Intelligent unloading machine suitable for box freight bagged materials

Abstract

The utility model provides an intelligent loading and unloading machine suitable for box type freight bagged materials, which comprises a crawler walking chassis and a power system, a telescopic conveying mechanism, a robot system, a swing arm conveying mechanism, a bag intelligent visual recognition control system, a quick-change clamp structure, an electrical control system, a walking guide correction system and a remote control handle. The quick-change clamp mechanism is composed of a jaw material taking mechanism and a suction cup material taking mechanism, can quickly change the target according to the stacking condition of the bag, has reliable grabbing, flexible movement, is convenient for intelligent control and has high adaptability, so that the unloading machine can not only unload the bags on the shed car, truck and other transport tools, but also be used for the transfer of bagged goods in the warehouse. The bag visual recognition intelligent system can accurately identify the position and posture of the bag, and can assist the airborne mechanical equipment to efficiently load and unload the car, greatly reducing the labor intensity of the loading and unloading workers and improving the loading and unloading efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to an automatic loading device, in particular to an intelligent unloading machine suitable for box freight bagged materials. BACKGROUND

[0002] At present, the transportation of bagged goods in the industry mainly adopts train shed cars or van trucks and other means. Because the bags are prone to deformation and irregular in shape, manual operation is required when stacking, and the spacing between the bags cannot be controlled. During transportation, the bags will change position under the braking and starting of the whole vehicle, so it is very difficult to achieve mechanized handling during unloading. At present, the unloading and warehouse handling of bagged goods in the industry mainly rely on manual lifting, carrying, and simple conveying equipment assisted by forklifts, which has high labor intensity, poor working environment, low loading and unloading efficiency, high labor cost for enterprises, low production efficiency, and low automation level.

[0003] In order to realize efficient loading and unloading of train shed cars and facilitate material handling, the prior art uses an automatically controlled mechanical hand to clamp the material, but the size cannot be adjusted or the adjustment range is limited, which may cause insufficient applicability to materials of different sizes. SUMMARY

[0004] To solve the above problems, the present application provides an intelligent unloading machine suitable for box freight bagged materials, which uses a mechanical arm to grab the bags, is reliable in grabbing, flexible in movement, easy to control intelligently, and has high adaptability, so that the unloading machine can not only unload bags on transport tools such as shed cars and trucks, but also be used for transferring bagged goods in warehouses. It is equipped with a bag visual recognition intelligent system, which can accurately identify the position and posture of the bags, and can assist the on-board mechanical equipment to efficiently load and unload the vehicle, greatly reducing the labor intensity of the loading and unloading workers and improving the loading and unloading efficiency.

[0005] The present application adopts the following technical solutions to solve the technical problems:

[0006] An intelligent unloading machine suitable for box freight bagged materials, comprising a tracked walking chassis and a power system, a telescopic conveying mechanism, a robot system, a quick-change clamp structure, an arm swinging conveying mechanism, a bag intelligent visual recognition control system, a walking guide correction system, an electrical control system and a remote control handle.

[0007] The tracked walking chassis and power system include tracked wheels, an external toothed rotary support and a vehicle frame. The vehicle frame is an integrated frame, and the vehicle frame is connected to the external toothed ring flange on the external toothed rotary support. The external toothed ring is engaged with the pinion, and the pinion is connected to the power input, so that the entire vehicle body is adjusted in angle or rotated under the adjustment of the external toothed rotary support.

[0008] The telescopic conveying mechanism includes several stages of movable output belts. The telescopic conveying mechanism serves as a material output component of the unloading machine, which can be telescoped in the length direction of the unloading machine to ensure that the unloading machine itself can move in a smaller space and can adapt to different length sizes of working environments such as box cars and trucks in a working state.

[0009] The robot system comprises a six-axis industrial robot, a robot mounting seat, and a quick-change clamp mechanism equipped at the end of the robot. The quick-change clamp mechanism is a quick-change structure composed of a gripper material taking mechanism and a suction cup material taking mechanism, which can quickly change the grasping target according to the stacking condition of the bag.

[0010] The swing arm conveying mechanism comprises two stages of telescopic conveying mechanisms, which serve as material input components of the unloading machine. The swing arm conveying mechanism is hingedly fixed to the front end of the telescopic conveying mechanism, can realize overall up-and-down swing, and can guarantee that it is as close as possible to the bag material by cooperating with the telescoping of the swing arm conveying mechanism itself.

[0011] The bag intelligent visual recognition control system adopts an intelligent visual algorithm to accurately identify the position and posture of the bag, and the electrical control system controls the quick-change clamp structure to take material according to the identification result of the bag intelligent visual recognition control system.

[0012] The walking guide correction system adopts a composite guide system and cooperates with RFID radio frequency identification technology to control the walking chassis and power system, realizes autonomous guidance and real-time correction of the unloading machine.

[0013] The telescopic conveying mechanism and the robot system mounting seat are seated on the vehicle frame and are integrated with the vehicle frame.

[0014] The electrical control system is electrically connected with the tracked walking chassis and power system, the telescopic conveying mechanism, the robot system, the swing arm conveying mechanism, the bag intelligent visual recognition control system, and the walking guide correction system, and controls the operation of each component of the unloading machine.

[0015] The remote control handle communicates with the electrical control system and is used for remotely controlling the action of the loading and unloading machine.

[0016] Preferably, the bag intelligent visual recognition control system comprises left and right cameras arranged at the end of the robot, an image acquisition and processing module, a joint sensor and a controller; the controller comprises a joint controller and a power amplifier; the image acquisition and processing module comprises a feedback operation module, a pose estimation module, a feature detection module and an image acquisition module; the joint sensor acquires the angle and displacement parameters of each joint of the robot and sends them to the controller; the image acquisition module transmits the images acquired by the cameras to the feature detection module, the feature detection module extracts the image feature information of the overall range and the key feature information of the shape and position of the bag, and then transmits the feature information to the pose estimation module; the pose estimation module estimates the accurate pose of the end of the robot and the pose of the bag to be grabbed according to the above feature information and transmits them to the feedback operation module; the feedback operation module transmits the target pose of the end of the robot, the pose of the target bag and the feedback pose information to the controller; the controller integrates the feedback information of the feedback operation module and the joint sensor to obtain the target point of the grabbing structure, and sends instructions to the robot and the end grabbing structure thereof.

[0017] In addition, a working method of an intelligent unloading machine suitable for box freight bagged materials is provided, in particular:

[0018] S1: Before the unloading machine works, it is kept in a non-working state and automatically drives along the center line of the carriage to a certain distance in front of the bag. During the driving process, the whole vehicle is moved by the double-track wheels. During the movement, the attitude of the vehicle body is adjusted by the differential action of the double-track wheels and the rotation action of the rotary support.

[0019] S2: The grabbing sequence of the bag is in an S-shaped order from top to bottom. When the unloading machine works, it starts from the highest point of each section. The swing arm conveying mechanism swings towards the bag side, and the telescopic conveying of the swing arm conveying mechanism is extended to the maximum length. At this time, the front end of the swing arm conveying system approaches the bag stack. After the bag intelligent visual recognition control system determines the position of the bag, the end tool of the industrial robot is quickly replaced by the gripper taking mechanism. The gripper clamps the bag from the exposed end of the bag and drags the stack down, then the gripper releases the bag for the next grabbing, and at the same time, the bag falls to the swing arm conveying mechanism by gravity, and then is conveyed to the telescopic conveying mechanism by the swing arm conveying mechanism, and finally is output from the telescopic conveying mechanism to the downstream equipment.

[0020] S3: The unloading machine takes materials along the sequence from top to bottom. As the height of the bag stack decreases to a certain height, the swing arm conveying mechanism adjusts the length and angle to adapt to the height of the bag stack to complete the material taking work. If necessary, the unloading machine as a whole can also move forward and backward to adjust the distance between the front end of the swing arm and the bag. During the whole unloading process, the unloading machine adjusts the state.

[0021] S4: When the bag is taken to the lower bag, the swing arm conveying mechanism is lowered to the lowest position as a whole, and the clamping jaw material taking mechanism is limited in operation space and operation mode. After the bag position is determined by the intelligent visual recognition control system of the bag, the end tool quick change of the industrial robot is changed into the suction cup material taking mechanism. After the suction cup material taking mechanism grabs the bag from above the bag, it is translated and dragged to the swing arm conveying mechanism, then conveyed to the telescopic conveying mechanism by the swing arm conveying mechanism, and finally output from the telescopic conveying mechanism to the downstream equipment.

[0022] S5: After the unloading machine completes the first section bag material taking, the whole vehicle advances under the action of the track wheel, and the telescopic conveying mechanism is elongated to adapt to the connection of the downstream equipment, so as to start the second section unloading work. Repeat this process until the unloading work in this direction is completed.

[0023] S6: After completing the unloading work in one direction, the industrial robot is retracted, the swing arm conveying mechanism is retracted and becomes upright, the telescopic conveying mechanism is retracted, and the whole vehicle returns to the state. The whole vehicle is retreated to the door, and the differential driving of the two track wheels is completed at the door to realize the 180° turning of the whole vehicle in place, and then the bag disassembly work on the other side of the vehicle door is started.

[0024] The beneficial effects of the present application are as follows:

[0025] (1) The track walking chassis and the power system adopt rubber track wheels, which can adapt to various running environments and complex road conditions. Two track wheels are respectively driven by servo motors, and the two track wheels are controlled by differential to realize the forward movement, backward movement and turning of the whole vehicle. The rotary support is used between the track wheel chassis and the vehicle frame, which can further adjust the vehicle angle during walking, so that the whole vehicle has stronger adaptability to the walking space.

[0026] (2) The two-stage or multi-stage telescopic conveying mechanism is based on the ordinary belt conveyor and increases the telescopic mechanism, so that the conveyor can freely stretch and retract in the length direction, and the telescopic length can be adjusted as required. When the telescopic machine belt is running, the belt will deviate to one side, causing belt wear and reducing the service life of the equipment.

[0027] (3) The bag intelligent visual recognition control system adopts intelligent visual algorithm and has good environmental adaptability, can accurately identify the position and posture of the bag, and well assists the airborne mechanical equipment to complete the bag loading and unloading operation efficiently. Through the left and right cameras and the image acquisition and processing module, the bag intelligent visual recognition control system can realize high-precision bag shape and position information extraction. This visual recognition capability can accurately locate the bag to ensure that the robot can accurately grasp the bag and avoid errors and losses. In addition, the feature detection module can extract key feature information of the bag, quickly switch the gripper material taking mechanism and the suction cup material taking mechanism according to the bag stacking condition, and realize self-adaptive grasping mode. This enables the robot to operate flexibly according to different bag characteristics and working environment, improving the work efficiency and success rate. Furthermore, the posture estimation module can accurately estimate the posture of the robot end through analyzing the feature information. This is very important for the fine control of the robot, which can ensure that the robot end maintains the correct posture during operation, thereby avoiding material slipping, collision and other situations, and improving the success rate of grasping and placing. Finally, the feedback operation module can transmit the target posture and the actual feedback posture information to the controller. The controller integrates the feedback information for operation control and sends instructions to the robot and its end grasping structure. This feedback control mechanism can adjust the robot's action in real time to make it more accurate in task execution and dynamically adjust according to the actual situation. This control method has strong flexibility and adaptability, and can be set and adjusted according to different types of bag materials, suitable for loading and unloading operations in various fields such as warehousing, logistics, production line, etc. At the same time, through the optimization and upgrading of software algorithm, it can also cope with more complex bag shapes and stacking conditions, improving the intelligent level of the system.

[0028] (4) The walking guide correction system adopts a composite guide system and cooperates with RFID radio frequency identification technology to realize autonomous guidance and real-time correction of the whole machine. Especially in narrow spaces such as shed carriages, the walking posture control of the vehicle is a key technical point to determine whether the unloading machine can work smoothly.

[0029] (5) The industrial robot and its end tool six-axis industrial robot has sufficient degrees of freedom to complete various actions, the quick-change clamp mechanism is composed of a gripper material taking mechanism and a suction cup material taking mechanism, which can quickly change the grasping target according to the stacking condition of the bag, and has reliable grasping, flexible movement, easy intelligent control and high adaptability. BRIEF DESCRIPTION OF DRAWINGS

[0030] Figure 1 is a schematic diagram of the overall structure of the intelligent unloading machine in the working state suitable for box cargo bagged materials.

[0031] Figure 2 is a schematic diagram of the intelligent unloading machine in the working state suitable for box cargo bagged materials.

[0032] Figure 3 is a schematic diagram of the running state of the intelligent unloading machine for box freight bagged materials.

[0033] Figure 4 is a structural schematic diagram of the gripper material taking mechanism of the intelligent unloading machine.

[0034] Figure 5 is a structural schematic diagram of the suction cup material taking mechanism of the intelligent unloading machine.

[0035] Figure 6 is a schematic diagram of the intelligent unloading machine moving to the working position.

[0036] Figure 7 is a schematic diagram of the bag grabbing sequence of the intelligent unloading machine.

[0037] Figure 8 is a schematic diagram of the gripper material taking mechanism of the intelligent unloading machine starting to unload.

[0038] Figure 9 is a schematic diagram of the gripper material taking mechanism of the intelligent unloading machine working.

[0039] Figure 10 is a schematic diagram of the suction cup material taking mechanism of the intelligent unloading machine working.

[0040] Figure 11 is a flowchart of the working method of the intelligent unloading machine.

[0041] Figure 12 is a control principle diagram of the bag intelligent visual recognition control system of the intelligent unloading machine.

[0042] Reference signs

[0043] 1 walking chassis and power system, 2 telescopic conveying mechanism, 3 robot system, 4 quick-change clamp structure, 5 swing arm conveying mechanism, 6 bag intelligent visual recognition control system, 7 walking guidance and correction system, 8 electrical control system, 9 remote control handle, 10 bag intelligent visual recognition control system controller, 11 joint controller, 12 power amplifier, 13 feedback operation module, 14 attitude estimation module, 15 feature detection module, 16 image acquisition module, 17 joint sensor, 18 camera DETAILED DESCRIPTION

[0044] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application. In addition, the technical features involved in each embodiment of the present application described below can be combined with each other as long as they do not conflict with each other.

[0045] As shown in Figures 1-10 , an intelligent unloading machine suitable for box freight bagged materials, comprising a crawler chassis and power system 1, a telescopic conveying mechanism 2, a robot system 3, a quick-change clamp structure 4, a swing arm conveying mechanism 5, a bag intelligent visual recognition control system 6, a walking guide correction system 7, an electrical control system 8 and a remote control handle 9.

[0046] The crawler chassis and power system 1 comprises crawler wheels, an external toothed rotary support and a vehicle frame. The vehicle frame is a one-piece frame, and the vehicle frame is connected to the flange of the external toothed ring on the external toothed rotary support. The external toothed ring is engaged with the pinion gear, and the pinion gear is connected to the power input, so that the entire vehicle body can be adjusted in angle or rotated under the adjustment of the external toothed rotary support.

[0047] The telescopic conveying mechanism 2 comprises a plurality of stages of movable output belts. The telescopic conveying mechanism 2 is used as a material output component of the unloading machine, which can be telescopic in the length direction of the unloading machine to ensure that the unloading machine itself can move in a smaller space, and at the same time can adapt to different length dimensions of working environment such as shed cars and trucks in working state.

[0048] The robot system 3 comprises a six-axis industrial robot, a robot mounting seat, and a quick-change clamp mechanism 4 equipped at the end of the robot. The quick-change clamp mechanism 4 is a quick-change structure composed of a gripper material taking mechanism 10 and a suction cup material taking mechanism 11, which can quickly change the target according to the stacking condition of the bag.

[0049] The swing arm conveying mechanism 5 comprises two stages of telescopic conveying mechanisms, which are used as material input components of the unloading machine. The swing arm conveying mechanism is hingedly fixed to the front end of the telescopic conveying mechanism, and can realize overall up-down swing. In combination with the telescopic length of the swing arm conveying mechanism, it can ensure that it is as close to the bag material as possible.

[0050] The bag intelligent visual recognition control system 6 adopts intelligent visual algorithm to accurately identify the position and posture of the bag. The electrical control system 8 controls the quick-change clamp structure 4 according to the identification result of the bag intelligent visual recognition control system 6.

[0051] The walking guide correction system adopts a composite guide system and cooperates with RFID radio frequency identification technology to control the crawler chassis and power system 1, so as to realize the autonomous guidance and real-time correction of the unloading machine.

[0052] The telescopic conveying mechanism 2 and the robot system 3 mounting seat are both seated on the vehicle frame and are integrated with the vehicle frame.

[0053] The electrical control system 8 is electrically connected with the tracked chassis and power system 1, the telescopic conveying mechanism 2, the robot system 3, the swing arm conveying mechanism 5, the bag intelligent visual recognition control system 6 and the walking guide correction system 7, and controls the operation of each part of the unloading machine.

[0054] The remote control handle 9 communicates with the electrical control system 8 and is used for remotely controlling the action of the loading and unloading machine.

[0055] The bag intelligent visual recognition control system 6 comprises left and right cameras 18 arranged at the end of the robot, an image acquisition and processing module, a joint sensor 17 and a controller 10; the controller 10 comprises a joint controller 11 and a power amplifier 12; the image acquisition and processing module comprises a feedback operation module 13, a pose estimation module 14, a feature detection module 15 and an image acquisition module 16; the joint sensor 17 acquires the angle and displacement parameters of each joint of the robot and sends them to the controller 10; the image acquisition module 16 transmits the image acquired by the camera 18 to the feature detection module 15, which extracts the image feature information of the overall range and the key feature information of the shape and position of the bag, and then transmits the feature information to the pose estimation module 14; the pose estimation module 14 estimates the accurate pose of the end of the robot and the pose of the bag to be grabbed according to the above feature information and transmits them to the feedback operation module 13; the feedback operation module 13 transmits the target pose of the end of the robot, the pose of the target bag and the feedback pose information to the controller 10; the controller 10 integrates the feedback information of the feedback operation module 13 and the joint sensor 17 to obtain the target point of the grabbing structure, and sends instructions to the robot and the end grabbing structure thereof.

[0056] In the feature detection and pose estimation, the image processing is performed around the region of interest and the feature detection is performed. The coordinate (u, v) of the center point of the boundary box around the feature represented in the image plane is detected. The three-dimensional pose of the feature represented in the world coordinate is determined in the pose estimation stage. For the left and right stereo cameras of the present application, the disparity method is used to extract the feature.

[0057] The controller 10 controls the robot motion using real-time information from vision, uses computer vision techniques to extract features from the real world, and infers how the robot should move to make the image features converge to the target pose. As shown in Figure 12 The camera 18 is connected to the moving end effector, and conversion is performed between the end effector coordinate system and the camera coordinate system.

[0058] The purpose of the vision-based control method is to minimize the error e(t), which is shown in formula (1):

[0059] e(t) = s(m(t), a) - s *(1)

[0060] m(t) is a set of image measurements, s(m(t), a) is a system that takes additional knowledge (a) into account from the measurements of the visual features k, s * is a vector of expected values of the features. The robot end-effector is a position-based visual servoing control, where s consists of the pose of the features estimated from the image measurements.

[0061] The relationship between the time variation of s and the camera motion is established by a velocity controller: v c = (v c , ω c ), where v c is the instantaneous linear velocity of the origin of the camera coordinate system, and ω c is the instantaneous angular velocity of the camera coordinate system. The relationship between v c and is:

[0062]

[0063] where L s ∈ R k×6 is the interaction matrix. Using equation (1) and equation (2), the relationship between the time variation of the error and the camera velocity is calculated:

[0064]

[0065] where L e is equal to L s . The input of the controller is set to v c , and the goal is to make the error decrease exponentially, i.e., to: The function formula of the controller is:

[0066] v c = -λL e + e = -λL e + (s - s * ) (4)

[0067] where L e + ∈ R 6×k is the pseudo-inverse matrix of L e . Since L e is not easy to obtain, an approximate value is used to represent L e , and the function formula of the controller can be rewritten as:

[0068]

[0069] The cycle stops and the error of the recognition control system changes over time as follows:

[0070]

[0071] If then the exponential error is not desired.

[0072] In position-based visual servoing control, s is defined with camera pose and reference coordinates, a is the intrinsic parameters of the camera and the 3D model of the object. where t is the transformation and is the rotation angle axis is the parameterized value, F c , F d , and F o is the current camera coordinate system, the desired camera coordinate system and the reference coordinate system of the robot end, in position-based visual servoing control, it is assumed that The superscript in front of the coordinate represents the coordinate system represented by the coordinate. At this time, s * = 0, e = s, and:

[0073]

[0074] where d R c is the rotation matrix, which specifies the vector of the current camera coordinate system to the desired coordinate system, is the interaction matrix. By decoupling the rotational and translational motion, the following control equation is obtained:

[0075]

[0076] By using a stereo vision system, the visual feature is represented by stacking the x and y coordinates of the point in the middle as the world coordinate point seen in the left and right images.

[0077] s = p s = (p1, p2) = (x l , y l , x r , y r ) (9)

[0078] where l and r subscripts represent left and right cameras, respectively. By repeating the above steps, two equations can be obtained:

[0079]

[0080] If the sensor frame rigidly connected with the stereo vision system is selected, the system can be written as:

[0081]

[0082] where the interaction matrix is determined by a spatial motion transformation matrix S that transforms the velocities expressed in the left and right camera coordinate systems into the sensor coordinate system. The S matrix is as follows:

[0083]

[0084] The numerical values of the matrix can be obtained by a stereo calibration step. Thus:

[0085]

[0086] The 3D coordinates of any point seen by the stereo camera pair, left and right cameras, are computed by triangulation putting it into the vector s.

[0087] In a specific embodiment, each camera is set as an independent node running at a frequency of 60 Hz, receiving two synchronized images from the two cameras, detecting the grasp point in each of them, and computing the disparity of the points in the bounding box to obtain the depth information, publishing the three-dimensional grasp point to the controller at a frequency of 30 Hz. Thus, the bag smart vision recognition control system control loop runs at 30 Hz. Assuming that each camera uses a pinhole model, is equipped with identical optics, and is located in the same plane with a 5 cm baseline separation, after stereo calibration, the following can be obtained:

[0088]

[0089] where K is the camera intrinsic matrix, P is the projection camera matrix, and the fourth column [T x T y 0] T is related to the position of the camera optical center in the left camera image. For the left camera, T x = T y = 0, the depth of the grasp point is calculated as:

[0090]

[0091] where B is the baseline obtained from the correct camera calibration , x l and x r are the feature points seen by the two cameras.

[0092] Thus, from the (u, v) pixel coordinates of the feature in the left camera image coordinates, the 3D point cl X o in the camera coordinates is computed by solving the projection as follows:

[0093]

[0094] Extracted 3D points cl X o represents s(m(t),a) in equation (1). cl X o is expressed as the current left stereo coordinate system F cl , mounted under the gripper, with a known transformation to the center point coordinate system tcp of the moving robot end effector.

[0095]

[0096] The above equation is a homogeneous transformation matrix from F cl to the world coordinate system of the robot base, where b R cl = b R tcp tcp R cl , b t cl = b t tcp tcp + b R tcp tcp t cl , the grasp point is expressed in the world coordinate as:

[0097] X0= H cl clX0. (18)

[0098] On the contrary, s * in equation (1) constitutes a three-dimensional point X o * , the depth d de is constructed, i.e. the distance in the z-axis of the stereo coordinate system to the center point of the robot end effector, and by repeating the calculation of equations (16) and (18) on the (u,v) coordinates on the image plane, which correspond to the target point of the grasping device. The controller 10 obtains the target point of the grasping structure according to the above operation and sends it to the robot and its end grasping structure.

[0099] In summary, the following function of the controller is constructed:

[0100] υ c = -K e = -K(X O -X O * )

[0101] where v c ∈ R 3 is the translational component of the velocity (v,v,v) only, e represents the position error, and K ∈ R 3×3Diagonal controller gain matrix. Calculate joint values for each step by inverse kinematics. This control law is used for the control diagram as shown in Figure 12

[0102] In summary, through the left and right cameras and image acquisition and processing module, the bag intelligent vision recognition control system can realize high-precision bag shape and position information extraction. This vision recognition capability can accurately locate the bag, ensuring that the robot can accurately grasp the bag, avoiding errors and losses. In addition, the feature detection module can extract key feature information of the bag, quickly switch the gripper material taking mechanism and the suction cup material taking mechanism according to the bag stacking situation, and realize adaptive grasping mode. This enables the robot to operate flexibly according to different bag characteristics and working environment, improving work efficiency and success rate. Furthermore, the pose estimation module can accurately estimate the pose of the robot end by analyzing the feature information. This is very important for the fine control of the robot, which can ensure that the robot end maintains the correct pose during operation, thereby avoiding material slipping, collision and other situations, and improving the success rate of grasping and placing. Finally, the feedback operation module can transmit the target pose and the actual feedback pose information to the controller. The controller integrates the feedback information for operation control and sends instructions to the robot and its end grasping structure. This feedback control mechanism can adjust the robot's action in real time, making it more accurate in task execution and dynamically adjusting according to actual conditions. This control method has strong flexibility and adaptability, and can be set and adjusted according to different types of bag materials, suitable for various fields of loading and unloading operations, such as warehousing, logistics, production line, etc. At the same time, through the optimization and upgrading of software algorithm, it can also cope with more complex bag shape and stacking situation, improving the intelligent level of the system.

[0103] As shown in Figure 11 , a working method of an intelligent unloading machine suitable for box freight bagged materials, specifically:

[0104] S1: Before the unloading machine works, keep the non-working state and automatically drive along the center line of the carriage to a certain distance in front of the bag, as shown in Figure 6 . During the driving process, the whole vehicle is moved by the double-track wheels. During the movement, the posture of the vehicle is adjusted by the differential action of the double-track wheels and the rotation action of the rotary support.

[0105] S2: The grasping sequence of the bag is as shown in Figure 7 ​The S-shaped material taking sequence from top to bottom is adopted. When the unloading machine is working, the unloading starts from the highest part of each section. The swing arm conveying mechanism swings towards the bag side, and the telescopic conveying of the swing arm conveying mechanism is extended to the longest, at this time, the front end of the swing arm conveying mechanism approaches the bag stack. After the bag position is determined by the bag intelligent visual recognition control system, the end tool of the industrial robot is quickly replaced by the gripper taking mechanism. The gripper clamps the bag from the exposed end of the bag and drags the stack, then the gripper releases the bag for the next grabbing, at the same time, the bag falls to the swing arm conveying mechanism by gravity, and then is conveyed to the telescopic conveying mechanism through the swing arm conveying mechanism, and finally is output from the telescopic conveying mechanism to the downstream equipment, such as Figure 8 as shown.

[0106] S3: The unloading machine takes material along the sequence from top to bottom, and as the stack height of the bag decreases to a certain height, the swing arm conveying mechanism adjusts the length and angle to adapt to the stack height of the bag to complete the material taking work. If necessary, the unloading machine as a whole can also move forward and backward to adjust the distance between the front end of the swing arm and the bag. During the whole unloading process, the unloading machine adjusts the state as Figure 9 shown.

[0107] S4: When the bag at a lower position is taken, the swing arm conveying mechanism is lowered to the lowest position as a whole, and the working space and working mode of the gripper taking mechanism are limited. After the bag position is determined by the bag intelligent visual recognition control system, the end tool of the industrial robot is quickly replaced by the suction cup taking mechanism. After the suction cup taking mechanism grabs the bag from above, it is translated and dragged to the swing arm conveying mechanism, and then is conveyed to the telescopic conveying mechanism through the swing arm conveying mechanism, and finally is output from the telescopic conveying mechanism to the downstream equipment. As Figure 10 shown.

[0108] S5: After the unloading machine completes the bag taking of the first section, the whole vehicle advances under the action of the track wheel, and the telescopic conveying mechanism is extended to adapt to the connection of the downstream equipment, so as to start the unloading work of the second section. Repeat the above steps until the unloading work in this direction is completed.

[0109] S6: After the unloading work in one direction is completed, the industrial robot is retracted, the swing arm conveying mechanism is retracted and becomes vertical, the telescopic conveying mechanism is retracted, and the whole vehicle returns to the Figure 6 state. When the train wagon is unloaded by the wagon shed worker, the whole vehicle is returned to the door at this time, and the whole vehicle is turned 180° in place by the differential driving of the two track wheels, and then the bag disassembly work on the other side of the door is started.

[0110] The beneficial effects of the present application are as follows:

[0111] (1) Tracked chassis and power system: The tracked chassis uses rubber track wheels, which can adapt to various operating environments and complex road conditions. The two track wheels are each equipped with a servo motor, and the two track wheels are controlled by differential speed to realize the forward, backward and turning of the whole vehicle. A slewing support is used between the tracked chassis and the frame, which can further fine adjust the vehicle angle during travel, making the whole vehicle more adaptable to the travel space.

[0112] (2) Two-stage or multi-stage telescopic conveyor mechanisms add a telescopic mechanism to the ordinary belt conveyor, allowing the conveyor to extend and retract freely in the length direction, and the extension length can be adjusted as required. During operation, the telescopic conveyor belt may deviate to the side, causing belt wear and reducing the service life of the equipment.

[0113] (3) The intelligent visual recognition control system for bags adopts intelligent visual algorithms, has good environmental adaptability, and can accurately identify the position and posture of bags, effectively assisting onboard mechanical equipment in efficiently completing bag loading and unloading operations. Through left and right cameras and image acquisition and processing modules, the intelligent visual recognition control system for bags can achieve high-precision extraction of bag shape and position information. This visual recognition capability can accurately locate bags, ensuring that the robot can accurately grasp bags and avoid errors and losses. In addition, the feature detection module can extract key feature information of bags and quickly switch between gripper and suction cup grasping mechanisms according to the bag stacking situation to achieve adaptive grasping mode. This allows the robot to operate flexibly according to different bag characteristics and working environment, improving work efficiency and success rate. Furthermore, the posture estimation module can accurately estimate the posture of the robot end by analyzing feature information. This is very important for the fine control of the robot, ensuring that the robot end maintains the correct posture during operation, thereby avoiding material slippage, collisions, etc., and improving the success rate of grasping and placement. Finally, the feedback calculation module can transmit the target posture and the actual feedback posture information to the controller. The controller performs calculations and controls by integrating feedback information, and sends instructions to the robot and its end effector. This feedback control mechanism can adjust the robot's movements in real time, enabling it to execute tasks more precisely and dynamically adjust according to actual conditions. This control method has strong flexibility and adaptability, and can be set and adjusted according to different types of bagged materials, making it suitable for loading and unloading operations in various fields, such as warehousing, logistics, and production lines. Furthermore, through software algorithm optimization and upgrades, it can also handle more complex bag shapes and stacking situations, improving the system's intelligence level.

[0114] (4) The walking guidance and correction system adopts a composite guidance system and is combined with RFID radio frequency identification technology to realize the autonomous guidance and real-time correction of the whole machine. Especially in narrow spaces such as boxcars, the control of the vehicle's walking posture is a key technical point that determines the smooth operation of the unloading machine.

[0115] (5) The six-axis industrial robot in the industrial robot and the end tool thereof has sufficient degrees of freedom to complete various actions, the quick-change clamp mechanism is composed of a jaw material taking mechanism and a suction cup material taking mechanism, can quickly change the target according to the stacking condition of the bag, has reliable grabbing, flexible movement, is convenient for intelligent control and has high adaptability.

[0116] Those skilled in the art will easily understand that the above description is only the preferred embodiment of the present application, and is not used to limit the present application, and any modification, equivalent replacement and improvement made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for operating an intelligent unloading machine suitable for bagged materials in box-type freight transport, applied to an intelligent unloading machine, wherein the intelligent unloading machine includes a tracked chassis and power system, a telescopic conveying mechanism, a robot system, a quick-change clamping structure, a swing-arm conveying mechanism, a bag intelligent vision recognition control system, a walking guidance and correction system, an electrical control system, and a remote control handle; the tracked chassis and power system include track wheels, an external toothed slewing bearing, and a frame; the frame is an integrated frame, the frame is connected to the external toothed ring flange on the external toothed slewing bearing, the external toothed ring meshes with a pinion, and the pinion... The gears provide power input, allowing the entire vehicle body to adjust its angle under the adjustment of the external gear rotary support; the telescopic conveyor mechanism includes several stages of relatively movable output belts that extend and retract along the length of the unloader to ensure the unloader itself can move within a small space, while adapting to the working environment of trucks of different lengths during operation; the robot system includes a six-axis industrial robot, a robot mounting base, and a quick-change gripper mechanism at the robot's end effector; this quick-change gripper mechanism is a quick-change structure composed of a gripper picking mechanism and a suction cup picking mechanism, which adjusts according to the stacking of bags. The system features a quick-change gripper; a swing-arm conveyor mechanism comprising a two-stage telescopic conveyor serving as the material input component of the unloading machine; the swing-arm conveyor mechanism is hinged and fixed to the front end of the telescopic conveyor mechanism, enabling overall up-and-down swinging, and in conjunction with the extension and retraction of its own length, ensuring it gets as close as possible to the bagged material; a bag intelligent vision recognition control system employs intelligent vision algorithms to accurately identify the bag's position and posture, and the electrical control system controls the quick-change gripper structure to pick up material based on the recognition results of the bag intelligent vision recognition control system; a walking guidance and correction system uses a composite guidance system combined with RFID radio frequency identification technology to control the tracked chassis and power system, achieving autonomous guidance and real-time correction of the unloading machine; the telescopic conveyor mechanism and robot system mounting base are both mounted on the frame, integrated with the frame; the electrical control system is electrically connected to the tracked chassis and power system, the telescopic conveyor mechanism, the robot system, the swing-arm conveyor mechanism, the bag intelligent vision recognition control system, and the walking guidance and correction system, controlling the operation of each component of the unloading machine; and a remote control handle that communicates with the electrical control system for remote control of the loading and unloading machine's actions; its key feature is... The operating method of the intelligent unloading machine includes the following steps: S1: Before the unloading machine starts working, it first maintains a non-working state and automatically travels along the center line of the car body to a certain distance in front of the bags; during the travel, the double track wheels drive the whole vehicle to move; during the movement, the differential action of the double tracks and the slewing action of the slewing support work together to control and adjust the vehicle's posture. S2: The bag grabbing sequence adopts an S-shaped grabbing sequence from top to bottom; when the unloading machine is working, it starts unloading from the highest point of each section; the swing arm conveyor swings towards the bag side, and the telescopic conveyor of the swing arm conveyor extends to its longest length. At this time, the front end of the swing arm conveyor system is close to the bag stack; the bag intelligent vision recognition control system extracts the bag shape and position information through left and right cameras and image acquisition and processing modules, identifies the bag position and posture, and after the bag intelligent vision recognition control system determines the bag position, the feature detection module extracts the key feature information of the bag, and switches the gripper grabbing mechanism and suction cup grabbing mechanism according to the bag stacking situation. The industrial robot end tool quickly switches to the gripper grabbing mechanism. The gripper grabs the bag from the exposed end of the bag and drags it down from the stack. After the gripper releases the bag for the next grabbing, the bag falls to the swing arm conveyor by its own weight, and is then transported to the telescopic conveyor through the swing arm conveyor, and finally output from the telescopic conveyor to the downstream equipment; S3: The unloading machine picks up materials in a top-to-bottom sequence. As the stacking height of the bags decreases to a certain height, the swing arm conveyor mechanism adjusts its length and angle to adapt to the stacking height of the bags and completes the receiving work. When necessary, the unloading machine as a whole can also move back and forth to adjust the distance between the front end of the swing arm and the bags. S4: When the bag is picked up at a lower position, the swing arm conveyor mechanism swings down to its lowest position, which restricts the working space and operation mode of the gripper picking mechanism. After the intelligent vision recognition control system for the bag determines the position of the bag, the end tool of the industrial robot quickly switches to the suction cup picking mechanism. The suction cup picking mechanism grabs the bag from above and moves it horizontally and drags it to the swing arm conveyor mechanism, then conveys it to the telescopic conveyor mechanism through the swing arm conveyor mechanism, and finally outputs it to the downstream equipment from the telescopic conveyor mechanism. S5: After the unloading machine completes the bagging of the first section, the whole vehicle moves forward under the action of the track wheels. At the same time, the telescopic conveyor extends to adapt to the connection of downstream equipment, thus starting the unloading work of the second section; repeat this process until the unloading operation in this direction is completed. S6: After completing the unloading operation in one direction, the industrial robot retracts, the swing arm conveyor retracts and becomes upright, and the telescopic conveyor retracts; at this time, the whole vehicle returns to the door and completes a 180° turn on the spot by traveling at the differential speed of the two track wheels at the door, and then begins the bag unloading operation on the other side of the door.

2. The operating method of the intelligent unloading machine for box-type bagged materials according to claim 1, characterized in that, The intelligent visual recognition control system for bags includes left and right cameras mounted at the end of the robot, an image acquisition and processing module, joint sensors, and a controller. The controller includes a joint controller and a power amplifier. The image acquisition and processing module includes a feedback calculation module, a posture estimation module, a feature detection module, and an image acquisition module. The joint sensors acquire the angle and displacement parameters of each joint of the robot and send them to the controller. The image acquisition module transmits the images acquired by the cameras to the feature detection module. The feature detection module extracts the overall image feature information and the key feature information of the bag's shape and position from the image, and then transmits the feature information to the posture estimation module. The attitude estimation module estimates the precise attitude of the robot's end effector and the attitude of the bag to be grasped based on the above feature information and transmits it to the feedback calculation module. The feedback calculation module transmits the target attitude of the robot's end effector, the attitude of the target bag, and the feedback attitude information to the controller. The controller integrates the feedback information from the feedback calculation module and the joint sensors to calculate the target point of the grasping structure and sends instructions to the robot and its end effector grasping structure.

Images