Intelligent path planning method and system for deicing vehicle mechanical arm spray head

Abstract

The application discloses an intelligent path planning method and system for a mechanical arm spray head of a deicing vehicle, and can realize automatic and accurate deicing on a complex three-dimensional surface of an airplane. The method comprises the following steps: target surface extraction and reconstruction are performed on a deicing area of the surface of the airplane to generate a continuous three-dimensional surface; the three-dimensional surface is mapped to a two-dimensional parameter plane by surface parameterization; a 2D waypoint sequence is generated on the two-dimensional parameter plane in combination with the effective spraying width of the nozzle; 3D space inverse mapping is performed based on the 2D waypoint sequence, the spray head posture of the deicing vehicle is aligned based on the result of the inverse mapping, and the three-dimensional target posture of the nozzle at each waypoint is obtained; the target posture of the nozzle is converted into specific angle instructions of each joint of the mechanical arm of the deicing vehicle; on the basis of the joint angle instructions, joint trajectory smoothing and S-type speed planning are performed, and finally the action path of the spray head of the mechanical arm of the deicing vehicle is generated.

Description

Technical Field

[0001] This application relates to the field of automated control and aviation ground service support equipment technology, and more specifically, to an intelligent path planning method and system for the nozzle of a de-icing truck robotic arm. Background Technology

[0002] In the air transport sector, aircraft surface de-icing during severe winter weather is a crucial step in ensuring flight punctuality and safety. Currently, airports typically use manually operated de-icing trucks with robotic arms for de-icing operations. Operators, located in the de-icing truck's control cabin or high-altitude basket, manually observe and operate control levers to adjust the extension and rotation of the robotic arm and the spray angle of the nozzles, spraying de-icing fluid onto the aircraft's wings, tail, and fuselage.

[0003] However, existing de-icing methods that rely on manual control of the robotic arm's nozzle trajectory have the following significant drawbacks: First, the quality and efficiency of the operation are highly dependent on human experience, making it difficult to ensure uniform coverage of the de-icing fluid and resulting in material waste. Aircraft wings, tail fins, and other components are extremely complex three-dimensional curved surfaces. Manual operation makes it difficult to maintain the optimal spray distance between the nozzle and the fuselage surface, the vertical spray attitude (normal vector alignment), and the uniform speed of movement on these complex surfaces at all times. This often leads to inconsistent row spacing and overlap in the robotic arm's movement trajectory, resulting in extremely uneven spraying. This not only causes significant waste of expensive and environmentally problematic de-icing fluid but may also leave significant flight safety hazards due to localized missed spraying.

[0004] Secondly, its poor obstacle avoidance capabilities in close proximity make it highly susceptible to scrapes and collisions in adverse weather conditions. De-icing operations are typically conducted in low-visibility environments such as snow, fog, and nighttime, severely obstructing the operator's vision and making it difficult to accurately judge the spatial distance between the nozzle and the aircraft. Furthermore, the aircraft surface is covered with vulnerable and sensitive areas that must not be sprayed or touched, including pitot tubes, static pressure holes, and engine air intakes. Manually operating a large, inertial robotic arm performing multi-degree-of-freedom movements in extremely close proximity to the aircraft has a very low tolerance for error; even slight misjudgments in spatial awareness or operational delays can easily lead to serious safety accidents, such as the robotic arm colliding with the aircraft.

[0005] Third, the work is physically demanding, and prolonged high-altitude operations can easily lead to fatigue risks. When dealing with widespread flight delays caused by extreme weather, the demand for de-icing operations surges. Operators need to maintain high concentration for extended periods, performing intensive micro-management, which can easily cause visual and muscle fatigue. This can lead to distorted operations and sluggish reactions, further amplifying the aforementioned risks of uneven spraying and collisions.

[0006] Therefore, how to overcome the limitations of traditional manual operation and realize automated and intelligent coverage path planning of the de-icing truck's robotic arm nozzle on the complex three-dimensional curved surface of an aircraft, while ensuring absolute collision safety, and achieving efficient, uniform, and economical automated operation, is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0007] The following provides a brief overview of one or more aspects to offer a basic understanding of them. This overview is not an exhaustive summary of all conceived aspects, nor is it intended to identify key or decisive elements of all aspects, nor to define the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form to prepare for the more detailed descriptions that follow.

[0008] The purpose of this invention is to solve the above-mentioned problems and provide an intelligent path planning method and system for the nozzle of the robotic arm of a de-icing truck. This system can achieve automated and precise de-icing on the complex three-dimensional curved surface of an aircraft. While ensuring the absolute collision safety of the aircraft, it can significantly improve the uniformity of de-icing fluid spraying and operational efficiency, and significantly reduce material waste and the risk of human operator fatigue.

[0009] The technical solution of this invention is as follows: This invention discloses an intelligent path planning method for the nozzle of a de-icing truck's robotic arm, the method comprising: Step S1: Extract and reconstruct the target surface of the de-icing area on the aircraft surface to generate a continuous three-dimensional surface; Step S2: For the continuous three-dimensional surface of the de-icing area of ​​the aircraft surface generated in Step S1, the three-dimensional surface is reduced to a two-dimensional parameter plane by surface parameterization. Step S3: On the two-dimensional parameter plane of step S2, a 2D coverage path, i.e. a 2D waypoint sequence, is generated by combining the effective spray width of the nozzle; Step S4: Based on the 2D waypoint sequence generated in step S3, perform 3D spatial inverse mapping, and align the nozzle attitude of the de-icing vehicle based on the result of the inverse mapping to obtain the three-dimensional target attitude of the nozzle at each waypoint. Step S5: Convert the target posture of the nozzle in step S4 into specific angle commands for each joint of the de-icing vehicle's robotic arm. Step S6: Based on the joint angle command in step S5, perform joint trajectory smoothing and speed planning to finally generate the movement path of the de-icing truck robotic arm nozzle.

[0010] According to an embodiment of the intelligent path planning method for the nozzle of the de-icing vehicle robotic arm of the present invention, step S1 further includes: High-precision 3D local point cloud data of the aircraft is obtained through an on-board multi-sensor fusion system. After denoising the point cloud, a segmentation algorithm is used to extract the specific area that needs to be de-iced. The discrete scanned point cloud is then fitted into a continuous three-dimensional surface using meshing technology.

[0011] According to an embodiment of the intelligent path planning method for the nozzle of the de-icing vehicle robotic arm of the present invention, the process of individually extracting the specific area requiring de-icing operation in step S1 further includes: Step S1.1: Point cloud preprocessing and rough screening of regions of interest; Step S1.2: Estimation of surface normal vector and curvature characteristics; Step S1.3: Region growing and segmentation based on physical constraints; Step S1.4: Combine local morphological optimization and anomaly removal using the RANSAC algorithm; Step S1.5: Output the final target point cloud.

[0012] According to an embodiment of the intelligent path planning method for the nozzle of the de-icing vehicle robotic arm of the present invention, in step S2, the complex three-dimensional wing / tail surface is unfolded and mapped to a flat two-dimensional parametric plane by using conformal mapping or a parameterization method based on energy optimization.

[0013] According to an embodiment of the intelligent path planning method for the nozzle of the de-icing vehicle robotic arm of the present invention, unfolding and mapping the complex three-dimensional wing / tail fin surface onto a flat two-dimensional parametric plane includes the following processing steps: Step S2.1: Extraction of surface topology boundaries and initialization of the mapping domain; Step S2.2: Construct the discrete Laplace-Beltrami operator; Step S2.3: Solve the global linear equation system to achieve preliminary dimensionality reduction; Step S2.4: Deformation relaxation based on energy functional optimization; Step S2.5: Establish a bidirectional mapping function.

[0014] According to an embodiment of the intelligent path planning method for the nozzle of the de-icing vehicle robotic arm of the present invention, in step S3, on the flat two-dimensional UV plane obtained in step S2, the effective coverage width of the spray on the aircraft surface is calculated based on the current nozzle physical parameters of the de-icing fluid and the preset optimal spraying height; the spacing parameters between adjacent scanning paths are set, and a series of dense and uniform 2D waypoint sequences are generated on the two-dimensional plane by using a reciprocating parallel traversal algorithm.

[0015] According to an embodiment of the intelligent path planning method for the nozzle of the de-icing vehicle robotic arm of the present invention, step S3 further includes: Step S3.1: Acquisition of jet physical parameters and dynamic height; Step S3.2: Geometric calculation of the theoretical projected footprint width; Step S3.3: Correction of effective coverage width based on edge attenuation; Step S3.4: Line spacing calculation and overlap rate compensation; Step S3.5: Determine the main scanning direction and boundary of the two-dimensional mapping domain; Step S3.6: Geometry construction of parallel scan lines in the ox-plowing style; Step S3.7: Spatial discretization sampling and waypoint generation; Step S3.8: Topological connection and reciprocating sequence assembly.

[0016] According to an embodiment of the intelligent path planning method for the nozzle of the de-icing vehicle robotic arm of the present invention, in step S4, the 2D waypoint sequence generated in step S3 is inversely mapped back to the real 3D physical space according to the parametric model to obtain the 3D spatial coordinates of each waypoint; then, normal vector calculation is performed: for each waypoint on the 3D curved surface, its corresponding surface normal vector is calculated, wherein nozzle attitude alignment means that the spray direction of the nozzle is forcibly constrained so that the nozzle must remain parallel to the normal vector at each waypoint, thereby solving the three-dimensional target attitude of the nozzle at each waypoint.

[0017] According to an embodiment of the intelligent path planning method for the nozzle of the de-icing vehicle robotic arm of the present invention, in step S5, after obtaining the target pose in the Cartesian coordinate system output in step S4, the target pose is converted into specific angle commands for each joint of the robotic arm through the inverse kinematic equation of the robotic arm.

[0018] According to an embodiment of the intelligent path planning method for the nozzle of the de-icing vehicle robotic arm of the present invention, step S5 further includes: Before issuing joint angle commands, collision avoidance and singularity avoidance checks are performed: verifying whether the planned joint sequence exceeds physical limits, whether it is in a singular posture that would cause a speed surge, and whether the robotic arm body will interfere with or collide with sensitive areas of the aircraft; if a potential risk is detected, it will automatically return to fine-tune waypoint attitude.

[0019] According to an embodiment of the intelligent path planning method for the nozzle of the de-icing vehicle robotic arm of the present invention, in step S6, the joint angle is smoothly interpolated using a fifth-order polynomial or a B-spline curve within the joint space, while timestamps are allocated and speed planning is performed using an S-curve.

[0020] This invention also discloses an intelligent path planning system for the nozzle of a de-icing truck's robotic arm, the system comprising: The 3D surface generation module extracts and reconstructs the target surface of the de-icing area on the aircraft surface, generating a continuous 3D surface. The 2D plane mapping module maps the continuous 3D surface of the aircraft surface de-icing area generated by the 3D surface generation module to a 2D parameter plane by parameterizing the surface. The 2D waypoint sequence generation module generates a 2D coverage path, i.e., a 2D waypoint sequence, on the 2D parameter plane generated by the 2D plane mapping module, combined with the effective spray width of the nozzle. The nozzle attitude alignment module performs 3D spatial inverse mapping based on the 2D waypoint sequence generated by the 2D waypoint sequence generation module, and performs nozzle attitude alignment of the de-icing truck based on the result of the inverse mapping to obtain the three-dimensional target attitude of the nozzle at each waypoint. The joint angle command module converts the target posture of the nozzle generated by the nozzle posture alignment module into specific angle commands for each joint of the de-icing truck robotic arm. The trajectory smoothing and speed planning module performs joint trajectory smoothing and speed planning based on the joint angle commands from the joint angle command module, ultimately generating the movement path of the de-icing truck's robotic arm nozzle.

[0021] The present invention also discloses a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the intelligent path planning method for the nozzle of the de-icing truck robotic arm as described above.

[0022] The present invention also discloses a computer program product, which, when executed by a processor, implements the steps of the intelligent path planning method for the nozzle of the de-icing truck robotic arm as described above.

[0023] Compared with the prior art, the present invention has the following advantages: 1. Achieve efficient and uniform coverage of de-icing fluid, significantly reducing waste of expensive materials.

[0024] To address the limitations of manual operation in maintaining optimal spraying distance, attitude, and uniform speed on complex curved surfaces, this invention innovatively introduces three-dimensional curved surface coverage path planning (CPP) technology. By parameterizing and reducing the complex three-dimensional wing / tail surface to a two-dimensional plane for "ox-plowing" path generation, and accurately calculating the effective width and overlap rate of the jet, it completely eliminates inconsistencies in row spacing and blind spots caused by manual operation. Simultaneously, during the three-dimensional inverse mapping process, the system forcibly constrains the nozzle attitude to align with the surface normal vector at all times, ensuring maximum impact force and optimal de-icing effect of the de-icing fluid. Combined with S-curve velocity planning, it ensures absolute uniform speed in the straight spraying section, thereby achieving perfect and uniform coverage of the de-icing fluid and fundamentally eliminating significant waste.

[0025] 2. Provides extremely robust system-level collision avoidance protection to ensure the absolute safety of aircraft.

[0026] Addressing the technical challenges of poor spatial perception and high risk of collisions in low visibility conditions, this solution eliminates reliance on human visual observation. The system reconstructs surfaces by extracting high-precision 3D point cloud data from the aircraft and incorporates rigorous collision avoidance and singularity avoidance checks before the inverse kinematics (IK) calculation stage. This allows the robotic arm to pre-determine whether its joint sequence will interfere with or collide with the aircraft when performing multi-degree-of-freedom movements near vulnerable and sensitive areas such as pitot tubes and engines, and to make timely attitude adjustments. Furthermore, by using B-spline curves to smoothly interpolate the joint trajectories, severe vibrations caused by frequent abrupt stops and starts of the robotic arm are effectively avoided, completely eliminating the safety hazard of collisions with the wing due to uncontrolled robotic arm vibration.

[0027] 3. Free up manpower and completely eliminate the derivative risks caused by fatigue work.

[0028] Addressing the challenges of surging flight de-icing demands during severe weather and the high risk of operator fatigue and operational errors due to prolonged high-intensity work, this invention achieves end-to-end automation from perception and planning to control. It completely liberates operators from the arduous "micro-manipulation" of joystick control within a high-altitude pod. This not only significantly reduces manual labor intensity and personnel training costs but also completely eliminates sluggish reactions and operational errors caused by human fatigue, greatly improving the overall efficiency and safety of large-scale airport de-icing operations under extreme weather conditions. Attached Figure Description

[0029] The above-described features and advantages of the present invention will be better understood after reading the following detailed description of embodiments of the present disclosure in conjunction with the accompanying drawings. In the drawings, components are not necessarily drawn to scale, and components having similar related characteristics or features may have the same or similar reference numerals.

[0030] Figure 1 A flowchart illustrating an embodiment of the intelligent path planning method for the nozzle of the de-icing vehicle robotic arm according to the present invention is shown.

[0031] Figure 2 It shows Figure 1 The flowchart shown is a separate extraction sub-flowchart for a specific area of ​​the de-icing operation in step S1.

[0032] Figure 3 It shows Figure 1 The sub-flowchart shown in step S2 is for unfolding the complex three-dimensional wing / tail surface and mapping it to a flat two-dimensional parametric plane.

[0033] Figure 4 It shows Figure 1The sub-flowchart of step S3 is shown.

[0034] Figure 5 It shows Figure 1 The flowchart shown is a sub-flowchart for calculating the surface normal vector corresponding to the waypoint in step S4.

[0035] Figure 6 A schematic diagram of an embodiment of the intelligent path planning system for the nozzle of the de-icing vehicle robotic arm of the present invention is shown. Detailed Implementation

[0036] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should be noted that the aspects described below with reference to the accompanying drawings and specific embodiments are merely exemplary and should not be construed as limiting the scope of protection of the present invention in any way.

[0037] Figure 1 The flowchart illustrates an embodiment of the intelligent path planning method for the nozzle of the robotic arm of a de-icing vehicle according to the present invention. Please refer to [link to relevant documentation]. Figure 1 The implementation steps of the method in this embodiment are described in detail below.

[0038] Step S1: Extract and reconstruct the target surface of the de-icing area on the aircraft surface to generate a continuous three-dimensional surface.

[0039] High-precision 3D local point cloud data of the aircraft ahead is acquired through an onboard multi-sensor fusion system (such as a self-heating solid-state lidar). After denoising the point cloud, segmentation algorithms such as Region Growing Algorithm (RANSAC) are used to extract specific areas requiring de-icing operations (such as the upper surface of the wing and the horizontal tail). Then, meshing techniques such as Poisson Surface Reconstruction are used to fit the discrete scanned point cloud into a continuous 3D surface.

[0040] This step provides an extremely accurate physical boundary model for subsequent path planning, completely eliminating the reliance on operator visual observation in adverse weather conditions and eliminating the risk of spatial distance misjudgment and collision caused by obstructed vision.

[0041] Combination Figure 2 As shown, the specific processing steps for individual extraction of a specific area during de-icing operations are as follows: Step S1.1: Point cloud preprocessing and rough screening of regions of interest (ROI).

[0042] After acquiring the raw 3D point cloud from the vehicle-mounted multi-sensor fusion system, a pass-through filter and statistical outlier removal (SOR) algorithm are first used to filter out ground point clouds and spatially suspended rain and snow noise points. Subsequently, based on the coarse global pose of the aircraft acquired during the previous chassis driving phase, a bounding box is set in 3D space to retain only the local point cloud of the target operating area (e.g., within the physical space of the left wing), forming the region of interest (ROI) point cloud, which significantly reduces the amount of subsequent computation.

[0043] Step S1.2: Surface normal vector and curvature feature estimation.

[0044] For each 3D data point in the ROI point cloud, its K-Nearest Neighbors (KNN) are searched, and a local covariance matrix is ​​constructed using Principal Component Analysis (PCA). This allows the calculation of the surface normal vector and local curvature for each point. Since the upper surface of the wing or horizontal tail that requires de-icing typically exhibits geometric characteristics of "normal vector pointing upwards (positive Z-axis)" and "gradual change in local curvature," these feature values ​​will serve as the core constraints for subsequent segmentation algorithms.

[0045] Step S1.3: Region Growing based on physical constraints.

[0046] Since the wing is a continuous and complex curved surface, this embodiment preferably uses a region growing algorithm based on normal and curvature constraints for fine extraction: Seed point selection: In the ROI point cloud, select the point whose normal vector is closest to the absolute vertical upward (positive Z-axis) and has the smallest local curvature as the initial seed point (usually located at the highest and gentlest part of the wing).

[0047] Growth criterion judgment: Starting from the seed point, traverse the adjacent points outwards, and calculate the angle between the normal vectors of the adjacent points and the seed point, as well as the curvature difference.

[0048] Region merging: Set a normal angle threshold (e.g., The normal vectors of adjacent points are compared with the angle threshold and the curvature threshold. When the angle between the normal vectors of adjacent points is less than the angle threshold and the curvature is less than the curvature threshold, the point is determined to belong to the same smooth surface and is incorporated into the target area. If the threshold is exceeded (for example, if it grows to the side of the fuselage or the edge of the flap, the normal vector changes abruptly), then growth in that direction is stopped.

[0049] Step S1.4: Combine local morphological optimization and anomaly removal using the RANSAC algorithm.

[0050] After region growing, the extracted wing surface point cloud may still contain a small number of non-target structures (such as small antennas protruding from the wing surface or distortion points caused by residual ice). In this case, the Random Sample Consensus (RANSAC) algorithm is introduced. For each extracted point cloud block, the RANSAC algorithm is used to fit a local two-dimensional planar or quadratic surface model. Outliers that deviate from the fitted surface by a distance greater than a set tolerance are forcibly removed.

[0051] Step S1.5: Output the final target point cloud.

[0052] After the above processing, the complex apron background, fuselage side and wing underside were successfully stripped away, and the final output was a clean and continuous 3D point cloud containing only the upper surface of the wing or tail (i.e. the area that strictly needs to be sprayed with de-icing fluid), which was then handed over to the subsequent steps for meshed surface reconstruction.

[0053] Step S2: For the continuous three-dimensional surface of the de-icing area of ​​the aircraft surface generated in Step S1, the three-dimensional surface is reduced to a two-dimensional parameter plane by surface parameterization.

[0054] Since the wing surface is a complex three-dimensional curved surface, it is extremely difficult to plan a uniform coverage path directly in 3D space and it is easy to generate computational singularities. Continuing from the continuous 3D curved surface generated in step S1, this step uses conformal mapping or an energy-optimized parametric method to unfold the complex three-dimensional wing / tail surface and map it onto a flat two-dimensional parametric plane (uv space).

[0055] This step is the cornerstone for enabling robotic arms to perform equidistant and uniform operations on complex curved surfaces. Without this step and directly projecting the plan in 3D space, de-icing fluid would be sprayed excessively densely in areas with large curvature, such as the leading edge of the wing, while insufficiently in smooth areas. Through conformal mapping and energy optimization, the extremely complex "3D curved surface equidistant path planning" is cleverly reduced to an extremely simple "2D plane straight line drawing" problem, fundamentally ensuring the absolute uniformity of de-icing fluid coverage and eliminating material waste.

[0056] Combination Figure 3 As shown, unfolding and mapping a complex three-dimensional wing / tail surface onto a flat two-dimensional parametric plane involves the following processing steps.

[0057] Step S2.1: Extraction of surface topology boundaries and initialization of mapping domain.

[0058] First, the half-edge data structure of the 3D triangular mesh model (i.e., the continuous 3D mathematical surface generated in step S1) is traversed to extract the open topological vertices of the wing / tail surface. To facilitate the subsequent generation of parallel "ox-plowing" straight paths, a target mapping domain is predefined in the 2D UV space, usually set as a standard rectangular or convex polygon region, and the extracted 3D mesh boundary vertices are fixedly mapped to the boundary of this 2D rectangle according to the chord length ratio.

[0059] Step S2.2: Construct the discrete Laplace-Beltrami operator.

[0060] After fixing the boundary vertices, the two-dimensional coordinates of the interior vertices need to be solved. A system of linear equations is constructed for each interior vertex on the 3D mesh. To achieve conformal mapping, that is, to preserve the local angles of the original surface as much as possible during the flattening process, cotangent weights are used to discretize the Laplacian operator. By calculating the cotangent values ​​of the interior angles of adjacent triangles as the weights of the connecting edges, a large sparse symmetric positive definite matrix is ​​constructed.

[0061] Step S2.3: Solve the global linear equation system to achieve preliminary dimensionality reduction.

[0062] Substitute the (u, v) coordinates of the boundary vertices into the positive definite matrix above as Dirichlet Boundary Conditions, and calculate the preliminary (u, v) coordinates of all internal vertices on the two-dimensional parametric plane by solving the sparse linear equation system (e.g., using the conjugate gradient method), thereby obtaining a preliminary flattened two-dimensional mesh.

[0063] Step S2.4: Energy-based optimization.

[0064] Since the wing is a non-developable surface (with non-zero Gaussian curvature), forced flattening will inevitably lead to local area stretching or compression. To prevent inconsistent spacing of the subsequently generated spray paths in three-dimensional space, an "As-Rigid-As-Possible" (ARAP) or similar energy optimization algorithm is introduced. A global energy functional is constructed that includes "angle distortion energy" and "area stretching energy," and the positions of vertices within the two-dimensional plane are adjusted through iterative optimization to find the optimal (u, v) coordinate distribution that minimizes the overall deformation energy.

[0065] Step S2.5: Establish a bijective mapping function.

[0066] After optimization, it was confirmed that the 2D mesh did not undergo triangle flipping or self-intersection, thus establishing a strict one-to-one correspondence (bijection) between the 3D physical space coordinate system (x, y, z) and the 2D parametric coordinate system (u, v). This mapping function (including the interpolation weights for forward expansion and reverse mapping) was encapsulated and saved for subsequent path generation.

[0067] Step S3: On the two-dimensional parameter plane of step S2, a 2D coverage path, i.e. a 2D waypoint sequence, is generated by combining the effective spray width of the nozzle.

[0068] On the flat two-dimensional UV plane obtained in step S2, the effective coverage width (i.e., footprint width) of the jet on the aircraft surface is calculated based on the current nozzle physical parameters of the de-icing fluid (such as the spray cone angle) and the preset optimal spray height (such as 1 meter from the fuselage). To prevent missed spraying, an overlap rate of 10%-20% is set between adjacent scanning paths. Based on this spacing parameter, a "bostrophedon" reciprocating parallel traversal algorithm is used to generate a series of dense and uniform 2D waypoint sequences on the two-dimensional plane.

[0069] This step, through algorithm-driven setting of overlap rate and standard line spacing, completely eliminates inconsistencies in line spacing and blind spots caused by manual joystick operation. It is key to achieving efficient and uniform coverage of de-icing fluid and significantly reducing waste of expensive materials.

[0070] Combination Figure 4 As shown, step S3 further includes the following processing steps.

[0071] Step S3.1: Acquisition of jet physical parameters and dynamic height.

[0072] First, read the hardware parameters of the currently mounted de-icing nozzles, including the preset calibration spray cone angle (set as θ). Simultaneously, based on the collision safety threshold of the current wing area and the de-icing fluid pump pressure, dynamically set the optimal vertical spraying height (set as θ) for the current waypoint. H ), that height H This is the vertical Euclidean distance from the nozzle center to the local tangent plane on the machine body surface.

[0073] Step S3.2: Geometric calculation of the theoretical projected footprint width.

[0074] Because the nozzle is strictly parallel to the surface normal, the projection of the jet onto the local tangential plane is a standard circle (or approximately an ellipse). Based on basic trigonometric functions, the theoretical projection footprint diameter (i.e., theoretical width) of this jet cone on the target surface is calculated. W Its geometric solution formula is: .

[0075] Step S3.3: Correction of effective coverage width based on edge attenuation.

[0076] In practical fluid dynamics applications, the density and impact pressure of the de-icing fluid at the edge of the jet cone naturally decrease, failing to completely break up the ice layer. Therefore, an effective pressure coefficient (denoted as k, typically ranging from 0.70 to 0.90, obtained through factory calibration or feedback from a dynamic flow pressure sensor) is introduced. The theoretical width is then corrected for shrinkage to calculate the truly effective coverage width for de-icing (denoted as k). ): .

[0077] Step S3.4: Line spacing calculation and overlap rate compensation.

[0078] To absolutely prevent gaps (blind spots) from appearing between parallel scanning paths in a "plowing" manner, an overlap rate (denoted as ρ, typically set to 10% - 20%) must be maintained between adjacent scanning paths. Based on the effective coverage width and overlap rate, the precise row spacing (denoted as ρ) for generating parallel paths on the two-dimensional parametric plane is finally calculated. D ): .

[0079] Step S3.5: Determine the main scanning direction and boundary of the two-dimensional mapping domain.

[0080] In the two-dimensional uv parameter plane generated in step S2, the longest geometric axis of the two-dimensional mapping domain (i.e., the flattened wing surface) is first calculated using principal component analysis (PCA) or the bounding rectangle algorithm. To minimize the number of U-turns (reversals) by the robotic arm at the boundary and improve operational efficiency, the main scanning direction is set to be parallel to the longest axis (e.g., the u-axis direction), and the vertical direction is set to the stepping reversal direction (v-axis direction).

[0081] Step S3.6: Geometry construction of boustrophedon parallel scan lines.

[0082] In the v-axis direction, with spacing D Using a step size, draw rays parallel to the u-axis from one boundary of the mapping domain to the other. By finding the intersection points of these rays with the outer contour boundary of the two-dimensional mapping domain, extract a set of a series of mutually parallel effective scan line segments located inside the mapping domain. .

[0083] Step S3.7: Spatial discretization sampling and waypoint generation.

[0084] In order to convert continuous line segments into position commands that can be executed by a servo motor, a fixed spatial sampling resolution is set. (For example, 0.05 meters). For each valid scan line segment Along the direction of the line segment Discretized interpolation sampling is performed at intervals to generate a series of two-dimensional discrete coordinate points.

[0085] Step S3.8: Topological connection and reciprocating sequence assembly.

[0086] Based on the "ox-plowing" (snake-like) traversal topological logic, all generated discrete coordinate points are assembled in an ordered manner: traversing odd-numbered scan lines (such as...) The sampling points are sequentially stored into the global waypoint sequence in ascending order of the u-coordinate (from left to right); a reversing connection trajectory (such as a semicircular arc or spline curve) is generated between the end point of the current row and the start point of the next row for smooth transition; the even-numbered scan lines (such as...) are traversed. The sampling points are stored sequentially into the global waypoint sequence in the direction of decreasing u-coordinate (from right to left).

[0087] Finally, the output is a complete sequence of two-dimensional waypoint coordinates containing a clear order of execution. .in, (Point Sequence): Represents the total set of waypoint sequences. Subscript seqIt was emphasized that this is absolutely not a disordered set of points (like scattered sand), but rather an array or queue with a strict temporal order (like a string of pearls). The robotic arm must execute the commands strictly in a left-to-right sequence. : Represents each individual waypoint in the sequence. This is the starting point for the de-icing operation; This is the final stage of the de-icing operation; subscript This represents the order in which the robotic arm's nozzles arrive at these points. : Represents the first The precise mathematical coordinates of each waypoint in the two-dimensional parameter plane (i.e., the two-dimensional space after the 3D wing is flattened like an orange peel in step S2). and Just like on a two-dimensional plane shaft and Axis coordinates. At this point, these coordinates are still based on the "dimensionality-reduced 2D drawing." In subsequent step S4, these coordinates will be further... Coordinates are converted into three-dimensional coordinates in the real physical world through a mapping function. .

[0088] This step logically achieves a perfect "seamless splicing" of the de-icing fluid. Through strict boundary interception and alternating direction assembly, the complex coverage task is transformed into a series of ordered digital waypoints. This not only completely avoids missed spraying (a 10%-20% overlap rate serves as a safety redundancy, absorbing mechanical and wind drift errors), but also greatly improves the de-icing efficiency of the robotic arm by sweeping along the longest axis (minimizing the number of reversals), providing the most fundamental data support for subsequent mapping back to three-dimensional space for dynamic calculations.

[0089] Step S4: Based on the 2D waypoint sequence generated in step S3, perform 3D spatial inverse mapping, and align the nozzle attitude of the de-icing vehicle based on the inverse mapping result to obtain the three-dimensional target attitude of the nozzle at each waypoint.

[0090] The 2D waypoint sequence generated in step S3 is inversely mapped back to the real 3D physical space according to the parametric model to obtain the 3D spatial coordinates (x, y, z) of each waypoint. Then, normal vector calculation is performed: for each waypoint on the 3D surface, its corresponding surface normal vector is calculated. Nozzle attitude alignment refers to forcibly constraining the nozzle's spray direction so that it must remain parallel to the normal vector at each waypoint (i.e., the nozzle is always perpendicular to the fuselage surface), thereby solving for the three-dimensional target attitude (Roll, Pitch, Yaw) of the nozzle at each waypoint, where Roll is the roll angle, Pitch is the pitch angle, and Yaw is the yaw angle.

[0091] Normal vector alignment solves the technical challenge of maintaining a vertical spray posture on complex curved surfaces at all times during manual operation, ensuring that the de-icing fluid can break the ice layer with the maximum physical impact force, achieving the best de-icing effect.

[0092] Combination Figure 5 As shown, the specific processing steps for calculating the corresponding surface normal vector for each waypoint on the 3D surface are as follows.

[0093] Step S4.1: Obtain the vertex normal vector of the target face (Vertex Normal Extraction).

[0094] For any waypoint generated in step S3, after determining the target 3D triangle patch in which it is located, first retrieve the three vertices of that patch. The smooth vertex normal vectors pre-calculated during the Poisson reconstruction stage (or by area-weighted averaging of the normal vectors of adjacent faces) are denoted as follows: .

[0095] Step S4.2: Barycentric Normal Interpolation.

[0096] To ensure absolutely smooth nozzle attitude when crossing grid patch boundaries, the planar geometric normal of the triangular patch is not used directly. Instead, the same barycentric coordinates obtained when mapping two-dimensional waypoints to three-dimensional space are used. The initial interpolated normal vector at the waypoint is obtained by performing linear weighted interpolation on the normal vectors of the three vertices. The calculation formula is as follows: .

[0097] Step S4.3: Normalization of normal vectors.

[0098] Since the length of the vector obtained by linear weighted interpolation is usually not 1, it cannot be directly used for subsequent pose matrix construction. Therefore, the initial interpolated normal vector must be normalized to obtain a standard unit normal vector with a length of 1. : .

[0099] Step S4.4: Orientation Consistency Check.

[0100] Aircraft wings have two curved surfaces, inner and outer. To prevent the interpolated normal vector from pointing inwards (which would incorrectly instruct the robotic arm to spray from inside the fuselage outwards), an orientation verification mechanism is introduced. A reference observation vector is defined. (Usually, the line-of-sight direction of the vehicle-mounted radar sensor or the positive Z-axis of the global coordinate system is used). Calculation and The dot product: If the dot product is greater than 0, the current normal vector is retained; if the dot product is less than 0, it means the normal is pointing inwards, and it is reversed (i.e., ...). This ensures that the normal vector always faces outwards from the aircraft.

[0101] Step S4.5: Nozzle attitude matrix construction.

[0102] Obtain the accurate outward unit normal vector Then, combining the preset forward direction vector of the ox-plowing scanning (e.g., the X-axis of the nozzle coordinate system), the complete orthogonal three-dimensional rotation matrix or quaternion of the nozzle at that waypoint is constructed through vector cross product, thus obtaining the target attitude of the nozzle at that waypoint. .

[0103] Step S5: Convert the target posture of the nozzle in step S4 into specific angle commands for each joint of the de-icing vehicle's robotic arm.

[0104] After obtaining the target pose (x, y, z, Roll, Pitch, Yaw) in Cartesian coordinates output in step S4, it is converted into specific angle commands for each joint of the robotic arm using the inverse kinematics (IK) equations. , , ..., ).

[0105] Ideally, a rigorous collision avoidance and singularity avoidance check should be performed before issuing joint angle commands. This verifies whether the planned joint sequence exceeds physical limits, whether it is in a singular posture that would cause a surge in speed, and whether the robotic arm itself will interfere with or collide with sensitive areas of the aircraft (such as the engine, pitot tube, etc.). If a potential risk is detected, the system will automatically return to the fine-tuned waypoint attitude.

[0106] This step involves a robust, system-level collision avoidance system built at the control layer. It directly addresses the low tolerance for errors in close-range human intervention, ensuring the absolute safety of the aircraft.

[0107] Step S6: Based on the joint angle command in step S5, perform joint trajectory smoothing and S-shaped speed planning to finally generate the movement path of the de-icing truck robotic arm nozzle.

[0108] If the original waypoint sequence, after IK calculation and safety verification, is executed directly, the broken lines at the connection points will cause the robotic arm to frequently stop and start abruptly. Therefore, within the joint space, a fifth-order polynomial or B-spline curve is used to smoothly interpolate the joint angles. Simultaneously, timestamps are assigned, and an S-curve is used for velocity planning. This not only ensures smooth acceleration and deceleration during reversals (U-turns) but also forces the nozzle to maintain an absolutely uniform speed during straight spraying sections.

[0109] The smoothing process effectively avoids safety hazards caused by violent shaking of the robotic arm; while the uniform speed control further ensures the absolute uniformity of the spraying volume. At this point, the entire process is automated, completely freeing up manpower and eliminating the risks of fatigue and operational deformation caused by prolonged high-intensity manual labor.

[0110] Figure 6 The principle of an embodiment of the intelligent path planning system for the nozzle of a de-icing truck robotic arm according to the present invention is shown. (Reference) Figure 6 The system in this embodiment includes: a three-dimensional surface generation module, a two-dimensional plane mapping module, a 2D waypoint sequence generation module, a nozzle attitude alignment module, a joint angle command module, and a trajectory smoothing and velocity planning module.

[0111] The 3D surface generation module is configured to extract and reconstruct the target surface of the de-icing area on the aircraft surface to generate a continuous 3D surface.

[0112] High-precision 3D local point cloud data of the aircraft ahead is acquired through an onboard multi-sensor fusion system (such as a self-heating solid-state lidar). After denoising the point cloud, segmentation algorithms such as Region Growing Acoustics (RANSAC) are used to extract specific areas requiring de-icing operations (such as the upper surface of the wing and the horizontal tail). Then, meshing techniques such as Poisson Surface Reconstruction are used to fit the discrete scanned point cloud into a continuous 3D surface.

[0113] This module provides an extremely accurate physical boundary model for subsequent path planning, completely eliminating the reliance on operator visual observation in adverse weather conditions and eliminating the risk of spatial distance misjudgment and collision caused by obstructed vision.

[0114] For details on the internal processing of the 3D surface generation module, please refer to [link / reference]. Figure 1 The specific description of step S1 in the method embodiment shown will not be repeated here.

[0115] The 2D plane mapping module is configured to: map the continuous 3D surface of the aircraft surface de-icing area generated by the 3D surface generation module to a 2D parameter plane by reducing the 3D surface to a 2D parameter plane through surface parameterization.

[0116] Since the wing surface is a complex three-dimensional curved surface, it is extremely difficult to plan a uniform coverage path directly in 3D space and it is easy to generate computational singularities. Continuing from the continuous 3D curved surface generated in step S1, this step uses conformal mapping or an energy-optimized parametric method to unfold the complex three-dimensional wing / tail surface and map it onto a flat two-dimensional parametric plane (uv space).

[0117] This module is the cornerstone for solving the problem of equidistant and uniform operation of robotic arms on complex curved surfaces. Without this step and directly projecting the plan in 3D space, de-icing fluid would be sprayed too densely in areas with large curvature, such as the leading edge of the wing, while insufficiently in smooth areas. Through conformal mapping and energy optimization, the extremely complex "3D curved surface equidistant path planning" problem is cleverly reduced to an extremely simple "2D plane straight line drawing" problem, fundamentally ensuring the absolute uniformity of de-icing fluid coverage and eliminating material waste.

[0118] For details on the internal processing of the 2D plane mapping module, please refer to [link / reference]. Figure 1 The specific description of step S2 in the method embodiment shown will not be repeated here.

[0119] The 2D waypoint sequence generation module is configured to generate a 2D coverage path, i.e., a 2D waypoint sequence, on the 2D parameter plane generated by the 2D plane mapping module, combined with the effective spray width of the nozzle.

[0120] On the flat 2D UV plane obtained by the 2D plane mapping module, the effective coverage width (i.e., footprint width) of the jet on the aircraft surface is calculated based on the current nozzle physical parameters of the de-icing fluid (such as the spray cone angle) and the preset optimal spray height (such as 1 meter from the fuselage). To prevent missed spraying, an overlap rate of 10%-20% is set between adjacent scanning paths. Based on this spacing parameter, a "boss ploughing" parallel traversal algorithm is used to generate a series of dense and uniform 2D waypoint sequences on the 2D plane.

[0121] This module uses algorithms to force the setting of overlap rate and standard line spacing, completely eliminating the inconsistency in line spacing and blind spots caused by manual joystick operation. It is the key to achieving efficient and uniform coverage of de-icing fluid and significantly reducing the waste of expensive materials.

[0122] This module logically achieves a seamless "seamless splicing" of de-icing fluid. Through strict boundary interception and alternating direction assembly, the complex coverage task is transformed into a series of ordered digital waypoints. This not only completely avoids missed spraying (a 10%-20% overlap rate serves as a safety redundancy, absorbing mechanical and wind drift errors), but also greatly improves the de-icing efficiency of the robotic arm by sweeping along the longest axis (minimizing the number of reversals), providing the most fundamental data support for subsequent mapping back to three-dimensional space for dynamic calculations.

[0123] For details on the internal processing of the 2D waypoint sequence generation module, please refer to [link / reference]. Figure 1 The specific description of step S3 in the method embodiment shown will not be repeated here.

[0124] The nozzle attitude alignment module is configured as follows: based on the 2D waypoint sequence generated by the 2D waypoint sequence generation module, a 3D spatial inverse mapping is performed, and the nozzle attitude of the de-icing vehicle is aligned based on the result of the inverse mapping to obtain the three-dimensional target attitude of the nozzle at each waypoint.

[0125] The 2D waypoint sequence generated by the 2D waypoint sequence generation module is inversely mapped back to the real 3D physical space according to the parametric model to obtain the 3D spatial coordinates (x, y, z) of each waypoint. Then, normal vector calculation is performed: for each waypoint on the 3D surface, its corresponding surface normal vector is calculated. Nozzle attitude alignment refers to forcibly constraining the nozzle's spray direction so that it must remain parallel to the normal vector at each waypoint (i.e., the nozzle is always perpendicular to the fuselage surface), thereby solving for the three-dimensional target attitude (Roll, Pitch, Yaw) of the nozzle at each waypoint, where Roll is the roll angle, Pitch is the pitch angle, and Yaw is the yaw angle.

[0126] Normal vector alignment solves the technical challenge of maintaining a vertical spray posture on complex curved surfaces at all times during manual operation, ensuring that the de-icing fluid can break the ice layer with the maximum physical impact force, achieving the best de-icing effect.

[0127] For details on the internal processing of the nozzle attitude alignment module, please refer to [link / reference]. Figure 1 The specific description of step S4 in the method embodiment shown will not be repeated here.

[0128] The joint angle command module is configured to convert the target posture of the nozzle generated by the nozzle posture alignment module into specific angle commands for each joint of the de-icing truck robotic arm.

[0129] After obtaining the target pose (x, y, z, Roll, Pitch, Yaw) in Cartesian coordinates output by the nozzle pose alignment module, it is converted into specific angle commands for each joint of the robotic arm using the inverse kinematics (IK) equations. , , ..., ).

[0130] Ideally, a rigorous collision avoidance and singularity avoidance check should be performed before issuing joint angle commands. This verifies whether the planned joint sequence exceeds physical limits, whether it is in a singular posture that would cause a surge in speed, and whether the robotic arm itself will interfere with or collide with sensitive areas of the aircraft (such as the engine, pitot tube, etc.). If a potential risk is detected, the system will automatically return to the fine-tuned waypoint attitude.

[0131] This module is a robust, system-level collision avoidance system built at the control layer. It directly addresses the low tolerance for errors in close-range manual operations, ensuring the absolute safety of the aircraft.

[0132] The trajectory smoothing and speed planning module is configured to perform joint trajectory smoothing and speed planning based on the joint angle commands of the joint angle command module, and finally generate the movement path of the de-icing truck robotic arm nozzle.

[0133] If the original waypoint sequence, after IK calculation and safety verification, is executed directly, the broken lines at the connection points will cause the robotic arm to frequently stop and start abruptly. Therefore, within the joint space, a fifth-order polynomial or B-spline curve is used to smoothly interpolate the joint angles. Simultaneously, timestamps are assigned, and an S-curve is used for velocity planning. This not only ensures smooth acceleration and deceleration during reversals (U-turns) but also forces the nozzle to maintain an absolutely uniform speed during straight spraying sections.

[0134] The smoothing process effectively avoids safety hazards caused by violent shaking of the robotic arm; while the uniform speed control further ensures the absolute uniformity of the spraying volume. At this point, the entire process is automated, completely freeing up manpower and eliminating the risks of fatigue and operational deformation caused by prolonged high-intensity manual labor.

[0135] The present invention also discloses a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements as follows: Figure 1 The steps of the intelligent path planning method for the nozzle of the de-icing truck's robotic arm are shown. Specific details are described in the foregoing method embodiments and will not be repeated here.

[0136] This invention also discloses a computer program product, which, when executed by a processor, implements the following: Figure 1 The steps of the intelligent path planning method for the nozzle of the de-icing truck's robotic arm are shown. Specific details are described in the foregoing method embodiments and will not be repeated here.

[0137] Although the methods described above are illustrated and depicted as a series of actions for the sake of simplicity, it should be understood and appreciated that these methods are not limited by the order of the actions, as some actions may occur in a different order and / or concurrently with other actions from the illustrations and descriptions herein or not illustrated and described herein but which may be understood by those skilled in the art, according to one or more embodiments.

[0138] Those skilled in the art will further appreciate that the various illustrative logic blocks, modules, circuits, and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented as electronic hardware, computer software, or a combination of both. To clearly illustrate this interchangeability between hardware and software, the various illustrative components, blocks, modules, circuits, and steps are described above in a generalized manner in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system. Those skilled in the art may implement the described functionality in different ways for each specific application, but such implementation decisions should not be construed as departing from the scope of the invention.

[0139] The various illustrative logic blocks, modules, and circuits described in conjunction with the embodiments disclosed herein can be implemented or performed using a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. The general-purpose processor may be a microprocessor, but in alternatives, it may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors cooperating with a DSP core, or any other such configuration.

[0140] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of both. The software module may reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor such that the processor can read and write information to / from the storage medium. In an alternative, the storage medium may be integrated into the processor. The processor and storage medium may reside in an ASIC. The ASIC may reside in a user terminal. In an alternative, the processor and storage medium may reside as discrete components in the user terminal.

[0141] In one or more exemplary embodiments, the described functionality may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software as a computer program product, the functionality may be stored or transmitted as one or more instructions or code on or through a computer-readable medium. A computer-readable medium includes both computer storage media and communication media, encompassing any medium that facilitates the transfer of a computer program from one location to another. A storage medium may be any available medium accessible to a computer. By way of example and not limitation, such a computer-readable medium may include RAM, ROM, EEPROM, CD-ROM or other optical disc storage, disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and is accessible to a computer. Any connection is also legitimately referred to as a computer-readable medium. For example, if the software is transmitted from a website, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of a medium. As used in this article, disk and disc include compact discs (CDs), laser discs, optical discs, digital multi-purpose discs (DVDs), floppy disks, and Blu-ray discs. Disks typically reproduce data magnetically, while discs reproduce data optically using lasers. Combinations of these should also be included within the scope of computer-readable media.

[0142] The prior description of this disclosure is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to this disclosure will be apparent to those skilled in the art, and the general principles defined herein may be applied to other variations without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not intended to be limited to the examples and designs described herein, but should be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent path planning method for the nozzle of a de-icing truck's robotic arm, characterized in that, The methods include: Step S1: Extract and reconstruct the target surface of the de-icing area on the aircraft surface to generate a continuous three-dimensional surface; Step S2: For the continuous three-dimensional surface of the de-icing area of ​​the aircraft surface generated in Step S1, the three-dimensional surface is reduced to a two-dimensional parameter plane by surface parameterization. Step S3: On the two-dimensional parameter plane of step S2, a 2D coverage path, i.e. a 2D waypoint sequence, is generated by combining the effective spray width of the nozzle; Step S4: Based on the 2D waypoint sequence generated in step S3, perform 3D spatial inverse mapping, and align the nozzle attitude of the de-icing vehicle based on the result of the inverse mapping to obtain the three-dimensional target attitude of the nozzle at each waypoint. Step S5: Convert the target posture of the nozzle in step S4 into specific angle commands for each joint of the de-icing vehicle's robotic arm. Step S6: Based on the joint angle command in step S5, perform joint trajectory smoothing and speed planning to finally generate the movement path of the de-icing truck robotic arm nozzle.

2. The intelligent path planning method for the nozzle of the de-icing truck robotic arm according to claim 1, characterized in that, Step S1 further includes: High-precision 3D local point cloud data of the aircraft is obtained through an on-board multi-sensor fusion system. After denoising the point cloud, a segmentation algorithm is used to extract the specific area that needs to be de-iced. The discrete scanned point cloud is then fitted into a continuous three-dimensional surface using meshing technology.

3. The intelligent path planning method for the nozzle of the de-icing truck robotic arm according to claim 2, characterized in that, The process of individually extracting specific areas requiring de-icing operations in step S1 further includes: Step S1.1: Point cloud preprocessing and rough screening of regions of interest; Step S1.2: Estimation of surface normal vector and curvature characteristics; Step S1.3: Region growing and segmentation based on physical constraints; Step S1.4: Combine local morphological optimization and anomaly removal using the RANSAC algorithm; Step S1.5: Output the final target point cloud.

4. The intelligent path planning method for the nozzle of the de-icing truck robotic arm according to claim 1, characterized in that, In step S2, the complex three-dimensional wing / tail surface is unfolded and mapped onto a flat two-dimensional parametric plane using conformal mapping or an energy-optimized parametric method.

5. The intelligent path planning method for the nozzle of the de-icing truck robotic arm according to claim 4, characterized in that, Unfolding and mapping complex 3D wing / tail surfaces onto a flat 2D parametric plane involves the following processing steps: Step S2.1: Extraction of surface topology boundaries and initialization of the mapping domain; Step S2.2: Construct the discrete Laplace-Beltrami operator; Step S2.3: Solve the global linear equation system to achieve preliminary dimensionality reduction; Step S2.4: Deformation relaxation based on energy functional optimization; Step S2.5: Establish a bidirectional mapping function.

6. The intelligent path planning method for the nozzle of the de-icing truck robotic arm according to claim 1, characterized in that, In step S3, on the flat two-dimensional UV plane obtained in step S2, the effective coverage width of the jet on the aircraft surface is calculated based on the current nozzle physical parameters of the de-icing fluid and the preset optimal spraying height; the spacing parameters between adjacent scanning paths are set, and a series of dense and uniform 2D waypoint sequences are generated on the two-dimensional plane using a reciprocating parallel traversal algorithm.

7. The intelligent path planning method for the nozzle of the de-icing truck robotic arm according to claim 6, characterized in that, Step S3 further includes: Step S3.1: Acquisition of jet physical parameters and dynamic height; Step S3.2: Geometric calculation of the theoretical projected footprint width to obtain the theoretical projected footprint diameter of the jet cone on the target surface. The calculation formula is: ; Step S3.3: Correction of effective coverage width based on edge attenuation; Step S3.4: Line spacing calculation and overlap rate compensation; Step S3.5: Determine the main scanning direction and boundary of the two-dimensional mapping domain; Step S3.6: Geometry construction of parallel scan lines in the ox-plowing style; Step S3.7: Spatial discretization sampling and waypoint generation; Step S3.8: Topological connection and reciprocating sequence assembly.

8. The intelligent path planning method for the nozzle of the de-icing truck robotic arm according to claim 1, characterized in that, In step S4, the 2D waypoint sequence generated in step S3 is inversely mapped back to the real 3D physical space according to the parametric model to obtain the 3D spatial coordinates of each waypoint; then, normal vector calculation is performed: for each waypoint on the 3D surface, its corresponding surface normal vector is calculated, where nozzle attitude alignment means that the nozzle's spray direction is forcibly constrained so that the nozzle must remain parallel to the normal vector at each waypoint, thereby solving the three-dimensional target attitude of the nozzle at each waypoint.

9. The intelligent path planning method for the nozzle of the de-icing truck robotic arm according to claim 1, characterized in that, In step S5, after obtaining the target pose in the Cartesian coordinate system output in step S4, the target pose is converted into specific angle commands for each joint of the robotic arm through the inverse kinematics equations of the robotic arm.

10. The intelligent path planning method for the nozzle of the de-icing truck robotic arm according to claim 1, characterized in that, Step S5 also includes: Before issuing joint angle commands, collision avoidance and singularity avoidance checks are performed: verifying whether the planned joint sequence exceeds physical limits, whether it is in a singular posture that would cause a speed surge, and whether the robotic arm body will interfere with or collide with sensitive areas of the aircraft; if a potential risk is detected, it will automatically return to fine-tune waypoint attitude.

11. The intelligent path planning method for the nozzle of the de-icing truck robotic arm according to claim 1, characterized in that, In step S6, within the joint space, the joint angles are smoothed by using a fifth-order polynomial or a B-spline curve, while timestamps are assigned and velocity planning is performed using an S-curve.

12. An intelligent path planning system for the nozzle of a de-icing truck's robotic arm, characterized in that, The system includes: The 3D surface generation module extracts and reconstructs the target surface of the de-icing area on the aircraft surface, generating a continuous 3D surface. The 2D plane mapping module maps the continuous 3D surface of the aircraft surface de-icing area generated by the 3D surface generation module to a 2D parameter plane by parameterizing the surface. The 2D waypoint sequence generation module generates a 2D coverage path, i.e., a 2D waypoint sequence, on the 2D parameter plane generated by the 2D plane mapping module, combined with the effective spray width of the nozzle. The nozzle attitude alignment module performs 3D spatial inverse mapping based on the 2D waypoint sequence generated by the 2D waypoint sequence generation module, and performs nozzle attitude alignment of the de-icing truck based on the result of the inverse mapping to obtain the three-dimensional target attitude of the nozzle at each waypoint. The joint angle command module converts the target posture of the nozzle generated by the nozzle posture alignment module into specific angle commands for each joint of the de-icing truck robotic arm. The trajectory smoothing and speed planning module performs joint trajectory smoothing and speed planning based on the joint angle commands from the joint angle command module, ultimately generating the movement path of the de-icing truck's robotic arm nozzle.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent path planning method for the nozzle of the de-icing truck robotic arm as described in any one of claims 1-11.

14. A computer program product, characterized in that, When the program is executed by the processor, it implements the steps of the intelligent path planning method for the nozzle of the de-icing truck robotic arm as described in any one of claims 1-11.

Images