Method and System for Motion Planning of the Manipulator of a Curtain Wall Cleaning Robot

Abstract

The present invention provides a method and system for the motion planning of the robotic arm of a curtain wall cleaning robot, which relates to the technical field of the motion planning of cleaning robots. By integrating multi-source sensing data and an adaptive control algorithm, the present invention solves the problems of dynamic stability and precise force application in high-altitude curtain wall cleaning. The pollution recognition model based on the weighted analysis of the HSV color space can quantify the severity of different pollution types and provide a reliable basis for differential cleaning strategies; the multi-joint dynamic wind pressure compensation mechanism combined with the feedforward control of wind speed changes improves the trajectory tracking accuracy of the robotic arm in a strong wind environment; the closed-loop self-calibration system ensures the stability of long-term operation through real-time offset detection and parameter optimization. Thus, the balance among cleaning efficiency, curtain wall protection, and energy consumption control is achieved, and it is more applicable to the densely populated areas of urban high-rise buildings and chemical pollution-sensitive areas, providing an innovative solution for the reliable operation of intelligent cleaning equipment under complex high-altitude working conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of motion planning for cleaning robots, and specifically to a method and system for motion planning of the robotic arm of a curtain wall cleaning robot. Background Technique

[0002] With the large-scale application of high-rise building curtain walls, automated cleaning robots have gradually replaced manual high-altitude operations. However, existing curtain wall cleaning robots face two major technical bottlenecks: First, in complex wind field environments (such as strong side winds and instantaneous gusts caused by the urban canyon effect), the robotic arm generates trajectory deviations due to wind pressure disturbances, resulting in a decrease in cleaning coverage rate and even collision risks; Second, traditional cleaning parameters are fixed and cannot dynamically adjust the brush pressure according to the type (such as dust accumulation, oil stain attachment, chemical corrosion) and severity of the pollution on the curtain wall surface, causing insufficient cleaning or excessive wear.

[0003] In the prior art, the publication number CN114795033A discloses a method and system for real-time planning and control of a robotic arm for indoor three-dimensional cleaning, including a mobile chassis module, a robotic arm module, and a vision module; after receiving a task instruction, the robotic arm module starts to work, and the vision module transmits the information of the object to be cleaned and the surrounding environment information to the robotic arm module. If the robotic arm has completed the preset cleaning path, then the work process ends and the task is completed; if the robotic arm has not completed the preset cleaning path, it continues to move under the preset path, plans the pose of the robotic arm at the next moment according to the environmental information and the robotic arm state obtained from the vision module, and controls the robotic arm to reach this pose. Although it can constrain the position movement of the robotic arm, it is not applicable to high-altitude environments, and pollution recognition mostly uses grayscale analysis, which cannot distinguish the types of pollutants, resulting in poor targeting of cleaning strategies. In addition, the traditional closed-loop calibration system relies on regular downtime detection and cannot meet the continuous operation requirements, seriously restricting the cleaning efficiency and safety.

[0004] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure, and therefore it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for motion planning of the robotic arm of a curtain wall cleaning robot to solve the problems raised in the above background art.

[0006] To achieve the above purpose, the present invention provides the following technical solutions:

[0007] A method for motion planning of the robotic arm of a curtain wall cleaning robot, the specific steps include:

[0008] S1: Collect the wind pressure data on the surface of the robot arm and the surface image of the curtain wall, and generate a surface pollution index based on the surface image of the curtain wall;

[0009] S2: Generate a dynamic compensation amount based on the wind pressure data on the surface of the robot arm, correct the preset initial motion path according to the dynamic compensation amount, and dynamically adjust the brush head pressure of the robot arm based on the surface pollution index of the curtain wall;

[0010] S3: Repeat steps S1 - S2, detect the offset of the robot arm during movement, and use the closed-loop verification mechanism to calibrate the corrected motion path in real time until the overall cleaning of the curtain wall is completed.

[0011] Preferably, the method for generating the surface pollution index based on the surface image of the curtain wall is as follows:

[0012] Convert the collected surface image into the HSV color space, and calculate the pollution degree value of each pixel point therein. The calculation method is:

[0013] ;

[0014] In the formula , , , respectively represent the pollution degree value, hue value, saturation, and lightness of the pixel point. , , all represent model weights, all three are greater than 0, and ;

[0015] Then calculate the surface pollution index according to the pollution degree value of each pixel point in the surface image. The calculation method is:

[0016] ;

[0017] In the formula represents the surface pollution index, represents the total number of pixel points in the surface image, represents the th pollution degree value of the pixel point, and the subscript represents the index of the pixel point, and .

[0018] Preferably, the wind pressure data includes the pressure difference value and the wind speed change rate. The calculation method for generating the dynamic compensation amount based on the wind pressure data on the surface of the robot arm is:

[0019] ;

[0020] In the formula Represents the dynamic compensation amount of the th joint of the robotic arm, in radians, , are respectively the wind pressure torque coefficient of the th joint of the robotic arm and the wind speed change sensitivity coefficient of the whole robot, Represents the pressure difference at the th detection point, Represents the effective detection area of the th detection point, Represents the distance from the th detection point to the th joint of the robotic arm, Represents the wind speed change rate, , respectively represent the indices of the detection point and the joint, Represents the total number of detection points.

[0021] Preferably, the generation logic of the wind pressure torque coefficient is:

[0022] Measure the offset angles of each joint of the robotic arm at a constant wind speed, and perform fitting using the least squares method. The fitting equation is expressed as:

[0023] ;

[0024] In the formula represents the offset angle of the th joint of the robotic arm during the th measurement. The subscript represents the index of the measurement times, represents the total number of measurements, represents the first fluctuation coefficient, .

[0025] Preferably, the calculation method of the wind speed change sensitivity coefficient is:

[0026] ;

[0027] In the formula represents the second fluctuation coefficient, , , , respectively represent the moment of inertia components of the robotic arm in three directions.

[0028] Preferably, the method for correcting the preset initial motion path according to the dynamic compensation amount is:

[0029] Perform coordinate compensation on the trajectory points on the initial motion path:

[0030] ;

[0031] wherein and respectively represent the horizontal and vertical coordinates of the th trajectory point, represents the third fluctuation coefficient, , represents the dynamic compensation amount of the base joint of the robot manipulator, represents the end attitude angle of the robot manipulator.

[0032] Preferably, the method for dynamically adjusting the brush head pressure of the robot manipulator based on the surface pollution index of the curtain wall is as follows:

[0033] When the surface pollution index satisfies , it is considered that the area is lightly polluted, and the brush head pressure is set to the initial pressure;

[0034] When the surface pollution index satisfies , it is considered that the area is heavily polluted, and the calculation method of the brush head pressure is as follows:

[0035] ;

[0036] wherein represents the pollution threshold, , and respectively represent the brush head pressure and the initial pressure, and both represent empirical coefficients, , .

[0037] Preferably, the method for real-time calibration of the corrected motion path using a closed-loop verification mechanism is as follows:

[0038] Collect the actual motion path of the robot manipulator, calculate the offset based on the actual motion path, and the calculation method is:

[0039] ;

[0040] wherein represents the offset, represents the total number of trajectory points, and respectively represent the horizontal and vertical coordinates of the th trajectory point in the actual motion path;

[0041] When the offset continuously satisfies three times, update the wind pressure torque coefficient of each joint of the robot manipulator and the wind speed change sensitivity coefficient of the whole robot, where represents a preset offset threshold.

[0042] The robotic arm motion planning system for a curtain wall cleaning robot adopts the above-mentioned robotic arm motion planning method for a curtain wall cleaning robot, and specifically includes:

[0043] A data acquisition module, which is used to acquire the wind pressure data on the surface of the robotic arm and the surface image of the curtain wall;

[0044] A dynamic compensation module, which is used to generate a dynamic compensation amount for the wind pressure data on the surface of the robotic arm;

[0045] A trajectory optimization module, which is used to correct the preset initial motion path and adjust the brush head pressure;

[0046] An offset monitoring module, which is used to monitor and calibrate the corrected motion path in real time.

[0047] Compared with the prior art, the beneficial effects of the present invention are:

[0048] By integrating multi-source sensing data and an adaptive control algorithm, the present invention innovatively solves the problems of dynamic stability and precise force application in high-altitude curtain wall cleaning. The pollution recognition model based on weighted analysis in the HSV color space can accurately quantify the severity of different pollution types, providing a reliable basis for differential cleaning strategies; the multi-joint dynamic wind pressure compensation mechanism combined with feedforward control of wind speed changes significantly improves the trajectory tracking accuracy of the robotic arm in strong wind environments; the closed-loop self-calibration system ensures the stability of long-term operation through real-time offset detection and parameter optimization. This solution achieves a balance among cleaning efficiency, curtain wall protection, and energy consumption control, and is especially suitable for urban high-rise building dense areas and chemical pollution sensitive areas, providing an innovative solution for the reliable operation of intelligent cleaning equipment under high-altitude complex working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 is a schematic diagram of the overall method flow of the present invention;

[0050] Figure 2 is a schematic diagram of the overall module structure of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] To make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to specific embodiments.

[0052] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention shall have the ordinary meanings understood by those of ordinary skill in the field to which the present invention pertains. The "first", "second" and similar terms used in the present invention do not denote any order, quantity or importance, but are only used to distinguish different components. The terms such as "comprising" or "including" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. The terms such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "upper", "lower", "left" and "right" are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0053] Embodiment:

[0054] Please refer to Figures 1 to 2 , the present invention provides a technical solution:

[0055] A method for motion planning of the robotic arm of a curtain wall cleaning robot, the specific steps including:

[0056] S1: Collect the wind pressure data on the surface of the robotic arm of the robot, the wind pressure data including the pressure difference value and the wind speed change rate, as well as the surface image of the curtain wall, and generate a surface pollution index according to the surface image of the curtain wall;

[0057] The method for generating a surface pollution index according to the surface image of the curtain wall is as follows:

[0058] Convert the collected surface image into the HSV color space, calculate the pollution degree value of each pixel point therein, and the calculation method is:

[0059] ;

[0060] In the formula , , , respectively represent the pollution degree value, hue value, saturation and lightness of the pixel point, , , all represent model weights, all three are greater than 0, and , the model weights can be determined by the historical data regression analysis method and can also be adjusted according to user requirements. For example, when there is more oil stain pollution on the curtain wall, the value of can be appropriately increased to focus on the hue (H) feature for recognition.

[0061] Then calculate the surface pollution index according to the pollution degree value of each pixel point in the surface image, and the calculation method is:

[0062] ;

[0063] wherein represents the surface contamination index, represents the total number of pixel points in the surface image, represents the contamination degree value of the th pixel point, and the subscript represents the index of the pixel point, and .

[0064] In this embodiment, the wind pressure data can be collected by arranging a differential pressure sensor and a top ultrasonic anemometer on the surface of the robot manipulator, and a pixel industrial camera is used to capture the surface image of the curtain wall.

[0065] In this step, compared with the traditional gray analysis method, the HSV three-channel weighted calculation can accurately distinguish pollution types such as dust accumulation (high S value) and chemical corrosion (low V value), and can greatly improve the accuracy of pollution identification. Moreover, the dynamic weight mechanism can also be optimized through historical data training to avoid the subjective deviation of manually setting weights, providing a reliable basis for subsequent cleaning parameter adjustment.

[0066] S2: Generate a dynamic compensation amount based on the wind pressure data on the surface of the robot manipulator, correct the preset initial motion path according to the dynamic compensation amount, and dynamically adjust the brush pressure of the robot manipulator based on the surface contamination index of the curtain wall.

[0067] The calculation method for generating the dynamic compensation amount based on the wind pressure data on the surface of the robot manipulator is:

[0068] ;

[0069] wherein represents the dynamic compensation amount of the th joint of the robot manipulator, with the unit of radian, , are respectively the wind pressure torque coefficient of the th joint of the robot manipulator and the wind speed change sensitivity coefficient of the whole robot, represents the pressure difference value of the th detection point, represents the effective detection area of the th detection point, represents the distance from the th detection point to the th joint of the robot manipulator, represents the wind speed change rate, , respectively represent the indexes of the detection point and the joint, Indicates the total number of detection points.

[0070] The generation logic of the wind pressure torque coefficient is as follows:

[0071] Measure the offset angles of each joint of the robot manipulator at a constant wind speed, and use the least squares method for fitting. The fitting equation is expressed as:

[0072] ;

[0073] In the formula represents the offset angle of the th joint of the robot manipulator during the th measurement. The subscript represents the index of the measurement times, represents the total number of measurements, represents the first fluctuation coefficient, .

[0074] The calculation method of the wind speed change sensitivity coefficient is as follows:

[0075] ;

[0076] In the formula represents the second fluctuation coefficient, , , , respectively represent the moment of inertia components of the robot manipulator in three directions. These three moment of inertia components can be calculated based on the moment of inertia parameters of the manipulator or obtained through simulation using a 3D software, such as the mass analysis module in SolidWorks.

[0077] In this embodiment, when calculating the initial wind pressure torque coefficient, fitting can be performed through a wind tunnel calibration experiment, and subsequent updates can be made based on the pressure difference values and joint offset angles collected during actual operation. By measuring the wind pressure torque coefficient for each joint of the robot manipulator separately, the influence of wind pressure on different joints can be better described. For example, at the base joint (q = 1), the influence of wind pressure is relatively large, while at the end joint (q = 6), the influence of wind pressure is relatively small. Further, considering manufacturing costs and response speed, the robot is only equipped with an ultrasonic anemometer at the top. Therefore, the wind speed change sensitivity coefficient is considered for the entire robot.

[0078] The method for correcting the preset initial motion path according to the dynamic compensation amount is as follows:

[0079] Perform coordinate compensation on the trajectory points on the initial motion path:

[0080] ;

[0081] In the formula and respectively represent the horizontal and vertical coordinates of the th trajectory point, represents the third fluctuation coefficient, , represents the dynamic compensation amount of the base joint of the robot manipulator, represents the end attitude angle of the robot manipulator, which can be read through the robot's encoder.

[0082] In this embodiment, when performing coordinate compensation, the base joint of the robot manipulator is used as the reference because the base joint has the greatest influence on the displacement of the end of the robot manipulator. Therefore, in the application scenario of curtain wall cleaning that does not require ultra-high precision, such compensation can increase the adjustment speed and response speed.

[0083] The method for dynamically adjusting the brush head pressure of the robot manipulator based on the surface pollution index of the curtain wall is as follows:

[0084] When the surface pollution index satisfies , it is considered that the area is lightly polluted, and the brush head pressure is set to the initial pressure;

[0085] When the surface pollution index satisfies , it is considered that the area is heavily polluted, and the calculation method of the brush head pressure is as follows:

[0086] ;

[0087] In the formula represents the pollution threshold, , and respectively represent the brush head pressure and the initial pressure, and both represent empirical coefficients, , .

[0088] In this step, when compensating for the offset angles of multiple joints, the differential compensation strategy can not only reduce energy consumption but also improve the trajectory accuracy of the end compared with the traditional uniform compensation method. When performing coordinate compensation, using the base joint of the robot manipulator, which has the greatest influence on the displacement of the end of the robot manipulator, as the reference can increase the adjustment speed and response speed in the application scenario of curtain wall cleaning that does not require ultra-high precision. The pressure-pollution degree coupling model breaks through the traditional constant-pressure cleaning mode, automatically increases the pressure in the heavily polluted area, and has higher cleaning efficiency than the uniform force application scheme.

[0089] S3: Repeat steps S1 - S2 to detect the offset of the robotic arm during movement, and use the closed - loop verification mechanism to perform real - time calibration on the corrected movement path until the overall cleaning of the curtain wall is completed.

[0090] The method for performing real - time calibration on the corrected movement path using the closed - loop verification mechanism is as follows:

[0091] Collect the actual movement path of the robotic arm, and calculate the offset based on the actual movement path. The calculation method is:

[0092] ;

[0093] In the formula represents the offset, represents the total number of trajectory points, 、 respectively represent the horizontal and vertical coordinates of the th trajectory point in the actual movement path;

[0094] When the offset continuously satisfies for three consecutive times, update the wind pressure torque coefficient of each joint of the robotic arm and the wind speed change sensitivity coefficient of the whole robot, where represents the preset offset threshold.

[0095] In this step, the actual movement path of the robotic arm can be collected by adding a small amount of fluorescent agent to the cleaner and then using a laser tracker. The closed - loop verification system realizes full - automatic calibration, which can reduce manual intervention.

[0096] The robotic arm motion planning system of the curtain wall cleaning robot adopts the above - mentioned robotic arm motion planning method for curtain wall cleaning, and specifically includes:

[0097] Data acquisition module, which is used to collect the wind pressure data on the surface of the robotic arm and the surface image of the curtain wall;

[0098] Dynamic compensation module, which is used to generate a dynamic compensation amount based on the wind pressure data on the surface of the robotic arm;

[0099] Trajectory optimization module, which is used to correct the preset initial movement path and adjust the brush head pressure;

[0100] Offset monitoring module, which monitors and calibrates the corrected movement path in real time.

[0101] In summary, by integrating multi-source sensing data with an adaptive control algorithm, the present invention innovatively solves the problems of dynamic stability and precise force application in high-altitude curtain wall cleaning. The pollution recognition model based on weighted analysis in the HSV color space can accurately quantify the severity of different pollution types, providing a reliable basis for differential cleaning strategies; the multi-joint dynamic wind pressure compensation mechanism combined with feedforward control of wind speed changes significantly improves the trajectory tracking accuracy of the robotic arm in strong wind environments; the closed-loop self-calibration system ensures the stability of long-term operation through real-time offset detection and parameter optimization. This solution achieves a balance among cleaning efficiency, curtain wall protection, and energy consumption control, and is particularly suitable for urban high-rise building clusters and chemical pollution-sensitive areas, providing an innovative solution for the reliable operation of intelligent cleaning equipment under complex high-altitude working conditions.

[0102] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to obtain a formula that is closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0103] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any arbitrary combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0104] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units. They may be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0105] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and all of them should be covered by the protection scope of this application.

Claims

1. A motion planning method for a curtain wall cleaning robot arm, characterized in that: The specific steps include: S1: Collect wind pressure data on the surface of the robot arm and the surface image of the curtain wall, and generate a surface pollution index based on the surface image of the curtain wall; S2: Generate a dynamic compensation amount based on the wind pressure data on the surface of the robot arm, correct the preset initial motion path according to the dynamic compensation amount, and dynamically adjust the brush head pressure of the robot arm based on the surface pollution index of the curtain wall; The wind pressure data includes the pressure difference value and the wind speed change rate. The calculation method for generating the dynamic compensation amount based on the wind pressure data on the surface of the robot arm is: In the formula Indicates the robot arm The dynamic compensation amount of each joint, in radians, , The robot arm The wind pressure torque coefficient of each joint and the wind speed change sensitivity coefficient of the robot as a whole, Indicates The pressure difference value of each detection point, Indicates The effective detection area of ​​the detection points is Indicates The detection point to the robot arm The distance between the joints, represents the wind speed change rate, , Represent the indexes of detection points and joints respectively, Indicates the total number of detection points; The method for dynamically adjusting the brush head pressure of the robot arm based on the surface contamination index of the curtain wall is: When the surface pollution index satisfy When it is lightly polluted, the brush head pressure is set to the initial pressure; When the surface pollution index satisfy When it is severely polluted, the brush head pressure is calculated as follows: In the formula represents the pollution threshold, , , Represent the brush head pressure and initial pressure respectively, , Both represent empirical coefficients, , ; S3: Repeat steps S1 and S2 to detect the offset of the robot arm during movement, and use a closed-loop verification mechanism to perform real-time calibration on the corrected movement path until the overall cleaning of the curtain wall is completed.

2. The motion planning method for the mechanical arm of a curtain wall cleaning robot according to claim 1, characterized in that: The method for generating the surface pollution index based on the surface image of the curtain wall is: The collected surface image is converted into HSV color space, and the pollution value of each pixel is calculated. The calculation method is: In the formula , , , Respectively represent the pollution value, hue value, saturation and brightness of the pixel point. , , All represent model weights, all are greater than 0, and ; Then, the surface pollution index is calculated according to the pollution value of each pixel in the surface image. The calculation method is: In the formula Represents the surface contamination index, Represents the total number of pixels in the surface image, Indicates The pollution value of each pixel, subscript represents the index of the pixel, and .

3. The motion planning method for the mechanical arm of a curtain wall cleaning robot according to claim 1, characterized in that: The generation logic of the wind pressure torque coefficient is: The offset angles of each joint of the robot arm are measured under constant wind speed, and the least squares method is used for fitting. The fitting equation is expressed as: In the formula Indicates The robot arm is measuring The offset angle of the joint, subscript An index indicating the number of measurements, Indicates the total number of measurements, represents the first fluctuation coefficient, .

4. The motion planning method for the mechanical arm of a curtain wall cleaning robot according to claim 1, characterized in that: The calculation method of the wind speed change sensitivity coefficient is: In the formula represents the second fluctuation coefficient, , , , Represent the inertia components of the robot arm in three directions respectively.

5. The motion planning method for the mechanical arm of a curtain wall cleaning robot according to claim 1, characterized in that: The method for correcting the preset initial motion path according to the dynamic compensation amount is: Coordinate compensation for the trajectory points on the initial motion path: In the formula , Respectively represent The horizontal and vertical coordinates of the trajectory points, represents the third coefficient of fluctuation, , represents the dynamic compensation of the base joint of the robot arm, Represents the end attitude angle of the robot arm.

6. The motion planning method for the mechanical arm of a curtain wall cleaning robot according to claim 1, characterized in that: The method of real-time calibration of the corrected motion path using the closed-loop verification mechanism is: Collect the actual motion path of the robot arm and calculate the offset based on the actual motion path. The calculation method is: In the formula Indicates the offset, represents the total number of trajectory points, , Respectively represent the first The horizontal and vertical coordinates of the trajectory points; When the offset is met for 3 consecutive times When the wind pressure torque coefficient of each joint of the robot arm and the wind speed change sensitivity coefficient of the robot as a whole are updated, Indicates the preset offset threshold.

7. Curtain wall cleaning robot arm motion planning system, characterized by: The curtain wall cleaning robot mechanical arm motion planning system adopts the curtain wall cleaning robot mechanical arm motion planning method according to any one of claims 1 to 6, specifically comprising: A data acquisition module, the data acquisition module is used to collect wind pressure data on the surface of the robot mechanical arm and the surface image of the curtain wall; A dynamic compensation module, which is used to generate dynamic compensation amount based on wind pressure data on the surface of the robot arm; A trajectory optimization module, which is used to correct the preset initial motion path and adjust the brush head pressure; An offset monitoring module is used to monitor and calibrate the corrected motion path in real time.

Images