Building material intelligent detection robot path planning method and device based on improved JPS-ACO

Abstract

The invention provides a building material intelligent detection robot path planning method and device based on improved JPS-ACO, and belongs to the technical field of intelligent detection robot path planning. The method comprises the following steps: dividing a working environment of a robot into a grid map; adopting an improved JPS algorithm fused with an artificial potential field to optimize jump point generation; performing path optimization by adopting an improved ant colony algorithm; comprising the following steps: initializing ant colony pheromones, introducing an adaptive coefficient to improve a state transition probability, adding a death punishment mechanism, and performing pheromone enhancement on a high-quality path in a pheromone updating process; and the optimized path is smoothed by using a cubic uniform B-spline curve, a jump point is simplified as a control point, and a smooth path is generated. The requirement for movement stability of a robot carrying a detection instrument is met, and meanwhile collision with building material obstacles is avoided.

Description

Technical Field

[0001] This invention relates to a path planning method and device for a smart inspection robot for building materials based on an improved JPS-ACO, belonging to the field of smart inspection robot path planning technology. Background Technology

[0002] Intelligent testing robots for building materials need to move autonomously in laboratories or warehouses to complete tasks such as strength testing and appearance defect identification of building materials. Their path planning must meet the following requirements: High efficiency: The laboratory environment is narrow with aisle widths of 0.5-1.5m and building materials are densely stacked, requiring rapid path planning to shorten testing time; Safety: Keep away from piles of building materials, such as avoiding collisions with steel frame, cement stacks and high-altitude work areas, to ensure the safety of robots and testing equipment; Smoothness: The robot is equipped with testing instruments, such as ultrasonic testing instruments and stress sensors. The path must be smooth to avoid the instruments vibrating and affecting the testing accuracy.

[0003] Path planning is a crucial research topic in robotics, playing a key role in logistics, healthcare, aviation, and military fields. Path planning involves finding an optimal or near-optimal path from a starting point to a destination within a given environment, based on constraints. Jump Point Search (JPS) is a grid-based path planning algorithm. JPS doesn't require examining every grid point; it selects key nodes as jump points for evaluation, and connecting these optimal jump points forms a complete optimal path. Ant Colony Algorithm (AOC) is a biomimetic algorithm that mimics the foraging process of real ants. Ants release and leave pheromones along their paths, allowing other ants to choose their routes based on pheromone concentration. This positive feedback mechanism enables the search process to converge, ultimately achieving the goal of collaboratively finding the shortest path. In this process, multiple individuals perform parallel computation, improving computational power and efficiency, and exhibiting stability and robustness, making it widely used in path planning.

[0004] However, existing technologies have the following drawbacks: In a laboratory environment, the JPS algorithm is prone to generating a large number of useless jump points, such as jump points towards walls and building material piles, which leads to increased computation, longer search time, and increased memory usage. Ant colony algorithm has a uniform initial pheromone distribution, which makes it prone to aimless exploration in multi-target detection scenarios. It has a slow convergence speed (often more than 80 iterations), is prone to deadlock, and its path tends to approach obstacles. Traditional path smoothing methods, such as the simple B-spline method, do not consider laboratory passage constraints. The smoothed path may conflict with the building material pile or have abrupt curvature changes.

[0005] Therefore, it is necessary to design a path planning method that integrates and improves JPS and ACO to adapt to the special characteristics of building material testing scenarios. Summary of the Invention

[0006] The purpose of this invention is to provide a path planning method and device for a smart building material inspection robot based on an improved JPS-ACO, which is not only structurally robust and difficult to separate, but also convenient to construct.

[0007] To achieve the above objectives, the present invention employs the following technical solution: A path planning method for a smart building material inspection robot based on an improved JPS-ACO includes the following steps: The robot's working environment is divided into a grid map, marking obstacle points, passable areas, and important locations; An improved JPS algorithm incorporating artificial potential fields is used to optimize jump point generation. This includes calculating the gravitational and repulsive fields to determine the direction of the resultant force, guiding the JPS algorithm to search for jump points based on the direction of the resultant force, and eliminating redundant jump points to obtain a simplified jump point set. Path optimization is performed using an improved ant colony algorithm based on a simplified jump point set; this includes initializing ant colony pheromones, introducing adaptive coefficients to improve state transition probabilities, adding a death penalty mechanism, and reinforcing high-quality paths with pheromones during the pheromone update process. The optimized path is smoothed using a cubic uniform B-spline curve, and jump points are simplified into control points to generate a smooth path.

[0008] Preferably, the calculation of the gravitational and repulsive fields to determine the direction of the resultant force specifically includes: Calculate the gravitational field function and gravity to enhance the attractive force at key locations: , , in, Represents a gravitational field. This represents the gravitational proportional gain coefficient. This represents the current point's position vector. Represents the target point position vector. For gravity, This represents the negative gradient of the gravitational field; Calculate the repulsive field function and repulsive force, away from the obstacle point: , , in, Represents a repulsive field. This represents the repulsive force proportional gain coefficient. This represents the vector indicating the closest position of the obstacle to the current node. This represents the current point's position vector. This indicates the distance at which the repulsive force field of the obstacle influences. As a repulsive force, This represents the negative gradient of the repulsive field; Calculate the direction of the resultant force and guide the JPS algorithm to search for jump points along the direction of the resultant force; .

[0009] Preferably, the process of eliminating redundant jump points specifically includes: traversing the initial jump point set. Determine the two jump points and Check if the connecting line passes through obstacles; if not, delete the intermediate jump point. If one crosses over, it will cause Continue until the traversal is complete, and output the simplified jump point set.

[0010] Preferably, the pheromone initialization specifically involves: setting different initial pheromone concentrations based on the relationship between the number of simplified jump points and the number of other jump points, with higher initial pheromone concentrations set for simplified jump points, other jump points, and their adjacent points; the pheromone initialization formula is: , , In the formula, To simplify the initialization of concentration when the number of jump points is greater than other jump points, To simplify the initialization of concentration when the number of jump points is less than or equal to other jump points, express From node at any time Transfer to node pheromone concentration, Indicates the pheromone intensity factor. This represents the total number of ants. To simplify the number of jump points, This refers to the number of other hops, i.e., the number of hops on the optimized JPS algorithm path excluding simplified hops. As the initial pheromone factor, To optimize the path length of the JPS algorithm.

[0011] Preferably, the improved state transition probability introduces an adaptive coefficient in the node spacing of the heuristic function. An adaptive coefficient is added, which is calculated as follows: , , in, For grid State parameters, For grid State parameters, For grid Position parameters, For grid Position parameters, This represents the number of vertical nodes in the current node graph. For nodes To the target node Euclidean distance, For adaptive coefficients, Indicates the basic spacing between adjacent grid cells. This indicates the actual adjusted distance between adjacent grid cells after introducing the adaptive coefficient; The improved transition probability formula is: in, Indicates the first in the same generation An ant from the grid and grid The probability, and These are the current grid point and the grid point to be moved, respectively. Representing the same generation An ant, For the set of transferable nodes, For pheromone factors, As a heuristic functional factor, Indicate Ants from the grid To grid Visibility, express Time Node To the node The pheromone concentration between them Represents a grid The set of all reachable grid cells.

[0012] Preferably, the death penalty mechanism is as follows: when an ant is unable to move due to an obstacle, the pheromone concentration along its path is reduced, and the penalty formula is: , in, express Two nodes at time arrive The pheromone concentration between them.

[0013] Preferably, the pheromone update mechanism enhances the pheromone content of high-quality paths whose path length is less than 1.2 times the average length. The pheromone update formula for high-quality paths is as follows: , , , in, For optimal path length, For pheromone intensity, For the optimal path pheromone increment, The pheromone evaporation coefficient, for In the nth iteration Only high-quality ants are at the node To the node The increase in pheromones left on the road section For the number of ants, For iterative adjustment functions, , These are parameters that control the shape of the function; For non-premium paths, a decay factor is used to weaken pheromones, and the pheromone update formula is as follows: , in, A collection of high-quality paths, This is an additional attenuation factor.

[0014] Preferably, the adjustment parameter The value is 0.1. The value is -5, indicating the pheromone intensity. The value is 150, which is the additional attenuation factor. The value is 0.5.

[0015] Preferably, the specific method for generating the smooth path is as follows: Obtaining the third-order B-spline basis function based on the B-spline basis function formula: , in, It consists of four third-order B-spline basis functions; The optimized path is segmented, and a B-spline curve segment is generated for every four consecutive hop points, resulting in a cubic B-spline curve segment: , in, For points on a single cubic B-spline curve, These are the four control points for a single-segment curve; By combining the smoothing process of cubic B-spline curves, sharp inflection points in the path are eliminated.

[0016] The advantages of this invention are as follows: This invention addresses the specific characteristics of complex obstacle scenarios such as building material laboratories and warehouses by integrating optimized algorithms with scenario-adaptive design. It reduces unnecessary jumps in the JPS algorithm within the building material laboratory environment, lowering computational load and memory consumption; enhances the initial guidance of the ant colony algorithm in multi-target detection scenarios, improving convergence speed and ensuring path safety; and optimizes path smoothing to meet the motion stability requirements of the robot carrying detection instruments while avoiding conflicts with building material obstacles. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0018] Figure 1 To improve the JPS-ACO path planning flowchart; Figure 2 A comparison of traditional ACO and improved JPS-ACO path planning; Figure 3 A comparison of the convergence curves of the traditional ACO and the improved JPS-ACO; Figure 4 A comparison of path length metrics between traditional ACO and improved JPS-ACO; Figure 5 A comparison of runtime metrics between traditional ACO and the improved JPS-ACO; Figure 6 A comparison of the convergence iteration count between traditional ACO and the improved JPS-ACO; Figure 7 A comparison of the total number of iterations between traditional ACO and the improved JPS-ACO; Figure 8 A comparison of ant count metrics between traditional ACO and the improved JPS-ACO; Performance indicators in the figure: Path length: The total length of the final path; Execution time: Algorithm execution time; Convergence iteration count: The number of iterations required to find the optimal solution; Total number of iterations: The maximum number of iterations the algorithm can perform; Ant count: The number of ants used in each iteration. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] like Figure 1 As shown, a path planning method for a smart building material inspection robot based on an improved JPS-ACO includes the following steps: S1: Divide the robot's working environment into a grid map, marking obstacle points, passable areas, and important locations; S2: An improved JPS algorithm that integrates artificial potential fields is used to optimize jump point generation. This includes calculating the gravitational and repulsive fields to determine the direction of the resultant force, guiding the JPS algorithm to search for jump points based on the direction of the resultant force, and eliminating redundant jump points to obtain a simplified jump point set. S3: Path optimization is performed using an improved ant colony algorithm based on a simplified jump point set; this includes initializing ant colony pheromones, introducing adaptive coefficients to improve state transition probabilities, adding a death penalty mechanism, and strengthening high-quality paths with pheromones during the pheromone update process. S4: The optimized path is smoothed using a cubic uniform B-spline curve, and the jump points are simplified into control points to generate a smooth path.

[0021] As a refinement of the above embodiment, step S1 takes a building materials laboratory as an example. Obstacles include piles of steel bars, piles of cement test blocks, universal testing machines, and walls (boundary grids); passable areas include main passageways and testing stations; important locations include the starting point, robot charging piles, steel bar testing stations, cement testing stations, equipment testing stations, and the endpoint.

[0022] As a refinement of the above embodiments, step S2 specifically includes: S201: Artificial Potential Field Fusion Direction Guidance To address the challenges posed by multiple obstacles in the laboratory environment, the artificial potential field parameters were adjusted as follows: To enhance the attractiveness of target detection points, To enhance the repulsive force of building material piles, To cover the safety perimeter of the building materials pile. Calculate the current point. Direction of the resultant force: (1) The gravitational field function and gravity enhance the attraction of the target detection point, pointing it toward the nearest target detection point: , , in, Represents a gravitational field. This represents the gravitational proportional gain coefficient. This represents the current point's position vector. Represents the target point position vector. For gravity, This represents the negative gradient of the gravitational field; (2) Repulsive field function and repulsive force: away from various building material obstacles, such as the pile of steel bars.

[0023] , , in, Represents a repulsive field. This represents the repulsive force proportional gain coefficient. This represents the vector indicating the closest position of the obstacle to the current node. This represents the current point's position vector. This indicates the distance at which the repulsive force field of the obstacle influences. As a repulsive force, This represents the negative gradient of the repulsive field; (3) Resultant force, JPS searches for jump points along reasonable directions: .

[0024] S202: Redundant jump point removal Traverse the initial jump point set ,from Begin, make a judgment and Check if the connection passes through obstacles; if not, delete the intermediate node. ,renew If one travels through time, Repeat the judgment until Output simplified jump point set , To simplify the total number of back jump points.

[0025] Final Simplified Jump Point Set The number of nodes is significantly reduced, thus lowering the computational complexity of subsequent ant colony algorithms.

[0026] The artificial potential field method iteratively plans the motion trajectory, and the resultant force direction guides JPS to prioritize searching along the main channel direction, such as the positive x-axis. Starting from the starting point, each time a selected jump point is reached, the next direction calculated by the artificial potential field method is used as the reference direction for the JPS algorithm.

[0027] By integrating artificial potential fields and the JPS algorithm, invalid jump points to areas such as building material piles and walls can be effectively filtered out, reducing redundant calculations and significantly lowering memory usage and search time. A simplified jump point pheromone initialization strategy provides explicit guidance for the ant colony algorithm, avoiding aimless exploration, greatly improving convergence speed, and adapting to the high-efficiency operation requirements of multi-target detection scenarios.

[0028] As a refinement of the above embodiments, the improved ant colony algorithm in step S3 is as follows: S301: Pheromone Initialization Traditional AOC algorithms employ an equalization strategy to set the initial pheromone concentration, meaning all walkable grid cells on the raster map are set to the same constant. This makes them prone to blind searching in the early stages of pathfinding. To address this blind searching problem, an initial pheromone concentration strategy based on an improved JPS is proposed, setting higher initial pheromone concentrations for simplified jump points, other jump points, and their neighbors.

[0029] The formula for initializing pheromones is: , , In the formula, To simplify the initialization of concentration when the number of jump points is greater than other jump points, To simplify the initialization of concentration when the number of jump points is less than or equal to other jump points, express From node at any time Transfer to node pheromone concentration, Indicates the pheromone intensity factor. This represents the total number of ants. To simplify the number of jump points, This refers to the number of other hops, i.e., the number of hops on the optimized JPS algorithm path excluding simplified hops. As the initial pheromone factor, To optimize the path length of the JPS algorithm.

[0030] S302: Improved state transition probability To adapt to the dense distribution of obstacles such as steel bars and cement test blocks in the building materials laboratory, a distance adaptive coefficient is introduced to adjust the search direction probability, prioritizing the search along the channel direction, such as the direction of the guide rail steel bars to reduce invalid lateral movement.

[0031] In the grid (node) spacing of the heuristic function An adaptive coefficient has been added, and the formula is as follows: , , in, For grid State parameters, For grid State parameters, For grid Position parameters, For grid Position parameters, This represents the number of vertical nodes in the current node graph. For nodes To the target node Euclidean distance, For adaptive coefficients, Indicates the basic spacing between adjacent grid cells. This indicates the actual adjusted distance between adjacent grid cells after introducing the adaptive coefficient; An adaptive coefficient is used to increase or decrease the distance between adjacent grids. Specifically, this coefficient decreases the grid spacing when the target grid is located on the side of the current grid closer to the starting point, and increases the grid spacing when the target grid is located on the side of the current grid farther from the starting point. The purpose is to increase the spacing to facilitate vertical traversal first, guiding the overall traversal trend to the right; decreasing the spacing ensures immediate coverage of the left side during traversal, thereby reducing the probability of encountering untraversed areas on the left side of the current grid, thus avoiding unnecessary turning and detours in subsequent traversals. Through these improvements, horizontal transfers during traversal can be limited, invalid path searches can be minimized, path length and equivalent motion costs can be significantly reduced, and robot operation efficiency can be improved.

[0032] The improved transition probability formula is: in, Indicates the first in the same generation An ant from the grid and grid The probability, and These are the current grid point and the grid point to be moved, respectively. Representing the same generation An ant, a heuristic function Used to represent Ants at all times arrive Visibility, For the set of transferable nodes, For pheromone factors, As a heuristic functional factor, Indicate Ants from the grid To grid Visibility, express Time Node To the node The pheromone concentration between them Represents a grid The set of all reachable grid cells.

[0033] To address the deadlock problem that easily occurs in narrow laboratory passages, which leads to a large number of ant deaths, reduces the diversity of pathfinding, and affects the quality of converged paths, a death penalty mechanism is proposed. When an ant is unable to continue moving due to an obstacle, the pheromone concentration along its path is reduced to decrease the probability of subsequent ants repeatedly falling into deadlock.

[0034] The penalty formula is: , in, express Two o'clock arrive The pheromone concentration between them.

[0035] The introduced ant death penalty mechanism specifically addresses the deadlock problem, reduces the interference of invalid paths on subsequent exploration, ensures the robot's continuous and stable operation in complex obstacle layouts, and guarantees the continuity of detection tasks.

[0036] S303: Pheromone Update Mechanism In traditional ant colony optimization algorithms, pheromones are left inversely proportional to path length after each iteration, without distinguishing path quality. This can easily lead to pheromones from suboptimal paths interfering with subsequent iterations. Therefore, this paper proposes an improved strategy: only short paths are selected for pheromone updates; and an iterative adjustment function is added. To enhance the algorithm's early-stage random perturbation (to escape local optima) and accelerate later-stage convergence; a local pheromone decay factor is introduced for unselected paths. This further reduces its impact, ultimately leading to an improved pheromone update formula.

[0037] Iterative adjustment function as follows: , in, This represents the current iteration number. , In order to adjust the parameters, during the experiment The value is 0.1. The value is -5, which makes the algorithm more efficient in the early stages. Approaching 1 in the later stages encourages exploration, while approaching 0 in the later stages accelerates convergence.

[0038] Only high-quality paths with a path length less than 1.2 times the average length are retained during iterations, and their pheromone increment is: , in, For the optimal path pheromone increment, The formula for updating pheromones along a high-quality path is as follows: , in, This indicates the length of the path that meets the criteria. For the optimal path length, This represents the number of ants, and Q is the pheromone constant, which is set to 150 here. The pheromone evaporation coefficient, for In the nth iteration Only high-quality ants are at the node To the node The increase in pheromones left on the road section This refers to the number of ants.

[0039] For suboptimal paths that are not selected, an additional attenuation factor is applied. To mitigate its impact, the pheromone update formula is as follows: , in, A collection of high-quality paths, This is an additional attenuation factor.

[0040] S304: Path smoothing processing: A cubic uniform B-spline curve is used to smooth the path, simplifying jump points as control points. Inflection points are eliminated through piecewise fitting. The expression for the k-th order B-spline function is: , In the formula, These are control points, used to define the direction and limits of the spline curve; For the first indivual B-order spline basis functions, Let be a point on a k-order B-spline curve.

[0041] The basis functions have the following de Boer-Cox recurrence relation: , in, For the number to be calculated k-order basis functions; Indicates the order of the basis functions; It is a node vector.

[0042] To balance smoothness and complexity, we selected... The basis functions of the cubic B-spline curve are obtained as follows: , in, It consists of four 3rd-order B-spline basis functions.

[0043] The segment of the cubic B-spline curve is: , in, For points on a single cubic B-spline curve, These are the four control points of a single-segment curve.

[0044] The path is segmented, and a B-spline curve is generated every four consecutive jump points. Adjacent curves maintain second-order continuity at the connection points to ensure smooth turning of the robot when moving in the narrow passage between piles of steel bars and piles of cement test blocks, thus avoiding errors in the detection instrument due to vibration.

[0045] By combining the smoothing process of cubic B-spline curves, sharp inflection points in the path are eliminated, making the motion trajectory continuous and stable. This avoids detection errors caused by vibration in the mounted ultrasonic detectors, stress sensors, and other equipment, ensuring data accuracy and meeting the stringent requirements for building material quality testing.

[0046] The algorithm design fully considers the obstacle distribution characteristics of building material scenarios (such as densely packed building materials and narrow passages). Through parameter adjustment and strategy optimization, it can be flexibly adapted to different scenarios such as laboratories, warehouses, and construction sites. It can be applied to various building material testing tasks without major modifications, providing reliable technical support for the large-scale application of intelligent testing robots.

[0047] It should be noted that: such as Figures 2-8 As shown, the method described in this invention demonstrates advantages in all performance indicators, with an average improvement rate of 46.6%.

[0048] Example 2 This disclosure also provides a path planning device for a building material intelligent inspection robot based on an improved JPS-ACO, including a processor and a memory. Optionally, the device may further include a communication interface and a bus. The processor, communication interface, and memory can communicate with each other via the bus. The communication interface can be used for information transmission. The processor can call logical instructions in the memory to execute the path planning method for the building material intelligent inspection robot based on the improved JPS-ACO described in the above embodiments.

[0049] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium.

[0050] Memory, as a computer-readable storage medium, can be used to store software programs and computer-executable programs, such as the program instructions / modules corresponding to the methods in the embodiments of this disclosure. The processor executes the program instructions / modules stored in the memory to perform functional applications and data processing, thereby realizing the path planning method for the intelligent building material inspection robot based on the improved JPS-ACO in the above embodiments.

[0051] The memory may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the terminal device. Furthermore, the memory may include high-speed random access memory and may also include non-volatile memory.

[0052] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A path planning method for a smart building material inspection robot based on an improved JPS-ACO, characterized in that, Includes the following steps: The robot's working environment is divided into a grid map, marking obstacle points, passable areas, and important locations; An improved JPS algorithm incorporating artificial potential fields is used to optimize jump point generation. This includes calculating the gravitational and repulsive fields to determine the direction of the resultant force, guiding the JPS algorithm to search for jump points based on the direction of the resultant force, and eliminating redundant jump points to obtain a simplified jump point set. Path optimization is performed using an improved ant colony algorithm based on a simplified jump point set. This includes initializing ant colony pheromones, introducing adaptive coefficients to improve state transition probabilities, adding a death penalty mechanism, and reinforcing high-quality paths with pheromones during the pheromone update process. The optimized path is smoothed using a cubic uniform B-spline curve, and jump points are simplified into control points to generate a smooth path.

2. The path planning method for a building material intelligent inspection robot based on the improved JPS-ACO as described in claim 1, characterized in that, The calculation of the gravitational and repulsive fields to determine the direction of the resultant force specifically includes: Calculate the gravitational field function and gravity to enhance the attractive force at key locations: ， ， in, Represents a gravitational field. This represents the gravitational proportional gain coefficient. Represents the current point position vector. Represents the target point position vector. For gravity, This represents the negative gradient of the gravitational field; Calculate the repulsive field function and repulsive force, away from the obstacle point: ， ， in, Represents a repulsive field. This represents the repulsive force proportional gain coefficient. This represents the vector indicating the closest position of the obstacle to the current node. Represents the current point position vector. This indicates the distance at which the repulsive field of the obstacle influences. As a repulsive force, This represents the negative gradient of the repulsive field; Calculate the direction of the resultant force and guide the JPS algorithm to search for jump points along the direction of the resultant force; 。 3. The path planning method for a building material intelligent inspection robot based on the improved JPS-ACO as described in claim 1, characterized in that, The process of eliminating redundant jump points specifically includes: traversing the initial jump point set. Determine the two jump points and Check if the connecting line passes through obstacles; if not, delete the intermediate jump point. If one crosses over, it will cause Continue until the traversal is complete, then output the simplified jump point set.

4. The path planning method for a building material intelligent inspection robot based on the improved JPS-ACO as described in claim 1, characterized in that, The pheromone initialization specifically involves: based on the relationship between the number of simplified jump points and the number of other jump points, setting different initial pheromone concentrations, with higher initial pheromone concentrations set for simplified jump points, other jump points, and their adjacent points; the pheromone initialization formula is: ， ， In the formula, To simplify the initialization of concentration when the number of jump points is greater than other jump points, To simplify the initialization of concentration when the number of jump points is less than or equal to other jump points, express From node at any time Transfer to node pheromone concentration, Indicates the pheromone intensity factor. This represents the total number of ants. To simplify the number of jump points, This refers to the number of other hops, i.e., the number of hops on the optimized JPS algorithm path excluding simplified hops. As the initial pheromone factor, To optimize the path length of the JPS algorithm.

5. The path planning method for a building material intelligent inspection robot based on the improved JPS-ACO as described in claim 4, characterized in that, The improved state transition probability introduces an adaptive coefficient, which is applied to the node spacing of the heuristic function. An adaptive coefficient is added, which is calculated as follows: ， ， in, For grid State parameters, For grid State parameters, For grid Position parameters, For grid Position parameters, This represents the number of vertical nodes in the current node graph. For nodes To the target node Euclidean distance, For adaptive coefficients, Indicates the basic spacing between adjacent grid cells. This indicates the actual adjusted distance between adjacent grid cells after introducing the adaptive coefficient; The improved transition probability formula is: in, Indicates the first in the same generation An ant from the grid and grid The probability, and These are the current grid point and the grid point to be moved, respectively. Representing the same generation An ant, For the set of transferable nodes, For pheromone factors, As a heuristic functional factor, express Ants from the grid To grid Visibility, express Time Node To the node The pheromone concentration between them Represents a grid The set of all reachable grid cells.

6. The path planning method for a building material intelligent inspection robot based on the improved JPS-ACO as described in claim 5, characterized in that, The death penalty mechanism is as follows: when an ant is unable to move due to an obstacle, the pheromone concentration along its path is reduced, and the penalty formula is: ， in, express Two nodes at time arrive The pheromone concentration between them.

7. The path planning method for a building material intelligent inspection robot based on the improved JPS-ACO according to claim 5 or 6, characterized in that, The pheromone update mechanism enhances the pheromone content of high-quality paths whose path length is less than 1.2 times the average length. The pheromone update formula for high-quality paths is as follows: ， ， ， in, For optimal path length, For pheromone intensity, For the optimal path pheromone increment, The pheromone evaporation coefficient, for In the nth iteration Only high-quality ants are at the node To the node The increase in pheromones left on the road section For the number of ants, For iterative adjustment functions, , These are parameters that control the shape of the function; For non-premium paths, a decay factor is used to weaken pheromones, and the pheromone update formula is as follows: ， in, A collection of high-quality paths, This is an additional attenuation factor.

8. The path planning method for a building material intelligent inspection robot based on the improved JPS-ACO according to claim 7, characterized in that, The adjustment parameters The value is 0.

1. The value is -5, indicating the pheromone intensity. The value is 150, an additional attenuation factor. The value is 0.

5.

9. The path planning method for a building material intelligent inspection robot based on the improved JPS-ACO as described in claim 1, characterized in that, The specific method for generating a smooth path is as follows: Obtaining the third-order B-spline basis function based on the B-spline basis function formula: ， in, It consists of four third-order B-spline basis functions; The optimized path is segmented, and a B-spline curve segment is generated for every four consecutive hop points, resulting in a cubic B-spline curve segment: ， in, For points on a single cubic B-spline curve, These are the four control points for a single-segment curve; By combining the smoothing process of cubic B-spline curves, sharp inflection points in the path are eliminated.

10. A path planning device for a building material intelligent inspection robot based on an improved JPS-ACO, comprising a processor and a memory storing program instructions, characterized in that, The processor is configured to execute, when running the program instructions, the path planning method for a smart building material inspection robot based on any one of claims 1-9.

Images