortest distance between two turns, in order to ensure the accuracy of the algorithm, a flexible angle penalty is proposed, and the path planning adaptive elastic energy is set, and the angle between the path of the harness on a certain section and the direction along the x-axis or perpendicular to the x-axis is used as the parameter of the weight function of the path planning. The relationship between the penalty function f and the adaptive elastic energy M is as follows:

[0039]

[0040] Where M represents adaptive elastic energy, L represents path length, θ is the angle between the routing direction and the standard direction (along the x-axis or perpendicular to the x-axis), α is the angle between the routing direction and the clamp direction, and μ represents the debugging parameter.

[0041] If the algorithm routes all along the x-axis or perpendicular to the x-axis, then the total degree of multiple turns will inevitably be minimized, which is considered to be the route with the least turns. When routing according to the algorithm, by setting and adjusting the value of the adaptive elastic energy M, the algorithm will automatically choose the route with the least turns.

[0042] In other embodiments of the present application, Figure 1 -Attached Figure 2 The present invention is further described.

[0043] In view of the above-mentioned harness path optimization problem, the present invention provides a multivariate harness path optimization method in three-dimensional space. Under the wiring specifications such as collision interference constraints, minimum turning radius constraints, harness laying interval requirements, etc., find the harness path with the lowest total process cost and the shortest total length. The present invention adopts the elastic Astar path planning algorithm based on union-find set, and checks whether each point on the path belongs to the generated tree structure through union-find set to avoid the formation of loops in the harness path; then, the harness is collided with using BVH-tree (Bounding Volume Hierarchy Based On Tree, hierarchical bounding box tree), and a cubic B-spline curve is established to simulate the harness state; finally, adaptive elastic energy is set to improve the harness path penalty function. The present invention realizes automatic planning of harness paths without loops, interference, and optimal paths in three-dimensional space through algorithm iteration.

[0044] The technical solution adopted by the present invention is as follows:

[0045] A multi-element harness path optimization method in three-dimensional space includes the following technical means:

[0046] Step 1: Based on the existing clamp positions in three-dimensional space as the geometric constraints of the harness at this location, read the harness wiring relationship file and wiring parameters, use the priority queue to process the clamp position information, use the Astar path planning algorithm to search for the shortest path, and check the wiring path data through the union-find set to avoid loops in the path structure;

[0047] Step 2 uses a complex harness collision detection algorithm based on the harness flexible annular surface BVH-tree and establishes a cubic spline curve to more accurately simulate the cable state, avoid path interference, and ensure that the bending radius meets the requirements;

[0048] Step 3 uses an angle limitation algorithm based on adaptive elastic energy to perform flexible angle penalties, reduce the number of bad turns in step 2, automatically select the method with the least turns, and determine the optimal path.

[0049] Through the above steps, an optimal wiring harness path with no loops, no interference, and the minimum number of turns can be quickly found to complete the wiring of the wiring harness in three-dimensional space.

[0050] In the above technical solution, step 1 is specifically as follows:

[0051] According to the existing clamp positions in three-dimensional space as the geometric constraints of the harness here, the clamps are regarded as nodes in graph theory, and the node information is formed into a priority queue data structure; after reading the harness wiring relationship file, according to the clamp position data, the Astar algorithm is used to search for the shortest path of the harness, and the search is carried out from the starting end of the harness. The distance from the current clamp to the end point is used as a heuristic function, and the point closest to the end point of the path is processed first, so as to improve the efficiency of the algorithm. Since loops are not allowed when the harness is bundled, in order to avoid loops in the same harness path, when the Astar algorithm is used for search, each harness in the wiring relationship file is planned in turn. Assuming that the previous harnesses have formed a tree structure, when a new harness is added, each point on the harness path is checked by a union-find set to see if it belongs to the generated tree structure. If there are two or more intersections between this harness and the previous tree structure, the path that intersects with the tree structure is planned according to the original tree structure to avoid loops.

[0052] Furthermore, step 2 is as follows:

[0053] The cable has a flexible annular surface. A complex cable collision detection algorithm based on the flexible annular surface BVH-tree is proposed. BVH-tree is used for collision detection. For the harness path that does not meet the collision detection requirements, the clamp path is abandoned.

[0054] BVH-tree is a data structure used for collision detection. It divides the triangular facets on the aircraft model into two halves according to the x, y, and z axis coordinates in turn to form a binary tree structure. When performing collision detection, it can find the triangular facets that may collide in logarithmic time.

[0055] When detecting whether the harness collides with the physical model, four control points are constructed through information such as the starting point clamp, the end point clamp and its direction, and the lines between the control points are used as the basis for collision detection.

[0056] Read in the wiring parameter information to establish a cubic B-spline curve. The cubic B-spline curve can roughly estimate whether the bending radius is within the limit, simulate the cable status more accurately, estimate whether the bending radius is within the limit, and ensure that the bending radius of the harness path meets the wiring requirements.

[0057] Furthermore, the step 3 is specifically as follows:

[0058] According to the aircraft wiring specifications, the wiring harness path planning should meet the direction requirements, and the adaptive elastic energy angle limit algorithm is used for judgment and adjustment. First, the algorithm determines whether the wiring harness path has a turn, and then determines whether its turn angle is along the x-axis direction or perpendicular to the x-axis direction. Secondly, by taking the average of the y and z coordinates of all clamps, an approximate centerline of the aircraft can be obtained. For wiring near the center of the cabin, it is required to be along the y and z directions. If it is not in this direction, the turn should be changed, and the path planning gives priority to the path with fewer bad turns.

[0059] Since each path in the harness path optimization process involves multiple dimensions of data information such as distance, number of turns, number of bad turns, penalty value for excessive turns, number of consecutive turns, and the shortest distance between two turns, in order to ensure the accuracy of the algorithm, a flexible angle penalty is proposed, and the path planning adaptive elastic energy is set, and the angle between the path of the harness on a certain section and the direction along the x-axis or perpendicular to the x-axis is used as the parameter of the weight function of the path planning. The relationship between the penalty function f and the adaptive elastic energy M is as follows:

[0060]

[0061] Where M represents adaptive elastic energy, L represents path length, θ is the angle between the routing direction and the standard direction (along the x-axis or perpendicular to the x-axis), α is the angle between the routing direction and the clamp direction, and μ represents the debugging parameter.

[0062] If the algorithm routes all along the x-axis or perpendicular to the x-axis, then the total degree of multiple turns will inevitably be minimized, which is considered to be the route with the least turns. When routing according to the algorithm, by setting and adjusting the value of the adaptive elastic energy M, the algorithm will automatically choose the route with the least turns.

[0063] In other embodiments of the present application, the present application provides a multi-element wire harness path optimization method in three-dimensional space, comprising the following steps:

[0064] Step 1

[0065] Read in the existing clamp position information in three-dimensional space as the geometric constraint of the harness at this location.

[0066] Read the wiring harness relationship file as the starting and ending information of the wiring harness path planning.

[0067] The clamps are considered as nodes in graph theory, and the node information is formed into a priority queue data structure.

[0068] Search from the starting end of the harness, use the distance from the current clamp to the end point as a heuristic function, and give priority to the point closest to the end point of the path.

[0069] The Astar path planning algorithm is used to search for the shortest path. The A* (A-Star) algorithm is the most effective direct search method for solving the shortest path in a static road network. As a heuristic algorithm, it uses the evaluation function as the search constraint and finds the optimal path by continuously searching and approaching the destination node.

[0070] The specific implementation of the A* algorithm relies on two queues: the open queue and the closed queue. The open queue stores nodes waiting to be examined, and the closed queue stores nodes that have been examined and have relatively minimal costs. These nodes do not need to be checked again. Specifically, the pathfinder observes several adjacent nodes from the starting point, adds these points to the open queue, estimates the distance cost in turn, selects the node with the lowest estimated value to add to the closed queue, defines the forward and backward pointing relationship of the nodes, and moves to this node, and then repeats the above process until it reaches the target node. The final path is the process of starting from the target node and moving back to the starting node along the moving direction of each node's parent node.

[0071] The calculation formula of the current node n's valuation function f(n) is

[0072] f(n)=g(n)+h(n)

[0073] In the formula, the actual cost of the move is g(n) and the estimated cost is h(n). Among them, f(n) is the evaluation function from the initial point to the target point via node n, and g(n) is the actual cost from the initial node to the n node in the state space, which is called the actual cost of the move. h(n) is the estimated cost of the shortest path from n to the target node, also known as the estimated cost. Generally speaking, the larger the value of h(n), the greater the heuristic information, which can reduce the number of nodes expanded during the search process, thereby speeding up the search.

[0074] In order to meet the requirements of various cable laying processes, the objective function in the A* algorithm is improved, and process evaluation parameters such as the minimum bending radius bend(n), circuit electromagnetic isolation emc(n), and high temperature area cable protection cover(n) are added. Calculate the length of each path to form the total process cost of the cable path f(n), which is expressed as follows:

[0075] f(n)=g(n)+h(n)+bend(n)+emc(n)+cover(n)

[0076] The cable process cost is taken as the optimization target, the discretized wiring space environment is used as the path search queue, and the A* heuristic path search optimization algorithm is used to generate the wiring path of the single-source cable. In order to improve the accuracy of collision interference detection during the search process, a cubic non-uniform B-spline curve is used to preliminarily simulate the geometric shape of the cable in sections, and more refined discrete points are selected on the wiring space surface as the control points of the cable spline curve of this section to refine the preliminary shape of the cable path planning.

[0077] Check the routing path data through union-find to avoid loops in the path structure;

[0078] Since loops are not allowed when the harnesses are bundled, in order to avoid loops in the same harness path, when using the Astar algorithm to search, each harness in the wiring relationship file is path-planned in turn. Assuming that the previous harnesses have already formed a tree structure, when a new harness is added, a union-find set is used to check whether each point on the harness path belongs to the generated tree structure. If this harness has two or more intersections with the previous tree structure, the path that intersects with the tree structure is planned according to the original tree structure to avoid loops.

[0079] According to step 1, the next clamp set that satisfies the current clamp to avoid loops and the shortest path can be obtained.

[0080] Step 2

[0081] According to the set of next clamps to which the current clamp of the wiring harness can be connected obtained in step 1, an interference check is performed on the path formed by each clamp, and the optimal path clamp point is further determined.

[0082] When routing a harness within a routing area, the geometry model needs to be checked for three types of collisions: ① collisions between the harness and aircraft components; ② collisions between different branches of the same harness; and ③ collisions between the harness being routed and any previously generated harness.

[0083] A complex wire harness collision detection algorithm based on the wire harness flexible annular surface BVH-tree is adopted, and collision detection is performed using BVH-tree.

[0084] BVH-tree is a data structure used for collision detection. It divides the triangular facets on the aircraft model into two halves according to the x, y, and z axis coordinates in turn to form a binary tree structure. When performing collision detection, it can find the triangular facets that may collide in logarithmic time.

[0085] When detecting whether the harness collides with the physical model, four control points are constructed through information such as the starting point clamp, the end point clamp and its direction, and the lines between the control points are used as the basis for collision detection.

[0086] The bending radius of the wiring refers to the distance from the center of the circle to the inner surface of the bend. To prevent cable damage, the bending radius must not exceed the allowable range. The minimum allowable bending radius of the wiring harness is determined by the product of its diameter and the allowable bending radius ratio, that is,

[0087]

[0088] Where D bundle is the outer diameter of the bundle segment, ζ is the allowable bending radius ratio, which mainly depends on the bundle material.

[0089] Read in the wiring parameter information and establish a cubic B-spline curve. The cubic B-spline curve can roughly estimate whether the bending radius is within the limit, simulate the cable status more accurately, estimate whether the bending radius is within the limit, and ensure that the bending radius of the harness path meets the wiring requirements.

[0090] According to step 2, the next clamp set can be obtained, in which the current clamp satisfies the requirements of avoiding collision and the bending radius meets the requirements.

[0091] Step 3

[0092] According to the aircraft wiring specifications, the wiring harness path planning should meet the direction requirements. For the set of clamps that meet step 2, it is necessary to add angle judgment and use the angle limitation algorithm of adaptive elastic energy for judgment and adjustment.

[0093] If the cable path turns, it will determine whether the angle is along the x-axis or perpendicular to the x-axis. If it is not in this direction, it is considered a bad turn, and the path planning will give priority to the path with fewer bad turns. The y and z coordinates of all clamps are averaged to obtain an approximate centerline of the aircraft. For the routing of the center attachment of the cabin, it is required to be along the y and z directions. In addition, the algorithm gives appropriate penalties based on the angle of deviation.

[0094] With the gradual advancement of cable path optimization methods, each path has multiple dimensions of data information, such as distance, number of turns, number of bad turns, penalty value for excessive turns, number of consecutive turns, shortest distance between two turns, etc. The simplest method is to compare these dimensional information according to priority. However, cable path optimization involves many dimensions, which not only increases the complexity of comparison, but also requires manual determination of priorities, which also reduces the accuracy of the algorithm to a certain extent. To this end, a flexible angle penalty is proposed, which takes the angle between the path of the harness on a certain section and the standard direction into the weight function of path planning. The weight function f is a two-dimensional function of the angle θ between the routing direction and the standard direction and the length L of the path segment. Using an exponential function for penalty is a common method. The contribution of the length L is large at first and then small. Taking the logarithm of the length L, the contribution of the angle is basically linear, and the contribution of the length and angle are multiplicative. The expression is as follows:

[0095] f=eθ*ln(L)=L θ

[0096] Considering that the value of θ is too small, and taking the weight function f and the path length L into consideration, we get the following formula:

[0097] f=L 5θ+1

[0098] Considering the correlation between the routing direction and the clamp direction, let the angle between the routing direction and the clamp direction be α, θ and α are two independent angles, and define the single-segment routing metric function as follows:

[0099]

[0100] Add α to the original position of θ, and the penalty function f is expressed as follows:

[0101]

[0102] In fact, the algorithm did not achieve better results using this penalty function. Compared with the original penalty function, the result changes relatively little, which shows that the status of α and θ is different. θ is the angle between the routing direction and the standard direction. The longer the routing, the worse the cable path. But α is different. The shorter L is, the more obvious the distortion effect caused by α. If it is a very long line, even if there is a clamp in the middle of the cable that is perpendicular to the routing direction, then the sufficient length can straighten the line, so it is not so obvious visually; however, once the length of the line is too short, the distortion is visually catastrophic.

[0103] Therefore, the longer L is, the greater the weight provided by θ; the shorter L is, the greater the weight provided by α. Therefore, considering α and θ separately, we get the following expression:

[0104]

[0105] Where M represents adaptive elastic energy. The larger M is, the more the algorithm prioritizes reducing α. After a lot of debugging, M is given a larger value. If the algorithm routes all along the standard direction, then there must be a routing method that minimizes the total degree of multiple turns, which is considered to be the method with the least turns. Even if the adaptive elastic energy M is set to a large value, all methods with the least number of turns need to receive the same penalty on a certain section of the cable, so it will not affect the result. In addition, if the adaptive elastic energy M is set to a large value, the algorithm will automatically choose the method with the least turns when routing.

[0106] According to step 3, the optimal clamp path that meets the multi-wire harness constraint conditions can be obtained. The application case effect is shown in the following figure. Figure 2 .

Claims

1. A multi-element beam path optimization method in three-dimensional space, characterized in that: The method comprises the following steps: Step 1: Based on the existing clamp positions in three-dimensional space as the geometric constraints of the harness at this location, read the harness wiring relationship file and wiring parameters, use the priority queue to process the clamp position information, use the Astar path planning algorithm to search for the shortest path, and check the wiring path data through the union-find set to avoid loops in the path structure; Step 2 uses a complex harness collision detection algorithm based on the BVH-tree of the harness flexible annular surface and establishes a cubic spline curve to more accurately simulate the cable state, avoid path interference, and ensure that the bending radius meets the requirements; Step 3 uses an angle limitation algorithm based on adaptive elastic energy to perform flexible angle penalties, reduce the number of bad turns in step 2, automatically select the method with the least turns, and determine the optimal path.

2. The method according to claim 1, characterized in that: In step 1, the Astar algorithm is used to search for the shortest path of the harness, starting from the starting end of the harness, taking the distance from the current clamp to the end point as the heuristic function, and giving priority to the point closest to the end point of the path, so as to improve the efficiency of the algorithm.

3. The method according to claim 2, characterized in that When searching with the Astar algorithm, the wiring harness path is checked against the tree structure plan of the existing wiring harnesses through a union-find operation to avoid loops.

4. The method according to claim 3, characterized in that: In step 2, a complex cable collision detection algorithm based on the flexible torus BVH-tree is used.

5. The method according to claim 4, characterized in that BVH-tree is a data structure used for collision detection.

6. The method according to claim 5, characterized in that BVH-tree divides the triangular facets on the aircraft model into two halves according to the x, y, and z axis coordinates in turn, forming a binary tree structure. When performing collision detection, the triangular facets that may collide can be found in logarithmic time.

7. The method according to claim 6, characterized in that When detecting whether the cable collides with the model, a cubic B-spline curve is established to simulate the cable more accurately, estimate whether the bending radius is within the limit, and serve as the basis for collision detection.

8. The method according to claim 7, characterized in that In step 3, the angle limitation algorithm of adaptive elastic energy is used for judgment and adjustment, a flexible angle penalty is proposed, and the path planning adaptive elastic energy is set. When routing, by setting and adjusting the value of adaptive elastic energy M, the algorithm will automatically choose the method with the least turns.

Images