A network path optimization method and system

Abstract

This invention discloses a network path optimization method and system. The method includes: adjusting the adaptive weights of the heuristic function in the underlying model LPA* algorithm to intelligently change the magnitude of the heuristic function's influence with iteration; simultaneously, introducing a bias p in the K-value calculation to improve the algorithm's search efficiency; and performing unbalanced allocation of initial pheromones for the ant colony algorithm based on the initial path calculated by the underlying model. The upper-layer model applies the ACO algorithm to plan fiber optic network paths on a grid map to obtain the globally optimal network path. This invention, through the design of a hierarchical algorithm, provides the ant colony algorithm with prior information including pre-selected paths, enhancing the traditional ant colony algorithm's ability to guide pre-selected areas in path planning. This enables faster search for feasible paths, effectively solving the problems of slow search speed and susceptibility to local optima in the traditional ACO algorithm, and significantly improving the overall performance of the fiber optic network planning system.

Description

Technical Field

[0001] This invention belongs to the field of path planning technology, specifically relating to a network path optimization method and system. Background Technology

[0002] Network path optimization aims to find the optimal path based on network topology, node locations, and service requirements, while satisfying network performance, reliability, and scalability requirements. However, fiber optic network planning is a complex combinatorial optimization problem characterized by high nonlinearity, multiple constraints, and multiple objectives. Conventional optimization methods often employ graph theory or Dijkstra's algorithm, but these methods suffer from high computational complexity, are prone to getting trapped in local optima, and exhibit slow convergence speeds in complex 3D scenarios, resulting in low efficiency. Summary of the Invention

[0003] To address the aforementioned issues, this invention proposes a network path optimization method and system. The method maps network paths onto a grid map and searches for a locally optimal path on the grid map as the initial pheromone setting for the ant colony algorithm. This enhances the guiding ability of the pre-selected area in path planning, enabling the method to search for feasible paths more quickly, further improving algorithm performance, and performing better when handling complex scenarios.

[0004] A network path optimization method for achieving one of the objectives of this invention includes the following steps:

[0005] S1. Obtain the actual shortest distance g* from each node to the starting point start on the grid map; obtain the rhs value of each node on the grid map, where the rhs value of a node is the sum of the distances from each of its neighboring nodes to that node and the g* value of the neighboring nodes themselves, and the value with the smallest sum is the rhs value of the node; the grid map is a map based on the network topology and node mapping;

[0006] S2. Based on the g* and rhs values ​​of each node, perform path optimization on the raster map to obtain a locally optimal network path;

[0007] S3. Set the initial pheromone concentration for the ant colony algorithm based on the nodes and / or links on the locally optimal network path; the initial pheromone concentration is the pheromone concentration at the 0th iteration.

[0008] S4. Use the ant colony algorithm to plan network paths on the grid map to obtain the globally optimal network path.

[0009] A network path optimization system for achieving the second object of the present invention includes a first parameter acquisition module, a second parameter acquisition module, a local optimal path optimization module, an initial pheromone concentration setting module, and a global optimal path optimization module;

[0010] The first parameter acquisition module is used to obtain the actual shortest distance g* from each node on the grid map to the starting point start;

[0011] The second parameter acquisition module is used to obtain the rhs value of each node on the grid map. The rhs value of the node is the sum of the distances from each adjacent node of the node to the node and the g* value of the adjacent node itself. The smallest value of the sum is the rhs value of the node;

[0012] The local optimal path optimization module is used to perform path optimization on the grid map according to the g* value and rhs value of each node to obtain a locally optimal network path;

[0013] The initial pheromone concentration setting module is used to set the initial pheromone concentration of the ant colony algorithm according to the nodes and / or links on the locally optimal network path;

[0014] The global optimal path optimization module is used to perform network path planning on the grid map using the ant colony algorithm to obtain a globally optimal network path.

[0015] A non-transitory computer-readable storage medium for achieving the third object of the present invention, on which a computer program is stored, wherein the computer program realizes any step of the network path optimization method when executed by a processor.

[0016] Beneficial effects

[0017] The present invention proposes a hierarchical algorithm framework based on ACO-LPA*. This method makes the role of the heuristic function change intelligently with the iteration by adjusting the adaptive weight of the heuristic function in the underlying model LPA* algorithm, and introduces a bias p in the calculation of the K value, thereby improving the search efficiency of the algorithm.

[0018] Combined with the initial path calculated by the improved LPA* algorithm of the underlying model, the initial pheromone of the ant colony algorithm is unevenly distributed; the upper layer applies the ACO algorithm to perform fiber optic network path planning on the grid map to obtain a globally optimal network path.

[0019] This invention, through the design of a hierarchical algorithm, provides ant colony optimization (ACO) with prior information including pre-selected paths. This effectively enhances the traditional ACO's ability to guide pre-selected areas in path planning, enabling the algorithm to search for feasible paths more quickly and thus improving its performance. It effectively solves the problems of slow search speed and susceptibility to local optima in traditional fiber optic network planning methods, and has significant practical application value for fiber optic line planning in closed areas. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating an embodiment of the method described in this invention;

[0021] Figure 2 This is a schematic diagram of the grid map described in this invention. Detailed Implementation

[0022] The following detailed embodiments are provided to explain the technical solutions of the claims of this invention, so that those skilled in the art can understand the claims. The scope of protection of this invention is not limited to the following specific embodiments. Any modifications made by those skilled in the art that incorporate the technical solutions of the claims but differ from the following detailed embodiments are also within the scope of protection of this invention.

[0023] S1. Obtain the actual shortest distance g* from each node to the starting point start on the grid map; obtain the rhs value of each node on the grid map, where the rhs value of a node is the sum of the distances from each of its neighboring nodes to that node and the g* value of the neighboring nodes themselves, and the value with the smallest sum is the rhs value of the node; the grid map is a map based on the network topology and node mapping;

[0024] In this embodiment, taking a closed fiber optic planning area as an example, a grid method is used to determine the grid map. The method includes:

[0025] Suppose there are N nodes and M links in the fiber optic network, and the grid map is a square matrix of R rows, with each grid cell being a square with a side length of 1. Then we have:

[0026] The total number of raster cells on the raster map is R×R. Each raster cell on the raster map can be represented by a two-dimensional coordinate (i,j), where i = 1, 2, ..., R, j = 1, 2, ..., R. Each raster cell on the raster map can be represented by a unique number B, where B = 1, 2, ..., R×R. The relationship between the number B and the coordinate (i,j) is as follows:

[0027] B = (i-1) × R + j

[0028] Each grid cell on the raster map has a label t(B), where t(B) ∈ {0, 1, ..., R × R}. t(B) represents the node or link number in the fiber optic network corresponding to that grid cell, or it may be an obstacle point. Specifically:

[0029] If t(B) = 0, it means that the grid cell is an obstacle point, that is, it does not correspond to any node or link.

[0030] If t(B)∈{1,…,R×R}, then it means that the grid cell corresponds to the node or link numbered t(k) in the optical fiber network.

[0031] In this context, a node is a node within an optical distribution box, a link is a "connection" between adjacent optical distribution boxes, and an obstacle is a point on the network that is impassable.

[0032] It should be noted that the map in this embodiment is an R×R square grid map. In fact, a non-square map can be made into a square grid map by adding grid cells. The newly added grid cells are set as obstacle points.

[0033] The actual shortest distance g from node s to the starting point start * (s) Calculate based on the actual network topology plan;

[0034] It should be noted that, in this invention, the distance calculations, except for the heuristic function h(s,goal) which uses Euclidean distance, all other distances are calculated using Manhattan distance.

[0035] The methods for calculating the node RHS value include:

[0036]

[0037] In the formula:

[0038] pred(s): Represents the set of adjacent nodes of node s in the raster map;

[0039] g * (s): The actual shortest distance from node s to the starting point start. The actual shortest distance is calculated based on the grid map. For example, the distance of one grid in the vertical direction is 1, and the distance of one step in the diagonal direction is 1.414.

[0040] s'; represents the adjacent node of node s;

[0041] d(s',s): The distance between nodes s' and s;

[0042] For fiber optic path planning in grid maps, it's crucial to avoid skipped or missed nodes when traversing switch nodes. Manhattan distance is a simple and efficient four-way distance calculation method that avoids missing nodes during the search. It simply adds the absolute differences between the x and y coordinates of two points. Furthermore, compared to Euclidean distance, the Manhattan distance calculation method is more efficient because it doesn't require square root operations. Therefore, the method for calculating the distance between nodes s' and s includes:

[0043] d(s',s)=|x s′ -x s |+|y s′ -y s |

[0044] (x s′ ,y s′ ): Represents the coordinates of node s' on the raster map;

[0045] (x s ,y s ): Represents the coordinates of node s on the raster map.

[0046] S2. Based on the g* and rhs values ​​of each node, perform path optimization on the raster map to obtain a locally optimal network path;

[0047] In this embodiment, the method for obtaining a locally optimal network path from the starting node start to the target node goal by performing path optimization on a raster map includes the following steps:

[0048] S201. Set the g value and rhs value of each node in the grid map to infinity, and set the rhs value of the starting node start to 0. At this time, the rhs value and g value of the starting point start are not equal, that is, the starting point is in a locally inconsistent state, so the starting point start is inserted into the priority queue U to complete the initialization. The g value of the node is used to represent the expected shortest distance from the node in the grid map to the starting point start.

[0049] S202. Update the g value of the child node set of the current node based on the g* value and rhs value of each node; including the following steps:

[0050] S202.1 Calculate the g* value and rhs value of the child nodes of each node in the priority queue U, and add all child nodes with g values ​​greater than rhs values ​​to the priority queue U;

[0051] S202.2 Select the node S with the smallest K value from the updated priority queue U. min As the new parent node, let S min The value of g is equal to S minThe rhs value; this will become the new parent node S min Remove from priority queue U; if the K value of the node with the smallest K value is greater than the K value of the target node goal, or the rhs value of the target node goal is equal to the g value and is not infinite, then end; otherwise, proceed to the next step.

[0052] The methods for calculating the K value include:

[0053]

[0054] In the formula:

[0055] h(s,goal): The distance from node s to the target node goal (i.e., the shortest straight-line distance without considering obstacles); in this embodiment, it is the Euclidean distance;

[0056] w: Adaptive weight value;

[0057] Methods for calculating the K value also include:

[0058]

[0059] In the formula:

[0060] h(s,goal): The distance from node s to the target node goal (i.e., the shortest straight-line distance without considering obstacles); in this embodiment, it is the Euclidean distance;

[0061] w: Adaptive weight value;

[0062] P: Additional bias value.

[0063] S202.3. Add the child nodes of the node with the smallest K value whose g value is greater than the rhs value to the priority queue U, and return to step S202.2.

[0064] S203. Starting from the target node goal, move towards the parent node with the smallest g value until you reach the initial node start. The path formed during this movement is a locally optimal shortest path in the fiber optic path planning from the starting node start to the target node goal.

[0065] In another embodiment, in step S202.2, in order to better adapt to different search stages, such as using a larger weight in the early stages of the search to speed up the search, and gradually reducing the weight of the heuristic function value to finely adjust the pheromone as the search approaches the optimal solution, the method for calculating the weight value w includes:

[0066]

[0067] ||start,goal||: represents the second normal form distance from the starting node start to the target node goal;

[0068] In another embodiment, step S202.2, the method for calculating the additional bias value p includes:

[0069] Let the coordinates of the current node s be (x s ,y s The coordinates of the target node (goal) are (x... g ,y g The starting node's coordinates are (x0, y0). To distinguish nodes with the same K value, the heuristic function h(s) can be the initial-target vector A = (x0 - x0) / (x0 - x0) / (y0 - x0). g ,y0-y g ) and the current-target vector B = (x s -x g ,y s -y g The bias value p is constructed from the cross product of vectors. Methods for calculating the cross product include:

[0070] A×B=((x0-x g )×(y s -y g )-(x s -x g )×(y0-y g ))

[0071] The methods for calculating the additional bias value p include:

[0072]

[0073] map_length is the maximum value of the length or width of the raster map; in this embodiment, map_length is equal to R;

[0074] In one embodiment, the method for determining the minimum K value in step S202.3 includes:

[0075] ifk1(s)<=k1(s')or(k1(s)==k1(s')and k2(s)<=k2(s'))

[0076] K(s) <= K(s');

[0077] That is: first compare the k1 values ​​in K. If the k1 values ​​are not equal, the node with the smallest k1 value is the node with the smallest K value; if the k1 values ​​are equal, then further compare the size of the k2 values, and the node with the smallest k2 value is the node with the smallest K value.

[0078] Preferably, for the fiber optic path planning problem in a grid map, it is necessary to avoid skipping and missing nodes when traversing switch nodes. Manhattan distance is a simple and effective four-way distance calculation method that can avoid the problem of missing nodes during the search. It only requires adding the absolute values ​​of the differences between the x and y coordinates of two points. In addition, compared with Euclidean distance, the Manhattan distance calculation method is more efficient because it does not require square root operations. Therefore, the calculation method of d(s',s) in the RHS value includes:

[0079] d(s',s)=|x s' -x s |+|y s' -y s |

[0080] Preferably, when the environment map changes, for example, some nodes are overloaded or are released, the RHS value and priority of the affected nodes need to be recalculated. If the G value and RHS value of the affected node are equal, the node is removed from the priority queue; otherwise, the node is added to the priority queue U. The process is repeated until the nodes around the changed grid are locally consistent. Finally, the new shortest path is obtained by connecting the paths according to step 3.

[0081] S3. Set the initial pheromone concentration for the ant colony algorithm based on the nodes and / or links on the locally optimal network path;

[0082] Traditional ACO algorithms require a uniform initial pheromone distribution, which can lead to blindness in the initial stages and slow convergence. To address these shortcomings, this embodiment combines the optimized initial path output by LPA* to guide a non-uniform initial pheromone distribution, improving the pre-selection capability of candidate regions during path planning. This improved method facilitates adjustment of the search speed from the current point to the target point, increases the strength of heuristic information, and effectively reduces the blindness and arbitrariness of the ACO algorithm in the early stages of the search, thereby enhancing global search capability and convergence speed.

[0083] Based on the output of the locally optimal network path The method for setting the initial pheromone for the ant colony algorithm is shown in the following formula:

[0084]

[0085] In the formula:

[0086] τ ij (0): The pheromone concentration from node i to node j at the 0th iteration;

[0087] C0: The initial pheromone concentration on the locally optimal network path, which is set according to actual needs. In this embodiment, the value is 6, but it is not limited to this.

[0088] ζ is the initial deviation pheromone decay coefficient, which is set through experimental parameter tuning;

[0089] The set of nodes on the locally optimal network path;

[0090] Q(j): represents the sum of Manhattan distances between node j and every node on the locally optimal network path. The calculation method for Q(j) is shown in the following formula:

[0091] In the formula:

[0092] (x j ,y j (x) represents the coordinates of point j; j' y j' () represents the coordinates of point j'.

[0093] Preferably, when C0-ζ×Q(j) is less than a set value, then τ is set... ij (0) equals this setting value.

[0094] S4. Use the ant colony algorithm to plan network paths on the grid map to obtain the globally optimal network path.

[0095] For example Figure 2 In the raster map shown, optical intersection 1 (7, 0) is the starting point and optical intersection 20 (20, 16) is the ending point. A locally optimized network path obtained above is as follows: Figure 2 As shown by the underlined lines, the coordinates of the nodes on this network path are shown in Table 1 below:

[0096]

[0097] Table 1 Network Path Node Coordinates

[0098] (1) Parameter initialization; Set the ant colony algorithm parameters according to the actual situation of the current fiber optic planning instance, as shown in Table 2 below:

[0099]

[0100] Table 2. Ant Colony Algorithm Parameter Values

[0101] (2) Place 50 ants at the starting point light intersection 1 and let them search. With a relatively better initial path as a clue, guide the ant colony to start searching from a relatively more accurate direction.

[0102] During the search process, ants use a positive feedback mechanism to choose the best path under the influence of pheromones. The more pheromones on a path, the greater the probability that an ant will choose that path. Ultimately, this will cause most ants to tend towards the best path, thus yielding the optimal path.

[0103] Specifically, the probability of the ant's next search node is determined by the path transition probability, and the next node is selected by roulette wheel selection.

[0104] The formula for the path transition probability of ants is as follows:

[0105]

[0106] In the formula:

[0107] The probability of the k-th ant moving from node i to every possible next node j;

[0108] τ ij (t): The pheromone concentration from node i to node j at the t-th iteration;

[0109] η ij (t): The heuristic function from node i to node j at the t-th iteration;

[0110] α: Pheromone importance coefficient; the value in this embodiment is shown in Table 2 above;

[0111] β: Heuristic importance coefficient; in this embodiment, the value is shown in Table 2 above;

[0112] v k : The set of all possible nodes that ant k can choose next when it is at node i.

[0113] (3) Update the pheromone matrix. After each iteration, the pheromone matrix is ​​updated according to the following formula:

[0114] τ ij (t+1)=(1-ρ)τ ij (t)+Δτ ij (t)

[0115]

[0116] In the formula:

[0117] ρ: Volatilization coefficient of pheromones;

[0118] Δτ ij (t): The increment of pheromone concentration from node i to node j at the t-th iteration;

[0119] N: represents the total number of ants;

[0120] The increase in pheromone concentration between nodes i and j for the k-th ant in the t-th iteration compared to the previous iteration;

[0121] (4) Determine the termination condition. If the maximum number of iterations has been reached or the results of multiple consecutive iterations are consistent, stop the iteration, indicating that a shortest path has been found, and output the fiber optic route map. Otherwise, return to step 2 to continue the search.

[0122] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0123] This application embodiment also provides a network path optimization system, including a first parameter acquisition module, a second parameter acquisition module, a local optimal path optimization module, an initial pheromone concentration setting module, and a global optimal path optimization module;

[0124] The first parameter acquisition module is used to obtain the actual shortest distance g* from each node on the raster map to the starting point start;

[0125] The second parameter acquisition module is used to obtain the RHS value of each node on the grid map. The RHS value of a node is the sum of the distance from each of its neighboring nodes to the node and the g* value of the neighboring node itself. The value with the smallest sum is the RHS value of the node.

[0126] The local optimal path optimization module is used to perform path optimization on the raster map based on the g value and rhs value of each node to obtain a locally optimal network path;

[0127] The initial pheromone concentration setting module is used to set the initial pheromone concentration of the ant colony algorithm based on the nodes and / or links on the locally optimal network path;

[0128] The global optimal path optimization module is used to plan network paths on the grid map using the ant colony algorithm to obtain the globally optimal network path.

[0129] In another embodiment, the local optimal path optimization module includes a parameter initialization module, which is used to set the g value of all nodes in the grid map to infinity, set the rhs value of the starting node start to 0, and set the priority queue U to be a queue that only includes the starting node start; the g value of the node is used to represent the expected shortest distance from the node in the grid map to the starting point start.

[0130] In another embodiment, the local optimal path optimization module further includes a first estimated distance update module, which updates the g value of the child nodes of the current node in the raster map according to the g* value and rhs value of each node; the g value is used to represent the estimated shortest distance from the node in the raster map to the starting point start.

[0131] In another embodiment, the local optimal path finding module further includes a path construction module, which moves from the target node goal to the parent node with the smallest sum of g(s') and d(s',s) until it moves to the initial node start. The path connected during this movement is a locally optimal shortest path in the fiber optic path planning from the starting node start to the target node goal.

[0132] In another embodiment, the first estimated distance update module further includes a K-value calculation module, wherein the K-value is used to select the next node to be searched.

[0133] In another embodiment, the first estimated distance update module includes a first priority queue update module, which is used to obtain the g* value and rhs value of the child nodes of each node in the priority queue U, and add all child nodes with g* values ​​greater than rhs values ​​to the priority queue U.

[0134] In another embodiment, the first estimated distance update module includes a second priority queue update module, which selects the node with the smallest K value from the updated priority queue U as the new parent node, updates the g value of the node with the smallest K value to equal its rhs value, and deletes the node with the smallest K value from the priority queue U.

[0135] In another embodiment, the K-value calculation module includes a weight calculation module for calculating an adaptive weight value w; and an additional bias calculation module for calculating a bias value for the distance from a node to the target node goal.

[0136] This application also provides a computer-readable storage medium storing a computer program, which includes program instructions that, when executed by a processor, implement the various steps of the method described in this invention, which will not be elaborated further here.

[0137] The computer-readable storage medium can be the data transmission apparatus or the internal storage unit of a computer device provided in any of the foregoing embodiments, such as the hard disk or memory of the computer device. The computer-readable storage medium can also be an external storage device of the computer device, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., provided on the computer device.

[0138] Furthermore, the computer-readable storage medium may include both internal storage units and external storage devices of the computer device. The computer-readable storage medium is used to store the computer program and other programs and data required by the computer device. The computer-readable storage medium can also be used to temporarily store data that is to be output or has already been output.

[0139] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0140] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0141] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0142] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0143] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

[0144] The contents not described in detail in this specification are existing technologies known to those skilled in the art.

Claims

1. A network path optimization method, characterized in that, Includes the following steps: S1. Obtain the actual shortest distance g* from each node to the starting point start on the grid map; obtain the rhs value of each node on the grid map, where the rhs value of a node is the sum of the distances from each of its neighboring nodes to that node and the g* values ​​of the neighboring nodes themselves, and the value with the smallest sum is the rhs value of the node; the grid map is a map based on the network topology and node mapping; S2. Based on the g* and rhs values ​​of each node, perform path optimization on the raster map to obtain a locally optimal network path; S3. Set the initial pheromone concentration for the ant colony algorithm based on the nodes and / or links on the locally optimal network path; S4. Use the ant colony algorithm to plan network paths for the grid map to obtain the globally optimal network path; Step S2 involves obtaining a locally optimal network path, including: S201. Set the g value and rhs value of each node in the grid map to infinity, and set the rhs value of the starting node start to 0; set the priority queue U to be a queue that only includes the starting node start; the g value of the node is used to represent the expected shortest distance from the node in the grid map to the starting point start. S202. Update the g value of each current node's child nodes in the raster map based on the g value and rhs value of each node; S203. Starting from the target node goal, move towards the parent node with the smallest g value until you reach the initial node start. The path formed during this movement is a locally optimal shortest path in the fiber optic path planning from the starting node start to the target node goal. The method for calculating the initial pheromone concentration in step S3 includes: ； In the formula: : The pheromone concentration from node i to node j at the 0th iteration; C0: The pheromone concentration on the locally optimal network path; The initial deviation pheromone decay coefficient; The set of nodes on the locally optimal network path; Q(j): represents the sum of the Man distances between node j and each node on the locally optimal network path.

2. The network path optimization method as described in claim 1, characterized in that, In step S202, the method for updating the g value of the child nodes of each current node in the raster map includes calculating the K value of the node, wherein the K value is used to select the node to be searched in the next step, and the calculation method includes: ； In the formula: K ( s ): The K value of node s; h ( s , goal ): The distance from node s to the target node goal; w: Adaptive weight value.

3. The network path optimization method as described in claim 1, characterized in that, In step S202, the method for updating the g value of the child nodes of each current node in the raster map includes calculating the K value of the node, wherein the K value is used to select the node to be searched in the next step, and the calculation method includes: ； In the formula: K ( s ): The K value of node s; h ( s , goal ): The distance from node s to the target node goal; w: Adaptive weight value; P: Additional bias value.

4. The network path optimization method as described in claim 3, characterized in that, The method for calculating the additional bias value p includes: ； ； ( x s ,y s ): The coordinates of the current node; ( x g ,y g ): The coordinates of the target node (goal); ( x 0 ,y 0): The coordinates of the starting node; map _ length This represents the maximum length or width of the raster map.

5. The network path optimization method as described in claim 2 or 3, characterized in that, The method for calculating the adaptive weight value w includes: ； In the formula: : Represents the L2 distance between the starting node start and the target node goal.

6. The network path optimization method as described in claim 2 or 3, characterized in that, In step S202, the method for updating the g value of the node to be searched in the raster map includes: S202.

1. Obtain the g* value and rhs value of the child nodes of each node in the priority queue U, and add all child nodes with g values ​​greater than rhs values ​​to the priority queue U. S202.2 Select the node with the smallest K value from the updated priority queue U as the new parent node, and update the g value of the node with the smallest K value to be equal to its rhs value; delete the node with the smallest K value from the priority queue U; if the K value of the node with the smallest K value is greater than the K value of the target node goal, or the rhs value of the target node goal is equal to the g value and is not infinite, then end; otherwise, proceed to the next step. S202.

3. Add all child nodes of the node with the smallest K value whose g value is greater than the rhs value to the priority queue U, and return to step S202.

2.

7. A network path optimization system employing the method described in claim 1, characterized in that, It includes a first parameter acquisition module, a second parameter acquisition module, a local optimal path optimization module, an initial pheromone concentration setting module, and a global optimal path optimization module; The first parameter acquisition module is used to obtain the actual shortest distance g* from each node on the raster map to the starting point start; The second parameter acquisition module is used to obtain the RHS value of each node on the grid map. The RHS value of a node is the sum of the distance from each of its neighboring nodes to the node and the g* value of the neighboring node itself. The value with the smallest sum is the RHS value of the node. The local optimal path optimization module is used to perform path optimization on the raster map based on the g* value and rhs value of each node to obtain a locally optimal network path. The initial pheromone concentration setting module is used to set the initial pheromone concentration of the ant colony algorithm based on the nodes and / or links on the locally optimal network path. The global optimal path optimization module is used to plan network paths on the grid map using the ant colony algorithm to obtain the globally optimal network path.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the network path optimization method as described in any one of claims 1 to 6.

Images