Van truck cargo unloading robot control system and method

Abstract

The invention relates to the technical field of cargo unloading control technologies, and discloses a control system and method for a cargo unloading robot of a van truck. A van truck cargo unloading robot control system comprises an acquisition module, a main control module, an execution module and a light supplementing module. A control method of a van truck cargo unloading robot comprises the steps that multi-mode three-dimensional point clouds of a truck carriage and cargoes are obtained, the multi-mode three-dimensional point clouds are preprocessed, and ordered three-dimensional point clouds are obtained; performing path planning on the ordered three-dimensional point cloud through improved RRT * to obtain an initial path trajectory; optimizing the initial path trajectory to generate a predicted path trajectory; and based on the predicted path trajectory, data analysis is carried out through adaptive feedback PID control, and optimal control parameters of the mechanical arm are obtained. The unloading efficiency is effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of freight unloading control, in particular to a box car freight unloading robot control system and method. BACKGROUND

[0002] In the field of logistics and freight transportation, box car freight unloading operations have long relied on manual or simple equipment, which has problems such as low efficiency, high labor intensity, and poor precision. With the development of automation technology, unloading robots have gradually been applied, but existing control methods have significant shortcomings:

[0003] (1) The path planning of traditional unloading robots does not fully consider the characteristics of the narrow space of the box car compartment and the complex stacking of the goods, which can easily cause the robot arm to collide with the compartment wall;

[0004] (2) There is a lack of coordination control, and the double-robot arms interfere with each other, causing path failure or operation interruption;

[0005] (3) Existing control methods cannot adapt to the needs of robots in different unloading scenarios, resulting in large trajectory tracking errors, making it difficult to guarantee the precision of goods grabbing and connection, and affecting the continuity of the unloading process.

[0006] Based on the above-mentioned unloading robot operation efficiency and stability, it cannot meet the modern logistics large-scale and efficient unloading needs, so there is an urgent need for a box car freight unloading robot control system and method that can optimize the path planning of the unloading robot. SUMMARY

[0007] (I) Technical problems solved

[0008] To solve the problems of the prior art, the present application provides a box car freight unloading robot control system.

[0009] (II) Technical solutions

[0010] To solve the above problems, the present application provides the following technical solutions:

[0011] A box car freight unloading robot control system, comprising:

[0012] The acquisition module includes a first detection unit, a second detection unit, an identification unit and a visual perception unit; the first detection unit is used to detect whether the third conveying unit has goods, the weight of the goods and the distance of the goods; the second detection unit is used to detect whether the first conveying unit has goods, the weight of the goods and the distance of the goods; the visual perception unit is used to obtain the position of the goods and the state of the compartment space in real time; the identification unit is used to identify whether there is an obstacle in front of the robot;

[0013] The master module receives data information of the collection module, performs data analysis, and sends control instructions to the execution module.

[0014] The execution module receives the control instructions of the master module and executes actions; the execution module includes a walking execution module, a lifting execution module, a mechanical arm grabbing execution module, and a conveying execution module.

[0015] The lifting execution module includes a first lifting unit, a second lifting unit, a third lifting unit, and a fourth lifting unit; the first lifting unit is used to lift the front end of the conveying execution module, the second lifting unit is used to lift the mechanical arm grabbing execution module, the third lifting unit is used to lift the gantry, and the fourth lifting unit is used to lift the rear end of the front end of the conveying execution module.

[0016] The mechanical arm grabbing execution module includes a first driving unit, a first horizontal joint mechanical arm, a first machine picking unit, a second driving unit, a second horizontal joint mechanical arm, and a second machine picking unit; the first driving unit is used to drive the operation of the first horizontal joint mechanical arm; the second driving unit is used to drive the operation of the second horizontal joint mechanical arm; the first machine picking unit is located at the end of the first horizontal joint mechanical arm and is used to pick up goods; the second machine picking unit is located at the end of the second horizontal joint mechanical arm and is used to pick up goods.

[0017] The conveying execution module includes the first conveying unit, the second conveying unit, the third conveying unit, the fourth conveying unit, and a conveying driving unit.

[0018] The light supplementing module is located in the middle of the upper cross beam of the gantry; the light supplementing module is used to provide light compensation.

[0019] The collection module transmits data to the master module through wired or wireless communication, the master module transmits data to the execution module through wired communication, and the master module transmits data to the light supplementing module through wired communication.

[0020] Preferably, the first detection unit includes a first weighing sensor array and a first laser sensor; the first weighing sensor array is installed below the first conveying unit and is used to weigh the goods; and the first laser sensor is installed on the side of the first conveying unit and is used to measure the distance of the goods.

[0021] The second detection unit includes a second weighing sensor array and a second laser sensor; the second weighing sensor array is installed below the third conveying unit and is used to weigh the goods; and the second laser sensor is installed on the side of the third conveying unit and is used to measure the distance of the goods.

[0022] The visual perception unit comprises a first visual perception unit and a second visual perception unit; the first visual perception unit is installed on the left side of the gantry, and the first visual perception unit comprises a first laser radar sensor and a first binocular camera;

[0023] The second visual perception unit is installed on the right side of the gantry, and the second visual perception unit comprises a second laser radar sensor and a second binocular camera;

[0024] The identification unit comprises a third laser radar sensor, and the third laser radar sensor is installed in front of the second conveying unit.

[0025] Preferably, the first picking unit comprises a first picking motor and a second vacuum suction cup; the second picking unit comprises a second picking motor and a second vacuum suction cup; the first driving unit is a first driving motor; and the second driving unit is a second driving motor.

[0026] Preferably, the light supplementing module comprises an environment perception unit and a light supplementing unit; the environment perception unit is used for detecting the light intensity of the surrounding environment and transmitting the detected information to the main control module; and the light supplementing unit is used for receiving the control instruction of the main control module, and the light supplementing unit illuminates through an LED lamp.

[0027] A box truck cargo unloading robot control method, comprising:

[0028] Obtaining a multi-modal three-dimensional point cloud of a truck carriage and cargo, preprocessing the multi-modal three-dimensional point cloud to obtain an ordered three-dimensional point cloud, and constructing an environment model;

[0029] Performing path planning on the ordered three-dimensional point cloud through an improved RRT* to obtain an initial path trajectory;

[0030] Optimizing the initial path trajectory to generate a predicted path trajectory, a predicted speed of a mechanical arm, and a predicted acceleration of the mechanical arm;

[0031] Based on the data information of the predicted path trajectory, the predicted speed of the mechanical arm, and the predicted acceleration of the mechanical arm, performing data analysis through adaptive feedback PID control to obtain optimal control parameters of the mechanical arm.

[0032] Preferably, the multi-modal three-dimensional point cloud of the truck carriage and the cargo is obtained, the multi-modal three-dimensional point cloud is preprocessed to obtain an ordered three-dimensional point cloud, and an environment model is constructed, specifically comprising:

[0033] Scanning the surface of the truck carriage and the cargo through a laser radar sensor to generate an original three-dimensional point cloud;

[0034] Obtaining color, texture, and angle of incidence information of the cargo through a binocular camera;

[0035] The laser radar sensor is data-fused with data information collected by the binocular camera to obtain a multi-modal three-dimensional point cloud;

[0036] The multi-modal three-dimensional point cloud is filtered to obtain an ordered three-dimensional point cloud;

[0037] An environment model containing a cargo pose and an obstacle distribution is constructed according to the ordered three-dimensional point cloud.

[0038] Preferably, the ordered three-dimensional point cloud is path planned by the improved RRT* to obtain an initial path trajectory, specifically including:

[0039] Partition constraint, the work area of the first horizontal joint robot arm is the right half area, with the robot center axis as the boundary, and the boundary is set as X≥0; the work area of the first horizontal joint robot arm is the left half area, with the robot center axis as the boundary, and the boundary is set as X<0;

[0040] Each node coordinate in the path is detected in real time, and when the first horizontal joint robot arm enters the left half area or the second horizontal joint robot arm enters the right half area, it is determined as a border crossing behavior, and the path priority is reduced by 50%;

[0041] Obstacle avoidance, constructing a distance cost for the obstacle, and marking the obstacle area as an impassable area to avoid the obstacle;

[0042] According to the partition constraint and the obstacle avoidance, path planning is performed, and the specific steps are as follows:

[0043] A. Taking the initial pose of the robot arm as the starting point and the target point as the end point, a root node of a path tree is constructed, and all the grabbing poses of the cargos are extracted from the ordered three-dimensional point cloud as a target set for path planning;

[0044] B. Random sampling is performed in the grabbing pose set to generate a path tree;

[0045] C. The path segment cost of adjacent path points in the path tree is calculated, and if the cost of a new path segment is lower than that of an old path segment, the path tree is updated;

[0046] D. All paths reaching the target point in the path tree are traversed, the total path cost is calculated, the path with the minimum total cost is obtained, and an initial path trajectory is obtained.

[0047] Preferably, the initial path trajectory is optimized to generate a predicted path trajectory, a predicted speed of the robot arm, and a predicted acceleration of the robot arm, specifically including:

[0048] The initial path trajectory is optimized into a continuous and smooth trajectory based on a B-spline basis function;

[0049] The continuous smooth trajectory is mapped to an actual time dimension to obtain a shortest-time prediction path trajectory, and a robot arm prediction speed and a robot arm prediction acceleration are obtained.

[0050] Preferably, the data information based on the prediction path trajectory, the robot arm prediction speed and the robot arm prediction acceleration is analyzed by adaptive feedback PID control to obtain optimal control parameters of the robot arm, and specifically includes:

[0051] The prediction path trajectory is compared with the real path trajectory to calculate an error between the prediction path trajectory and the real path trajectory.

[0052] According to the error, a PID parameter is adjusted in real time, a feedforward compensation coefficient is introduced, a feedforward compensation amount is calculated, and the feedforward compensation amount and a feedback adjustment amount are superimposed to generate a preliminary control instruction.

[0053] According to the preliminary control instruction, a new real path trajectory is obtained, the error is recalculated, and the PID parameter is adjusted again.

[0054] When a PID parameter change rate of 50 continuous time points is less than 1%, the optimal control parameter is output.

[0055] (Three) beneficial effects

[0056] Compared with the prior art, the application provides a box truck cargo unloading robot control system and method, which has the following beneficial effects:

[0057] 1. The application strictly divides the double robot arm operation area and punishes the out-of-bound path to avoid robot arm collision, adapts to the narrow space operation of the box truck, and ensures the safety of the equipment and the cargo; the distance cost and the repulsive field are constructed for the obstacles to automatically avoid the path of the obstacles such as the compartment wall and the cargo pile, reduce the collision risk, and improve the path planning success rate and the operation stability.

[0058] 2. The application analyzes the error feedback and multi-round iterative training through adaptive feedback PID control to continuously adapt the control parameters to the operation scene, so that even in the face of differences in truck models and changes in cargo stacking, the optimal control instruction can still be stably output, and the adaptability of the robot to complex working conditions is improved.

[0059] 3. The application optimizes the initial path trajectory with time and efficiency as the target, converts the discrete initial path into a continuous smooth trajectory, reduces the number of robot arm acceleration and deceleration, reduces the risk of equipment wear and cargo shaking, shortens the total time of robot arm movement, and improves the overall efficiency of the logistics unloading link.

[0060] Additional aspects and advantages of the application will be in part apparent and in part pointed out hereinafter in the description. BRIEF DESCRIPTION OF DRAWINGS

[0061] The above and / or additional aspects and advantages of the present application will become apparent and more readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings in which:

[0062] Figure 1 It is a schematic diagram of a box truck cargo unloading robot control system of the present application;

[0063] Figure 2 It is a schematic diagram of a box truck cargo unloading robot control system structure of the present application;

[0064] Figure 3 It is a schematic diagram of a box truck cargo unloading robot control system structure of the present application;

[0065] Figure 4 It is a schematic diagram of a box truck cargo unloading robot control system structure of the present application;

[0066] Figure 1 is a schematic diagram of a box truck cargo unloading robot control system of the present application; DETAILED DESCRIPTION

[0067] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0068] The terms "first", "second" in the description and claims of the present application can explicitly or implicitly include one or more of the features. In the description of the present application, unless otherwise specified, the meaning of "a plurality of" is two or more. In addition, "and / or" in the specification and claims means at least one of the connected objects, and the character " / ", generally means that the front and rear associated objects are in an "or" relationship.

[0069] In the description of the present application, it should be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application.

[0070] In the description of the present application, it should be noted that unless otherwise specified and limited, the terms "mounting", "connection", "connection" should be understood broadly, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium; it can be the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0071] Please refer to Figures 1-3 The present application provides a new technical solution: a box truck cargo unloading robot control system, comprising:

[0072] The acquisition module 1, the main control module 2, the execution module 3 and the light supplementing module 4; the main control module 2 is connected with the execution module 3 through the CAN bus.

[0073] The acquisition module 1 comprises a first detection unit 110, a second detection unit 120, an identification unit and a visual perception unit.

[0074] The first detection unit 110 comprises a first weighing sensor array 111 and a first laser sensor 112.

[0075] The second detection unit 120 comprises a second weighing sensor array 121 and a second laser sensor 122.

[0076] The visual perception unit comprises a first visual perception unit 140 and a second visual perception unit 150; the first visual perception unit 140 comprises a first laser radar sensor 141 and a first binocular camera 142; and the second visual perception unit 150 comprises a second laser radar sensor 151 and a second binocular camera 152.

[0077] The recognition unit comprises a third laser radar sensor 130.

[0078] The execution module 3 comprises a walking execution module, a lifting execution module, a mechanical arm grabbing execution module 330, and a conveying execution module 340.

[0079] The lifting execution module comprises a first lifting unit 321, a second lifting unit 322, a third lifting unit 323, and a fourth lifting unit 324.

[0080] The mechanical arm grabbing execution module 330 comprises a first driving unit, a first horizontal joint mechanical arm 332, a first mechanical picking unit 333, a second driving unit, a second horizontal joint mechanical arm 335, and a second mechanical picking unit 336.

[0081] The conveying execution module 340 comprises the first conveying unit 341, the second conveying unit 342, the third conveying unit 343, the fourth conveying unit 344, and a conveying driving unit.

[0082] The light supplement module 4 comprises an environment perception unit 410 and a light supplement unit 420.

[0083] The environment perception unit 410 is a light intensity sensor that continuously monitors the light intensity in the vehicle cabin. When the detected light intensity is lower than the preset threshold, a signal is sent to the main control module 2 to trigger the light supplement unit 420 to work.

[0084] A box truck cargo unloading robot control method comprises:

[0085] Obtaining a multi-modal three-dimensional point cloud of the truck cabin and the cargo, preprocessing the multi-modal three-dimensional point cloud to obtain an ordered three-dimensional point cloud, and constructing an environment model;

[0086] An improved RRT* is used to plan a path for the ordered three-dimensional point cloud to obtain an initial path trajectory;

[0087] The initial path trajectory is optimized to generate a predicted path trajectory, a predicted mechanical arm speed, and a predicted mechanical arm acceleration;

[0088] Based on the data information of the predicted path trajectory, the predicted mechanical arm speed, and the predicted mechanical arm acceleration, data analysis is performed through adaptive feedback PID control to obtain optimal mechanical arm control parameters.

[0089] In the application, the multi-modal three-dimensional point cloud of the truck compartment and the goods is acquired, the multi-modal three-dimensional point cloud is preprocessed to obtain an ordered three-dimensional point cloud, and an environment model is constructed, specifically comprising:

[0090] The surface of the truck compartment and the goods is scanned by a laser radar sensor to generate an original three-dimensional point cloud;

[0091] The color, texture and angle of incidence information of the goods are acquired by a binocular camera;

[0092] The data information acquired by the laser radar sensor and the binocular camera is fused to obtain a multi-modal three-dimensional point cloud;

[0093] The multi-modal three-dimensional point cloud is filtered to obtain an ordered three-dimensional point cloud;

[0094] An environment model containing the pose of the goods and the distribution of obstacles is constructed according to the ordered three-dimensional point cloud.

[0095] In a specific implementation, the mathematical expression of the three-dimensional point cloud is:

[0096]

[0097] In formula (1), C is a set of goods point clouds, O is a set of obstacle point clouds, X is a coordinate in the width direction of the compartment, Y is a coordinate in the direction of the unloading robot, and Z is a height coordinate; the working partition of the first horizontal joint robot arm is X>0, the working partition of the second horizontal joint robot arm is X<0, and X=0 is the midpoint of the first horizontal joint robot arm and the second horizontal joint robot arm;

[0098] The three-dimensional point cloud is filtered to obtain effective goods point clouds C' and effective obstacle point clouds O'.

[0099] In the application, the ordered three-dimensional point cloud is path planned by an improved RRT* to obtain an initial path trajectory, specifically comprising:

[0100] Partition constraint, the working area of the first horizontal joint robot arm is the right half area, the boundary is set as X>0 with the robot central axis as the boundary; the working area of the first horizontal joint robot arm is the left half area, the boundary is set as X<0 with the robot central axis as the boundary;

[0101] The coordinates of each node in the path are detected in real time, when the first horizontal joint robot arm enters the left half area or the second horizontal joint robot arm enters the right half area, it is determined as a border crossing behavior, and the priority of the path is reduced by 50%;

[0102] Obstacle avoidance, the distance cost of the obstacle is constructed, and the obstacle area is marked as an impassable area to avoid the obstacle;

[0103] According to the partition constraint and obstacle avoidance, path planning is performed, and the specific steps are as follows:

[0104] A. Taking the initial pose of the mechanical arm as the starting point and the target point as the end point, a root node of a path tree is constructed, and a set of all the grabbing poses of the goods is extracted from the ordered three-dimensional point cloud as a target set of path planning;

[0105] B. Random sampling is performed in the set of grabbing poses to generate a path tree;

[0106] C. The path segment cost of adjacent path points in the path tree is calculated, and if the cost of a new path segment is lower than that of an old path segment, the path tree is updated;

[0107] D. All paths reaching the target point in the path tree are traversed, the total path cost is calculated, the path with the minimum total cost is obtained, and the initial path trajectory is obtained.

[0108] In the specific implementation, the path of the grabbing pose of the goods is Q={q0,q1,q i ,…,q M |q M ∈C'}.

[0109]

[0110] In formula (2), σ is a boundary punishment coefficient, τ(q) is a boundary judgment function, γ is a decay coefficient, d min is the shortest distance from q i , q i+1 to the obstacle, μ is the repulsive force weight, q is the grabbing pose coordinate of the goods, and M is the number of nodes.

[0111] The mathematical expression of the total path cost is:

[0112]

[0113] In formula (3), c is the total path cost.

[0114] The optimal path Q'={q'0,q1',…,q' i ,…,q' M |q' M ∈C'} satisfies Q'=min{c}.

[0115] In the present application, the initial path trajectory is optimized to generate a predicted path trajectory, a predicted speed of the mechanical arm and a predicted acceleration of the mechanical arm, and the specific steps are as follows:

[0116] The initial path trajectory is optimized to a continuous and smooth trajectory based on a B-spline basis function;

[0117] Map the continuous smooth trajectory to the actual time dimension to obtain the shortest time-consuming prediction path trajectory, and obtain the robot prediction speed and the robot prediction acceleration.

[0118] In specific implementation, the mathematical expression for generating the prediction smooth trajectory is:

[0119]

[0120] In formula (4), s(t) is the prediction smooth trajectory, N i,b (u(t)) is the B-spline basis function;

[0121] The mathematical expression for deriving the velocity and acceleration of the smooth trajectory is:

[0122]

[0123] In formula (5), v(t) is the prediction speed, and a(t) is the prediction acceleration.

[0124] In the application, the data information based on the prediction path trajectory, the robot prediction speed and the robot prediction acceleration is analyzed through adaptive feedback PID control to obtain the optimal control parameters of the robot, specifically including:

[0125] The prediction path trajectory is compared with the real path trajectory to calculate the error between the prediction path trajectory and the real path trajectory.

[0126] According to the error, the PID parameters are adjusted in real time, a feedforward compensation coefficient is introduced, a feedforward compensation amount is calculated, the feedforward compensation amount and the feedback adjustment amount are superimposed to generate a preliminary control instruction.

[0127] According to the preliminary control instruction, a new real path trajectory is obtained, the error is recalculated, and the PID parameters are adjusted again.

[0128] When the PID parameter change rate of 50 consecutive time points is less than 1%, the optimal control parameters are output.

[0129] In specific implementation, the mathematical expression for outputting the control instruction through adaptive feedback PID control is:

[0130]

[0131] In formula (6), the error term e(t) = s(t)-s'(t), wherein s'(t) is the real trajectory, K p (t) is the proportional gain, K i (t) is the integral gain, K d (t) is the differential gain, K f is the feedforward coefficient.

[0132] The mathematical expression of the adaptive parameter adjustment is:

[0133]

[0134] In the formula (7), K p0 is a proportional gain base value, K i0 is an integral gain base value, K d0 is a differential gain base value, v max is a maximum speed.

[0135] According to U(t) output by the control unit, the control parameters of the first horizontal joint mechanical arm Φ1, the joint angle of the second horizontal joint mechanical arm Φ2, the height of the first lifting unit h1, the height of the second lifting unit h2, the height of the third lifting unit h3, and the height of the fourth lifting unit h4 are obtained.

[0136] In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "exemplary embodiment", "example", "specific example", or "some examples" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the exemplary description of the above terms does not necessarily mean the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0137] Although the embodiments of the present application have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made thereto without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A box truck cargo unloading robot control system, characterized by, The application relates to a robot for picking up goods in a vehicle, which comprises a collecting module, a main control module and an executing module. The collecting module comprises a first detecting unit, a second detecting unit, an identifying unit and a visual perception unit; the first detecting unit is used for detecting whether the third conveying unit exists, the weight of the goods and the distance of the goods; the second detecting unit is used for detecting whether the first conveying unit exists, the weight of the goods and the distance of the goods; the visual perception unit is used for acquiring the position of the goods and the state of the vehicle space in real time; and the identifying unit is used for identifying whether there is an obstacle in front of the robot. The main control module receives the data information of the collecting module, carries out data analysis and sends control instructions to the executing module. The executing module receives the control instructions of the main control module and executes actions; the executing module comprises a walking executing module, a lifting executing module, a mechanical arm grabbing executing module and a conveying executing module. The walking executing module is used for driving the robot to move. The lifting executing module comprises a first lifting unit, a second lifting unit, a third lifting unit and a fourth lifting unit; the first lifting unit is used for lifting the front end of the conveying executing module; the second lifting unit is used for lifting the mechanical arm grabbing executing module; the third lifting unit is used for lifting the gantry; and the fourth lifting unit is used for lifting the rear end of the front end of the conveying executing module. The mechanical arm grabbing executing module comprises a first driving unit, a first horizontal joint mechanical arm, a first machine picking unit, a second driving unit, a second horizontal joint mechanical arm and a second machine picking unit; the first driving unit is used for driving the first horizontal joint mechanical arm to run; the second driving unit is used for driving the second horizontal joint mechanical arm to run; the first machine picking unit is located at the tail end of the first horizontal joint mechanical arm and is used for picking up goods; and the second machine picking unit is located at the tail end of the second horizontal joint mechanical arm and is used for picking up goods. The conveying executing module comprises the first conveying unit, the second conveying unit, the third conveying unit, the fourth conveying unit and a conveying driving unit. A light supplementing module is located in the middle of the upper cross beam of the gantry; the light supplementing module is used for providing light compensation. The collecting module transmits data to the main control module through wired or wireless communication; the main control module transmits data to the executing module through wired communication; and the main control module transmits data to the light supplementing module through wired communication.

2. The control system for a boxcar unloading robot according to claim 1, wherein, The first detecting unit comprises a first weighing sensor array and a first laser sensor; the first weighing sensor array is installed below the first conveying unit and is used for weighing the goods; and the first laser sensor is installed on the side of the first conveying unit and is used for measuring the distance of the goods. The second detecting unit comprises a second weighing sensor array and a second laser sensor; the second weighing sensor array is installed below the third conveying unit and is used for weighing the goods; and the second laser sensor is installed on the side of the third conveying unit and is used for measuring the distance of the goods. The visual perception unit comprises a first visual perception unit and a second visual perception unit; the first visual perception unit is installed on the left side of the gantry, and the first visual perception unit comprises a first laser radar sensor and a first binocular camera; The second visual perception unit is installed on the right side of the gantry, and the second visual perception unit comprises a second laser radar sensor and a second binocular camera; The identification unit comprises a third laser radar sensor, and the third laser radar sensor is installed in front of the second conveying unit.

3. The boxcar cargo unloading robot control system of claim 1, wherein, The first picking unit comprises a first picking motor and a second vacuum suction cup; the second picking unit comprises a second picking motor and a second vacuum suction cup; the first driving unit is a first driving motor; and the second driving unit is a second driving motor.

4. The boxcar cargo unloading robot control system of claim 1, wherein, The light supplementing module comprises an environment perception unit and a light supplementing unit; the environment perception unit is used for detecting the light intensity of the surrounding environment and transmitting the detected information to the main control module; and the light supplementing unit is used for receiving the control instruction of the main control module, and the light supplementing unit illuminates through an LED lamp.

5. A box truck cargo unloading robot control method for the box truck cargo unloading robot control system of claims 1-4, comprising: obtaining a multi-modal three-dimensional point cloud of the truck compartment and the cargo, preprocessing the multi-modal three-dimensional point cloud to obtain an ordered three-dimensional point cloud, and constructing an environment model; path planning for the ordered three-dimensional point cloud by an improved RRT* to obtain an initial path trajectory; optimizing the initial path trajectory to generate a predicted path trajectory, a predicted speed of the robot arm, and a predicted acceleration of the robot arm; based on the data information of the predicted path trajectory, the predicted speed of the robot arm, and the predicted acceleration of the robot arm, performing data analysis by adaptive feedback PID control to obtain optimal control parameters of the robot arm.

6. The method of claim 5, wherein the robot is a boxcar unloading robot. The multi-modal three-dimensional point cloud of the truck compartment and the cargo is obtained, the multi-modal three-dimensional point cloud is preprocessed to obtain an ordered three-dimensional point cloud, and an environment model is constructed, specifically comprising: scanning the surface of the truck compartment and the cargo by a laser radar sensor to generate an original three-dimensional point cloud; obtaining color, texture, and angle of incidence information of the cargo by a binocular camera; performing data fusion on the data information collected by the laser radar sensor and the binocular camera to obtain a multi-modal three-dimensional point cloud; performing filtering processing on the multi-modal three-dimensional point cloud to obtain an ordered three-dimensional point cloud; constructing an environment model containing the pose of the cargo and the distribution of obstacles according to the ordered three-dimensional point cloud.

7. The method of claim 6, wherein the robot is a boxcar unloading robot. The ordered three-dimensional point cloud is path planned by an improved RRT* to obtain an initial path trajectory, specifically comprising: partition constraint, the work area of the first horizontal joint robot arm is the right half area, the boundary is set as X≥0 with the robot central axis as the boundary; the work area of the first horizontal joint robot arm is the left half area, the boundary is set as X<0 with the robot central axis as the boundary; real-time detection of the coordinates of each node in the path, when the first horizontal joint robot arm enters the left half area or the second horizontal joint robot arm enters the right half area, it is determined as a border crossing behavior, and the priority of the path is reduced by 50%. Obstacle avoidance, constructing distance cost for obstacles, and marking obstacle area as impassable area to avoid obstacles; According to the partition constraint and obstacle avoidance, path planning is carried out, and the specific steps are as follows: A. Taking the initial pose of the mechanical arm as the starting point and the target point as the end point, the root node of the path tree is constructed, and the set of all grabbing poses of the goods is extracted from the ordered three-dimensional point cloud as the target set of path planning; B. Random sampling is carried out in the set of grabbing poses to generate a path tree; C. The path segment cost of adjacent path points in the path tree is calculated, and if the cost of the new path segment is lower than that of the old path segment, the path tree is updated; D. All paths reaching the target point in the path tree are traversed, the total path cost is calculated, the path with the minimum total cost is obtained, and the initial path trajectory is obtained.

8. The method of claim 7, wherein the robot is a boxcar unloading robot. The initial path trajectory is optimized to generate a predicted path trajectory, a predicted speed of the mechanical arm and a predicted acceleration of the mechanical arm, specifically including: Based on the B-spline basis function, the initial path trajectory is optimized to a continuous smooth trajectory; The continuous smooth trajectory is mapped to the actual time dimension to obtain the shortest time-consuming predicted path trajectory, and the predicted speed of the mechanical arm and the predicted acceleration of the mechanical arm are obtained.

9. The method of claim 8, wherein the robot is a boxcar unloading robot. The data information based on the predicted path trajectory, the predicted speed of the mechanical arm and the predicted acceleration of the mechanical arm is analyzed through adaptive feedback PID control to obtain the optimal control parameters of the mechanical arm, specifically including: The predicted path trajectory is compared with the real path trajectory to calculate the error between the predicted path trajectory and the real path trajectory; According to the error, the PID parameters are adjusted in real time, and a feedforward compensation coefficient is introduced to calculate the feedforward compensation amount, and the feedforward compensation amount and the feedback adjustment amount are superimposed to generate a preliminary control instruction; According to the preliminary control instruction, a new real path trajectory is obtained, the error is recalculated, and the PID parameters are adjusted again; When the change rate of the PID parameters of the continuous 50 time points is less than 1%, the optimal control parameters are output.

Images