An automated spray method and system that self-adjusts spray attitude

Abstract

The application provides an automatic spraying method and system for self-adaptive adjustment of spraying posture, relates to the technical field of automatic construction, and in particular relates to automatic spraying operation. The spraying method specifically comprises the following steps: controlling a UAV to perform stable flight tracking according to a preset global path waypoint; in the stable flight tracking process, a five-point laser sensing array continuously collects synchronous data frames of an actual target surface; a posture correction angle is solved in real time based on the synchronous data frames; a two-axis holder is controlled to execute the posture correction angle to adjust the spraying posture of a spraying unit; a flatness error is calculated in real time based on the synchronous data frames; the flatness error is compared with a preset flatness threshold value to make a spraying decision; when the flatness error is not greater than the flatness threshold value, it is determined that the actual surface is flat, and one-time spraying is performed; when the flatness error is greater than the flatness threshold value, it is determined that the actual surface is complex, and adaptive zoned spraying is performed.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of automated construction technology, in particular to automated spraying operation. More specifically, the present application relates to an automated spraying method and system capable of self-adaptively adjusting the spraying posture, which can perceive the local geometry of the target surface in real time during the spraying process and self-adaptively adjust the spraying posture of the spraying unit, especially suitable for unmanned spraying operation on large-scale infrastructure with complex three-dimensional curved surface features, such as highway bridges, tunnels or large structural components. BACKGROUND

[0002] In the fields of maritime engineering, large equipment manufacturing and energy industry (such as wind turbine towers, large storage tanks), surface coating protection of large structural components is a key process to ensure the structural integrity and prolong the service life. Taking large ships as an example, the hull, such as the freeboard, the vicinity of the waterline and the bottom, is exposed to the high-salt and high-humidity marine corrosion environment for a long time, and at the same time faces the serious challenge of marine bio-attachment. Therefore, it is necessary to apply high-performance anti-corrosion and anti-fouling coatings to the hull regularly. These coating operations have very high quality requirements, and the uniformity of coating thickness and adhesion directly determine the maintenance cycle and operating cost of the ship.

[0003] Traditionally, the spraying operation of large ships, whether new or repaired, highly depends on manual work, supplemented by large-area scaffolding, aerial work platforms or baskets and other auxiliary equipment. This mode has inherent defects that are difficult to overcome. First, the safety risk is extremely high, and workers operate in a swaying environment on the dock or at sea, which is prone to safety accidents such as falling. Second, the operation efficiency is low, and the surface area of the ship is huge, and it takes a lot of time and labor cost to build and move the scaffolding covering the hull, especially the large-curvature areas such as the bow and stern. Third, the uniformity of the spraying quality is difficult to guarantee. Large ships, especially the bulbous bow at the bow and the area near the propeller at the stern, present a complex three-dimensional free curved surface (hyperboloid) morphology. When workers manually hold the spray gun, it is difficult to maintain a constant distance and vertical spraying angle between the spray gun and the work surface on such a large-scale and continuously changing surface.

[0004] In order to solve the above problems, automated spraying technology has been introduced into the field of shipbuilding and maintenance. Some early automation attempts, such as robot systems based on ground tracks or large gantries, have achieved certain results in spraying standard planar segments in shipyards. However, these large fixed or track-based robot systems have a great limitation in terms of operation range and flexibility. For a fully formed whole ship, especially in areas with complex morphology, high altitude or large span on the hull shell, these devices are too long to reach and cannot achieve effective coverage.

[0005] In recent years, as a kind of high-mobility aerial platform, unmanned aerial vehicle system provides a new technical path for solving the automatic operation of large complex surface. In theory, unmanned aerial vehicle can carry a spraying unit, and by virtue of its flexible flight ability in three-dimensional space, it can fly to any position that is difficult for artificial or traditional robots to reach, thereby greatly reducing the dependence on scaffolding, improving the safety of operation and shortening the construction period. However, it is still a serious technical bottleneck to transform unmanned aerial vehicle from a simple flight platform to a precise spraying executor when facing large complex curved surface. Most of the existing unmanned aerial vehicle spraying schemes are equipped with only a single-point ranging sensor to maintain a single distance between the unmanned aerial vehicle and the target surface. This single-point ranging logic can barely maintain the distance when facing a plane, but it is completely unable to meet the requirements of high-quality spraying when facing the complex surface of the bow and stern of a large ship with significant curvature and inclination changes. Single-point ranging cannot perceive the local geometry of the target surface, especially the local vertical direction. Non-vertical spraying (i.e. non-normal spraying) will cause the spraying fan to produce distortion and irregular coverage on the workpiece, thereby causing serious unevenness in coating thickness, partial area over-thin leading to failure of protection, partial area over-thick leading to sagging and cracking, and sharply increasing paint mist rebound over-spraying, resulting in a large amount of waste of paint and environmental pollution. SUMMARY

[0006] The present application provides an automatic spraying method for self-adaptive adjustment of spraying posture, the method comprising:

[0007] controlling the unmanned aerial vehicle to perform stable flight tracking according to the preset global path waypoints;

[0008] during the stable flight tracking, continuously collecting synchronous data frames of the actual surface of the target by the five-point laser sensor array;

[0009] real-time solving the posture correction angle based on the synchronous data frames; controlling the two-axis gimbal to execute the posture correction angle to adjust the spraying posture of the spraying unit;

[0010] real-time calculating the flatness error based on the synchronous data frames; comparing the flatness error with the preset flatness threshold to make a spraying decision;

[0011] when the flatness error is not greater than the flatness threshold, determining that the actual surface is flat, and performing one-time spraying; when the flatness error is greater than the flatness threshold, determining that the actual surface is complex, and performing adaptive partition spraying.

[0012] The five-point laser sensor array comprises four corner lasers and one center laser ; the form of the synchronous data frames at moment is:

[0013] wherein, The four corner lasers are respectively located at Distance measured at any time For the center point laser in The actual center distance measured at any given time.

[0014] The steps for real-time calculation of attitude correction angles include: calculating the horizontal distance difference. and vertical distance difference :

[0015]

[0016]

[0017] Calculate the attitude correction angles, including the yaw correction angle. and pitch correction angle :

[0018] in, Given the known horizontal width of the sensor array, The known vertical height of the sensor array.

[0019] The steps for real-time calculation of smoothness error include: calculating the expected center distance. :

[0020] Calculate the smoothness error : .

[0021] The steps for performing a single spray coating include: activating the spray coating unit; while the spray coating unit remains on, the stable flight tracking step, the solution and control step, and the calculation and comparison step are executed in parallel and continuously to achieve coordinated spray coating flight.

[0022] The steps involved in performing adaptive zone spraying, which divides the area into multiple sub-regions, include: when the smoothness error is greater than a smoothness threshold... Calculate the horizontal distortion at time. and vertical twist :

[0023]

[0024] Will and With the distortion threshold Compare and make a partitioning decision based on horizontal two-partitioning, vertical two-partitioning, or four-quadrant partitioning.

[0025] The steps for controlling the drone to move sequentially and align with the sub-regions include: in the sensor array coordinate system Define the sub-region center offset vector corresponding to the partitioning decision; obtain the... Real-time drone body coordinate system Relative to the sensor array coordinate system rotation matrix The sub-region center offset vector is rotated using a rotation matrix. Transformed into an airframe offset vector; based on the UAV's current position when paused. Based on the body offset vector, calculate the hovering point of the corresponding sub-region. Control the drone to move to the hovering point in the sub-region. .

[0026] The steps for performing spot spraying include: while the drone is hovering, activating the spraying unit for a preset spot spraying duration. ; Based on sub-region area and nominal volumetric flow rate calculate:

[0027]

[0028] in, For the target coating thickness, This represents the number of sub-regions in the partition.

[0029] The steps for controlling a two-axis gimbal to perform attitude correction include: setting the pitch axis control error. and yaw axis control error :

[0030]

[0031] The PID control algorithm is used to calculate control commands, such as the control command for the pitch axis. :

[0032] in, For the pitch-axis PID controller, It is the integral variable.

[0033] The present invention also provides an automated spraying system that adapts to adjusting the spraying posture, the system comprising:

[0034] Central control module: Controls the UAV to perform stable flight tracking according to the preset global path waypoints;

[0035] Acquisition module: During stable flight tracking, it continuously acquires synchronous data frames of the target's actual surface through a five-point laser sensor array;

[0036] Attitude calculation module: calculates attitude correction angle in real time based on synchronous data frames; controls the two-axis gimbal to perform attitude correction angle to adjust the spraying attitude of the spraying unit;

[0037] Decision module: Calculates smoothness error in real time based on synchronous data frames; compares the smoothness error with the preset smoothness threshold to make a spraying decision;

[0038] When the smoothness error is not greater than the smoothness threshold, the actual surface is determined to be smooth, and a one-time spraying is performed; when the smoothness error is greater than the smoothness threshold, the actual surface is determined to be complex, and adaptive zone spraying is performed.

[0039] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the aforementioned automated spraying method that adaptively adjusts the spraying posture.

[0040] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned automated spraying method for adaptively adjusting spraying posture.

[0041] This invention provides an automated spraying method and system with adaptive spraying attitude adjustment, aiming to solve the fundamental contradiction in existing UAV spraying systems when dealing with large, complex three-dimensional curved surfaces such as large ship hulls, where it is difficult to balance flight safety, operational efficiency, and spraying quality. The beneficial effects of this invention are reflected in the following aspects:

[0042] This invention achieves a fundamental decoupling of flight platform stability and spraying attitude accuracy, significantly improving operational safety. In existing technologies, attempting to align the UAV's body attitude with the surface normal results in severe interference with flight stability due to the UAV's high inertia and slow response. This invention completely solves this problem through a separate design of global path flight and two-axis gimbal attitude adjustment. In this invention, the UAV's Flight Control System (FCS) is responsible for only one task: tracking a pre-planned, smooth global path with sufficient safety margin. The UAV body does not need to perform high-frequency flight attitude adjustments based on the local geometry of the target surface. This maximizes the protection of flight stability and operational safety for the UAV in complex windy environments such as shipyards.

[0043] Based on a dynamic reference for maintaining stable flight of an unmanned aerial vehicle (UAV) platform, this invention utilizes an integrated sensing and actuation subsystem to independently calculate the local pitch and yaw angles of the actual surface in real time using data from a rigidly connected five-point laser sensor array, particularly from the four corner lasers. This calculation speed is much faster than the inertia of the UAV body. Subsequently, this angular deviation is executed at high speed by a two-axis gimbal servo controller, achieving independent and rapid compensation for the spraying attitude of the spraying unit. This design ensures that the spraying direction of the spray gun remains aligned with the normal of the hull surface with large curvature changes.

[0044] The assumption of a locally approximate plane fails when dealing with welds, stiffeners, or sharp angles. This invention provides an independent verification basis through a "center point laser" in a "five-point" array. By comparing the expected center distance with the actual center distance, the smoothness error is calculated in real time, thereby quantifying the reliability of the calculation result. Decisions are made based on this reliability. This mechanism of verification before decision-making ensures that the spraying operation is only performed after confirming reliable alignment, fundamentally eliminating blind and erroneous spraying in areas of geometric abrupt change.

[0045] The system's collaborative spraying flight maximizes efficiency by maintaining normal alignment while flying at high speed. When encountering complex geometry that leads to unreliable solutions, the system doesn't simply skip the problem but automatically switches to an adaptive partitioned spraying strategy. Through pauses, partitioned decision-making, and sub-region cyclic processing, large-scale, unreliable regions are decomposed into smaller, verifiable sub-regions. For each sub-region, it re-executes the perception-solution-verification loop, and spot spraying is also performed only after the normal attitude has been confirmed as reliable. This quality-first recursive processing mechanism, combined with a safe exit function based on abnormal region logs, ensures overall operational efficiency while also possessing robustness in handling extremely complex geometry, achieving full-process automation and high-quality spraying. Attached Figure Description

[0046] Figure 1 This is a flowchart of an automated spraying process that adaptively adjusts the spraying posture. Detailed Implementation

[0047] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. This application can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0048] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this application, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number and aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0049] Additionally, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that practice can be carried out without these specific details.

[0050] This embodiment details the process of initializing the global path flight path in the automated spraying method of the present invention. The purpose of initializing the global path flight path is to provide a safe, stable, and repeatable macroscopic reference trajectory for the entire automated spraying operation. Before initializing the global path flight path, the system pre-acquires a coarse 3D mesh model of the target to be sprayed and a set of preset spraying process parameters, including the reference working distance and the nominal spraying coverage width. Based on the input 3D model and process parameters, a global flight path covering the entire target surface is calculated offline, and the UAV's flight control system loads and strictly executes the global path.

[0051] The specific steps for initializing the global path flight path include:

[0052] S1: Offline generation of global coverage path, that is, in the preparation stage of this invention, the central controller or background workstation first executes the global path generation algorithm.

[0053] S1.1: Input model and parameter definition, input a rough 3D mesh model of the target ship hull. The model It is a set of a large number of vertices and its normal vector The resulting triangular mesh set. Meanwhile, key process parameters include: baseline safety distance. The baseline safety distance is a preset scalar value, representing the expected average distance between the UAV body and the nominal surface of the ship's hull. The effective coverage width of a single spray, i.e., the spray width... Coating overlap rate .

[0054] S1.2: Generate the offset surface and the distance between the flight path to generate the UAV's flight path, based on... Calculate the offset surface. For any vertex on Its corresponding offset vertex :

[0055]

[0056] in, yes exist The unit normal vector of a point, all This constitutes an offset surface , The surface is the curved surface where the ideal trajectory of the drone's flight lies.

[0057] Next, the lateral spacing between two adjacent global paths, i.e., the flight path spacing, is calculated. :

[0058]

[0059] S1.3: Generate covering path waypoints. This invention employs an arc-shaped path planning method. A continuous path curve is generated on the surface. Sampling is performed along the path curve with a fixed path step size. Preferred, Meters discretize the path curve into global waypoints. .

[0060] each waypoint Each contains a three-dimensional position coordinate. and a three-dimensional attitude direction .here, Depend on Obtaining the coordinates of sampling points on the surface. The desired attitude represents the waypoint. In this embodiment, to ensure the smoothness and stability of the UAV flight, the desired attitude is... Specifically, the drone's Z-axis is set to point towards the Earth's center or the direction of gravity. Parallel, while aligning its X-axis (the fuselage X-axis, i.e., the direction of the nose) with the forward tangent direction of that point on the path. .

[0061] S1.4: Divide the spraying area; the central controller will control the entire global path. Divided into A series of consecutive path segments, each corresponding to a hull. A target spraying area on the surface (like ).

[0062] Execute global path flight

[0063] S2.1: Load the path and start the flight plan. After the UAV system starts, the central controller will load the global path and waypoints generated in S1. Loaded into the flight control system (FCS) of the drone.

[0064] S2.2: Stable flight tracking. After the operator issues the execution command, the UAV's FCS begins to strictly execute the path tracking task.

[0065] In this invention, the UAV's Flight Control System (FCS) is responsible for only one task—tracking a smooth, continuous global path with sufficient safety margin. The UAV does not need to frequently adjust its flight attitude based on localized complex changes on the hull surface. This greatly ensures the flight stability and operational safety of the UAV as a heavy-duty work platform in common complex environments such as shipyards. Simultaneously, the global path ensures that the UAV platform always moves stably within a controllable working area that is approximately parallel to the hull surface. This allows subsequent painting units to perform high-speed, precise attitude compensation within a small range on a relatively stable and predictable dynamic baseline, without having to deal with large swaying of the UAV body. This lays the foundation for achieving high-quality normal-aligned painting.

[0066] Next, this embodiment will describe in detail the process of real-time surface perception and attitude calculation in the automated spraying method of the present invention. In order to obtain complex surface information, continuous data acquisition of the target is required. While the UAV maintains stable flight, a five-point laser sensor array rigidly mounted on a two-axis gimbal is used to acquire local geometric data of the actual surface of the target hull in the spraying direction in front of the UAV.

[0067] To achieve the adaptive adjustment of this invention, this embodiment employs an integrated sensing and actuation subsystem. This subsystem is installed on the UAV airframe, but its motion control is decoupled from the UAV flight control system (FCS).

[0068] Two-axis gimbal integrating sensing and actuation subsystems Spraying unit Five-point laser sensor array Composition. It includes a two-axis gimbal. It is a servo-driven two-dimensional rotary mechanism, whose base is fixedly connected to the drone body. This two-axis gimbal It provides independent rotation with two degrees of freedom: pitch and yaw. (Spraying unit) Including spray guns Spraying unit Rigidly mounted on a two-axis gimbal The movable actuator. Five-point laser sensor array. Including four corner lasers ( ) and a center point laser ( ), These represent the top left, top right, bottom left, and bottom right focal lasers, respectively. The four corner lasers are rigidly mounted on a two-axis gimbal. Above, the center point laser is installed in the spraying unit. spray gun Above, with the spray gun They have the same position and posture.

[0069] UAV body coordinate system It is installed on the drone's body. FCS controls... Position in the global coordinate system and posture .

[0070] Spraying unit coordinate system Located in the spraying unit Its origin is set at the spray gun. At the nozzle, The axis is defined as the theoretical center of the spray gun and the direction of the spray. With two-axis gimbal relative to the movement sports.

[0071] Sensor array coordinate system Laser with origin set at center point launch point Axis is defined as The direction of laser beam emission.

[0072] because and Rigidly fixed, therefore and The relative transformation matrix between It is a constant, known value. Furthermore, this invention ensures the direction of the central laser beam through precise installation and calibration. With respect to the direction of the spray gun Strict parallelism, that is .

[0073] Four corner lasers The laser beams are all with They are launched parallel to the axis. coordinate system plane (i.e.) The projected coordinates on the plane are defined as follows: coordinate: , coordinate: , coordinate: , coordinate: .in, Given the known horizontal width of the sensor array, The known vertical height of the sensor array. Indicates in Coordinates in a coordinate system.

[0074] Step S3.2: Continuous Data Acquisition

[0075] Throughout the entire global path flight executed by the UAV's FCS, the central controller simultaneously performs high-frequency data acquisition using a five-point laser sensor array. A sampling frequency higher than the gimbal control frequency and the drone flight control frequency Continue working.

[0076] At any sampling time , array The five lasers in the middle ( Simultaneously, they launch towards the actual surface of the target hull along their respective launch directions. A laser beam is emitted. The laser beam encounters... The reflection occurs behind the surface, and the receivers of each laser capture the echo and calculate their respective flight times, thus obtaining five independent distance measurements.

[0077] Central controller in The synchronization data frame is acquired and encapsulated at all times. :

[0078]

[0079] in: , , as well as They represent , , , laser in The time measured along the edge Direction to The distance between surfaces.

[0080] The data acquisition process occurs when a drone is spraying data from a target area. Fly to the next The process is executed continuously throughout, generating a continuous data stream, which is then transmitted to the central controller.

[0081] Used for planning global paths It is sparse and coarse, possibly lacking local details such as welds, deformations, or attachments. Through high-frequency... Sampling obtains the actual physical surface of the ship's hull. The millimeter-level precision distance data provides a real online data foundation for attitude adjustment, which is a prerequisite for achieving high-quality spraying. In this embodiment, data acquisition and global path flight are executed in parallel. The UAV's FCS is not concerned with local surface details, but only with macroscopic stable flight; while It doesn't concern itself with the global path, only measuring the local geometry directly in front of it. This decoupled design ensures that the drone doesn't experience flight jitter from attempting to respond to surface details, significantly improving operational safety and stability. Output It's not a single distance value, but rather a collection of five spatially distributed synchronization distance points. These five points uniquely identify... Local surface faced by the coordinate system The center distance, pitch angle, and yaw angle.

[0082] Next, this embodiment details the process of calculating the local plane tilt angle in the automated spraying method of the present invention, and the real-time calculation of the local surface geometric information implied by the five data points. Specifically, it calculates the current sensor array coordinate system. Relative to the actual surface it faces Angular deviation of the local normal direction.

[0083] The central controller at each sampling time Received Then, immediately calculate the angular deviation in the horizontal direction.

[0084] Step S4.1: Calculate the horizontal angular deviation, i.e., the yaw angle;

[0085] The purpose of calculating the angular deviation in the horizontal direction is to calculate the actual surface... exist The relative tilt angle in the XZ plane of the coordinate system, which corresponds to the two-axis gimbal. Yaw angle that needs to be compensated.

[0086] S4.1.1 Calculate the average distance to the left and right:

[0087] Controller Extraction Data from the four corner points of the sensor array. Calculate the sensor array. Left side ( average distance and the right side ( The average distance .

[0088]

[0089]

[0090] S4.1.2 Calculate the horizontal distance difference:

[0091] calculate Horizontal distance difference at time :

[0092]

[0093] S4.1.3 Calculate the yaw correction angle:

[0094] Value and the known horizontal width of the sensor array These form two sides of a right triangle. The central controller calculates the yaw correction angle necessary to achieve normal alignment using the arctangent function. :

[0095]

[0096] when When the left side is farther away and the right side is closer (for example, when facing the right convex curved surface of the hull), it indicates that the distance is farther on the left and closer on the right (e.g., when facing the right convex curved surface of the hull). A positive value indicates the gimbal. You need to rotate to the right (positive yaw direction) to align with the normal.

[0097] S4.2: Calculate the vertical angular deviation, i.e., the pitch angle:

[0098] The purpose of calculating the angular deviation in the vertical direction is to calculate the actual surface. ,exist The relative tilt angle in the YZ plane of the coordinate system, which corresponds to the two-axis gimbal. The pitch angle that needs to be compensated.

[0099] S4.2.1 Calculate the average vertical distance:

[0100] Controller utilizes Computational sensor array upper side ( The average distance and the lower side ( The average distance .

[0101]

[0102]

[0103] S4.2.2 Calculate the vertical distance difference: Calculate Vertical distance difference at time :

[0104]

[0105] S4.2.3 Calculate the pitch correction angle: Value and the known vertical height of the sensor array These form two sides of another right triangle. The central controller calculates the pitch correction angle necessary to achieve normal alignment using the arctangent function. :

[0106]

[0107] when When the upper side is farther away and the lower side is closer (for example, when facing the curved surface of the lower part of the hull), it indicates that the upper side is farther away and the lower side is closer (for example, when facing the curved surface of the lower part of the hull). A positive value indicates the gimbal. It needs to be rotated upwards (in the positive pitch direction) to align with the normal.

[0108] S4.3: Output correction command. The central controller will use the values ​​calculated by S4.1 and S4.2 to perform the correction. and As a binary correction command vector, it is output to the two-axis gimbal. Servo controller.

[0109] This invention does not rely on a coarse 3D mesh model. To guide the attitude. Instead, the adjustment in this embodiment utilizes the actual surface sampled at high frequency. The data precisely decomposes the complex three-dimensional curved surface of large ships (such as bulbous bows or sterns) into two orthogonal angular deviation values ​​in a local manner. Meanwhile, the algorithm used in this embodiment only involves a few additions / subtractions, divisions, and two... The lightweight nature of this computation ensures that the central controller can complete the calculation within an extremely short latency (milliseconds), a processing speed far faster than the inertial response of the drone's airframe or a two-axis gimbal. The mechanical response enables high-bandwidth closed-loop control.

[0110] This embodiment will now describe in detail the gimbal attitude adjustment performed in the automated spraying method of the present invention. The purpose is to adjust the angle deviation calculated in S4 using a two-axis gimbal. The servo motion is compensated and eliminated in real time. This embodiment constructs a high-bandwidth closed-loop servo control system, which... and As an error input, drive The pitch and yaw motors are operated until the error converges to zero.

[0111] S5.1: Set control error;

[0112] The servo controller receives S4 in The correction angle calculated at each moment is directly defined as the current angle. Control error signal at any time:

[0113] Yaw axis control error :

[0114] Pitch axis control error :

[0115] Step S5.2: Execute PID servo control;

[0116] To achieve rapid and stable elimination of errors, the servo controller preferably adopts a dual-axis independent PID control algorithm.

[0117] Servo controller according to calculate The time should be applied to control output of the pitch axis motor :

[0118]

[0119] in, This indicates the control command for the pitch axis motor, i.e., the torque command. , which are the proportional, integral, and derivative gain coefficients of the pitch-axis PID controller, respectively. This represents the integral variable.

[0120] At the same time, the servo controller according to calculate The time should be applied to Control output of yaw motor :

[0121]

[0122] in, This indicates the control command for the yaw axis motor, i.e., the torque command. These are the proportional, integral, and derivative gain coefficients of the yaw axis PID controller, respectively.

[0123] Next, this embodiment details the calculation of the smoothness verification value in the automated spraying method of the present invention. The purpose is to perform real-time quantization verification of the local approximate plane assumption upon which the S4 algorithm relies, while the attitude correction angle is calculated using the four corner point data in S4 and the adjustment is performed in S5.

[0124] In the central controller Get it in real time While executing controls S4 and S5, the central controller simultaneously executes step S6:

[0125] S6.1: Calculate the expected center distance

[0126] S4 assumes a five-point laser sensor array. The actual surface being faced The region can be fitted by a local approximation plane. The purpose of S6.1 is to calculate, under this local approximation plane assumption, the region located at... Origin The theoretical distance value on the axis.

[0127] To obtain this theoretical value, the central controller adopts... time The arithmetic mean of the distances between the four corner points is used as the basis for calculation. Expected center distance at time :

[0128]

[0129] S6.2: Obtain the actual center distance

[0130] The central controller uses the same synchronization data frame as S6.1. In the middle, direct extraction from the center point laser Measured actual center distance .

[0131] S6.3: Calculate the smoothness error

[0132] Central controller calculation and The absolute difference between them is obtained Time smoothness error :

[0133]

[0134] Quantified the actual surface exist The vertical distance at the center point from the local approximate plane defined by the four corner points.

[0135] Step S4 itself cannot distinguish between a smooth ramp and a rough surface. In this embodiment, S6 introduces a center point. with corner average Independent comparisons solved this problem. In spraying applications on complex three-dimensional curved surfaces such as target ship hulls, when a five-point laser sensor array... Cross When there are welds, reinforcing ribs, or sharp corners on the surface, The local geometry deviates significantly from the local approximate plane assumption. In this case, the four corner points may fall within a smooth region, resulting in calculated... The performance is normal; however, the central laser... Located on a protrusion (leading to) ) or dent (causing) (above). This embodiment will immediately calculate a high... Value. This. The significance of the signal lies in the fact that it quantizes the output of S4. and And the reliability of the attitude adjustments performed by S5. A high Value means If the geometry of the area being addressed is too complex, attitude alignment based on a local approximate plane is unreliable, and painting should not be performed in this case. It is ensured that painting is only performed when the attitude alignment is confirmed as valid by S6 (i.e., ...). It is only executed when the surface is small enough, which fundamentally guarantees the coating quality of complex curved surfaces.

[0136] Next, this embodiment details the application of the verification result of S6 in the automated spraying method of the present invention. Time ( ) for the spraying unit It makes real-time adaptive decisions regarding starting and stopping.

[0137] Step S7: Spraying decision;

[0138] S7.1: Define the smoothness threshold. During system initialization, a key parameter, namely the smoothness threshold, is preset. . As a scalar value, it represents The acceptable upper limit. The physical meaning of the value is in the five-point laser sensor array of Within the sampling area, the actual surface area that the system can tolerate The maximum vertical distance from the local approximate plane. In spray painting applications on the target hull, this value should be less than the typical height of welds or stiffeners to ensure sensitive detection of these complex geometric features that are unsuitable for spraying.

[0139] S7.2: Perform real-time comparisons and decisions;

[0140] exist At any given moment, the central controller acquires the calculated... And immediately perform the following comparison:

[0141] Scenario 1 (Smooth): If

[0142] Central controller determines: Current Facing The region is flat, and the local approximate plane assumption of S4 holds. Therefore, the attitude commands solved by S4 and the attitude alignment executed by S5 are reliable.

[0143] The central controller sets an internal status flag: .

[0144] Scenario 2 (Complex): If

[0145] Central controller determines: Current The area exhibits significant local bulges or depressions.

[0146] Therefore, attitude alignment based on a local approximate plane is unreliable.

[0147] The central controller sets internal status flags: .

[0148] Step S8: Perform a one-time spraying of the flat area in Execution 1

[0149] This step is only determined in S7. It is executed and maintained at that time.

[0150] S8.1: Activate the spraying unit

[0151] Central controller to spraying unit The actuator valve sends an opening command to the spray gun. Begin spraying the paint.

[0152] S8.2: Performing coordinated spraying flights

[0153] exist While in the active state, the UAV's Flight Control System (FCS) continues to strictly perform stable flight tracking tasks, driving the UAV along the global path waypoints. Smoothly fly over the current target spraying area The S3-S4-S5 closed-loop operation continues. The PID controller continuously receives the error input from S4 and drives the two-axis gimbal. Real-time rotation to compensate for the effects of drone flight and Attitude deviation caused by surface curvature changes. Real-time verification continues. As long as the UAV is... When flying within the area Always maintain , It will always remain at 1, and the spraying command will always remain enabled.

[0154] S8.3: Regional Completion and Transition

[0155] When the drone's flight control system (FCS) detects that the drone has arrived When reaching the global pathpoint at the end of the region, the central controller sends... Send a shutdown command.

[0156] Subsequently, the FCS controlled the drone to fly along the path to the next target painting area. The system begins to perceive the new area and prepares to repeat the solution and decision-making process of S4-S7.

[0157] Next, this embodiment describes in detail the adaptive zone spraying process in the automated spraying method of the present invention.

[0158] When the judgment ,Right now At that time, adaptive zone spraying is triggered immediately.

[0159] S9: Execution Status 2: Adaptive Zone Spraying;

[0160] The purpose of adaptive zoning spraying is to pause the one-time spraying process of S8 when the relied-upon local approximate plane assumption fails, and instead execute a coarse-to-fine adaptive zoning processing strategy. This strategy involves adjusting the current complex target spraying area... The process is decomposed into multiple smaller sub-regions, and a perception-solution-verification loop (S3-S7) is performed independently for each sub-region in order to satisfy the local approximate plane assumption at a smaller scale, ultimately achieving high-quality point spray coverage of complex geometric surfaces.

[0161] Adaptive zoning spraying includes: (1) complexity analysis and zoning decision; (2) sub-region target point generation; (3) sub-region cyclic processing (movement-verification-point spraying).

[0162] S9.1: Spraying suppression and global path pause;

[0163] exist At any moment, the central controller immediately sets... Ensure the spraying unit Keep it off.

[0164] The central controller sends a pause and hover command to the drone's flight control system (FCS). The FCS immediately pauses the stable flight tracking mission and controls the drone to hover stably in its current position.

[0165] S9.2: Execute partitioning strategy decision

[0166] Central controller based on Analysis leads to The geometric origin of the data is used to determine which partitioning strategy to adopt. Indicates leading to This triggers the adaptive zone spraying process.

[0167] Define complex metrics:

[0168] Horizontal Twist Compare the left side ( to The vertical distance difference between the right and the right side ( to The vertical distance difference.

[0169]

[0170] Vertical Twist Compare the upper side ( to The horizontal distance difference between the two sides and the lower side ( to (Horizontal distance difference)

[0171]

[0172] Define the distortion threshold Used for determination Is there significant distortion or tilt within the coverage area?

[0173] IF AND

[0174] The surface will then exhibit significant distortion in the horizontal direction, for example, across a vertical weld or vertical stiffener.

[0175] Zoning decision Horizontal two-part division (left) ,right ).

[0176] IF AND

[0177] The surface will then exhibit significant distortion in the vertical direction, for example, across a horizontal weld seam.

[0178] Zoning decision Vertical two-partition (upper) ,Down ).

[0179] ELSE (IF AND OR (Extremely high)

[0180] The surface then exhibits complex distortions in both directions, such as at corners, or Extremely large, for example It hit a single, isolated protrusion.

[0181] Zoning decision Four-quadrant partitioning ( , , , ).

[0182] Step S9.3: Generate target points for the sub-region

[0183] The central controller is based on S9.2. Calculate the list of hovering points in the sub-regions that the drone's FCS needs to move to in sequence. .

[0184] calculate Offset vector in coordinate system:

[0185] (Left sub-region center offset)

[0186] (Right sub-region center offset)

[0187] (Center offset of the upper sub-region)

[0188] (Sub-region center offset)

[0189] according to ,and and Movable end alignment arrive rotation matrix Equivalent to .

[0190]

[0191]

[0192]

[0193] Generate FCS target point list FCS controls Drones need to be in Perform translation in the coordinate system.

[0194]

[0195] IF = Horizontal two-part division: in

[0196] IF = Vertical two-part partitioning: in

[0197] IF = Four-quadrant partitioning: in,

[0198] Step S9.4: Execute sub-region loop processing. The central controller enters the iteration loop and traverses the list. Hover points in each sub-region .

[0199] S9.4.1 Move to Sub-area: The central controller sends a move and hover command to the FCS, with the target point being... FCS controls the drone to move to And then it hovered steadily again.

[0200] S9.4.2 Re-verify:

[0201] drones in After stable hovering, the central controller reactivates the S3-S7 complete process: acquiring new data frames. Calculate the new attitude correction angle , .drive Adjust to the new normal orientation. Calculate the new smoothness error. .

[0202] Re-decision: IF (Sub-region is flat): Determination Corresponding sub-region Reliable, settings .

[0203] IF (The sub-region remains complex): Determination Corresponding sub-region Still unreliable, settings .

[0204] S9.4.3 Perform sub-area spraying or skip:

[0205] IF The sub-region passed the verification, and the attitude adjustment is reliable.

[0206] Execute Spot Spray: The central controller executes S8.1 and activates. Because the drone was in a hovering state ( The S8.2's cooperative flight was transformed into "timed spraying." The central controller maintained... Enable a preset spray duration. .

[0207] The calculation is based on the area of ​​the sub-region. , ,in The number of partitions is 2 or 4. . The nominal volumetric flow rate of the spray gun under standard pressure. After completion, the controller is turned off. .

[0208] IF Subregion Even after partitioning, the local approximate plane assumption still cannot be satisfied. Execution skip: Central controller remains. Close and then close the sub-region. The coordinates are recorded in the abnormal area log for subsequent manual inspection or processing.

[0209] Step S9.5: Restore global path flight

[0210] when All in the list After all processes have been completed, the central controller determines the complex cases. The area has been processed.

[0211] The central controller sends instructions to the FCS to make it start from the last Returning to the position when paused in S9.1 .

[0212] The central controller sends a path recovery command to the FCS. The FCS exits hovering mode, reactivates the stable flight tracking task, and controls the UAV to continue flying from its current position to the next global waypoint. .

[0213] The system exits S9 and returns to the standard loop of S2-S8.

[0214] This embodiment is the core adaptive mechanism of the present invention. By using a closed-loop process of pause-analysis-partitioning-movement-verification-spot spraying-recovery, it overcomes the limitations of S4 / S5 based on the local approximate plane assumption at complex geometric features such as welds, stiffeners, or sharp angles of large ships.

[0215] When the standard algorithm of S4 / S5 fails ( Instead of simply skipping the region, this invention switches to a more refined and reliable processing strategy.

[0216] Data-driven partitioning ensures that partitioning decisions (horizontal, vertical, or four-quadrant) are not blind but based on the actual geometric features that led to verification failures, improving the success rate of re-verification. Spot spraying ensures more precise control over coating thickness, even in hovering states. Simultaneously, skip logic and abnormal area logging ensure the system has a safe exit mechanism when facing recursive complexity (i.e., sub-regions remain complex), avoiding infinite loops or forced spraying under incorrect postures, thus achieving a quality-first principle on extremely complex surfaces.

[0217] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the aforementioned automated spraying method that adaptively adjusts the spraying posture.

[0218] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned automated spraying method for adaptively adjusting spraying posture.

[0219] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0220] In this specification, the same or similar parts between the various embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the descriptions of the embodiments described later are relatively simple, and relevant parts can be referred to the descriptions of the foregoing embodiments.

[0221] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An automated spraying method for adaptively adjusting spraying attitude, characterized in that the method... include: Control the drone to perform stable flight tracking according to the preset global path waypoints; During stable flight tracking, synchronous data frames of the target's actual surface are continuously collected through a five-point laser sensor array; The attitude correction angle is calculated in real time based on the synchronous data frame; the two-axis gimbal is controlled to perform attitude correction angle to adjust the spraying attitude of the spraying unit. The smoothness error is calculated in real time based on the synchronous data frame; the smoothness error is compared with the preset smoothness threshold to make a spraying decision; When the smoothness error is not greater than the smoothness threshold, the actual surface is determined to be smooth, and a one-time spraying is performed; When the smoothness error exceeds the smoothness threshold, the actual surface is determined to be complex, and adaptive zone spraying is performed.

2. The automated spraying method with adaptive adjustment of spraying attitude according to claim 1, characterized in that, The five-point laser sensor array includes four corner lasers. and a center point laser Synchronization data frames in The time format is: in, The four corner lasers are respectively located at Distance measured at any time For the center point laser in The actual center distance measured at any given time.

3. The automated spraying method for adaptively adjusting spraying attitude according to claim 2, characterized in that, The steps for real-time calculation of attitude correction angles include: calculating the horizontal distance difference. and vertical distance difference : Calculate the attitude correction angles, including the yaw correction angle. and pitch correction angle : in, Given the known horizontal width of the sensor array, The known vertical height of the sensor array.

4. The automated spraying method for adaptively adjusting spraying attitude according to claim 2, characterized in that, The steps for real-time calculation of smoothness error include: calculating the expected center distance. : Calculate the smoothness error : .

5. An automated spraying method for adaptively adjusting spraying posture according to claim 1 or 4, characterized in that, The steps for performing a single spray coating include: activating the spray coating unit; while the spray coating unit remains on, the stable flight tracking step, the solution and control step, and the calculation and comparison step are executed in parallel and continuously to achieve coordinated spray coating flight.

6. An automated spraying method for adaptively adjusting spraying posture according to claim 2 or 4, characterized in that, The steps involved in performing adaptive zone spraying, which divides the area into multiple sub-regions, include: when the smoothness error is greater than a smoothness threshold... Calculate the horizontal distortion at time. and vertical twist : Will and With the distortion threshold Compare and make a partitioning decision based on horizontal two-partitioning, vertical two-partitioning, or four-quadrant partitioning.

7. An automated spraying method for adaptively adjusting spraying attitude according to claim 6, characterized in that, The steps for controlling the drone to move sequentially and align with the sub-regions include: in the sensor array coordinate system Define the sub-region center offset vector corresponding to the partitioning decision; obtain the... Real-time drone body coordinate system Relative to the sensor array coordinate system rotation matrix The sub-region center offset vector is rotated using a rotation matrix. Transformed into an airframe offset vector; based on the UAV's current position when paused. Based on the body offset vector, calculate the hovering point of the corresponding sub-region. Control the drone to move to the hovering point in the sub-region. .

8. An automated spraying method for adaptively adjusting spraying attitude according to claim 7, characterized in that, The steps for performing spot spraying include: while the drone is hovering, activating the spraying unit for a preset spot spraying duration. ; Based on sub-region area and nominal volumetric flow rate calculate: in, For the target coating thickness, This represents the number of sub-regions in the partition.

9. An automated spraying method for adaptively adjusting spraying attitude according to claim 3, characterized in that, The steps for controlling a two-axis gimbal to perform attitude correction include: setting the pitch axis control error. and yaw axis control error : The control commands, specifically the pitch axis control commands, are calculated using a PID control algorithm. : in, For the pitch-axis PID controller, It is the integral variable.

10. An automated spraying system adaptable to adjusting spraying posture, characterized in that, The system includes: Central control module: Controls the UAV to perform stable flight tracking according to the preset global path waypoints; Acquisition module: During stable flight tracking, it continuously acquires synchronous data frames of the target's actual surface through a five-point laser sensor array; Attitude calculation module: calculates attitude correction angle in real time based on synchronous data frames; controls the two-axis gimbal to perform attitude correction angle to adjust the spraying attitude of the spraying unit; Decision module: Calculates smoothness error in real time based on synchronous data frames; compares the smoothness error with a preset smoothness threshold to make a spraying decision; When the smoothness error is not greater than the smoothness threshold, the actual surface is determined to be smooth, and a one-time spraying is performed; When the smoothness error exceeds the smoothness threshold, the actual surface is determined to be complex, and adaptive zone spraying is performed.

Images