A control method and system for an intelligent robotic arm used in the production of washing machine casings.

Abstract

This invention relates to the field of production control technology, specifically to a control method and system for an intelligent robotic arm used in the production of washing machine shells. The method includes: acquiring the robotic arm's axis parameters and visual images; constructing a multi-layered chain structure to determine the axis parameter chain; constructing a three-dimensional model of the robotic arm's motion; determining the clamping distance; sequentially adjusting and iterating the angle parameters of different layers of the axis parameter chain; determining the optimal iterative value for each axis after adjustment based on the changes in clamping distance under different iterations at the same layer; selecting candidate parameters from the angle parameters of different iterations based on the optimal iterative values; determining the control coefficients of each control chain; and selecting the optimal control chain based on the control coefficients to control the robotic arm's axes. This invention can obtain the optimal control path for the robotic arm, reduce vibration risks and energy consumption, and ensure accurate and stable clamping and positioning during the production of washing machine shells.

Description

Technical Field

[0001] This invention relates to the field of production control technology, specifically to an intelligent robotic arm control method and system for the production of washing machine shells. Background Technology

[0002] With the continuous improvement of industrial automation, various production equipment and systems are gradually developing towards intelligence. In the home appliance industry, especially in the production of washing machines, the manufacturing and assembly of the outer casing has always been a key link in improving production efficiency and quality control. In the production of washing machine casings, intelligent robotic arms are needed to position and clamp the stamped washing machine casings to facilitate subsequent production operations.

[0003] Regarding the control methods of intelligent robotic arms, traditional arm control systems mostly adopt a control method based on fixed programs. This method is difficult to respond flexibly to dynamically changing production environments and cannot effectively adapt to different shapes and sizes of washing machine shells to adjust to actual working conditions, resulting in low production efficiency and poor clamping accuracy and stability. Summary of the Invention

[0004] To address the technical problems of low production efficiency, poor clamping accuracy, and instability in related technologies, this invention provides an intelligent robotic arm control method and system for washing machine shell production. The specific technical solution adopted is as follows:

[0005] This invention proposes a control method for an intelligent robotic arm used in the production of washing machine casings. The method employs a multi-axis intelligent robotic arm to clamp the washing machine casing, with at least six axes. The method includes:

[0006] The robot arm's axis parameters and the visual image of the washing machine casing during the painting process are obtained, wherein the axis parameters are the angle parameters of different axes during operation;

[0007] A multi-layered chain structure is constructed, the number of layers of the chain structure being equal to the number of axes of the robot arm; the layers of the chain structure are sorted in order from farthest to near the end of the robot arm, and each node in the chain structure is an angle parameter of the axis, thereby obtaining an axis parameter chain;

[0008] A three-dimensional model of the robotic arm's motion is constructed. In the three-dimensional model, the clamping distance between the robotic arm's end effector and the washing machine casing to be clamped is determined based on visual images. The angle parameters of different layers of the axis parameter chain are adjusted and iterated sequentially. Based on the change in clamping distance under different iterations of the same layer, the optimal value of each axis after adjustment is determined. Candidate parameters are selected from the angle parameters of different iterations based on the optimal value of the iteration.

[0009] Based on the candidate parameters of all layer angle parameters, the control chain for axis control is determined. The control coefficient of each control chain is determined according to the clamping distance, the number of iterations of different layers, and the changes in angle parameters. The optimal control chain is selected based on the control coefficients to control the robot's axis.

[0010] Furthermore, the construction of the three-dimensional model of the robotic arm's motion includes:

[0011] A kinematic model of the robotic arm is established based on the DH parameter method and placed in a three-dimensional coordinate system that includes the spatial position of the washing machine casing to obtain a three-dimensional model.

[0012] Furthermore, the visual image is acquired by a binocular vision camera configured at the end effector of the robotic arm. Determining the clamping distance between the end effector of the robotic arm and the position to be clamped on the washing machine casing based on the visual image includes:

[0013] The YOLO3 algorithm is used to identify the washing machine shell in the visual image. Based on the principle of binocular vision, depth information is calculated through parallax to obtain the distance between the washing machine shell and the camera, which is used as the clamping distance between the end effector of the robot and the washing machine shell.

[0014] Furthermore, based on the changes in clamping distance under different iterations within the same layer, the adjusted iterative optimal value for each axis is determined, including:

[0015] Determine the allowable range of values ​​for the axial angle parameter of any layer;

[0016] Within the allowable range of values, the periodic analysis angle is determined, and the step size fluctuation is analyzed according to the beetle whisker algorithm to determine the fluctuation angle of each analysis angle.

[0017] All analysis angles and all fluctuation angles are combined in ascending order to form the iteration angles, simulating the clamping distance under the corresponding iteration angles. The iteration of the shaft angle parameter is in a fixed direction.

[0018] By combining the changes in clamping distance under different iterations, a control function is constructed, and the optimal value of each iteration is calculated.

[0019] Furthermore, the control function is:

[0020] In the formula, Indicates after the first After the nth iteration, traverse the sequence of axis parameters. Iterative optimization of parameter values; Indicates after the first After the nth iteration, traverse the sequence of axis parameters. The gripping distance of the robotic arm at each parameter value; Indicates after the first After the nth iteration, traverse the sequence of axis parameters. The gripping distance of the robotic arm at each parameter value; Indicates after the first After the nth iteration, traverse the sequence of axis parameters. For each parameter value, the rate of decrease in the gripping distance of the robotic arm is... Indicates after the first After the nth iteration, traverse the sequence of axis parameters. For each parameter value, the rate of decrease in the gripping distance of the robotic arm is... Let represent the sigmoid function, e represent the natural constant, and norm represent the linear normalization function.

[0021] Further, candidate parameters are selected from the angle parameters of different iterations based on the iterative preferred value, including:

[0022] The preset number of angle parameters with the largest iterative optimal value are selected as candidate parameters.

[0023] Furthermore, based on candidate parameters of all layer angle parameters, a control chain for achieving axis control is determined, including:

[0024] Based on the candidate parameters of all layer angle parameters, an adjustment tree structure is constructed;

[0025] Traverse the regulation tree structure to determine the path from the root node to the leaf node as the control chain.

[0026] Furthermore, based on the clamping distance corresponding to different control chains, the number of iterations at different layers, and the changes in angle parameters, the control coefficients of each control chain are determined, including:

[0027] The change in angle parameters of different layers from their initial values ​​before iteration to their values ​​under the corresponding control chain layers is used as an indicator of angle change.

[0028] Calculate the product of the angle change index of the same layer and the number of iterations of the corresponding layer, and use it as the penalty coefficient of the corresponding layer. Then, sum the penalty coefficients of all layers as the total penalty index.

[0029] Calculate the product of the clamping distance corresponding to the control chain and the total penalty index, and normalize the negative of the product to obtain the control coefficient.

[0030] Furthermore, the optimal control chain is selected based on the control coefficients to control the robot arm axis, including:

[0031] The control chain with the largest control coefficient is taken as the optimal control chain, and the robot arm axis is controlled based on the optimal control chain.

[0032] On the other hand, it also includes an intelligent robotic control system for the production of washing machine casings, the system including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of the method as described in any of the foregoing.

[0033] The present invention has the following beneficial effects:

[0034] This invention employs a multi-axis intelligent robotic arm to grip the outer casing of a washing machine. The gripping distance is analyzed using visual images, and the angle parameters of different axes are iteratively adjusted to determine the gripping distance between the robotic arm's end effector and the desired gripping position on the washing machine casing. The angle parameters of different layers of the axis parameter chain are sequentially adjusted and iterated. Based on the changes in gripping distance across different iterations within the same layer, an optimal iterative value is determined for each axis. Candidate parameters are then selected from the angle parameters across different iterations based on these optimal values. Finally, the axis control is determined based on the candidate parameters from all layers of angle parameters. The control chain is designed to determine the control coefficients of each control chain based on the clamping distance, iteration count of different layers, and changes in angle parameters. The optimal control chain is then selected based on the control coefficients to control the robot's axes. This allows for gradual convergence through chain-like parameter updates, preventing target deviation or loss of field of view. The hierarchical iteration based on axis parameter chains effectively conforms to the multiple solution characteristics of the robot's inverse kinematics. The robustness of parameter optimization is improved through the control function, and the optimal control path of the robot is obtained through the control coefficients, reducing vibration risks and energy consumption, and ensuring the accuracy and stability of clamping and positioning during the production of the washing machine shell. Attached Figure Description

[0035] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 A flowchart of an intelligent robotic arm control method for producing washing machine casings, provided in one embodiment of the present invention;

[0037] Figure 2 This is a schematic diagram of an intelligent robotic arm provided in one embodiment of the present invention;

[0038] Figure 3This is a schematic diagram of a shaft parameter chain provided in one embodiment of the present invention;

[0039] Figure 4 This is a schematic diagram of an adjustment tree structure provided in one embodiment of the present invention. Detailed Implementation

[0040] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of an intelligent robotic arm control method and system for washing machine casing production according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0041] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0042] The following description, in conjunction with the accompanying drawings, details a specific scheme for an intelligent robotic arm control method for producing washing machine casings provided by the present invention.

[0043] Please see Figure 1 The diagram illustrates a flowchart of an intelligent robotic arm control method for washing machine casing production according to an embodiment of the present invention. The method includes:

[0044] S101: Obtain the axis parameters of the robotic arm and the visual image of the washing machine shell during the painting process. The axis parameters are the angle parameters of different axes.

[0045] In the production of washing machine shells, intelligent robotic arms are needed to position and clamp the stamped washing machine shells to facilitate the next production operation. In terms of the control method of intelligent robotic arms, traditional arm control systems mostly adopt a control method based on fixed programs. This method is difficult to respond flexibly to the dynamically changing production environment and cannot effectively adapt to washing machine shells of different shapes and sizes to adjust to the actual working conditions, resulting in low production efficiency and automation level.

[0046] Currently, intelligent robotic arms typically consist of multiple robotic arms and axes. The movement of an intelligent robotic arm is essentially a change in the posture of the robotic arms caused by the rotational transformation of these axes. Therefore, the process of an intelligent robotic arm positioning itself on the washing machine casing involves adjusting the axis parameters of the robotic arms. This adjustment process requires clearly defining the current posture of the intelligent robotic arm and the position of the washing machine casing. (See also...) Figure 2 , Figure 2This is a schematic diagram of an intelligent robotic arm provided in one embodiment of the present invention; in this embodiment of the present invention, a multi-axis intelligent robotic arm is configured to clamp the outer shell of a washing machine, with at least 6 axes.

[0047] For a robotic arm, if it has fewer than 6 axes, it cannot achieve all poses in three-dimensional space, especially at the boundaries of the workspace, where the poses it can achieve are generally limited. Since the placement and size of the washing machine shell are not fixed in actual production, the workspace of the robotic arm needs to accommodate the changing placement and size of the washing machine shell as much as possible. Therefore, in the production of washing machine shells, in order to meet the production requirements, a 6-axis intelligent robotic arm is usually used to clamp the washing machine shell.

[0048] Select a binocular vision camera suitable for industrial environments and mount it at the front end of the robot arm. The position should ensure that it can capture a complete image of the washing machine casing. The binocular vision camera will collect video signals as visual images.

[0049] High-precision angle sensors are installed on each axis of each robotic arm to collect the angle parameters corresponding to each axis of the intelligent robotic arm in real time, which are used as axis parameters to reflect the posture of the intelligent robotic arm in space.

[0050] S102: Construct a multi-level chain structure, the number of levels of the chain structure being equal to the number of axes of the robot arm; sort the levels of the chain structure in order from farthest to near the end of the robot arm, and each node in the chain structure is the angle parameter of the axis, thus obtaining the axis parameter chain.

[0051] See Figure 3 , Figure 3 This is a schematic diagram of a shaft parameter chain provided in an embodiment of the present invention. The shafts are arranged from the base to the end, and each shaft corresponds to a layer of the shaft parameter chain. The element of each layer of the shaft parameter chain is the angle parameter corresponding to that layer of shaft.

[0052] S103: Construct a 3D model of the robotic arm's motion; in the 3D model, determine the clamping distance between the robotic arm's end effector and the washing machine casing to be clamped based on the visual image; sequentially adjust and iterate the angle parameters of different layers of the axis parameter chain, and determine the optimal value of each axis after adjustment based on the change in clamping distance under different iterations of the same layer; select candidate parameters from the angle parameters of different iterations based on the optimal value of the iteration.

[0053] Furthermore, in some embodiments of the present invention, constructing a three-dimensional model of the robotic arm motion includes: establishing a kinematic model of the robotic arm based on the DH parameter method, and placing it in a three-dimensional coordinate system containing the spatial position of the washing machine casing to obtain a three-dimensional model.

[0054] In this embodiment of the invention, the dimensions of each component of the robotic arm are obtained, and professional modeling software (such as ROS, MATLAB, Simulink, Gazebo, etc.) is used to establish a three-dimensional space and a digital structural model of the intelligent robotic arm. The digital structural model of the intelligent robotic arm is used as a virtual robotic arm. In addition, the visual images collected by the binocular vision camera and the axis parameters collected by the sensor are input into the virtual robotic arm. The visual images are used to calculate the position of the washing machine shell in three-dimensional space through computer vision algorithms, and the axis parameters provide the posture data of the robotic arm to obtain a three-dimensional model.

[0055] First, the YOLO3 algorithm is used to identify the washing machine shell in the visual image. Based on the principle of binocular vision, depth information is calculated through parallax to obtain the distance between the washing machine shell and the camera, which is used as the clamping distance between the end effector of the robot and the washing machine shell.

[0056] In this embodiment of the invention, the initial angles of each axis of the three-dimensional model are set to a safe position (usually 0°). The angle parameters of different layers of the axis parameter chain are adjusted and iterated sequentially. It should be noted that the chain iteration process is as follows: starting from the first layer of the axis parameter chain, the initial axis parameters are iterated. After each iteration, the iterated axis parameters are input into the corresponding axis of the virtual manipulator, causing the axis to rotate. The coordinate position of the end of the virtual manipulator in the three-dimensional model space after the axis rotation is obtained. Based on the coordinate position of the end of the virtual manipulator in the three-dimensional model space and the coordinate position of the washing machine shell, the clamping distance between the end of the virtual manipulator and the washing machine shell after adjustment is obtained.

[0057] Furthermore, in some embodiments of the present invention, the adjusted iterative optimal value of each axis is determined based on the change in clamping distance under different iterations of the same layer, including: determining the allowable range of angle parameters of the corresponding axis in any layer; within the allowable range, determining the periodic analysis angle, and analyzing the step size fluctuation according to the beetle whisker algorithm to determine the fluctuation angle of each analysis angle; using all analysis angles and all fluctuation angles in ascending order as the iterative angles to simulate the clamping distance under the corresponding iterative angle, wherein the iteration of the axis angle parameter is in a fixed direction; and constructing a control function based on the change in clamping distance under different iterations to calculate the adjusted iterative optimal value for each iteration.

[0058] It is understood that the adjustable angle of each axis is within a certain range. In this embodiment of the invention, the axis angle parameter is iterated in a fixed direction. The boundary angle within the adjustable range is taken as 0 degrees. Then, the allowable value range is determined by increasing the angle, such as [0°, 180°], that is, from [0°, 180°] is a fixed clockwise direction.

[0059] Uniform discrete sampling is used to sample within the allowable range of the shaft parameters, thereby generating a set of analytical angles within the allowable range. For example, the sampling step size of uniform discrete sampling is preset to... That is, data is collected every 10 degrees, thus collecting 19 analysis angles within [0°, 180°]. Since the overall robotic arm is a multi-axis linkage, fluctuation analysis is also required at different analysis angles.

[0060] Preset fluctuation step size As the step size parameter of the beetle whisker algorithm, it is used in conjunction with the beetle whisker algorithm to generate fluctuation parameter values. and And extended to multiple points, where , , for The left fluctuation value, for The right fluctuation value, it should be noted, is a parameter Increase from small to large ( (The subsequent control function is obtained through the elbow method) With parameters The inflection point during the change is used as the axis parameter corresponding to the inflection point, which is then adjusted to the corresponding layer number in the axis parameter chain.

[0061] In this embodiment of the invention, after determining the fluctuation angle by determining parameter a, all analysis angles and all fluctuation angles are used together as iteration angles in ascending order. By combining the analysis angles and all fluctuation angles, the search space for shaft parameters is expanded, and the fluctuation step size is used to control the fineness of the iteration of shaft parameters, thereby improving the dynamic adaptability of shaft parameter selection for the robot arm.

[0062] After determining the iteration angle, the clamping distance at the corresponding iteration angle is simulated in the 3D model. This facilitates the construction of a control function by combining the changes in clamping distance under different iterations, and the calculation of the optimal iteration value after each iteration adjustment. Furthermore, in some embodiments of the present invention, the control function is:

[0063]

[0064] In the formula, Indicates after the first After the nth iteration, traverse the sequence of axis parameters. Iterative optimization of parameter values; Indicates after the first After the nth iteration, traverse the sequence of axis parameters. The gripping distance of the robotic arm at each parameter value; Indicates after the first After the nth iteration, traverse the sequence of axis parameters. The gripping distance of the robotic arm at each parameter value; Indicates after the first After the nth iteration, traverse the sequence of axis parameters. For each parameter value, the rate of decrease in the gripping distance of the robotic arm is... Indicates after the first After the nth iteration, traverse the sequence of axis parameters. For each parameter value, the rate of decrease in the gripping distance of the robotic arm is... Let represent the sigmoid function, e represent the natural constant, and norm represent the linear normalization function.

[0065] Iterative optimization value The calculation is based on the influence of the axis parameter value iteration. When traversing the axis parameter sequence to iterate the parameters of the corresponding layer of the axis in the axis parameter chain, the clamping distance between the end of the virtual robot and the target position of the washing machine shell after iteration in the 3D model is considered. The larger the value, the better the proximity effect of the axis parameters obtained under the corresponding iteration result.

[0066] In the control function, the clamping distance This is the core indicator; the smaller the value, the closer the endpoint is to the target. However, since parameter changes during iteration may cause distance fluctuations, this method chooses to combine the current descent rate. To dynamically adjust; and the rate of decrease in clamping distance constitutes This demonstrates the degree of optimization in achieving a closer approximation of the clamping distance before and after iterative changes in the shaft parameters. If the clamping distance of the robotic arm is closer to the washing machine casing after iteration, it indicates a significant optimization of the clamping distance approach. This indicates that optimizing the clamping distance has a reverse effect, therefore Its purpose is to avoid relying solely on absolute distance while ignoring the dynamic iteration process of axis parameters.

[0067] in addition, As the cumulative decrease rate, it reflects the overall improvement trend of the parameter values. That is, the larger the sum of positive values, the better the historical performance of optimizing the clamping distance, and thus it can be used for subsequent fluctuation step size. Improve the overall approach effect of clamping distance when changes occur.

[0068] Candidate parameters are selected from the angle parameters of different iterations based on the iterative optimization value, including selecting a preset number of angle parameters with the largest iterative optimization value as candidate parameters.

[0069] The preset quantity refers to the number of candidate parameters during the analysis. Optionally, the preset quantity can be, for example, 5, meaning that in each layer, the 5 angle parameters with the largest iterative optimization value are selected as candidate parameters.

[0070] Fix any candidate parameter of the i-th layer in the axis parameter chain and input it into the robot model of the 3D model. Then, combine the (i+1)-th layer with the control function to obtain several candidate parameters of the corresponding layer. When i>6, stop iterating on the number of layers in the axis parameter chain.

[0071] S104: Determine the control chain for axis control based on the candidate parameters of all layer angle parameters. Determine the control coefficient of each control chain according to the clamping distance, the number of iterations of different layers and the changes of angle parameters. Select the optimal control chain based on the control coefficient to control the robot axis.

[0072] As can be seen from the above steps, each layer corresponds to different candidate parameters of the next layer under different candidate parameters, which is essentially a tree-like structural model. Therefore, in this embodiment of the invention, the control chain is used to organize and analyze the situation from the root node to the leaf node of different trees.

[0073] Furthermore, in some embodiments of the present invention, determining the control chain for shaft control based on candidate parameters of all layer angle parameters includes: constructing an adjustment tree structure based on candidate parameters of all layer angle parameters; and traversing the adjustment tree structure to determine the path from the root node to the leaf node in the adjustment tree structure as the control chain.

[0074] In each iteration of the axis parameter chain, the control function outputs multiple axis parameters corresponding to high T(Q,R) values. After the axis parameters of each layer are fixed, they serve as branch points to enter the next layer, thus forming a control tree structure. See [link to relevant documentation]. Figure 4 , Figure 4 This is a schematic diagram of an adjustment tree structure provided in one embodiment of the present invention.

[0075] The adjustment tree structure is a tree-like data structure used to represent multiple solutions (i.e., multiple feasible control paths) generated during the iteration of axis parameters. Each node corresponds to a candidate parameter of an axis, and each path (from the root node to the leaf node) represents a complete combination of axis parameters. The process of generating the adjustment tree is to allow multiple candidate parameter values ​​during iteration to expand the branches, thereby capturing the multiple solution characteristic of inverse kinematics.

[0076] The algorithm employs a depth-first search (DFS) approach to traverse the adjustment tree structure. Starting from the root node, it recursively visits each child node. When a leaf node is reached, the complete path (i.e., node sequence) from the root node to that leaf node is recorded, and this path is defined as a control chain. This process is repeated until all paths have been traversed, resulting in several control chains. The root node corresponds to the first layer of the axis parameter chain, representing the axis furthest from the robot's end effector. The leaf nodes correspond to the last layer of the axis parameter chain, representing the axis closest to the robot's end effector. Each node corresponds to a candidate parameter for an axis. Each path from the root node to a leaf node represents a complete control chain, i.e., a feasible sequence of axis parameter combinations.

[0077] After iterating the axis parameters of the robot, an adjustment tree structure with multiple paths is obtained, which is essentially a multiple solution problem for the robot. However, in the production of washing machine shells, in order to improve production efficiency, not every solution in the multiple solutions has the optimal axis parameters. Therefore, this method analyzes the adjustment path in the adjustment tree structure and quantifies the control coefficients to select the effective control path for the robot.

[0078] Then, for each control chain, the complete sequence of axis parameter combinations is applied to the virtual manipulator model to calculate the spatial distance between the virtual manipulator end and the target position on the washing machine casing, i.e., the clamping distance, which is used as the clamping distance corresponding to the control chain.

[0079] Furthermore, in some embodiments of the present invention, the control coefficient of each control chain is determined based on the clamping distance corresponding to different control chains, the number of iterations of different layers, and the change of angle parameters. This includes: taking the change of angle parameters of different layers from the initial value before iteration to the value under the corresponding control chain layer as the angle change index; calculating the product of the angle change index of the same layer and the number of iterations of the corresponding layer as the penalty coefficient of the corresponding layer, and taking the sum of the penalty coefficients of all layers as the total penalty index; calculating the product of the clamping distance corresponding to the control chain and the total penalty index, and normalizing the negative of the product to obtain the control coefficient.

[0080] The control coefficient describes the control effect of the corresponding control chain as input to the robotic arm during the production of washing machine casings. A higher control coefficient indicates better control; that is, a control chain with a larger control coefficient value can achieve precise positioning with higher efficiency and lower energy consumption, better meeting the needs of washing machine casing production. The calculation formula for the control coefficient can be illustrated as follows:

[0081]

[0082] in, Indicates the control coefficients of the control chain; Indicates the final clamping distance of the control chain; This indicates that the first step in obtaining the control chain is... In the process of determining the final axial parameters of the layer, the total number of iterations of the beetle beard algorithm is the number of iterations performed to reach the corresponding angle parameter from the initial 0 degrees. Indicates the first The angular change index of the corresponding axis of the layer; Let z represent an exponential function with the natural constant as the base, and z represent the total number of layers.

[0083] In the formula, This represents the final clamping distance of the control chain. The smaller the final clamping distance, the better the iteration effect. and For the penalty item corresponding to the axis, where This reflects the iterative time cost during the acquisition of shaft parameters. In other words, the more iterations, the more complex the shaft parameter optimization process, potentially implying convergence difficulties or parameter sensitivity issues. The larger the value, the smaller the control coefficient, and the less likely this type of control chain can be used in subsequent control processes; This reflects the extent of parameter adjustment of the axis. The larger the difference, the greater the range of axis movement, which may increase the energy consumption or vibration risk of the robot arm, making it easier to cause production risks during the clamping of the washing machine shell. In addition, this method evaluates the overall efficiency and stability of the control chain by accumulating the penalty terms of all axes in the control chain.

[0084] In this embodiment of the invention, the optimal control chain is selected based on the control coefficient to control the robot arm axis, including: selecting the control chain with the largest control coefficient as the optimal control chain, and controlling the robot arm axis based on the optimal control chain.

[0085] Since a larger control coefficient indicates a smaller final clamping distance and better overall efficiency and stability, the control chain with the largest control coefficient is selected as the best-performing control chain.

[0086] The final axis parameters of each node in the optimal control chain are extracted hierarchically. Each axis parameter is checked to ensure it is within the physical limits of the robotic arm. If it exceeds these limits, a suboptimal chain is activated. The suboptimal chain is a control chain whose control coefficient is second only to the corresponding control coefficient of the optimal control chain. The final axis parameters of each node are then sent sequentially to the robotic arm controller, driving each axis to rotate according to the final axis parameters. The end effector gradually approaches the center of the washing machine casing.

[0087] In this embodiment of the invention, the control process may also include a monitoring process. During the rotation of the robotic arm, the actual distance between the end effector and the target is continuously monitored by binocular vision. If the deviation exceeds a threshold (e.g., 5mm), the process is immediately interrupted and the adjustment tree is regenerated. When the end effector reaches the target position, the gripper closing command is triggered to complete the gripping of the washing machine shell. At the same time, the actual gripping accuracy is recorded for subsequent optimization.

[0088] This invention employs a multi-axis intelligent robotic arm to grip the outer casing of a washing machine. The gripping distance is analyzed using visual images, and the angle parameters of different axes are iteratively adjusted to determine the gripping distance between the robotic arm's end effector and the desired gripping position on the washing machine casing. The angle parameters of different layers of the axis parameter chain are sequentially adjusted and iterated. Based on the changes in gripping distance across different iterations within the same layer, an optimal iterative value is determined for each axis. Candidate parameters are then selected from the angle parameters across different iterations based on these optimal values. Finally, the axis control is determined based on the candidate parameters from all layers of angle parameters. The control chain is designed to determine the control coefficients of each control chain based on the clamping distance, iteration count of different layers, and changes in angle parameters. The optimal control chain is then selected based on the control coefficients to control the robot's axes. This allows for gradual convergence through chain-like parameter updates, preventing target deviation or loss of field of view. The hierarchical iteration based on axis parameter chains effectively conforms to the multiple solution characteristics of the robot's inverse kinematics. The robustness of parameter optimization is improved through the control function, and the optimal control path of the robot is obtained through the control coefficients, reducing vibration risks and energy consumption, and ensuring the accuracy and stability of clamping and positioning during the production of the washing machine shell.

[0089] On the other hand, an intelligent robotic arm control system for washing machine casing production is also provided. The system includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the aforementioned intelligent robotic arm control methods for washing machine casing production.

[0090] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0091] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for controlling an intelligent robot for washing machine shell production, characterized in that, The method comprises the following steps: Acquiring the shaft parameters of the manipulator and the visual image of the washing machine shell in the painting process, wherein the shaft parameters are the angle parameters of different shafts in operation; Constructing a multi-level chain structure, the number of levels of the chain structure being equal to the number of shafts of the manipulator; sorting the levels of the chain structure in order from far to near the end of the manipulator, and each level of the chain structure being the angle parameter of a shaft, thereby obtaining a shaft parameter chain; Constructing a three-dimensional model of the motion of the manipulator; in the three-dimensional model, determining the clamping distance between the end of the manipulator and the position to be clamped of the washing machine shell based on the visual image; adjusting and iterating the angle parameters of different levels of the shaft parameter chain in order, and determining the iteration optimal value of each shaft after adjustment according to the change of the clamping distance in the same level in different iterations; and selecting candidate parameters from the angle parameters in different iterations according to the iteration optimal value; Based on the candidate parameters of all levels of angle parameters, determining a control chain for realizing the control of the shafts, and determining the control coefficient of each control chain according to the clamping distance corresponding to different control chains, the number of iterations of different levels, and the change of the angle parameters; and controlling the shafts of the manipulator according to the optimal control chain selected according to the control coefficient; The method for determining the iteration optimal value of each shaft after adjustment comprises: Determining the allowable value range of the angle parameter of any level corresponding shaft; Within the allowable value range, determining periodic analysis angles, and analyzing the step fluctuation according to the tentacle algorithm to determine the fluctuation angle of each analysis angle; All analysis angles and all fluctuation angles are collectively arranged in order from small to large as iteration angles, and the clamping distance under the corresponding iteration angle is simulated, wherein the iteration of the angle parameter of the shaft is iterated towards a fixed direction; Combining the change of the clamping distance in different iterations, constructing a regulation function, and calculating the iteration optimal value after each iteration adjustment; The regulation function is: ; wherein, represents the iteration preferred value of the parameter value in the sequence of the axis body parameters after the first iteration; represents the iteration preferred value of the parameter value in the sequence of the axis body parameters after the first iteration; represents the gripping distance corresponding to the robot when the parameter value in the sequence of the axis body parameters is the first iteration; represents the gripping distance corresponding to the robot when the parameter value in the sequence of the axis body parameters is the first iteration; represents the gripping distance corresponding to the robot when the parameter value in the sequence of the axis body parameters is the first iteration; represents the descending rate of the gripping distance corresponding to the robot when the parameter value in the sequence of the axis body parameters is the first iteration; represents the descending rate of the gripping distance corresponding to the robot when the parameter value in the sequence of the axis body parameters is the first iteration; represents the descending rate of the gripping distance corresponding to the robot when the parameter value in the sequence of the axis body parameters is the first represents the sigmod function, e represents the natural constant, and norm represents the linear normalization function. The method for selecting candidate parameters comprises: The preset number of angle parameters with the largest iteration optimal value are selected as candidate parameters.

2. The intelligent robot control method for the production of a washing machine housing according to claim 1, characterized in that, The method for constructing a three-dimensional model of the motion of the manipulator comprises: Establishing a kinematic model of the manipulator based on the DH parameter method, and placing it in a three-dimensional coordinate system containing the spatial position of the washing machine shell to obtain a three-dimensional model.

3. The intelligent robot control method for the production of a washing machine housing according to claim 1, characterized in that, The visual image is collected by a binocular vision camera arranged at the end of the manipulator, and the clamping distance between the end of the manipulator and the position to be clamped of the washing machine shell is determined based on the visual image, which comprises: Using the YOLO3 algorithm to identify the washing machine shell in the visual image, and based on the binocular vision principle, calculating the depth information through parallax to obtain the distance between the washing machine shell and the camera as the clamping distance between the end of the manipulator and the washing machine shell.

4. The intelligent robot control method for the production of a washing machine housing according to claim 1, characterized in that, Based on the candidate parameters of all levels of angle parameters, the control chain for realizing the control of the shafts comprises: Based on the candidate parameters of all levels of angle parameters, constructing a regulation tree structure; Traversing the regulation tree structure to determine the path from the root node to the leaf node in the regulation tree structure as a control chain.

5. The intelligent robot control method for the production of a washing machine housing according to claim 1, characterized in that, According to the clamping distance corresponding to different control chains, the iteration number of different layers and the angle parameter change, the control coefficient of each control chain is determined, including: The angle parameter change of different layers from the initial value without iteration to the value under the corresponding control chain layer is taken as the angle change index; The product value of the angle change index of the same layer and the iteration number of the corresponding layer is taken as the penalty coefficient of the corresponding layer, and the sum value of the penalty coefficients of all layers is taken as the total penalty index; The product value of the clamping distance corresponding to the control chain and the total penalty index is calculated, the inverse number of the product value is normalized to obtain the control coefficient.

6. The intelligent robot control method for the production of a washing machine housing shell according to claim 1, characterized in that, According to the control coefficient, the optimal control chain is screened to control the robot shaft body, including: The control chain with the largest control coefficient is taken as the optimal control chain, and the robot shaft body is controlled based on the optimal control chain.

7. An intelligent robot control system for washing machine shell production, said system comprising a memory, a processor and a computer program stored in said memory and running on said processor, characterized in that, The processor executes the computer program to realize the steps of the method in any one of claims 1-6.

Images