A gantry robotic arm for loading and unloading vehicles and its control method

Abstract

This invention relates to the field of robotic arm control technology, and discloses a gantry robotic arm for loading and unloading vehicles and its control method. The method includes: collecting material information and images of the loading / unloading positions to construct an image simulation; determining the coordinates of the starting and ending points; determining the coordinates of obstacles; and determining the optimal path based on an improved heuristic search model. The method also includes: controlling the top pressure claw to move downwards to compress the material, and determining the initial downward pressure of the top pressure claw; collecting pressure data at the bottom of the loading / unloading arm in real time to determine if the material has moved; when movement occurs, determining an adjustment coefficient and adjusting the operation to achieve the adjusted downward pressure; and determining the adjustment coefficient based on the final downward pressure and the initial downward pressure, and storing the adjustment coefficient and the material characteristic set in a historical adjustment set. This application improves the adaptability, accuracy, and safety of gantry operations by integrating information acquisition, path planning, gripping control, and dynamic adjustment.

Description

Technical Field

[0001] This invention relates to the field of robotic arm control technology, and more specifically, to a gantry robotic arm for loading and unloading vehicles and its control method. Background Technology

[0002] With the continuous development of automation and intelligent technologies, loading and unloading operations in industrial production have gradually transformed towards intelligence and automation. Especially in warehousing, logistics, and manufacturing, trusses, due to their high stability, load-bearing capacity, and large working range, have significant advantages in complex loading and unloading tasks, improving work efficiency and safety.

[0003] However, in practical applications, manual operation is still required when using trusses for material loading and unloading. Firstly, the diversity and complexity of materials present challenges to loading and unloading operations. Different materials vary significantly in size, shape, and surface characteristics (such as smoothness and hardness), and traditional control methods rely heavily on operator experience. Even with automated control, current path planning technologies primarily depend on conventional path search algorithms, such as A* and Dijkstra's algorithm. While these algorithms perform well in static environments, they have limitations in dynamic environments (such as when obstacles are constantly changing). They often overlook the changes in forces acting on the material during movement and their feedback to the loading / unloading operator's pressure. This leads to problems such as unstable gripping, material slippage, and damage, affecting operational efficiency and safety.

[0004] Therefore, it is necessary to design a gantry robotic arm for loading and unloading vehicles and its control method to solve the problems existing in the current technology. Summary of the Invention

[0005] In view of this, the present invention proposes a gantry robotic arm for loading and unloading vehicles and its control method, aiming to solve the problems of reliance on human experience, time-consuming and labor-intensive operation, and unstable gripping, material slippage and damage caused by lack of feedback adjustment when using gantry robotic arms for loading and unloading goods.

[0006] In one aspect, the present invention proposes a control method for a gantry robotic arm used for loading and unloading vehicles, comprising:

[0007] Collect material information and image data of the loading / unloading location, construct an image simulation of the loading / unloading location, and determine the material loading / unloading location based on the material size information;

[0008] The starting and ending coordinates are determined based on the material loading / unloading location, the path image data from the starting point to the ending point is collected to determine the obstacle coordinates, and the optimal path of the material is determined based on the improved heuristic search model.

[0009] Control the loading and unloading hand to grip the material, control the top pressure claw to move downward to squeeze the material, and determine the initial downward pressure of the top pressure claw;

[0010] The material is transported using the initial downforce and the optimal path, and the pressure data at the bottom of the loading and unloading handle is collected in real time to determine whether the material has moved; when it is determined that the material has moved, the initial downforce is adjusted to operate with the adjusted downforce.

[0011] The adjustment coefficient is determined based on the final downforce and the initial downforce, and the adjustment coefficient and the set of material characteristics are stored in the historical adjustment set.

[0012] Furthermore, constructing an image simulation of the loading / unloading location, and determining the material loading / unloading location based on material size information, includes:

[0013] Image data of the loading / unloading location is acquired using an RGB camera and a depth camera;

[0014] An object detection algorithm is used to identify existing objects at the loading / unloading location;

[0015] Based on the material information, obtain the material dimensions and establish a material model and a model of the loading / unloading location;

[0016] Based on the material model and the loading / unloading position model, the placement method of the material at the loading / unloading position is obtained, and the placement process of the material is simulated in the model;

[0017] If there is no interference between the material model and the placement process and the placement result, the position of the material placement result in the model shall be taken as the material loading / unloading position.

[0018] Furthermore, when collecting path image data from the starting point to the ending point to determine obstacle coordinates, and determining the optimal path for the material based on an improved heuristic search model, the process includes:

[0019] S1: Define an open set and a closed set and initialize the cost function value. The open set stores the nodes to be explored, and initially only contains the starting point. The closed set stores the nodes that have been explored, and is initially empty.

[0020] S2: Gradually select the node with the lowest cost from the open set to extend the path, and gradually move closer to the key location coordinates;

[0021] S3: Explore all possible path directions for the current node, update the path cost, and if the neighbor node is not in the open set, add the neighbor node to the open set and record the parent node of the neighbor node; if the neighbor node is in the open set, but the new path cost is smaller, update the parent node and the path cost.

[0022] S4: Repeat step S3 until the current node coincides with the coordinates of the endpoint. Then, trace back from the endpoint coordinates to the parent node until the starting point coordinates, and use this path as the optimal path.

[0023] Furthermore, when exploring all possible path directions for the current node and updating the path cost, this includes:

[0024] The path cost is calculated using the following formula:

[0025]

[0026] Where f(n) represents the path cost, g(n) represents the cumulative path cost, h(n) represents the heuristic cost, e(n) represents the rate of change of velocity between the current node and its parent node, and r(n) represents the rate of change of the path direction angle between the node and its parent node;

[0027]

[0028] Where (x, y, z) represents the coordinates of the current node, (xt, yt, zt) represents the coordinates of the target node, and P represents the probability that the current node is occupied by an obstacle. When the path image data determines that there is an obstacle at the current node coordinates, P is 1; when the path image data determines that there is no obstacle at the current node coordinates, P is the historical probability that the current node was occupied by an obstacle.

[0029]

[0030] in, This represents the cumulative cost of the parent node. This represents the geometric distance between the current node and its parent node.

[0031] Furthermore, determining the initial downward pressure of the top pressure claw includes:

[0032] The initial downward pressure of the top pressure claw is determined based on the material weight and size data, and the initial downward pressure is calculated using the following formula:

[0033]

[0034] Where F represents the initial downward pressure, m represents the material weight data, g represents the gravitational acceleration, A represents the area of ​​contact between the bottom of the loading / unloading hand and the material, dmin represents the minimum thickness of the material, and L represents the maximum length of the material.

[0035] Furthermore, real-time collection of pressure data at the bottom of the loading / unloading handle, and determination of whether material movement has occurred based on the pressure data, includes:

[0036] Arrange all pressure data at the bottom of the loading and unloading handle into a two-dimensional matrix, marking non-zero points and zero pressure points; when the position of a non-zero point in the two-dimensional matrix changes, it is determined that the material has moved.

[0037] Furthermore, when it is determined that movement has occurred, adjusting the initial downforce includes:

[0038] The material characteristics are compared with the historical material characteristic set in the historical adjustment set to determine the adjustment coefficient and adjust the initial downforce; the historical adjustment set includes the historical material characteristic set and the historical adjustment coefficient, and the historical material characteristic set includes several sets of historical material characteristics;

[0039] When there is similar data for the material characteristics in the historical material characteristic set, the average historical adjustment coefficient is obtained based on the historical adjustment coefficient in the similar data, and the average historical adjustment coefficient is used as the adjustment coefficient.

[0040] When there is no similar data for the material feature in the historical material feature set, the adjustment coefficient is determined based on the material feature set;

[0041] The similar data refers to data where the difference between the historical material characteristics and the material characteristics is less than a% of the material characteristics.

[0042] Furthermore, determining the adjustment coefficient based on the set of material characteristics includes:

[0043] ;

[0044] Where K represents the adjustment coefficient, with a value range of (1, 1.6], μ represents the surface smoothness coefficient, Ls represents the remaining path length, L0 represents the total path length, V represents the material volume, L represents the material length, W represents the material width, and D represents the material thickness. Indicates the density of the material. It represents the elastic modulus of a material.

[0045] Furthermore, after operating with the adjusted downforce, it also includes:

[0046] Collect second pressure data to determine if the material has moved. If the material moves, adjust the downpressure based on the distance moved. When determining the final downpressure, the following steps are taken:

[0047] A second two-dimensional matrix is ​​established based on the second pressure data. When the position of a non-zero point in the second two-dimensional matrix changes, it is determined that the material has moved; when the position of a non-zero point in the second two-dimensional matrix does not change, it is determined that the material has not moved, and the adjusted downward pressure is taken as the final downward pressure.

[0048] When it is determined that the material has moved, the direction of movement and the distance of movement are obtained according to the second two-dimensional matrix. The adjustment coefficient is determined according to the distance of movement, and the adjusted downward pressure is adjusted a second time to determine the final downward pressure. The adjustment coefficient is proportional to the distance of movement, and the value range of the adjustment coefficient is (1, 1.2).

[0049] Compared with existing technologies, the advantages of this invention are as follows: By integrating material information acquisition, path planning, gripping force control, and dynamic adjustment mechanisms, it overcomes the limitations of traditional gantry robotic arms in material loading and unloading processes. Through comprehensive analysis of image data acquisition and material size information, the loading and unloading positions of the materials are determined, and based on an improved heuristic path search model, path selection is optimized to avoid interference from obstacles in the dynamic environment. By monitoring the pressure data at the bottom of the loading / unloading arm in real time, the downward pressure of the loading / unloading arm is dynamically adjusted to adapt to changes in factors such as material surface smoothness and remaining path length, ensuring stable gripping and preventing material slippage or damage. Through intelligent optimization of the pressure control strategy, the robotic arm can flexibly adjust under different materials and environmental conditions, improving the adaptability, operational accuracy, and safety of gantry operations, and enhancing the efficiency and reliability of automated loading and unloading operations.

[0050] On the other hand, this application also provides a gantry robot arm for loading and unloading vehicles, for applying the above-mentioned gantry robot arm control method for loading and unloading vehicles, including:

[0051] Support frame, top pressing mechanism, clamping mechanism, and data collection device;

[0052] The support frame includes a transverse fixing plate and a longitudinal fixing plate, and two transverse fixing plates and two longitudinal fixing plates are provided.

[0053] The acquisition device includes several image acquisition units, which are fixed on the support frame;

[0054] The top pressing mechanism includes a pressing cylinder and a top pressure claw. At least two pressing cylinders and two top pressure claws are provided. The pressing cylinders are fixedly mounted on the longitudinal fixing plate, and the top pressure claws are fixedly connected to the bottom of the pressing cylinders.

[0055] The clamping mechanism includes a transverse cylinder, a connecting rod, and a loading / unloading hand. The transverse cylinder is fixedly connected to the transverse fixed plate. One end of the transverse cylinder is connected to the connecting rod, and the connecting rod is rotatably connected to the transverse cylinder. The other end of the connecting rod is connected to the loading / unloading hand, and the loading / unloading hand is rotatably connected to the connecting rod. Several pressure sensors are evenly arranged on the loading / unloading hand.

[0056] It is understandable that the aforementioned gantry robotic arms for loading and unloading vehicles and their control methods have the same beneficial effects, and will not be elaborated upon here. Attached Figure Description

[0057] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0058] Figure 1 A flowchart of a gantry robotic arm control method for loading and unloading vehicles provided in an embodiment of the present invention;

[0059] Figure 2 This is a schematic diagram illustrating the application of the gantry robotic arm for loading and unloading vehicles provided in an embodiment of the present invention.

[0060] Figure 3 This is a schematic diagram of the structure of the gantry robotic arm for loading and unloading vehicles provided in an embodiment of the present invention;

[0061] Figure 4 This is a front view of the gantry robotic arm for loading and unloading vehicles provided in an embodiment of the present invention;

[0062] Figure 5 This is a side view of the gantry robotic arm for loading and unloading vehicles provided in an embodiment of the present invention.

[0063] In the diagram: 100, X-axis steel beam; 200, Y-axis steel beam; 300, Z-axis steel beam; 400, truss robotic arm; 411, transverse fixing plate; 412, longitudinal fixing plate; 421, downward pressure cylinder; 422, top pressure claw; 431, transverse cylinder; 432, connecting rod; 433, loading and unloading hand; 510, image acquisition unit; 610, pressure sensor; 700, material. Detailed Implementation

[0064] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0065] In some embodiments of this application, see Figure 1 As shown, a method for controlling a gantry robotic arm for loading and unloading vehicles includes:

[0066] S100: Collects material information and image data of the loading / unloading location, constructs an image simulation of the loading / unloading location, and determines the material loading / unloading location based on the material size information;

[0067] S200: Determine the coordinates of the starting and ending points based on the material loading / unloading locations, collect path image data from the starting point to the ending point to determine the coordinates of obstacles, and determine the optimal path for the material based on an improved heuristic search model;

[0068] S300: Controls the loading and unloading hand to grip the material and controls the top pressure claw to move downward to squeeze the material. The initial downward pressure of the top pressure claw is determined based on the material weight data and material size data.

[0069] S400: Transports materials with initial downforce and optimal path, and collects pressure data at the bottom of the loading and unloading handle in real time. Based on the pressure data, it determines whether the material has moved. When it is determined that the material has moved, it collects the surface smoothness coefficient and the remaining path length. It combines the surface smoothness coefficient, the remaining path length and the material size information to form the material characteristics. It compares the material characteristics with the historical material characteristics set in the historical adjustment set to determine the adjustment coefficient and adjust the initial downforce. It then operates with the adjusted downforce.

[0070] S500: Collects the second pressure data at the bottom of the loading and unloading handle, and judges again whether the material has moved based on the second pressure data. When the material moves, the adjusted down pressure is increased according to the movement distance to determine the final down pressure.

[0071] S600: Determine the adjustment coefficient based on the final downforce and the initial downforce, and store the adjustment coefficient and the material characteristic set into the historical adjustment set.

[0072] Specifically, image data acquisition technology is used to obtain image data of material information and the loading / unloading location, constructing an image simulation of the loading / unloading location. This image simulation is used to establish the spatial relationship between materials and locations in a virtual environment, thus providing accurate basic data for subsequent path planning. Material size information is also used to determine the loading / unloading location, including the acquisition of data such as material shape and size. Based on the material loading / unloading location, the coordinates of the starting and ending points are determined, and path image data from the starting point to the ending point is acquired. Through image analysis, possible obstacles in the path are identified, and path planning is performed based on the coordinates of these obstacles. To optimize path selection, an improved heuristic search model is adopted, enabling path planning to not only consider the shortest path but also adapt to changes in dynamic obstacles, thereby improving the flexibility and efficiency of path planning. After the path planning is determined, the robotic arm begins to perform the loading / unloading task. The loading / unloading arm grasps the material and controls the top pressure claw to determine the initial downward pressure based on the material's weight and size data. Then, the robotic arm begins to transport the material along the optimal path, acquiring real-time pressure data from the bottom of the loading / unloading arm to determine whether the material has slipped or shifted. If material movement is detected, the surface smoothness coefficient and remaining path length of the material are further collected. The surface smoothness coefficient can be obtained from the initial material information. This information is combined with the material's size information to form a material feature set. The material feature set is compared with historical material features in the historical adjustment set to determine the adjustment coefficient. The initial downforce is adjusted using the adjustment coefficient, and operation continues at the adjusted downforce. During transportation, the pressure data at the bottom of the loading / unloading handle is collected a second time to determine if the material continues to slide. If the material moves, the pressure of the loading / unloading handle is further increased based on the displacement to ensure stable material transportation. The adjustment coefficient is calculated based on the difference between this adjustment and the initial pressure, and this adjustment coefficient, along with the material features, is stored in the historical adjustment set to provide optimization basis for future tasks.

[0073] Understandably, by collecting path image data and combining it with dynamic obstacle information, efficient path planning can be performed in a real-time environment, avoiding the problem of insufficient adaptability to dynamic environments in traditional methods and improving operational efficiency. The downward pressure of the top pressure claw, through a real-time monitoring and dynamic adjustment mechanism, can be precisely adjusted according to material characteristics, effectively preventing material slippage and falling during transportation, thus ensuring the safety of loading and unloading operations. By storing historical adjustment coefficients and material characteristics in a historical adjustment set, the downward pressure adjustment strategy of the top pressure claw can be continuously optimized, improving adaptability and efficiency under different material and operational conditions. During actual transportation, pressure data from the loading and unloading hands is collected in real time, and pressure is adjusted according to the movement of the material to ensure that the material remains in a stable state, reducing damage or delays caused by material slippage or unstable gripping.

[0074] In some embodiments of this application, constructing an image simulation of the loading / unloading location and determining the material loading / unloading location based on material size information includes:

[0075] Image data of the loading / unloading location is acquired using an RGB camera and a depth camera;

[0076] An object detection algorithm is used to identify existing objects at the loading / unloading location;

[0077] Obtain material dimensions based on material information, and build material models and models of loading / unloading locations;

[0078] Based on the material model and the loading / unloading location model, obtain the placement method of the material at the loading / unloading location, and simulate the placement process of the material in the model;

[0079] If there is no interference between the material model and the placement process and the placement result, the position of the material placement result in the model shall be taken as the material loading / unloading position.

[0080] Specifically, RGB and depth cameras are used to acquire image data of the loading / unloading location. The RGB camera captures color information, providing the object's appearance features; the depth camera provides distance information between the object and the camera, accurately determining the object's three-dimensional spatial position. Object detection algorithms (such as convolutional neural network (CNN) based object detection techniques) are used to identify existing objects at the loading / unloading location. The presence of these existing objects may affect material placement; therefore, identifying and accurately locating their spatial positions is crucial. After obtaining the material's size information, a virtual model of the material is built based on this data. Simultaneously, a virtual three-dimensional model of the loading / unloading location is constructed, reflecting the spatial structure, size constraints, and potential obstacle distribution of the loading / unloading area. By combining the material model and the loading / unloading location model, the material placement process is simulated. Based on the material's size, shape, and the spatial structure of the loading / unloading location, the optimal placement method is predicted. For example, if the material is irregularly shaped, different placement postures and arrangements are simulated to find the most suitable position. The placement simulation also considers physical factors such as gravity and friction to determine whether the material can be stably placed in a specific location. During the simulated placement process, it checks whether the material interferes with existing objects at the loading / unloading location. If there is no interference, this location is determined as the final loading / unloading location. If interference exists, the material placement method is adjusted, and the simulation is repeated until a suitable and interference-free placement location is found.

[0081] Understandably, the combined acquisition of data from RGB and depth cameras allows for a comprehensive understanding of the spatial structure and actual dimensions of the materials at the loading / unloading location. This enables more accurate determination of the loading / unloading position, avoiding errors caused by a lack of spatial awareness in traditional methods. Object detection algorithms and collision detection mechanisms effectively prevent interference issues during loading and unloading. By accurately simulating the material placement process, the optimal placement position can be intelligently determined, reducing the risks and errors associated with traditional manual operations. In actual operation, the placement method and position can be adjusted in real time to achieve the optimal loading and unloading solution. By simulating placement in advance and eliminating potential interference risks, the safety of the material loading and unloading process is improved, preventing material damage or loading / unloading failures due to interference.

[0082] In some embodiments of this application, when collecting path image data from the starting point to the ending point to determine obstacle coordinates and determining the optimal path for materials based on an improved heuristic search model, the process includes:

[0083] S1: Define the open set and the closed set and initialize the cost function value. The open set stores the nodes to be explored, and initially only contains the starting point. The closed set stores the nodes that have been explored, and is initially empty.

[0084] S2: Gradually select the node with the lowest cost from the open set to expand the path, and gradually move closer to the key location coordinates;

[0085] S3: Explore all possible path directions for the current node, update the path cost, and if the neighbor node is not in the open set, add the neighbor node to the open set and record the parent node of the neighbor node; if the neighbor node is in the open set, but the new path cost is smaller, update the parent node and the path cost.

[0086] S4: Repeat step S3 until the current node coincides with the coordinates of the endpoint. Then, backtrack from the endpoint coordinates to the parent node until the starting point coordinates, and use this path as the optimal path.

[0087] In some embodiments of this application, when exploring all possible path directions for the current node and updating the path cost, the process includes:

[0088] The path cost is calculated using the following formula:

[0089]

[0090] Where f(n) represents the path cost, g(n) represents the cumulative path cost, h(n) represents the heuristic cost, e(n) represents the rate of change of velocity between the current node and its parent node, and r(n) represents the rate of change of the path direction angle between the node and its parent node;

[0091]

[0092] Where (x, y, z) represents the coordinates of the current node, (xt, yt, zt) represents the coordinates of the target node, and P represents the probability that the current node is occupied by an obstacle. When the path image data determines that there is an obstacle at the current node coordinates, P is 1; when the path image data determines that there is no obstacle at the current node coordinates, P is the historical probability that the current node was occupied by an obstacle.

[0093]

[0094] in, This represents the cumulative cost of the parent node. This represents the geometric distance between the current node and its parent node.

[0095] Understandably, by considering the rate of change of speed and the rate of change of direction angle along the path, the generated path is smoother, making it suitable for practical operation by devices such as robotic arms, avoiding sharp turns or drastic speed changes. Based on path image data and obstacle probability models, the path planning can be dynamically adjusted, updating the path in real time even if the obstacle distribution changes, thus enhancing robustness. Introducing the parameters e(n) and r(n) allows the algorithm to not only focus on path length but also optimize the dynamic characteristics of the path, making it suitable for complex path planning in three-dimensional space. Through cost function optimization and node update rules, the search space of the algorithm is reduced, making the calculation of the optimal path more efficient.

[0096] In some embodiments of this application, when determining the initial downward pressure of the top pressure claw based on material weight data and material size data, the initial downward pressure is calculated using the following formula:

[0097]

[0098] Where F represents the initial downward pressure, m represents the material weight data, g represents the gravitational acceleration, A represents the contact area between the loading / unloading hand and the material, dmin represents the minimum thickness of the material, and L represents the maximum length of the material.

[0099] In some embodiments of this application, when real-time pressure data is collected from the bottom of the loading and unloading handle and the material is determined to have moved based on the pressure data, the following steps are taken: arranging all the pressure data from the bottom of the loading and unloading handle into a two-dimensional matrix, marking non-zero points and zero pressure points; when the position of a non-zero point in the two-dimensional matrix changes, it is determined that the material has moved.

[0100] In some embodiments of this application, when comparing material characteristics with historical material characteristic sets in a historical adjustment set to determine adjustment coefficients for adjusting the initial downforce, the following steps are taken: the historical adjustment set includes a historical material characteristic set and historical adjustment coefficients; the historical material characteristic set includes several sets of historical material characteristics; when there are similar data of material characteristics in the historical material characteristic set, the average historical adjustment coefficient is obtained based on the historical adjustment coefficients in the similar data, and the average historical adjustment coefficient is used as the adjustment coefficient; when there are no similar data of material characteristics in the historical material characteristic set, the adjustment coefficient is determined based on the material characteristic set; similar data are data where the difference between the historical material characteristics and the material characteristics is less than a% of the material characteristics.

[0101] In some embodiments of this application, determining the adjustment coefficient based on the set of material characteristics includes:

[0102] ;

[0103] Where K represents the adjustment coefficient, with a value range of (1, 1.6], μ represents the surface smoothness coefficient, Ls represents the remaining path length, L0 represents the total path length, V represents the material volume, L represents the material length, W represents the material width, and D represents the material thickness. Indicates the density of the material. It represents the elastic modulus of a material.

[0104] Specifically, pressure sensors are placed at the bottom of the loading / unloading handle to acquire real-time pressure data at the contact points between the handle and the material, forming a two-dimensional matrix. Non-zero points: Pressure points detected by the sensors, indicating that force is applied at that point. Zero points: No pressure detected, indicating that no force is applied at that point. By comparing the pressure data matrices of two consecutive frames, changes in the arrangement of non-zero points (addition, reduction, or movement) indicate that material movement has occurred. The specific value of the similarity threshold 'a%' can be adjusted based on the actual application and sensitivity. After determining the adjustment coefficient, the initial downward pressure is adjusted according to the adjustment coefficient; the adjusted downward pressure is the product of the initial downward pressure and the adjustment coefficient.

[0105] Specifically, for example, the material weighs 50kg, the gravitational acceleration is 9.81m / s², the contact area between the loading / unloading hands and the material is 1.5m², the minimum thickness of the material is 0.5m, the maximum length of the material is 1.2m, and the calculated initial downward pressure is 234N. An adjustment coefficient is determined based on the material's characteristics and historical adjustment data. When there is no similar material characteristic data in the historical material characteristic set, the adjustment coefficient is determined based on the material characteristic set. Specifically, the surface smoothness coefficient is 0.5, the remaining path length is 2m, the total path length is 10m, the material volume is 0.4m³, the material length is 1.2m, the material width is 0.8m, the material thickness is 0.5m, the material density is 100kg / m³, and the material elastic modulus is 10MPa. The calculated adjustment coefficient K is approximately 2.1. Since 2.1 is greater than 1.6, K is set to its maximum value of 1.6, resulting in an adjusted downward pressure of 374.4N.

[0106] Understandably, the real-time pressure data matrix change detection method can quickly and accurately determine the movement of materials, avoiding judgment errors caused by lag or inaccuracies in traditional methods. Combining historical adjustment data with real-time feature calculations dynamically determines the adjustment coefficient, making adjustments more efficient and reducing the risk of material slippage or damage due to excessively high or low pressure. Introducing multi-dimensional material properties (geometric, mechanical, and frictional properties) into the formula calculation allows for adaptation to diverse materials and meets the needs of different material characteristics. By constructing and recalling historical adjustment sets, the pressure adjustment time for common materials is shortened, improving loading and unloading efficiency.

[0107] In some embodiments of this application, the determination of whether the material has moved is based on the second pressure data. When the material moves, the adjusted downward pressure is increased based on the moving distance to determine the final downward pressure. This includes: establishing a second two-dimensional matrix based on the second pressure data; determining that the material has moved when the position of a non-zero point in the second two-dimensional matrix changes; and determining that the material has not moved when the position of a non-zero point in the second two-dimensional matrix does not change, and using the adjusted downward pressure as the final downward pressure.

[0108] Specifically, when it is determined that the material has moved, the direction of movement and the distance of movement are obtained according to the second two-dimensional matrix. The adjustment coefficient is determined according to the distance of movement, and the adjusted downward pressure is adjusted a second time to determine the final downward pressure. The adjustment coefficient is proportional to the distance of movement, and the value range of the adjustment coefficient is (1, 1.2).

[0109] Specifically, when material moves, the movement is determined by changes in the values ​​in the matrix. For example, the initial data in the middle position (second row) of the pressure matrix is ​​[0, 15, 30, 25, 0], and the data in the second row changes to [0, 0, 15, 30, 25]. The interval distance of the pressure sensors represents the movement distance. The movement distance is compared with the preset first and second movement distances. Based on the comparison results, an adjustment coefficient is determined to make a secondary adjustment to the adjusted downward pressure. For example, if the material moves 0.15m, the first distance is 0.1m, and the second distance is 0.2m, which is between 0.1m and 0.2m, the adjustment coefficient is 1.1 when the movement distance is less than or equal to 0.1m, 1.15 when the movement distance is greater than 0.1m and less than or equal to 0.2m, and 1.2 when the movement distance is greater than 0.2m. Therefore, the adjustment coefficient of 1.15 is used to make a secondary adjustment to the adjusted downward pressure, and the final downward pressure is approximately 430N.

[0110] Understandably, by collecting secondary pressure data in real time and establishing a two-dimensional matrix analysis, it is possible to sensitively detect whether the material has moved during transportation. The addition of a verification mechanism for the first adjustment avoids errors relying on traditional static sensors or empirical rules, providing real-time feedback on material state changes and ensuring gripping stability. When material movement occurs, the adjustment coefficient is dynamically calculated, and the downward pressure is adjusted based on the actual distance the material has moved. The adjustment coefficient is proportional to the moving distance, avoiding over- or under-pressure, thus ensuring the stability and safety of the material during handling. If the adjusted downward pressure is determined based on the moving distance, the final downward pressure is the product of the adjusted downward pressure and the adjustment coefficient. Through this secondary pressure adjustment strategy, this solution can be dynamically optimized based on the real-time state of the material, ensuring the gripping stability of the material throughout the handling process, reducing the risk of material slippage, misalignment, or damage, thereby improving operational efficiency and safety.

[0111] The above embodiments address the limitations of traditional gantry robotic arms in material loading and unloading processes by integrating material information acquisition, path planning, gripping force control, and dynamic adjustment mechanisms. Through comprehensive analysis of image data acquisition and material size information, the loading and unloading positions are determined, and path selection is optimized based on an improved heuristic path search model to avoid interference from obstacles in the dynamic environment. By monitoring the pressure data at the bottom of the loading / unloading arm in real time, the downward pressure of the arm is dynamically adjusted to adapt to changes in factors such as material surface smoothness and remaining path length, ensuring stable gripping and preventing material slippage or damage. Through the accumulation and comparison of historical data, the pressure control strategy is intelligently optimized, enabling the robotic arm to flexibly adjust under different materials and environmental conditions, improving the adaptability, accuracy, and safety of gantry operations, and enhancing the efficiency and reliability of automated loading and unloading operations.

[0112] In another preferred embodiment based on the above embodiments, see [reference] Figure 2-5 As shown, this embodiment provides a gantry robotic arm for loading and unloading vehicles, used to apply the above-described gantry robotic arm control method for loading and unloading vehicles, including:

[0113] Support frame, top pressing mechanism, clamping mechanism, and data collection device;

[0114] The support frame includes a transverse fixing plate 411 and a longitudinal fixing plate 412, and two transverse fixing plates 411 and two longitudinal fixing plates 412 are provided.

[0115] The acquisition device includes several image acquisition units 510, which are fixed on a support frame;

[0116] The top pressing mechanism includes a pressing cylinder 421 and a top pressure claw 422. At least two pressing cylinders 421 are provided, and two top pressure claws 422 are provided. The pressing cylinders 421 are fixedly mounted on the longitudinal fixing plate 412, and the top pressure claws 422 are fixedly connected to the bottom of the pressing cylinders 421.

[0117] The clamping mechanism includes a transverse cylinder 431, a connecting rod 432, and a loading / unloading handle 433. The transverse cylinder 431 is fixedly connected to the transverse fixed plate 411. One end of the transverse cylinder 431 is connected to the connecting rod 432, and the connecting rod 432 is rotatably connected to the transverse cylinder 431. The other end of the connecting rod 432 is connected to the loading / unloading handle 433, and the loading / unloading handle 433 is rotatably connected to the connecting rod 432. Several pressure sensors 610 are evenly arranged on the loading / unloading handle 433.

[0118] Understandably, the truss robotic arm 400 is applied to a truss, which includes an X-axis steel beam 100, a Y-axis steel beam 200, and a Z-axis steel beam 300. The X-axis steel beam 100 is mounted on two parallel Y-axis steel beams 200, and the Z-axis steel beam 300 is movably connected to one end of the X-axis steel beam 100. The X-axis steel beam 100 can move parallel along the Y-axis, and the Z-axis steel beam 300 can move horizontally along the X-axis steel beam 100. The truss robotic arm 400 is installed at the bottom of the Z-axis steel beam 300 and is used for loading and unloading materials 700. The truss robotic arm 400 in this application is mainly used for loading and unloading top-opening vans.

[0119] Understandably, multiple image acquisition units 510 in the acquisition device are fixed on the support frame to acquire image data of the material to be loaded / unloaded, used to determine the position and size information of the material 700. This determines the material 700 to be gripped and its starting and ending points. The gripping mechanism, through the action of the transverse cylinder 431 and connecting rod 432, drives the loading / unloading hand 433 to grip the material 700. After the gripping action is completed, the top pressing mechanism operates, driving the pressing cylinder 421 to control the top pressure claw 422 to move downwards, squeezing the upper surface of the material 700, making the lower surface of the material 700 tightly fit with the loading / unloading hand 433. The loading / unloading hand 433 judges the stability of the material 700 and the gripping force through feedback from the pressure sensor 610. This ensures that the material 700 is stably pressed and prevents slippage or displacement. The pressure sensor 610 detects pressure changes at the bottom of the loading / unloading hand 433; if the material 700 moves, the pressing cylinder 421 in the top pressing mechanism is adjusted to adjust the downward pressure.

[0120] It is understood that in another embodiment of this application, the loading and unloading hand 433 adopts a suction cup structure. The loading and unloading hand 433 is attached to the upper surface of the material 700 by the top suction cup, and the material 700 is supported by the tray set at the bottom of the loading and unloading hand 433. The suction cup structure further prevents the material 700 from shifting its position during transportation.

[0121] Understandably, by integrating material information acquisition, path planning, gripping force control, and dynamic adjustment mechanisms, the limitations of traditional gantry robotic arms in material loading and unloading processes are overcome. Through comprehensive analysis of image data acquisition and material size information, the loading and unloading positions of the materials are determined, and based on an improved heuristic path search model, path selection is optimized to avoid interference from obstacles in the dynamic environment. By monitoring the pressure data at the bottom of the loading / unloading arm in real time, the downward pressure of the arm is dynamically adjusted to adapt to changes in factors such as material surface smoothness and remaining path length, ensuring stable gripping and preventing material slippage or damage. Through the accumulation and comparison of historical data, the pressure control strategy is intelligently optimized, enabling the robotic arm to flexibly adjust under different materials and environmental conditions, improving the adaptability, operational accuracy, and safety of gantry operations, and enhancing the efficiency and reliability of automated loading and unloading operations.

[0122] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program goods. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program goods embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0123] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program goods according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0124] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0125] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A control method for a gantry robotic arm used for loading and unloading vehicles, characterized in that, include: Collect material information and image data of the loading / unloading location, construct an image simulation of the loading / unloading location, and determine the material loading / unloading location based on the material size information; The starting and ending coordinates are determined based on the material loading / unloading location, the path image data from the starting point to the ending point is collected to determine the obstacle coordinates, and the optimal path of the material is determined based on the improved heuristic search model. Control the loading and unloading hand to grip the material, control the top pressure claw to move downward to squeeze the material, and determine the initial downward pressure of the top pressure claw; The material is transported using the initial downward pressure and the optimal path, and the pressure data at the bottom of the loading and unloading handle is collected in real time to determine whether the material has moved. When movement is detected, the initial downforce is adjusted to operate with the adjusted downforce. The adjustment coefficient is determined based on the final downforce and the initial downforce, and the adjustment coefficient and the set of material characteristics are stored in the historical adjustment set. Arrange all pressure data at the bottom of the loading and unloading handle into a two-dimensional matrix, marking non-zero points and zero pressure points; when the position of a non-zero point in the two-dimensional matrix changes, it is determined that the material has moved; The material characteristics are compared with the historical material characteristic set in the historical adjustment set to determine the adjustment coefficient and adjust the initial downforce; the historical adjustment set includes the historical material characteristic set and the historical adjustment coefficient, and the historical material characteristic set includes several sets of historical material characteristics; When there is similar data for the material characteristics in the historical material characteristic set, the average historical adjustment coefficient is obtained based on the historical adjustment coefficient in the similar data, and the average historical adjustment coefficient is used as the adjustment coefficient. When there is no similar data for the material feature in the historical material feature set, the adjustment coefficient is determined based on the material feature set; The similar data refers to data where the difference between the historical material characteristics and the material characteristics is less than a% of the material characteristics; Wherein, a% is a similarity threshold, used to determine whether the historical material characteristics are similar to the material characteristics; the similarity threshold is in percentage form, and its value ranges from 1% to 20%.

2. The method for controlling the gantry robotic arm of a loading and unloading vehicle according to claim 1, characterized in that, Constructing an image simulation of the loading / unloading location, and determining the material loading / unloading location based on material size information, includes: Image data of the loading / unloading location is acquired using an RGB camera and a depth camera; An object detection algorithm is used to identify existing objects at the loading / unloading location; Based on the material information, obtain the material dimensions and establish a material model and a model of the loading / unloading location; Based on the material model and the loading / unloading position model, the placement method of the material at the loading / unloading position is obtained, and the placement process of the material is simulated in the model; If there is no interference between the material model and the placement process and the placement result, the position of the material placement result in the model shall be taken as the material loading / unloading position.

3. The method for controlling the gantry robotic arm of a loading and unloading vehicle according to claim 2, characterized in that, When collecting path image data from the starting point to the ending point to determine obstacle coordinates, and determining the optimal path for materials based on an improved heuristic search model, the process includes: S1: Define an open set and a closed set and initialize the cost function value. The open set stores the nodes to be explored, and initially only contains the starting point. The closed set stores the nodes that have been explored, and is initially empty. S2: Gradually select the node with the lowest cost from the open set to extend the path, and gradually move closer to the key location coordinates; S3: Explore all possible path directions for the current node, update the path cost, and if the neighbor node is not in the open set, add the neighbor node to the open set and record the parent node of the neighbor node; if the neighbor node is in the open set, but the new path cost is smaller, update the parent node and the path cost. S4: Repeat step S3 until the current node coincides with the coordinates of the endpoint. Then, trace back from the endpoint coordinates to the parent node until the starting point coordinates, and use this path as the optimal path.

4. The method for controlling the gantry robotic arm of a loading and unloading vehicle according to claim 3, characterized in that, When exploring all possible path directions for the current node and updating path costs, the following should be included: The path cost is calculated using the following formula: Where f(n) represents the path cost, g(n) represents the cumulative path cost, h(n) represents the heuristic cost, e(n) represents the rate of change of velocity between the current node and its parent node, and r(n) represents the rate of change of the path direction angle between the node and its parent node; Where (x, y, z) represents the coordinates of the current node, (xt, yt, zt) represents the coordinates of the target node, and P represents the probability that the current node is occupied by an obstacle. When the path image data determines that there is an obstacle at the current node coordinates, P is 1; when the path image data determines that there is no obstacle at the current node coordinates, P is the historical probability that the current node was occupied by an obstacle. in, This represents the cumulative cost of the parent node. This represents the geometric distance between the current node and its parent node.

5. The method for controlling the gantry robotic arm of a loading and unloading vehicle according to claim 1, characterized in that, Determining the initial downward pressure of the top pressure claw includes: The initial downward pressure of the top pressure claw is determined based on the material weight and size data, and the initial downward pressure is calculated using the following formula: Where F represents the initial downward pressure, m represents the material weight data, g represents the gravitational acceleration, A represents the area of ​​contact between the bottom of the loading / unloading hand and the material, dmin represents the minimum thickness of the material, and L represents the maximum length of the material.

6. The method for controlling the gantry robotic arm of a loading and unloading vehicle according to claim 1, characterized in that, When determining the adjustment coefficient based on the set of material characteristics, the following are included: Where K represents the adjustment coefficient, with a value range of (1, 1.6], μ represents the surface smoothness coefficient, Ls represents the remaining path length, L0 represents the total path length, V represents the material volume, L represents the material length, W represents the material width, and D represents the material thickness. Indicates the density of the material. denoted by , where e represents the elastic modulus of the material, and e represents the base of the natural logarithm function.

7. The method for controlling the gantry robotic arm of a loading and unloading vehicle according to claim 6, characterized in that, After operating with the adjusted downforce, it also includes: Collect a second pressure data point to determine if the material has moved. If the material has moved, adjust the downward pressure based on the distance moved. When determining the final downward pressure, the following steps are taken: A second two-dimensional matrix is ​​established based on the second pressure data. When the position of a non-zero point in the second two-dimensional matrix changes, it is determined that the material has moved; when the position of a non-zero point in the second two-dimensional matrix does not change, it is determined that the material has not moved, and the adjusted downward pressure is taken as the final downward pressure. When it is determined that the material has moved, the direction of movement and the distance of movement are obtained according to the second two-dimensional matrix. The adjustment coefficient is determined according to the distance of movement, and the adjusted downward pressure is adjusted a second time to determine the final downward pressure. The adjustment coefficient is proportional to the distance of movement, and the value range of the adjustment coefficient is (1, 1.2).

8. A gantry robot arm for loading and unloading vehicles, used in applying the gantry robot arm control method for loading and unloading vehicles as described in any one of claims 1-7, characterized in that, include: Support frame, top pressing mechanism, clamping mechanism, and data collection device; The support frame includes a transverse fixing plate and a longitudinal fixing plate, and two transverse fixing plates and two longitudinal fixing plates are provided. The acquisition device includes several image acquisition units, which are fixed on the support frame; The top pressing mechanism includes a pressing cylinder and a top pressure claw. At least two pressing cylinders and two top pressure claws are provided. The pressing cylinders are fixedly mounted on the longitudinal fixing plate, and the top pressure claws are fixedly connected to the bottom of the pressing cylinders. The clamping mechanism includes a transverse cylinder, a connecting rod, and a loading / unloading hand. The transverse cylinder is fixedly connected to the transverse fixed plate. One end of the transverse cylinder is connected to the connecting rod, and the connecting rod is rotatably connected to the transverse cylinder. The other end of the connecting rod is connected to the loading / unloading hand, and the loading / unloading hand is rotatably connected to the connecting rod. Several pressure sensors are evenly arranged on the loading / unloading hand.

Images