Construction site data acquisition method, system and terminal based on Flink architecture

Abstract

The invention relates to a Flink architecture-based construction site data acquisition method, system and terminal, and relates to the field of data analysis, and the method comprises the following steps: obtaining a ground top view image taking a tower crane as a center; according to the ground top view image and the ground features, blocking object features are selected from the ground top view image, and a blocking object contour is determined; determining the position of a shelter according to the ground top view image, the shelter features and the reference point, and obtaining a sheltered ground image at the position of the shelter; cutting a ground splicing image from the shielded ground image according to the contour of the shielding object; splicing the ground spliced image to the position of a shielding object of the ground top view image, transparentizing the characteristics of the shielding object according to the contour of the shielding object, and outputting a complete ground image; matching a hoisting route of the tower crane according to the complete ground image, the cargo starting point position and the cargo ending point position; and controlling the tower crane to work according to the hoisting route of the tower crane. In this way, the view blind area of the tower crane can be eliminated, and therefore the working safety of the tower crane is improved.

Description

Technical Field

[0001] The present invention relates to the field of data analysis, and in particular to a construction site data collection method, system and terminal based on Flink architecture. Background Art

[0002] Tower crane is an important equipment on construction sites. It can quickly deliver various materials (such as steel bars, concrete, bricks, etc.) to the construction site to meet the needs of construction projects.

[0003] The height of a tower crane is usually high, and the operator sits in the cockpit on the top of the tower crane to control it. Due to the high height of the operator and the complex environment of the construction site, the tower crane has many blind spots when operating. If there are people in the blind spots blocked by obstructions on the ground, they cannot be discovered in time, which poses a great safety risk. Even if there is a safety officer to assist in remote guidance when the operator is operating the tower crane, due to factors such as information lag and subjective judgment errors of the personnel, the operation of the tower crane still has many uncertainties and dangers and is relatively cumbersome. Summary of the invention

[0004] In order to improve the safety of tower cranes at construction sites, the present invention provides a construction site data collection method, system and terminal based on Flink architecture.

[0005] In a first aspect, the present invention provides a construction site data collection method based on the Flink architecture, which adopts the following technical solution:

[0006] A construction site data collection method based on Flink architecture, including:

[0007] Obtain a ground bird's-eye view image centered on the tower crane;

[0008] According to the ground top view image and the preset ground features, the features of the occluder are selected from the ground top view image, and the outline of the occluder is determined;

[0009] Determine the location of the obstruction based on the ground top view image, the characteristics of the obstruction, and the preset reference point, and obtain the obstructed ground image at the location of the obstruction;

[0010] A ground stitching image is cropped from the occluded ground image according to the outline of the occluder;

[0011] The ground stitching image is stitched at the location of the occlusion in the ground overhead image, and the occlusion features are made transparent according to the occlusion contour to output a complete ground image;

[0012] Match the tower crane's lifting route based on the complete ground image, the preset cargo starting position, and the preset cargo end position;

[0013] Control the tower crane to operate according to its lifting route.

[0014] By adopting the above technical scheme, the blind area of ​​the tower crane or the position covered by the obstruction is photographed by a camera or drone and other equipment, and the captured image is then stitched into the ground overhead image, so that the obstructed area in the ground overhead image can be displayed, which is convenient for the tower crane to check the ground circumferential conditions in real time during the operation process to avoid unexpected situations; and the obstruction is made transparent, so that the position of the obstruction can still be determined from the ground overhead image, so that the tower crane lifting process can be adjusted in real time according to the position of the obstruction, which is more efficient.

[0015] Optionally, if there are obstacles in the lifting route, the tower crane needs to avoid the obstacles and correct the route. The lifting route correction methods include:

[0016] Determine the characteristics and locations of obstacles on the lifting route based on the complete ground image and the lifting route;

[0017] Determine the obstacle feature type according to the two complete ground images and obstacle features acquired within a preset unit time, where the obstacle feature types include fixed obstacles and mobile obstacles;

[0018] Based on fixed obstacles, the lifting height is determined according to the starting position and the end position of the cargo;

[0019] Determine the height of the obstacle based on the complete ground image and the characteristics of the obstacle;

[0020] Calculate the sum of the obstacle height and the preset redundant avoidance distance and define it as the elevated avoidance distance;

[0021] If the lifting height is greater than the lifting avoidance distance, when the cargo moves to the position of the obstruction, the tower crane is controlled to lift the cargo according to the lifting avoidance distance and continue to move along the lifting route;

[0022] If the lifting height is not greater than the lifting avoidance distance, the obstacle outline is determined based on the complete ground image and the characteristics of the obstacle;

[0023] Generate detour routes based on obstacle contours and redundant avoidance distances;

[0024] The lifting route is corrected according to the detour route so that when the cargo moves to the position of the obstruction, it moves along the detour route to bypass the obstruction feature and then continues to move along the lifting route.

[0025] By adopting the above technical solution, when the cargo encounters an obstacle during the lifting process, the detour method of the cargo at the obstacle is determined according to whether the cargo needs to be lifted during the lifting process. If it needs to be lifted, the cargo is lifted in advance to bypass the obstacle from above, and if it does not need to be lifted, it bypasses the obstacle from the side to avoid lifting the cargo and then putting it down again.

[0026] Optionally, during the lifting process, the cargo may be affected by wind and sway. Methods for reducing the swaying of the cargo and correcting the route under the action of wind include:

[0027] Obtain cargo image information, real-time wind speed and real-time wind direction;

[0028] Select cargo features from cargo image information and determine cargo feature outlines;

[0029] Determine the position of the minimum horizontal orthographic projection area and the minimum projection area based on the characteristic contour of the cargo;

[0030] Calculate the angle adjustment value of the horizontal rotation of the cargo based on the real-time wind direction and the position of the minimum horizontal projection area;

[0031] The cargo is controlled to rotate according to the angle adjustment value so that the position with the smallest horizontal projection area is aligned with the direction of the real-time wind direction;

[0032] Determine the height of the cargo from the ground based on the cargo image information and cargo features, and determine the length of the lifting rope based on the height of the cargo from the ground and the preset tower crane height;

[0033] Calculate the cargo swing angle based on the length of the sling, the minimum projection area and the real-time wind speed, and determine the cargo swing distance based on the cargo swing angle and the length of the sling;

[0034] Generate a wind bypass route based on the obstacle contour, redundant avoidance distance, and cargo swing distance;

[0035] The lifting route is corrected according to the wind detour route so that when the cargo moves to the obstruction position, it moves along the detour route to bypass the obstruction feature and then continues to move along the lifting route.

[0036] By adopting the above technical solution, if crosswinds are encountered during the lifting process, the crosswinds will drive the cargo to swing. By analyzing the crosswind parameters and cargo parameters, the minimum projection surface of the cargo can be rotated to face the wind direction. At this time, the swing amplitude of the cargo is minimized. Then the cargo lifting route is corrected, so that the cargo is not prone to collision with obstacles during lifting, which is safer.

[0037] Optionally, a windshield is preset on the sling, and the windshield can be unfolded to form a streamlined windshield surface; methods for further reducing the wind resistance of the cargo include:

[0038] Determine the overall height and width of the cargo based on cargo image information and cargo features;

[0039] Match the unfolded width of the wind deflector to the overall width of the cargo;

[0040] Controlling the windshield to unfold according to the unfolding width, and acquiring an unfolding image of the windshield;

[0041] Determine the windshield surface orientation of the windshield according to the unfolded image of the windshield and the reference point;

[0042] Determine the rotation angle of the windshield according to the real-time wind direction and the orientation of the windshield surface;

[0043] The wind shield is controlled to rotate according to its rotation angle until the wind shield surface is facing the real-time wind direction, and the wind shield is controlled to descend according to the overall height of the cargo to shield the cargo features.

[0044] By adopting the above technical solution, when the wind resistance cannot be reduced by rotating the cargo, the wind shield provided on the sling can be unfolded to block the cargo, and the streamlined surface of the wind shield can guide the wind, thereby reducing the wind resistance.

[0045] Optional solutions when the wind shield cannot completely shield the cargo include:

[0046] When the overall height of the cargo is greater than the preset windshield height, the difference between the overall height of the cargo and the windshield height is calculated and defined as the unobstructed height;

[0047] Match the unobstructed height with the corresponding supplementary baffle of the same height;

[0048] Determine the top position of the windshield according to the unfolded image of the windshield;

[0049] Controlling the expansion of the supplementary baffle according to the expansion width of the wind baffle, controlling the preset UAV to fly with the supplementary baffle, fixing the UAV to a parking seat preset on the suspension rope, and acquiring a hanging image of the supplementary baffle;

[0050] Determine the setting direction of the supplementary baffle according to the hanging image, and control the drone to rotate the setting direction of the supplementary baffle to be consistent with the direction of the windshield according to the direction of the windshield surface of the windshield;

[0051] Control the supplementary baffle to descend to the top position of the wind baffle to achieve docking between the two;

[0052] After docking, the force direction of the drone is determined according to the real-time wind direction, and the drone is controlled to apply force according to the force direction to help reduce the swing of the cargo.

[0053] By adopting the above technical solution, when the cargo is large and the wind shield cannot completely block the cargo, the system can control the drone to lift the supplementary baffle above the wind shield and dock it with the wind shield to expand the area of ​​the wind shield, and when the wind shield and the supplementary baffle are docked, the drone can be fixed on the drone's parking seat on the lifting rope, and force is applied to the cargo through the drone, thereby further reducing the swing of the cargo.

[0054] Optionally, when the characteristic type of the obstacle is a moving obstacle, the control method of the tower crane includes:

[0055] Based on the moving obstacles, the displacement and direction of the obstacles are determined according to the two complete ground images and obstacle features obtained in a unit time, and the speed of the obstacles is calculated;

[0056] Simulate the probabilistic movement paths of all moving obstacles based on the displacement direction, complete ground image and preset road planning routes;

[0057] Determine all the path overlap points of the two paths based on the probabilistic moving path and the lifting route;

[0058] Determine the cargo moving distance according to the path overlap point and the cargo starting point, and sort the path overlap points from near to far according to the cargo moving distance to obtain the overlap point sequence number;

[0059] According to the sequence number of the coincidence point, the moving distance of the obstacle is obtained according to the position of the obstacle and the coincidence point of the path, and the estimated encounter time is obtained according to the speed of the obstacle;

[0060] Determine the arrival time of lifting according to the moving distance of the goods and the preset lifting speed;

[0061] If and only if the estimated meeting time is consistent with the lifting arrival time, the lifting speed of the tower crane lifting cargo is determined according to the cargo moving distance, the estimated meeting time and the preset additional increase or decrease time;

[0062] The tower crane is controlled to lift the cargo according to the lifting speed to avoid moving obstacles.

[0063] By adopting the above technical solution, when there may be a moving obstacle on the cargo lifting route, the system can analyze the moving route of the moving obstacle and obtain all the overlapping points where the moving obstacle may overlap with the cargo lifting route. After determining the overlapping points, the system can slow down or speed up the cargo lifting speed, so that the cargo passes through the overlapping points in advance or delays passing through the overlapping points, avoiding collision between the cargo and the moving obstacle. Compared with changing the motion parameters of the cargo when the cargo is about to collide with the moving obstacle, the above method of estimating and changing the cargo lifting parameters in advance is safer.

[0064] Optionally, the speed-up or speed-down selection control method during the lifting process includes:

[0065] Determine the direction of cargo movement during cargo movement based on the lifting route;

[0066] Determine the cargo orientation type based on the cargo moving direction and the real-time wind direction. The cargo orientation types include downwind orientation and upwind orientation.

[0067] Based on the downwind direction, when the estimated meeting time falls within the lifting arrival time, the lifting speed is adjusted in an accelerated manner;

[0068] Based on the headwind direction, when the estimated meeting time falls within the lifting arrival time, the lifting speed is adjusted by decelerating.

[0069] By adopting the above technical solution, the wind direction and the inertia of the cargo are combined to select the overlapping points for the cargo to accelerate and decelerate. The force exerted by the wind on the cargo can be offset by the inertia of the cargo during acceleration and deceleration, so that the cargo swings less during acceleration and deceleration, thereby improving safety.

[0070] Optionally, the calculation formula for the cargo swing angle is:

[0071] in

[0072] Where I is the moment of inertia of the cargo, c is the damping coefficient, k is the stiffness coefficient of the suspension point, is the cargo swing angle at time t, S is the minimum projected area of ​​the cargo, V is the real-time wind speed, L is the length of the suspension rope, ω is the angular frequency, It is the initial phase.

[0073] In the second aspect, the present application provides a construction site data collection system based on the Flink architecture, which adopts the following technical solutions:

[0074] A construction site data collection system based on the Flink architecture, applying a construction site data collection method based on the Flink architecture, including:

[0075] An acquisition module is used to acquire a ground top view image, an obstructed ground image, cargo image information, real-time wind speed, real-time wind direction, a windshield deployment image, and a supplementary windshield hanging image;

[0076] A memory, used for storing a program of a construction site data collection method based on the Flink architecture;

[0077] The program in the processor memory can be loaded and executed by the processor to implement a construction site data collection method based on the Flink architecture.

[0078] In a third aspect, the present application provides a smart terminal, which adopts the following technical solution:

[0079] An intelligent terminal includes a memory and a processor. The memory stores a computer program that can be loaded by the processor and executes a construction site data collection method based on the Flink architecture.

[0080] In summary, the present application includes at least one of the following beneficial technical effects:

[0081] 1. Use cameras or drones to shoot the blind spots of the tower crane or the positions covered by the obstructions, and then stitch the images into the ground top-view images, so that the obstructed areas in the ground top-view images can be displayed, which is convenient for the tower crane to check the ground circumferential conditions in real time during the operation process to avoid unexpected situations; and make the obstructions transparent so that the positions of the obstructions can still be determined from the ground top-view images, so that the tower crane hoisting process can be adjusted in real time according to the positions of the obstructions, which is more efficient;

[0082] 2. If there is a crosswind during the lifting process, the system can control the cargo to rotate to the minimum projection surface facing the wind direction, at which time the cargo swing is minimized; when the wind resistance cannot be reduced by rotating the cargo, the wind shield provided on the lifting rope can be deployed to block the cargo, and the streamlined surface of the wind shield can guide the wind, thereby reducing wind resistance; when the cargo is large and the wind shield cannot completely block the cargo, the system can control the drone to lift the supplementary baffle above the wind shield and dock it with the wind shield to expand the area of ​​the wind shield, and after the wind shield and the supplementary baffle are docked, the drone can be fixed on the drone parking seat on the lifting rope, and the drone can apply force to the cargo, thereby further reducing the cargo swing;

[0083] 3. When there may be mobile obstacles on the cargo lifting route, the system can analyze the moving route of the mobile obstacles and obtain all the overlapping points where the mobile obstacles may overlap with the cargo lifting route. After determining the overlapping points, the system can slow down or accelerate the lifting speed of the cargo, so that the cargo passes through the overlapping points in advance or delays passing through the overlapping points, avoiding collision between the cargo and the mobile obstacles. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] Figure 1 It is a method flow chart of a construction site data collection method based on Flink architecture according to an embodiment of the present invention;

[0085] Figure 2 is a method flow chart of a method for correcting a lifting route according to an embodiment of the present invention;

[0086] Figure 3is a flow chart of a method for reducing cargo swaying and correcting a route under wind force according to an embodiment of the present invention;

[0087] Figure 4 It is a flow chart of a method for further reducing the wind resistance of cargo according to an embodiment of the present invention. DETAILED DESCRIPTION

[0088] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0089] The embodiment of the present application discloses a construction site data collection method based on the Flink architecture. The environment of the construction site is detected by a camera, and the blind spot position of the tower crane is image-completed. The tower crane is then controlled based on the completed image. The system can effectively solve the wind force, fixed obstacles, and mobile obstacles that the tower crane may encounter during the lifting process.

[0090] Reference Figure 1 ,A construction site data collection method based on Flink architecture includes the following steps:

[0091] Step S100: Acquire a ground top view image centered on the tower crane.

[0092] The ground top view image is an image obtained by a camera installed on the top of the tower crane, which takes the top of the tower crane as the center and looks down at the ground conditions around the tower crane. If there are obstructions at the construction site, there will be blind spots in the ground top view image, making it impossible to see the ground conditions behind the obstructions.

[0093] Step S101: select features of an obstruction from the ground top view image according to the ground top view image and preset ground features, and determine the outline of the obstruction.

[0094] Ground features include the ground and objects that can be seen on the ground.

[0095] The occlusion feature refers to an occlusion that blocks the camera's field of view and forms a blind spot. The occlusion contour refers to the outline of the occlusion feature in the ground top view image.

[0096] By analyzing the ground top view image and the ground features, the ground features are removed from the ground top view image, and the remaining image is the occluder feature. After the occluder feature in the ground top view image is determined, the occluder outline of the occluder feature in the ground top view image can be analyzed at the same time.

[0097] Step S102: determining the position of the obstruction according to the ground top view image, the characteristics of the obstruction and the preset reference point, and obtaining the obstructed ground image at the position of the obstruction.

[0098] The reference point is a reference point set by the technicians. It is a fixed location point at the construction site and is used to determine the image ratio when the camera takes an image. The distance between the reference point and the camera is a fixed value and will not be elaborated here.

[0099] The image scale can be determined based on the ground top view image and the reference points, and the position of the occluding feature in the actual construction site can be determined based on the image scale and the position of the occluding feature in the ground top view image. After the position of the occluding feature is determined, the system can obtain the ground conditions behind the occluding feature.

[0100] In this embodiment, the ground conditions behind the obstruction are collected by using a drone to collect terrain images. After the location of the obstruction is determined, the system drives the drone to the location of the obstruction and shoots the ground conditions and objects on the ground behind it, thereby forming an obstructed ground image. The obstructed ground image can be uploaded to the system through the drone, and then the image is processed and calculated by the Flink architecture system. The obstructed ground image is an image of the ground conditions that is obstructed by the obstruction features obtained by the drone and cannot be obtained by the camera on the tower crane. The Flink architecture is a core component of a distributed processing engine. It supports stateful computing of bounded and unbounded data streams, and supports high-throughput, low-latency, high-performance stream processing and batch processing.

[0101] Step S103: cropping a ground stitching image from the blocked ground image according to the outline of the blocking object.

[0102] After the Flink-based system processes the occluded ground image into an image of the direction in which the camera views the occluder's features, the system then crops a ground stitching image that is consistent with the occluder's outline from the occluded ground image. The ground stitching image is an image cropped from the occluded ground image based on the occluder's outline.

[0103] Step S104: stitching the ground stitching image to the position of the occluder in the ground bird's-eye view image, and making the occluder features transparent according to the occluder contour, and outputting a complete ground image.

[0104] The ground stitching image is stitched with the ground overhead image so that the ground stitching image overlaps with the occlusion features in the ground overhead image. At this time, the ground overhead image can present the complete ground situation. In order to retain the occlusion features in the ground overhead image, the occlusion features in the ground overhead image are made transparent. The transparent occlusion features and the ground stitching image are at the same position of the ground overhead image. At this time, both the occlusion features and the ground situation behind the occlusion features can be seen.

[0105] The complete ground image refers to an image that contains the complete ground situation and transparent occlusion features.

[0106] Step S105: matching the tower crane's hoisting route according to the complete ground image, the preset cargo starting position, and the preset cargo end position.

[0107] The starting point of the cargo is the location where the cargo begins to be lifted according to the construction plan of the construction site, and the key location of the cargo is the target location to which the cargo needs to be lifted according to the construction plan of the construction site, which will not be elaborated here.

[0108] The lifting route is the path that the goods move when the tower crane lifts the goods.

[0109] After obtaining the complete ground image, the system plans the best route from the cargo starting point to the cargo end point by determining the cargo starting point and cargo end point in the complete ground image and according to the actual ground conditions in the complete ground image.

[0110] Step S106: Control the tower crane to perform operations according to the hoisting route of the tower crane.

[0111] After determining the lifting route, the system controls the tower crane to lift the goods according to the lifting route.

[0112] Reference Figure 2 There may be obstacles on the ground and the obstacles can be identified from the complete ground image. If an obstacle is encountered during the hoisting route, the tower crane needs to avoid the obstacle and correct the route. The hoisting route correction method includes the following steps:

[0113] Step S200: determining the characteristics and positions of obstacles on the lifting route according to the complete ground image and the lifting route.

[0114] The feature of an obstacle refers to an obstacle located on the ground that will have an obstruction effect on the lifting of goods. The position of the obstacle refers to the position of the obstacle feature on the ground at the actual construction site.

[0115] The hoisting route is analyzed based on the complete ground image. Except for the road surface on the hoisting route, the remaining features are obstacle features. After the obstacle features are determined, the position of the obstacle on the ground at the actual construction site can be determined based on the position of the obstacle features in the complete ground image and the image ratio.

[0116] Step S201: determining the obstacle feature type according to the two preceding and succeeding complete ground images acquired within a preset unit time and the obstacle features, wherein the obstacle feature types include fixed obstacles and mobile obstacles.

[0117] The unit time is the time interval set by the technicians to repeatedly obtain complete ground images. Since the ground conditions may change, the ground conditions need to be continuously updated, which will not be elaborated here.

[0118] By analyzing the characteristics of obstacles in the two complete ground images, if the position of the obstacle does not change, it means that the obstacle is a fixed obstacle. If the position of the obstacle changes, it means that the obstacle is a mobile obstacle. The system will use different solutions to deal with fixed obstacles and mobile obstacles.

[0119] Step S202: Based on the fixed obstacles, the lifting height is determined according to the starting position and the end position of the cargo.

[0120] The lifting height refers to the height to which the cargo is lifted when it is lifted to the cargo destination according to the actual construction plan. The lifting height can be determined based on the cargo starting position and cargo destination, both of which are fixed positions.

[0121] Step S203: Determine the height of the obstacle according to the complete ground image and the obstacle features.

[0122] The height of the obstacle refers to the height of the obstacle feature itself. The height of the obstacle feature in the complete ground image can be determined by the complete ground image, and the actual obstacle height of the obstacle feature can be determined according to the image ratio.

[0123] Step S204: Calculate the sum of the obstacle height and the preset redundant avoidance distance, and define it as the elevated avoidance distance.

[0124] The redundant avoidance distance refers to the distance that the cargo maintains from the obstacle when bypassing the obstacle feature in order to prevent the cargo from colliding with the obstacle. The redundant avoidance distance is a reference value set by the technicians and will not be elaborated here.

[0125] When cargo encounters an obstacle during lifting, if the cargo needs to be bypassed over the obstacle, the cargo needs to be lifted. The lifting avoidance distance is the height that the cargo needs to be lifted to bypass the top of the obstacle. The lifting avoidance distance is the sum of the obstacle height and the redundant avoidance distance.

[0126] Step S2051: If the lifting height is greater than the lifting avoidance distance, when the cargo moves to the position of the obstacle, the tower crane is controlled to lift the cargo according to the lifting avoidance distance and continue to move along the lifting route.

[0127] In order to ensure that the tower crane does not need to lower the cargo again after lifting it, it is necessary to judge the lifting height and the lifting avoidance distance.

[0128] If the lifting height is greater than the lifting avoidance distance, it means that after the goods are lifted by the lifting avoidance distance, there is no need to lower the goods during the subsequent lifting process. When an obstacle of this height is needed, the system controls the goods to be lifted, bypass the top of the obstacle, and then continue to move along the preset lifting route.

[0129] Step S2052: If the lifting height is not greater than the lifting avoidance distance, the outline of the obstacle is determined based on the complete ground image and the characteristics of the obstacle.

[0130] If the lifting height is not greater than the lifting avoidance distance, it means that the height of the cargo's final position is lower than the height of the obstacle feature. At this time, the cargo cannot be bypassed from the top of the obstacle, and the system will control the cargo to bypass from the side of the obstacle. In this case, the system needs to first determine the obstacle outline. The obstacle outline is the horizontal outline of the obstacle feature in the horizontal direction. The obstacle outline can be determined by image analysis and recognition of the complete ground image.

[0131] Step S20521: Generate a detour route based on the obstacle outline and the redundant avoidance distance.

[0132] The detour route refers to the route that the tower crane uses to control the cargo to bypass the side of an obstacle. When bypassing the side of an obstacle, it is also necessary to have a redundant avoidance distance between the cargo and the obstacle. Therefore, the detour route can be generated based on the outline of the obstacle and the redundant avoidance distance. The detour route is formed by expanding the redundant avoidance distance outward from the outline of the obstacle.

[0133] Step S20522: Correct the lifting route according to the detour route so that when the cargo moves to the position of the obstacle, it moves along the detour route to bypass the obstacle feature and then continues to move along the lifting route.

[0134] After determining the detour route, when the cargo encounters this type of obstacle, the system controls the tower crane to move the cargo around the obstacle along the detour route, and then installs the preset lifting route to continue moving.

[0135] Reference Figure 3 During the lifting process, the cargo may be affected by wind and swing. The method of reducing the swing of the cargo and correcting the route under the action of wind includes the following steps:

[0136] Step S300: Obtain cargo image information, real-time wind speed and real-time wind direction.

[0137] The cargo image information refers to an image obtained by photographing the position of the cargo hoisted by the tower crane through a camera, and the cargo image information includes an image of the cargo being hoisted by the tower crane. The camera is also a camera on the top of the tower crane.

[0138] The real-time wind speed and real-time wind direction are both parameters of the ambient wind at the construction site. The tower crane is equipped with an anemometer, which can measure the real-time wind speed and real-time wind direction of the ambient wind and upload them to the system.

[0139] Step S301: Select cargo features from cargo image information and determine cargo feature outlines.

[0140] The cargo features refer to the cargo being lifted by the tower crane. The cargo image information has an image of the cargo features. Therefore, the background other than the cargo features in the cargo image information can be removed to obtain the cargo person features.

[0141] The cargo feature profile refers to the profile of the cargo feature itself, and the cargo feature profile can be obtained by identifying and analyzing the cargo features in the cargo image information.

[0142] Step S302: Determine the position of the minimum horizontal orthographic projection area and the minimum projection area according to the characteristic contour of the cargo.

[0143] After determining the characteristic contour of the cargo, the system can identify and analyze the horizontal orthographic projection of the cargo circumference along the characteristic contour of the cargo, thereby determining the position of the minimum horizontal orthographic projection area in the characteristic contour of the cargo, and determining the minimum projection area at the position where the minimum horizontal orthographic projection area is located. The minimum projection area refers to the projection area of ​​the cargo feature at the position where the minimum horizontal orthographic projection area is located.

[0144] Step S303: Calculate the angle adjustment value of the horizontal rotation of the cargo according to the real-time wind direction and the position of the minimum horizontal projection area.

[0145] In this embodiment, in order to reduce the wind resistance on the cargo, the position where the cargo's horizontal projection area is the smallest is aligned with the wind direction, so that the area of ​​separation between the cargo and the ambient wind is reduced, and the wind resistance is also reduced accordingly.

[0146] The angle adjustment value refers to the angle required to rotate the minimum horizontal projection area position of the cargo to the wind direction position. The angle adjustment value can be directly calculated based on the real-time wind direction and the position of the minimum horizontal projection area position.

[0147] Step S304: Control the cargo to rotate according to the angle adjustment value so that the position with the smallest horizontal projection area is aligned with the direction of the real-time wind direction.

[0148] After determining the angle adjustment value, the system rotates the lifting rope through the tower crane to drive the cargo to rotate to adjust the angle of the cargo.

[0149] Step S305: determining the height of the cargo from the ground according to the cargo image information and cargo features, and determining the length of the lifting rope according to the height of the cargo from the ground and a preset tower crane height.

[0150] The height of cargo above the ground refers to the distance between the cargo currently being lifted and the ground. Based on the cargo image information and cargo features, the distance between the cargo features in the cargo image information and the ground can be determined, and then the actual height of cargo above the ground between the cargo features and the ground can be determined based on the image ratio.

[0151] The tower crane height is the actual height designed and determined when the tower crane is erected. It is a fixed standard value and will not be described here. The hoisting rope is all the cables used by the tower crane to lift the cargo. The length of the hoisting rope is the current length of the hoisting rope when the tower crane lifts the cargo to the height above the ground. The length of the hoisting rope is the difference between the tower crane height and the cargo height above the ground.

[0152] Step S306: Calculate the cargo swing angle according to the length of the lifting rope, the minimum projection area and the real-time wind speed, and determine the cargo swing distance according to the cargo swing angle and the length of the lifting rope.

[0153] The cargo swing angle is the angle generated when the cargo swings due to the influence of wind. The cargo swing angle can be calculated by the length of the lifting rope, the minimum projection area and the real-time wind speed. The specific calculation formula will not be described here and will be introduced in the subsequent embodiments.

[0154] The cargo swing distance is the horizontal displacement distance of the cargo when it is swung by wind. The cargo swing distance can be determined by the cargo swing angle and the length of the sling according to trigonometric functions.

[0155] Step S307: Generate a wind bypass route according to the obstacle outline, redundant avoidance distance and cargo swing distance.

[0156] The wind bypass route is the route that the system regenerates when the goods bypass the side of the obstacle after considering the impact of wind on the goods.

[0157] After determining the cargo swing distance, the system adds the effect of the cargo swing distance to the obstacle outline and the redundant avoidance distance, thereby regenerating the wind detour route. The generation method is the same as step S20521 and will not be repeated here.

[0158] Step S308: Correct the hoisting route according to the wind detour route so that when the cargo moves to the location of the obstacle, it moves along the detour route to bypass the obstacle feature and then continues to move along the hoisting route.

[0159] At a construction site with ambient wind, when cargo encounters an obstacle, the system will lift the cargo according to the wind detour route so that the cargo bypasses the obstacle.

[0160] Reference Figure 4 The sling is provided with a windshield which can be unfolded to form a streamlined windshield surface; the method for further reducing the wind resistance of the cargo includes the following steps:

[0161] Step S400: determining the overall height and width of the cargo according to the cargo image information and cargo features.

[0162] The overall height of the cargo refers to the height of the cargo being hoisted. The overall width of the cargo refers to the width of the cargo being hoisted. Based on the cargo image information, the height and width values ​​of the cargo features in the cargo image information can be determined first, and then the actual overall height and overall width of the cargo features can be determined based on the image ratio.

[0163] Step S401: matching the unfolded width of the wind deflector according to the overall width of the cargo.

[0164] In this embodiment, a windshield is provided on the sling rope, and the windshield is formed by two hinged plates, and the surface of the windshield is streamlined. The windshield can rotate circumferentially on the sling rope, and can be raised and lowered vertically on the sling rope, and the two plates of the windshield can also be relatively unfolded. When the goods need to be shielded by the windshield, the system controls the windshield to descend and shield the goods, and allows the ambient wind to flow along the streamlined surface of the windshield, thereby reducing wind resistance.

[0165] The unfolded width refers to the overall width of the windshield after it is unfolded. The unfolded width is determined by the overall width of the cargo. The windshield can completely cover the cargo after it is unfolded, so the unfolded width of the windshield is equal to the overall width of the cargo.

[0166] Step S402: controlling the windshield to be unfolded according to the unfolded width, and acquiring an unfolded image of the windshield.

[0167] The windshield deployment image refers to an image of the windshield after being deployed, obtained by a camera.

[0168] Step S403: determining the windshield surface orientation of the windshield according to the unfolded image of the windshield and the reference points.

[0169] The windshield orientation refers to the direction in which the streamlined surface of the windshield faces. The orientation of the reference point is fixed in the windshield unfolded image, and the windshield orientation can be determined by comparing the windshield with the reference point in the windshield unfolded image.

[0170] Step S404: determining the rotation angle of the wind shield according to the real-time wind direction and the orientation of the wind shield surface.

[0171] In order to make the windshield surface completely face the wind direction of the ambient wind, the windshield surface of the windshield needs to be rotated to face the wind direction of the ambient wind. The windshield rotation angle is the angle that the windshield needs to rotate when the windshield surface rotates to the direction of the real-time wind direction.

[0172] Step S405: controlling the wind shield to rotate to a direction where the wind shield surface is directly opposite to the real-time wind direction according to the wind shield rotation angle, and controlling the wind shield to descend according to the overall height of the cargo to shield the cargo features.

[0173] The system first unfolds the wind shield, then rotates the wind shield according to the rotation angle of the wind shield, and finally blocks the wind shield in front of the cargo according to the overall height of the cargo, thereby reducing wind resistance.

[0174] When the wind shield cannot completely cover the cargo, the solution includes the following steps:

[0175] Step S500: When the overall height of the cargo is greater than the preset windshield height, the difference between the overall height of the cargo and the windshield height is calculated and defined as the unblocked height.

[0176] The height of the windshield refers to the height of the windshield body before and after the windshield is unfolded, and the height of the windshield is a fixed size of the windshield.

[0177] The width of the windshield can be determined according to the overall width of the cargo. However, when the overall height of the cargo is too high, the windshield can only block a part of the lower side of the cargo and cannot effectively reduce wind resistance. Therefore, it is necessary to first determine the overall height of the cargo to see if the windshield needs to be processed.

[0178] The unobstructed height refers to the height of the part of the cargo that is not blocked by the windshield. The unobstructed height is the difference between the overall height and the windshield height.

[0179] Step S501: matching the unobstructed height with a corresponding supplementary baffle of the same height.

[0180] In this embodiment, the portion of the cargo not blocked by the windshield is blocked by the supplementary baffle. The supplementary baffle is lifted from the ground to the top of the windshield by the drone and docked with the windshield.

[0181] The supplementary baffle is selected according to the uncovered height of the cargo, and it only needs to be able to cover the uncovered part of the cargo.

[0182] Step S502: determining the top position of the windshield according to the unfolded image of the windshield.

[0183] In this embodiment, the top of the windshield has a slot and a plug for plugging with the supplementary baffle, so before the windshield and the supplementary baffle are connected, the top position of the windshield needs to be determined. The top position of the windshield can be analyzed and identified from the unfolded image of the windshield.

[0184] Step S503: Control the expansion of the supplementary baffle according to the expansion width of the wind deflector, control the preset UAV to fly with the supplementary baffle, fix the UAV to a parking seat preset on the suspension rope, and obtain a hanging image of the supplementary baffle.

[0185] Before docking the windshield and the supplementary baffle, the supplementary baffle is first unfolded, and the unfolded width of the supplementary baffle is consistent with the unfolded width of the windshield, and then lifted by the drone. In this embodiment, a parking seat for the drone to be fixed and parked is provided on the suspension rope. When the drone lifts the supplementary baffle, the system can control the drone to automatically park on the parking seat. When the drone is fixed on the parking seat, the angle of the supplementary baffle is not consistent with the windshield, and the angle of the supplementary baffle needs to be adjusted.

[0186] The supplementary baffle hanging image refers to an image obtained by a camera when the UAV is parked on the parking seat and the supplementary baffle is hung under the UAV.

[0187] Step S504: determining the setting direction of the supplementary baffle according to the hanging image, and controlling the drone to rotate the setting direction of the supplementary baffle to be consistent with the direction of the windshield surface according to the direction of the windshield surface of the windshield.

[0188] The setting direction refers to the current orientation of the streamlined surface of the supplementary baffle. The method for determining the setting direction of the supplementary baffle is consistent with the method for determining the windshield surface orientation of the windshield in step S403, and the setting direction of the supplementary baffle can be determined according to the hanging image.

[0189] When the setting direction of the supplementary baffle and the windshield surface direction of the windshield are determined, the system can control the supplementary baffle to rotate the setting direction to be consistent with the direction of the windshield surface.

[0190] Step S505: Control the supplementary baffle to descend to the top position of the wind shield to achieve docking between the two.

[0191] When the supplementary baffle is set in the same direction as the windshield of the wind shield, the system controls the drone to lower the supplementary baffle to the top of the wind shield, so that the supplementary baffle and the wind shield are docked and fixed, so that the supplementary baffle and the wind shield can work together to shield the cargo to reduce wind resistance.

[0192] Step S506: After docking, the force application direction of the drone is determined according to the real-time wind direction, and the drone is controlled to apply force according to the force application direction to help reduce the swing of the cargo.

[0193] After the drone docks the correction baffle with the wind shield, it does not need to leave the parking seat, but continues to stay on the parking seat. At this time, the drone can be used as an auxiliary tool to apply thrust to the cargo, thereby reducing the swing of the cargo.

[0194] The force applied by the drone is in the opposite direction to the real-time wind direction, so the force applied by the drone to the cargo can partially offset the wind force.

[0195] In step S201, the characteristic type of the obstacle is identified and judged. When the characteristic type of the obstacle is a moving obstacle, the control method of the tower crane includes the following steps:

[0196] Step S600: Based on the moving obstacle, the displacement and direction of the obstacle are determined according to the two complete ground images obtained in a unit time and the characteristics of the obstacle, and the speed of the obstacle is calculated.

[0197] The displacement of the obstacle refers to the distance that the moving obstacle moves per unit time. The displacement direction refers to the direction of movement of the moving obstacle. By identifying the two complete ground images, the position of the obstacle is determined from each of them, and then the distance between the positions of the obstacle at two different times is magnified by the image ratio to obtain the actual displacement of the obstacle. The displacement direction can also be identified and determined based on the change in the position of the obstacle in the two complete ground images.

[0198] The speed of the obstacle refers to the average speed of the moving obstacle when it moves. The speed of the obstacle is the quotient of the displacement of the obstacle and the unit time.

[0199] Step S601: simulate the probabilistic moving paths of all moving obstacles according to the displacement direction, the complete ground image and the preset road planning route.

[0200] The road planning route refers to all road surfaces in the complete ground image that can be used for the movement of mobile obstacles. The road planning route is collected and entered into the system database in advance and can be directly called, so it will not be described in detail here.

[0201] The probabilistic moving path refers to all possible paths that a mobile obstacle may take when it moves. In the complete ground image, all routes in the displacement direction of the mobile obstacle can be matched from the road planning route, and this route is the probabilistic moving path of the mobile obstacle.

[0202] Step S602: Determine all the path overlap points of the two paths based on the probabilistic moving path and the lifting route.

[0203] The path overlap point refers to the position where the lifting route of the goods during the lifting process overlaps with the probabilistic moving path of the moving obstacle. The actual meaning of this position is the position point where the goods may collide with the moving obstacle during the lifting process.

[0204] The corresponding probabilistic moving path and lifting route are found in the complete ground image respectively, and the position where the two routes overlap is the path coincidence point.

[0205] Step S603: determining the cargo moving distance according to the path overlap points and the cargo starting point position, and sorting the path overlap points from near to far according to the cargo moving distance to obtain the overlap point sequence number.

[0206] The cargo moving distance refers to the distance that the cargo needs to move from the cargo starting point to the corresponding path overlap point. The cargo moving distance can be obtained by finding the path overlap point and the cargo starting point in the lifting route and calculating the distance between the two points along the lifting route.

[0207] Since there may be multiple probabilistic moving paths for the moving obstacle, there may also be multiple path overlap points. In this embodiment, all the path overlap points are first sorted according to the order in which they may collide with the goods, that is, the moving distance of the goods, to obtain the overlap point sequence number, and then each path overlap point is analyzed in turn according to the overlap point sequence number.

[0208] Step S604: Obtain the moving distance of the obstacle according to the obstacle position and the path coincidence point in sequence according to the coincidence point number, and obtain the estimated encounter time according to the obstacle speed.

[0209] The obstacle movement distance refers to the distance that the moving obstacle moves from the obstacle position to the path overlap point. The positions of two points are determined in the complete ground image, and then the distance between the two points on the probabilistic moving path is determined based on the probabilistic moving path.

[0210] The estimated encounter time refers to the time it takes for a moving obstacle to move from the obstacle position to the path overlap point, and is also the time when the moving obstacle may encounter the cargo. The estimated encounter time is the quotient of the obstacle's moving distance and the obstacle's speed.

[0211] Step S605: Determine the lifting arrival time according to the cargo moving distance and the preset lifting speed.

[0212] The lifting arrival time refers to the time required for the cargo to move from the cargo starting position to the path overlap point at the preset lifting speed. The lifting speed is the moving speed of the tower crane lifting cargo set by the technician, which will not be described here.

[0213] The lifting arrival time is the quotient of the cargo moving distance and the lifting speed.

[0214] Step S606: If and only if the estimated meeting time is consistent with the lifting arrival time, determine the lifting speed of the tower crane for lifting the goods according to the moving distance of the goods, the estimated meeting time and the preset additional increase or decrease time.

[0215] When analyzing each path overlap point according to the overlap point sequence number, if the estimated encounter time at the path overlap point is inconsistent with the lifting arrival time, it means that the moving obstacle will not collide with the cargo at the path overlap point, so there is no need to consider the path overlap point. If the estimated encounter time is consistent with the lifting arrival time, it means that the moving obstacle may collide with the cargo at the path overlap point. Therefore, in order to avoid the two colliding at the path overlap point, the cargo movement parameters need to be adjusted in advance, mainly the lifting speed when the tower crane is lifting the cargo.

[0216] Both slowing down or accelerating the cargo can prevent the cargo from colliding with the moving obstacle. How to choose acceleration or deceleration will not be described here, but will be described in detail in subsequent embodiments.

[0217] The additional increase or decrease time is the amount of time set by the technician to increase or decrease the estimated encounter time to avoid the moving obstacle and the cargo. It will not be described in detail here. By adding or decreasing a certain amount of time to the estimated encounter time, the cargo can arrive at the path overlap point earlier or later. The lifting speed can be calculated and determined based on the cargo moving distance and the adjusted time.

[0218] Step S607: Control the tower crane to lift the cargo according to the lifting speed to avoid the moving obstacles.

[0219] After determining the lifting speed, the system controls the tower crane to lift the cargo at the lifting speed, which can effectively avoid all possibilities of collision with the moving blocking block.

[0220] The selective control method for increasing speed or decreasing speed during lifting includes the following steps:

[0221] Step S700: determining the direction of cargo movement during cargo movement according to the lifting route.

[0222] The cargo moving direction refers to the moving direction of the cargo in the complete ground image when the cargo is lifted by the tower crane. The cargo moving direction can be determined according to the lifting route, and the cargo moving direction is the tangent direction of a specific position point on the lifting route.

[0223] Step S701: determining the cargo orientation type according to the cargo moving direction and the real-time wind direction, where the cargo orientation type includes a downwind orientation and an upwind orientation.

[0224] By comparing the direction of cargo movement with the real-time wind direction, it is possible to determine whether the cargo is moving with or against the wind.

[0225] Step S702: Based on the downwind direction, when the estimated meeting time falls within the lifting arrival time, the lifting speed is adjusted in an accelerated manner.

[0226] If the cargo is moving in the direction of the wind, the ambient wind can drive the cargo to swing in the direction of the cargo's advance. During the acceleration process, the cargo has a tendency to swing backward. Therefore, when the cargo and the moving obstacle may collide at the point where the path overlaps, the cargo is accelerated. At this time, the thrust of the ambient wind on the cargo and the inertia of the cargo during acceleration can offset each other, so that the cargo swings less when moving and is safer.

[0227] Step S703: Based on the headwind direction, when the estimated encounter time falls within the lifting arrival time, the lifting speed is adjusted in a decelerating manner.

[0228] Similar to step S702, the moving direction of the cargo is against the wind. When the cargo and the moving obstacle may collide at the point where the path overlaps, the cargo is decelerated. At this time, the thrust of the ambient wind on the cargo and the inertia of the cargo during deceleration can offset each other, so that the cargo has a smaller swing amplitude when moving and is safer.

[0229] The method of dealing with sudden strong winds during the tower crane lifting of goods includes the following steps:

[0230] Step S800: Determine the current location of the cargo based on the complete ground image and preset cargo features.

[0231] The current location of the cargo refers to the location of the cargo when it encounters strong winds during the lifting process. The current location of the cargo can be identified from the complete ground image. The cargo lifting route and the cargo on the lifting route at any time can be obtained from the complete ground image.

[0232] Step S801: within the preset wind-shielding range of the current location of the cargo, determine the features of obstacles that can provide wind shelter for the cargo based on the complete ground image and fixed obstacles, and number the searched obstacle features to obtain an obstacle number.

[0233] The wind shelter range is the distance range of obstacles set by technicians to provide a wind shelter environment for the goods in time. Within this range, the tower crane can quickly move the goods to the designated wind shelter location, which will not be elaborated here.

[0234] By identifying the complete ground image within the windproof range of the current location of the cargo, all the obstruction features around the cargo can be obtained. The searched obstruction features are first numbered, and each obstruction feature needs to be analyzed according to the obstruction numbering system to determine whether there is an obstruction feature that can be used to shield the cargo from the wind.

[0235] Step S802: Analyze the obstacle features of each obstacle number according to the complete ground image and the real-time wind direction to obtain the windshield area of ​​the obstacle features.

[0236] The windproof surface area refers to the projected area of ​​the back of the obstruction feature in the upwind direction. Cargo placed behind the obstruction feature can be protected from the wind.

[0237] By identifying and distinguishing the features of obstacles in the complete ground image, the windshield area can be obtained.

[0238] Step S803: Determine whether there is an obstacle feature whose windshield area is larger than the minimum projection area of ​​the cargo.

[0239] The size of the windshield area will affect the windshield capacity. When the windshield area is larger than the minimum projected area of ​​the cargo, it can block the wind for the cargo to the greatest extent. Therefore, it is necessary to determine whether the windshield area of ​​each obstacle feature is larger than the minimum projected area of ​​the cargo, so as to select the obstacle feature with the largest windshield area.

[0240] Step S804: If there is an obstacle feature whose windshield area is larger than the minimum projection area of ​​the cargo, determine the obstacle number of the obstacle feature with the largest windshield area, and determine the back wind shelter point position of the obstacle feature based on the complete ground image and the obstacle number.

[0241] If there is an obstacle feature whose windshield area is larger than the minimum projection area of ​​the cargo, the system will filter out all the obstacle features that meet the requirements, and then select the obstacle feature with the largest windshield area from the filtered obstacle features.

[0242] The back wind shelter point position refers to the position of the back of the obstacle feature that can be windproof at the actual construction site. The designated obstacle feature can be determined first according to the obstacle number, and then the back wind shelter point position of the obstacle feature can be identified according to the complete ground image and the obstacle feature.

[0243] Step S805: Control the cargo to move from its current position to a wind-sheltered point at the back for wind shelter.

[0244] After determining the location of the rear wind shelter point, the system controls the tower crane to lift the cargo to the rear wind shelter point.

[0245] Step S806: If there is no obstacle feature whose windshield area is larger than the minimum projection area of ​​the cargo, determine the obstacle feature closest to the current position of the cargo based on the complete ground image, and determine the back wind shelter point position of the obstacle feature.

[0246] If there is no obstacle feature whose wind-blocking surface area is larger than the minimum projection area of ​​the cargo, the system will directly select the obstacle feature closest to the current position of the cargo, and then control the tower crane to lift the cargo to the back sheltered point of the obstacle feature.

[0247] Step S807: Control the cargo to move from its current position to the wind-sheltered point at the back, and control the adsorption device preset on the sling to be adsorbed on the obstacle feature.

[0248] When the cargo is located at the back windproof point, the system will control the adsorption device to adsorb the cargo to the barrier feature, so that the cargo is connected to the barrier feature and is not easy to shake. In this embodiment, the adsorption device is set on the hanging rope, and can extend the adsorption claw under the control of the system and be fixed on the barrier feature.

[0249] The calculation formula for the cargo swing angle is:

[0250] in

[0251] Where I is the moment of inertia of the cargo, c is the damping coefficient, k is the stiffness coefficient of the suspension point, is the cargo swing angle at time t, S is the minimum projected area of ​​the cargo, V is the real-time wind speed, L is the length of the suspension rope, ω is the angular frequency, It is the initial phase.

[0252] In the above formula, I, c, and k are known constants. Therefore, the cargo swing angle can be calculated based on the obtained S, which is the minimum projected area of ​​the cargo, V, which is the real-time wind speed, and L, which is the length of the suspension rope.

[0253] Based on the same inventive concept, an embodiment of the present invention provides a construction site data collection system based on the Flink architecture, and applies a construction site data collection method based on the Flink architecture, including:

[0254] The acquisition module is used to obtain ground bird's-eye view images, obstructed ground images, cargo image information, real-time wind speed, real-time wind direction, windshield deployment images, and supplementary baffle hanging images.

[0255] The memory is used to store a program of a construction site data collection method based on the Flink architecture.

[0256] The program in the processor memory can be loaded and executed by the processor to implement a construction site data collection method based on the Flink architecture.

[0257] Based on the same inventive concept, an embodiment of the present invention provides an intelligent terminal, including a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and execute a construction site data collection method based on the Flink architecture.

[0258] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A construction site data collection method based on Flink architecture, characterized in that: include: Obtain a ground bird's-eye view image centered on the tower crane; According to the ground top view image and the preset ground features, the features of the occluder are selected from the ground top view image, and the outline of the occluder is determined; Determine the location of the obstruction based on the ground top view image, the characteristics of the obstruction, and the preset reference point, and obtain the obstructed ground image at the location of the obstruction; A ground stitching image is cropped from the occluded ground image according to the outline of the occluder; The ground stitching image is stitched at the location of the occlusion in the ground overhead image, and the occlusion features are made transparent according to the occlusion contour to output a complete ground image; Match the tower crane's lifting route based on the complete ground image, the preset cargo starting position, and the preset cargo end position; Control the tower crane to operate according to its lifting route.

2. According to a construction site data collection method based on Flink architecture according to claim 1, it is characterized in that: If there are obstacles in the lifting route, the tower crane needs to avoid the obstacles and correct the route. The lifting route correction methods include: Determine the characteristics and locations of obstacles on the lifting route based on the complete ground image and the lifting route; Determine the obstacle feature type according to the two complete ground images and obstacle features acquired within a preset unit time, where the obstacle feature types include fixed obstacles and mobile obstacles; Based on fixed obstacles, the lifting height is determined according to the starting position and the end position of the cargo; Determine the height of the obstacle based on the complete ground image and the characteristics of the obstacle; Calculate the sum of the obstacle height and the preset redundant avoidance distance and define it as the elevated avoidance distance; If the lifting height is greater than the lifting avoidance distance, when the cargo moves to the position of the obstruction, the tower crane is controlled to lift the cargo according to the lifting avoidance distance and continue to move along the lifting route; If the lifting height is not greater than the lifting avoidance distance, the obstacle outline is determined based on the complete ground image and the characteristics of the obstacle; Generate detour routes based on obstacle contours and redundant avoidance distances; The lifting route is corrected according to the detour route so that when the cargo moves to the position of the obstruction, it moves along the detour route to bypass the obstruction feature and then continues to move along the lifting route.

3. According to a construction site data collection method based on Flink architecture according to claim 2, it is characterized in that: During the lifting process, the cargo may be affected by wind and sway. Methods to reduce the swaying of the cargo and correct the route under the action of wind include: Obtain cargo image information, real-time wind speed and real-time wind direction; Select cargo features from cargo image information and determine cargo feature outlines; Determine the position of the minimum horizontal orthographic projection area and the minimum projection area based on the characteristic contour of the cargo; Calculate the angle adjustment value of the horizontal rotation of the cargo based on the real-time wind direction and the position of the minimum horizontal projection area; The cargo is controlled to rotate according to the angle adjustment value so that the position with the smallest horizontal projection area is aligned with the direction of the real-time wind direction; Determine the height of the cargo from the ground based on the cargo image information and cargo features, and determine the length of the lifting rope based on the height of the cargo from the ground and the preset tower crane height; Calculate the cargo swing angle based on the length of the sling, the minimum projection area and the real-time wind speed, and determine the cargo swing distance based on the cargo swing angle and the length of the sling; Generate a wind bypass route based on the obstacle contour, redundant avoidance distance, and cargo swing distance; The lifting route is corrected according to the wind detour route so that when the cargo moves to the obstruction position, it moves along the detour route to bypass the obstruction feature and then continues to move along the lifting route.

4. According to a construction site data collection method based on Flink architecture according to claim 3, it is characterized in that: The sling is provided with a windshield which can be unfolded to form a streamlined windshield surface. Methods for further reducing the wind resistance of the cargo include: Determine the overall height and width of the cargo based on cargo image information and cargo features; Match the unfolded width of the wind deflector to the overall width of the cargo; Controlling the windshield to unfold according to the unfolding width, and acquiring an unfolding image of the windshield; Determine the windshield surface orientation of the windshield according to the unfolded image of the windshield and the reference point; Determine the rotation angle of the windshield according to the real-time wind direction and the orientation of the windshield surface; The wind shield is controlled to rotate according to its rotation angle until the wind shield surface is facing the real-time wind direction, and the wind shield is controlled to descend according to the overall height of the cargo to shield the cargo features.

5. A construction site data collection method based on Flink architecture according to claim 4, characterized in that: When the wind shield cannot completely cover the cargo, the solutions include: When the overall height of the cargo is greater than the preset windshield height, the difference between the overall height of the cargo and the windshield height is calculated and defined as the unblocked height; Match the unobstructed height with the corresponding supplementary baffle of the same height; Determine the top position of the windshield according to the unfolded image of the windshield; Controlling the expansion of the supplementary baffle according to the expansion width of the wind baffle, controlling the preset UAV to fly with the supplementary baffle, fixing the UAV to a parking seat preset on the suspension rope, and acquiring a hanging image of the supplementary baffle; Determine the setting direction of the supplementary baffle according to the hanging image, and control the drone to rotate the setting direction of the supplementary baffle to be consistent with the direction of the windshield according to the direction of the windshield surface of the windshield; Control the supplementary baffle to descend to the top position of the wind baffle to achieve docking between the two; After docking, the force direction of the drone is determined according to the real-time wind direction, and the drone is controlled to apply force according to the force direction to help reduce the swing of the cargo.

6. A construction site data collection method based on Flink architecture according to claim 3, characterized in that: When the obstacle feature type is a moving obstacle, the control method of the tower crane includes: Based on the moving obstacles, the displacement and direction of the obstacles are determined according to the two complete ground images and obstacle features obtained in a unit time, and the speed of the obstacles is calculated; The probabilistic movement paths of all moving obstacles are simulated based on the displacement direction, complete ground image and preset road planning routes; Determine all the path overlap points of the two paths based on the probabilistic moving path and the lifting route; Determine the cargo moving distance according to the path overlap point and the cargo starting point, and sort the path overlap points from near to far according to the cargo moving distance to obtain the overlap point sequence number; According to the sequence number of the coincidence point, the moving distance of the obstacle is obtained according to the position of the obstacle and the coincidence point of the path, and the estimated encounter time is obtained according to the speed of the obstacle; Determine the arrival time of lifting according to the moving distance of the goods and the preset lifting speed; If and only if the estimated meeting time is consistent with the lifting arrival time, the lifting speed of the tower crane lifting cargo is determined according to the cargo moving distance, the estimated meeting time and the preset additional increase or decrease time; The tower crane is controlled to lift the cargo according to the lifting speed to avoid moving obstacles.

7. A construction site data collection method based on Flink architecture according to claim 6, characterized in that: The selective control methods for increasing or decreasing speed during lifting include: Determine the direction of cargo movement during cargo movement based on the lifting route; Determine the cargo orientation type based on the cargo moving direction and the real-time wind direction. The cargo orientation types include downwind orientation and upwind orientation. Based on the downwind direction, when the estimated meeting time falls within the lifting arrival time, the lifting speed is adjusted in an accelerated manner; Based on the headwind direction, when the estimated meeting time falls within the lifting arrival time, the lifting speed is adjusted by decelerating.

8. A construction site data collection method based on Flink architecture according to claim 3, characterized in that: The calculation formula for the cargo swing angle is: in Where I is the moment of inertia of the cargo, c is the damping coefficient, k is the stiffness coefficient of the suspension point, is the cargo swing angle at time t, S is the minimum projected area of ​​the cargo, V is the real-time wind speed, L is the length of the suspension rope, ω is the angular frequency, It is the initial phase.

9. A construction site data collection system based on Flink architecture, applying a construction site data collection method based on Flink architecture as described in any one of claims 1 to 8, characterized in that: include: An acquisition module is used to acquire a ground top view image, an obstructed ground image, cargo image information, real-time wind speed, real-time wind direction, a windshield deployment image, and a supplementary windshield hanging image; A memory, used for storing a program of a construction site data collection method based on the Flink architecture; The program in the processor memory can be loaded and executed by the processor to implement a construction site data collection method based on the Flink architecture.

10. An intelligent terminal, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executes a construction site data collection method based on the Flink architecture as described in any one of claims 1 to 8.

Images