Intelligent control method and system for fully automatic unloading

Abstract

The application discloses an intelligent control method and system for full-automatic plate unloading, and relates to the technical field of intelligent manufacturing. The method comprises the following steps: a mechanical arm performs a grabbing operation according to optimal grabbing parameters, and analyzes the grabbing stability of a PCB (Printed Circuit Board) in real time; the grabbing force and angle of the mechanical arm are dynamically adjusted according to the grabbing stability of the PCB; the speed of unloading the PCB is initialized using an adaptive control algorithm based on the adjusted grabbing force and angle of the mechanical arm; the unloading operation is performed according to the initialized unloading speed, the actual motion trajectory of the PCB is monitored in real time, and the unloading risk is evaluated; and the unloading speed is dynamically adjusted according to the evaluation result. The application analyzes the grabbing stability in real time through an energy balance equation, dynamically adjusts the grabbing force and angle, and ensures the stability of the grabbing process; the unloading speed is initialized using an adaptive control algorithm, and the unloading speed is dynamically adjusted using a nonlinear mapping and a dynamic adjustment mechanism, thereby ensuring the smoothness and safety of the unloading process.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing technology, and in particular to an intelligent control method and system for fully automated unloading. Background Technology

[0002] With the rapid development of industrial automation and intelligent manufacturing, traditional PCB unloading operations mainly rely on manual labor or semi-automatic equipment, resulting in low efficiency, poor accuracy, and easy damage to PCBs. In recent years, fully automated PCB unloading systems based on machine vision and robot control technologies have gradually become a research hotspot. In existing technologies, machine vision technology is widely used for PCB positioning and inspection, while robot control technology is used to achieve precise gripping and unloading operations. However, existing methods still have significant shortcomings in areas such as gripping parameter prediction, gripping stability analysis, and unloading speed control.

[0003] The main shortcomings of the existing technology are: First, the prediction of the captured parameters lacks deep integration of PCB board geometric features and time series data, resulting in insufficient prediction accuracy and inability to adapt to dynamic changes in complex production environments; Second, the control of the unloading speed lacks a dynamic adjustment mechanism based on real-time monitoring and risk assessment, making it difficult to ensure the smoothness and safety of the unloading process. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides an intelligent control method for fully automatic unloading to solve the problems of insufficient accuracy in predicting grasping parameters and poor stability in the unloading process.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides an intelligent control method for fully automatic board unloading, comprising: acquiring visual data of the PCB board and extracting geometric features; using a temporal fusion mechanism based on the geometric features to predict the optimal gripping parameters of the robotic arm; the visual data including the position coordinates, angle information, contour shape, surface texture, and edge data of the PCB board; the robotic arm performing a gripping operation according to the optimal gripping parameters and analyzing the gripping stability of the PCB board in real time; dynamically adjusting the gripping force and angle of the robotic arm according to the gripping stability of the PCB board; initializing the unloading speed using an adaptive control algorithm based on the adjusted gripping force and angle of the robotic arm; performing the unloading operation according to the initialized unloading speed; monitoring the actual movement trajectory of the PCB board in real time and performing an unloading risk assessment; and dynamically adjusting the unloading speed according to the assessment results.

[0008] As a preferred embodiment of the fully automatic unloading intelligent control method of the present invention, the specific steps for extracting geometric features are as follows:

[0009] The visual data of the PCB board is preprocessed through noise reduction, contrast enhancement, edge sharpening and image segmentation.

[0010] Based on the preprocessed visual data of the PCB board, contour detection algorithm, Hough transform, GLCM and edge detection algorithm are used to extract the contour features, angle features, surface texture features and edge features of the PCB board.

[0011] As a preferred embodiment of the intelligent control method for fully automatic unloading described in this invention, the step of predicting the optimal grasping parameters of the robotic arm based on geometric features using a temporal fusion mechanism includes the following specific steps.

[0012] Organize geometric features according to time series to construct time series data;

[0013] Based on time-series data, a bidirectional long short-term memory network is constructed using two LSTM units, and the time-series features of the PCB board are extracted through the bidirectional long short-term memory network.

[0014] Geometric features and temporal features are fused temporally using a dynamic interaction mechanism based on graph neural networks to generate geometric temporal features;

[0015] Based on geometric temporal characteristics, nonlinear transformations and enhancements are performed using hyperbolic orthogonal functions and exponential functions, respectively, and the Sigmoid function is used for mapping to generate the optimal grasping parameters y of the robotic arm.

[0016] As a preferred embodiment of the fully automatic board unloading intelligent control method of the present invention, the robotic arm performs the gripping operation according to the optimal gripping parameters and analyzes the gripping stability of the PCB board in real time. The specific steps are as follows:

[0017] During the process of the robotic arm grasping the PCB board according to the optimal grasping parameters, the status data of the PCB board is collected.

[0018] Based on the force data of the PCB board, by analyzing the dynamic energy interaction during the gripping process, an energy balance equation is established, and the gripping stability index S(t) of the PCB board is calculated.

[0019] Based on historical crawled data, a stability standard S1 is defined through statistical analysis.

[0020] When S(t)≥S1, the robotic arm grasping process is considered stable; otherwise, the robotic arm grasping process is considered unstable.

[0021] As a preferred embodiment of the fully automatic board unloading intelligent control method of the present invention, the specific steps of dynamically adjusting the gripping force and angle of the robotic arm based on the gripping stability of the PCB board are as follows:

[0022] Based on the deviation between the grasping stability index S(t) and the stability standard S1, the grasping force is dynamically adjusted according to the principle of energy balance. ˙

[0023] The gripping angle is dynamically adjusted by using the frictional loss E(t) between the robotic arm and the PCB board and the vibration velocity b(t) of the PCB board as adjustment factors.

[0024] As a preferred embodiment of the intelligent control method for fully automatic unloading described in this invention, the unloading speed is initialized using an adaptive control algorithm based on the adjusted gripping force and angle of the robotic arm. The specific steps are as follows:

[0025] An adaptive control algorithm is developed based on MRAC and RL learning mechanisms.

[0026] The adjusted gripping force and angle of the robotic arm are used as adjustment factors for the adaptive control algorithm;

[0027] The adaptive control algorithm uses MRAC to initially calculate the lower plate speed and optimizes it through the RL learning mechanism to generate the initial lower plate speed V.

[0028] As a preferred embodiment of the fully automatic board unloading intelligent control method of the present invention, the steps of performing the unloading operation according to the initial unloading speed, monitoring the movement trajectory of the PCB board in real time, conducting unloading risk assessment, and dynamically adjusting the unloading speed according to the assessment results are as follows.

[0029] During the unloading process, the actual movement trajectory of the PCB board is collected, and the deviation ΔH between the actual movement trajectory and the expected movement trajectory is calculated.

[0030] Based on statistical analysis of the motion trajectory deviation of historical PCB boards, a deviation threshold H1 is defined.

[0031] When ΔH≥H1, there is a risk in the process of lowering the plate; otherwise, there is no risk in the process of lowering the plate.

[0032] Based on the risk assessment results, a nonlinear mapping and dynamic adjustment mechanism is adopted to dynamically adjust the lowering speed.

[0033] Secondly, this invention provides a fully automatic intelligent control system for unloading PCBs, comprising: a gripping parameter prediction module, a robotic arm adjustment module, an unloading speed initialization module, and an unloading speed adjustment module; the gripping parameter prediction module is used to collect visual data of the PCB and extract geometric features, and predict the optimal gripping parameters of the robotic arm based on the geometric features using a temporal fusion mechanism; the robotic arm adjustment module is used to have the robotic arm perform gripping operations according to the optimal gripping parameters, and to analyze the gripping stability of the PCB in real time, and dynamically adjust the gripping force and angle of the robotic arm according to the gripping stability of the PCB; the unloading speed initialization module is used to initialize the unloading speed using an adaptive control algorithm based on the adjusted gripping force and angle of the robotic arm; the unloading speed adjustment module is used to perform the unloading operation according to the initialized unloading speed, monitor the actual movement trajectory of the PCB in real time, perform unloading risk assessment, and dynamically adjust the unloading speed according to the assessment results.

[0034] Thirdly, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the fully automatic board unloading intelligent control method as described in the first aspect of the present invention.

[0035] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the intelligent control method for fully automatic board unloading as described in the first aspect of the present invention.

[0036] The beneficial effects of this invention are as follows: by collecting visual data of the PCB board and extracting geometric features, the optimal grasping parameters are predicted using a time-series fusion mechanism to ensure the accuracy and adaptability of the grasping operation; the grasping operation is executed based on the optimal grasping parameters, and the grasping stability is analyzed in real time through the energy balance equation, and the grasping force and angle are dynamically adjusted to ensure the stability of the grasping process; the board removal speed is initialized by combining an adaptive control algorithm, and the board removal speed is dynamically adjusted by nonlinear mapping and dynamic adjustment mechanisms to ensure the smoothness and safety of the board removal process. Attached Figure Description

[0037] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. 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.

[0038] Figure 1 This is a flowchart of the fully automatic unloading intelligent control method in Example 1.

[0039] Figure 2This is a schematic diagram of the fully automatic unloading intelligent control system in Example 1.

[0040] Figure 3 This is a flowchart of the timing fusion mechanism in the fully automatic board unloading intelligent control method of Example 1.

[0041] Figure 4 This is a flowchart illustrating the dynamic adjustment of the unloading speed in the fully automatic unloading intelligent control method of Example 1. Detailed Implementation

[0042] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0043] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0044] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0045] Example 1, referring to Figures 1-4 This is the first embodiment of the present invention, which provides a fully automatic board unloading intelligent control method, including the following steps:

[0046] S1. Collect visual data from the PCB board and extract geometric features. Based on the geometric features, use a temporal fusion mechanism to predict the optimal grasping parameters of the robotic arm.

[0047] Visual data includes the PCB board's position coordinates, angle information, outline shape, surface texture, and edge data.

[0048] It should be noted that a camera mounted above the work area captures real-time images of the PCB board, ensuring full coverage of the operating area to obtain comprehensive information. Position coordinates and angle information are used to identify key points on the PCB board using feature point detection algorithms in image processing, and their position and angle relationships relative to preset reference points are calculated. The outline shape is determined by edge detection algorithms analyzing areas of significant brightness change in the image, thus accurately delineating the PCB board's shape. Surface texture is extracted from the image using contrast enhancement and texture analysis algorithms to identify surface characteristics and potential defects. Edge data is processed by specialized edge detection algorithms to precisely locate the PCB board edges.

[0049] The visual data of the PCB board is preprocessed through noise reduction, contrast enhancement, edge sharpening and image segmentation.

[0050] Furthermore, the visual data of the PCB board is optimized through denoising processing. A Gaussian filter is used to remove random noise from the image, ensuring that the value of each pixel in the image is influenced by its surrounding pixels, thereby smoothing the image and reducing false features caused by noise. For example, when detecting minute defects on the PCB board, the denoising step can effectively reduce interference introduced by sensor or environmental factors.

[0051] The contrast enhancement process utilizes histogram equalization technology to adjust the grayscale distribution of an image, thereby expanding the brightness differences between different areas of the image and improving the overall clarity of visual data. Specifically, for PCB boards where there may be variations in color depth, contrast enhancement can make surface textures more apparent, facilitating the accurate identification of various features, such as solder joints or component markings.

[0052] Edge sharpening enhances the display of PCB board outlines and details by applying the Laplacian operator to highlight boundary information in an image. For example, when identifying fine edges of a PCB board, edge sharpening increases the contrast between the edge and the background, making previously blurry edges clear and sharp, which is beneficial for accurate measurement and positioning.

[0053] Image segmentation uses specific algorithms to divide the visual data of a PCB board into multiple meaningful parts or regions for individual analysis. For example, threshold segmentation can be used to divide an image into target regions (such as the PCB board itself) and background regions based on a preset grayscale threshold.

[0054] Based on the preprocessed visual data of the PCB board, contour detection algorithm, Hough transform, GLCM and edge detection algorithm are used to extract the contour features, angle features, surface texture features and edge features of the PCB board.

[0055] Organize geometric features according to time series to construct time series data;

[0056] Furthermore, the geometric features in the continuously acquired multi-frame visual data of the PCB board need to be labeled and recorded. The geometric features in each frame, including position coordinates, angle information, contour shape, surface texture, and edge data, are assigned a corresponding timestamp to reflect how these features change over time. For example, within an automated unloading operation cycle, from the moment the robotic arm approaches the PCB board until it completes the gripping process, images are acquired at fixed time intervals (e.g., every 10 milliseconds), and all geometric features at that moment are extracted. Then, these geometric features are concatenated according to time sequence to form one or more time-series data on the evolution of specific geometric features over time.

[0057] Based on time-series data, a bidirectional long short-term memory network is constructed using two LSTM units, and the time-series features of the PCB board are extracted through the bidirectional long short-term memory network.

[0058] Furthermore, the first LSTM unit is used to propagate the time series data forward in chronological order to capture the temporal dependencies of past moments, and the second LSTM unit is used to propagate the time series data backward in reverse chronological order to capture the temporal dependencies of future moments.

[0059] The specific process includes: First, the first LSTM unit performs forward propagation on the data in chronological order. This means starting from the earliest time point and gradually working backward to the latest time point, thereby capturing the temporal dependencies of past moments. For example, in analyzing the process of a PCB board moving from its initial position to the grasping point, the first LSTM unit processes the geometric features corresponding to each frame of the image sequentially, learning the trends of changes in information such as position coordinates and angles over time. Simultaneously, the second LSTM unit processes the same time-series data in reverse chronological order, i.e., backtracking from the latest time point to the earliest time point, to capture the temporal dependencies of future moments. Taking the same PCB board movement process as an example, the second LSTM unit can identify the impact of subtle changes near the completion of the grasping stage on future operation steps.

[0060] A bidirectional long short-term memory network is constructed by concatenating the hidden states of the first LSTM unit and the second LSTM unit at each time step, and trained using backpropagation and gradient descent based on historical time series data.

[0061] The specific process involves constructing a Bidirectional Long Short-Term Memory (Bi-LSTM) network by concatenating the hidden states of the first and second LSTM units at each time step. This allows for the comprehensive capture of forward and backward dependencies in the time-series data of geometric features. Specifically, when processing the process of a PCB board moving from its initial position to the grasping point, for the geometric features corresponding to each frame, the first LSTM unit extracts information from past moments via forward propagation, while the second LSTM unit extracts information from future moments via backward propagation. At the end of each time step, the hidden states generated by the two LSTM units are concatenated to form a vector that comprehensively represents the information before and after that time step. Based on these concatenated hidden states and historical time-series data, the gradient of the loss function relative to the network weights is calculated using the backpropagation algorithm, and the network weights are updated using gradient descent to minimize the prediction error. For example, in optimizing the gripping parameters of a robotic arm, through iterative training, the Bi-LSTM network can more accurately predict the optimal gripping parameters for different situations, thereby improving the accuracy and efficiency of the operation.

[0062] Geometric features and temporal features are fused temporally using a dynamic interaction mechanism based on graph neural networks to generate geometric temporal features;

[0063] Specifically, geometric and temporal features are constructed as a graph structure, where each node represents a specific geometric or temporal feature, and edges represent the dependencies between these features. For example, in processing PCB board gripping operations, geometric features such as position coordinates and angle information, along with temporal features extracted from a Bidirectional Long Short-Term Memory (Bi-LSTM) network, together form the nodes of the graph. The edges between nodes reflect how geometric and temporal features influence each other, such as how position changes affect the choice of gripping angle. Using Graph Neural Networks (GNNs), the complex interactions between these features can be captured, and the feature representation of each node is dynamically updated in each iteration through graph convolution operations. Specifically, graph convolution operations allow nodes to aggregate information from their neighboring nodes, thereby better understanding the interactions between local features and the long-range dependencies in the overall graph structure. This process continuously optimizes the combination of each geometric and temporal feature, ultimately generating geometric-temporal features that integrate temporal and spatial information.

[0064] Based on geometric temporal characteristics, nonlinear transformations and enhancements are performed using hyperbolic orthogonal functions and exponential functions, respectively, and a sigmoid function is used for mapping to generate the optimal grasping parameters y for the robotic arm. The expression is as follows:

[0065]

[0066] Where y is the optimal capture parameter, n is the total number of geometric time-series features, i is the index variable of the geometric time-series features, σ is the Sigmoid function, and α is the index variable of the geometric time-series features. i is the weight coefficient of the i-th geometric time series feature, W1 is the weight matrix of the geometric time series feature after nonlinear transformation (through the tanh function), W2 is the weight matrix of the geometric time series feature after enhancement (through the exp function), b1 is the bias term of the geometric time series feature after nonlinear transformation, b2 is the bias term of the geometric time series feature after enhancement, h i It is the i-th geometric temporal feature;

[0067] It should be noted that after constructing the geometric temporal features, in order to determine the degree of influence of each feature on the optimal grasping parameters y of the generated robotic arm, α is automatically learned through an optimization algorithm during the training process. i The value of α. This process considers the relationship between time-series features and the final crawling effect, ensuring that features crucial to the crawling operation receive higher weights. i The value of α is arbitrary real number in the initial stage, but after the Softmax function is used to convert α... i Transform into probabilistic form such that all α i The value of is between 0 and 1, and satisfies conditions.

[0068] The optimal gripping parameters include gripping force, gripping angle, and gripping position.

[0069] S2. The robotic arm performs the gripping operation according to the optimal gripping parameters and analyzes the gripping stability of the PCB board in real time. Based on the gripping stability of the PCB board, the robotic arm dynamically adjusts the gripping force and angle.

[0070] During the process of the robotic arm grasping the PCB board according to the optimal grasping parameters, the status data of the PCB board is collected.

[0071] The status data of the PCB board includes gripping force, displacement, torque, angle change, linear velocity, angular velocity, elastic deformation, and vibration velocity;

[0072] Based on the force data of the PCB board, an energy balance equation is established by analyzing the dynamic energy interaction during the gripping process, and the gripping stability index S(t) of the PCB board is calculated. The expression is as follows:

[0073]

[0074] Where S(t) represents the grasping stability index at time t, F(t) represents the grasping force applied by the robotic arm at time t, d(t) represents the displacement of the robotic arm at time t, M(t) represents the torque applied by the robotic arm at time t, θ(t) represents the change in angle of the robotic arm at time t, E(t) represents the friction loss at time t, q represents the mass of the PCB board, g(t) represents the linear velocity of the PCB board at time t, I represents the moment of inertia of the PCB board, ω(t) represents the angular velocity of the PCB board at time t, and k represents the moment of rotation of the PCB board at time t. ˙ The elastic coefficient, x(t) represents the elastic variable of the PCB board at time t, c represents the damping coefficient of the PCB board, and b(t) represents the vibration velocity of the PCB board at time t.

[0075] It should be noted that the process of establishing the energy balance equation through dynamic energy interaction is as follows: First, considering that the force and torque applied by the robotic arm to the PCB board produce linear displacement and angular changes respectively, according to the principles of physics, these actions can be converted into work done on the PCB board. That is, F(t)·d(t) represents the linear work caused by the gripping force, while M(t)·θ(t) represents the rotational work caused by the torque. These two parts together constitute the total energy input to the PCB board. However, in actual operation, due to factors such as friction, some energy will be lost. This lost energy is deducted from the total energy in the form of frictional loss E(t). On the other hand, in order to evaluate whether the PCB board is stable in this process, its kinetic energy state, including linear kinetic energy, needs to be considered. Rotational kinetic energy elastic potential energy and the kinetic energy caused by vibration Integrating all the above factors, an energy balance equation is formed, where the numerator is the net energy input from the outside (i.e., (F(t)·d(t)+M(t)·θ(t))-E(t)), and the denominator is the total kinetic energy of the PCB board. The gripping stability index S(t) is calculated using this ratio.

[0076] It should also be noted that the mass q of the CB board is obtained by measuring the actual weight of the PCB board; the moment of inertia I of the PCB board is determined by calculating the geometry and mass distribution of the PCB board; the elastic coefficient k of the PCB board is obtained by experimentally determining the deformation and recovery characteristics of the PCB board under stress; and the damping coefficient c of the PCB board is determined by analyzing the energy dissipation characteristics of the PCB board during vibration. The linear velocity g(t), angular velocity ω(t), elastic variable x(t), and vibration velocity b(t) of the PCB board are obtained by real-time acquisition of the motion state and vibration data of the PCB board during the grasping process using sensors.

[0077] Based on historical crawled data, a stability standard S1 is defined through statistical analysis.

[0078] When S(t)≥S1, the robotic arm grasping process is considered stable; otherwise, the robotic arm grasping process is considered unstable.

[0079] Based on the deviation between the grasping stability index S(t) and the stability standard S1, the grasping force is dynamically adjusted according to the energy balance principle, and the expression is:

[0080] F'(t)=F(t)+ΔF·(S1-S(t));

[0081] Where F′(t) is the adjusted gripping force, and ΔF represents the adjustment step size of the gripping force;

[0082] It should be noted that the process of dynamically adjusting the gripping force based on the deviation between the gripping stability index and the stability standard, according to the principle of energy balance, relies on the energy assessment and optimization of the current gripping state. When insufficient stability of the gripping operation is detected, it means that the currently applied force is not effectively maintaining the PCB board in the expected stable state. At this time, it is necessary to increase the gripping force to compensate for this instability and ensure that the robotic arm can firmly and safely grip the PCB board. Conversely, if the gripping is too stable or even exceeds the necessary range, the gripping force is appropriately reduced to prevent potential damage caused by excessive constraint. This process not only considers the forces and torques directly acting on the PCB board, but also incorporates the effects of factors such as friction loss and vibration velocity. By integrating this information and making real-time adjustments, the aim is to minimize energy consumption while ensuring the safety and stability of the gripping process. ˙

[0083] The frictional loss E(t) between the robotic arm and the PCB board, and the vibration velocity b(t) of the PCB board are used as adjustment factors to dynamically adjust the gripping angle. The expression is as follows:

[0084]

[0085] Where θ′(t) is the adjusted gripping angle, and Δθ represents the adjustment step size of the gripping angle.

[0086] It should be noted that the principle of optimizing the gripping process by dynamically adjusting the gripping angle, based on the frictional loss between the robotic arm and the PCB board and the vibration velocity of the PCB board, lies in precisely controlling the angle at which the robotic arm contacts the PCB board to adapt to different operating conditions and ensure optimal gripping results. During the gripping process, frictional loss reflects the energy loss between the contact surfaces, while vibration velocity indicates the dynamic behavior of the PCB board at the moment of gripping. Using these two factors as adjustment factors, the energy efficiency and stability under the current gripping configuration can be evaluated in real time, and the gripping angle can be adjusted accordingly. When the frictional loss is large or the vibration velocity exceeds expectations, it indicates that the existing gripping angle may not be suitable for the current conditions, and corresponding adjustments are needed to reduce energy loss and suppress unnecessary vibration, thereby ensuring that the gripping process is both stable and efficient. This adaptive adjustment mechanism based on actual operating parameters can effectively improve the accuracy and safety of the robotic arm's gripping of the PCB board, while also helping to extend the equipment's service life.

[0087] S3. Based on the adjusted gripping force and angle of the robotic arm, the lowering speed is initialized using an adaptive control algorithm.

[0088] An adaptive control algorithm is developed based on MRAC and RL learning mechanisms.

[0089] Furthermore, relying on pre-defined ideal behavioral standards, the system dynamically adjusts strategies to narrow the gap between the actual operating state and these standards by evaluating the differences in real time, ensuring high-efficiency performance under known conditions. Simultaneously, leveraging a learning-based operational mechanism, the system continuously learns and accumulates experience from numerous operational examples, enabling it to explore and implement optimal operational strategies in uncertain or changing work environments. This approach identifies different working conditions and continuously optimizes the decision-making process based on feedback, quickly responding to and adapting to new situations. The final intelligent adaptive control algorithm combines the advantages of both methods: on the one hand, it handles foreseeable changes through precise dynamic adjustments; on the other hand, it utilizes a learning mechanism to cope with unknown or unpredictable situations. This algorithm automatically adjusts key parameters such as gripping force and lowering speed, ensuring not only operational stability and safety but also significantly improving overall efficiency and flexibility.

[0090] The adjusted gripping force and angle of the robotic arm are used as adjustment factors for the adaptive control algorithm;

[0091] It should be noted that the gripping force is adjusted based on the deviation between the gripping stability index and the preset standard to ensure sufficient gripping stability under different conditions while avoiding damage caused by excessive force. Simultaneously, the gripping angle is further fine-tuned using factors such as friction loss and vibration velocity to compensate for any factors that may lead to instability. This method allows the adaptive control algorithm to make precise adjustments based on feedback from actual operation, improving not only the stability and safety of the gripping process but also enhancing its adaptability to various working conditions.

[0092] The adaptive control algorithm uses MRAC to initially calculate the lower plate velocity and then optimizes it through the RL learning mechanism to generate the initial lower plate velocity V, expressed as:

[0093]

[0094] Where V is the initial lowering velocity, g(F′(t),θ′(t)) is the preliminary estimate of the lowering velocity mapped by the nonlinear mapping function, η is the weight coefficient of the RL learning mechanism, and Q(s,a) is the state-action value function in the RL learning mechanism. This represents the expected cumulative reward for taking action a in state s.

[0095] It should be noted that η, as a weighting coefficient in the RL learning mechanism, is defined based on the need to balance reinforcement learning feedback with traditional control strategies. The value of η determines the degree of influence of the experience gained from the reinforcement learning mechanism (i.e., the state-action value function) on the final decision. Specifically, η is used to adjust the emphasis of the reinforcement learning part on future expected returns, so as to gradually introduce the optimized strategy learned through learning while adhering to the preset behavioral standards. The value of η ranges from 0 to 1, where a value close to 0 indicates a greater tendency to rely on the initially set behavioral standards and precise dynamic adjustment strategies, while a value close to 1 indicates a high degree of emphasis on the experience accumulated through reinforcement learning and the optimal strategy explored.

[0096] The expression for the nonlinear mapping function is:

[0097]

[0098] Where k1 represents the scaling factor, used to adjust the magnitude of the mapping result, and r is a parameter that controls the mapping width;

[0099] The expression for the state-action value function is:

[0100]

[0101] Where s is the current state, a is the current action, m is the time step, and r is the current action. mγ is the instantaneous reward at time step m, γ is the discount factor, R represents the expected value, s0 represents the initial state, and a0 represents the initial action.

[0102] S4. Perform the board removal operation according to the initial board removal speed, monitor the actual movement trajectory of the PCB board in real time, conduct a board removal risk assessment, and dynamically adjust the board removal speed according to the assessment results.

[0103] During the unloading process, the actual movement trajectory of the PCB board is collected, and the deviation ΔH between the actual movement trajectory and the expected movement trajectory is calculated.

[0104] Specifically, the expected motion trajectory is defined based on a series of coordinate points that the PCB board needs to traverse from its starting position to its target position. These coordinate points take into account factors such as the robotic arm's mobility, speed planning, and obstacles in the operating environment. The expected motion trajectory is represented by a series of continuous time functions, with each time point corresponding to the spatial position the PCB board should be in. For example, in three-dimensional space, the expected motion trajectory can be represented in three-dimensional coordinate form. The actual motion trajectory is determined by real-time data acquisition, including the actual position of the robotic arm at each moment. The deviation ΔH between the actual and expected motion trajectories is calculated, i.e., comparing the difference between the actual and expected positions at each time point, to evaluate the accuracy and stability of the entire board unloading process.

[0105] Based on statistical analysis of the motion trajectory deviation of historical PCB boards, a deviation threshold H1 is defined.

[0106] When ΔH≥H1, there is a risk in the process of lowering the plate; otherwise, there is no risk in the process of lowering the plate.

[0107] Based on the risk assessment results, a nonlinear mapping and dynamic adjustment mechanism is adopted to dynamically adjust the lowering speed, as expressed in the following expression:

[0108]

[0109] Where V′(t) is the dynamically adjusted lowering speed, V(t) is the lowering speed at the current time t, ΔV represents the adjustment step size of the lowering speed, u is the scaling parameter of the stability deviation, and p is the scaling parameter of the vibration deviation.

[0110] It should be noted that the hyperbolic tangent function tanh is used to convert this stability deviation into an adjustment factor, which smoothly reflects the change process from an unstable to a stable state. Simultaneously, considering the potential vibration issues during robotic arm operation, the exponential function exp is used to quantify the vibration deviation caused by the difference between the actual vibration velocity b(t) and the friction loss E(t), and the effect of larger deviations is further emphasized through a squared term. These two adjustment factors are combined to dynamically calculate the adjustment amount of the lowering speed. Thus, when a large stability or vibration deviation is detected, the adjusted lowering speed V′(t) will decrease accordingly to increase operational stability; conversely, when the deviation is small, the lowering speed can be appropriately increased to optimize efficiency. The entire process achieves precise control of the lowering speed through nonlinear mapping and dynamic adjustment mechanisms, ensuring that efficiency is maximized while guaranteeing operational safety.

[0111] This embodiment also provides a fully automatic intelligent control system for unloading PCBs, including: a gripping parameter prediction module, a robotic arm adjustment module, an unloading speed initialization module, and an unloading speed adjustment module. The gripping parameter prediction module collects visual data of the PCB and extracts geometric features, then uses a temporal fusion mechanism based on these geometric features to predict the optimal gripping parameters for the robotic arm. The robotic arm adjustment module executes the gripping operation based on the optimal gripping parameters and analyzes the gripping stability of the PCB in real time, dynamically adjusting the gripping force and angle of the robotic arm according to the gripping stability. The unloading speed initialization module initializes the unloading speed using an adaptive control algorithm based on the adjusted gripping force and angle of the robotic arm. The unloading speed adjustment module executes the unloading operation according to the initialized unloading speed, monitors the actual movement trajectory of the PCB in real time, performs unloading risk assessment, and dynamically adjusts the unloading speed based on the assessment results.

[0112] This embodiment also provides a computer device applicable to the intelligent control method for fully automatic board removal, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the intelligent control method for fully automatic board removal as proposed in the above embodiment.

[0113] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0114] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the intelligent control method for fully automatic board unloading as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0115] In summary, this invention achieves the following: by acquiring visual data of the PCB board and extracting geometric features, using a temporal fusion mechanism to predict optimal grasping parameters, ensuring the accuracy and adaptability of the grasping operation; by executing the grasping operation based on the optimal grasping parameters, and by analyzing the grasping stability in real time through the energy balance equation, dynamically adjusting the grasping force and angle, ensuring the stability of the grasping process; and by initializing the board removal speed using an adaptive control algorithm, and by dynamically adjusting the board removal speed using a nonlinear mapping and dynamic adjustment mechanism, ensuring the smoothness and safety of the board removal process.

[0116] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A fully automatic board unloading intelligent control method, characterized in that: comprising, Collecting visual data of the PCB and extracting geometric features, using a time sequence fusion mechanism based on the geometric features to predict the optimal grabbing parameters of the mechanical arm, the visual data including the position coordinates, angle information, contour shape, surface texture and edge data of the PCB; The mechanical arm performs grabbing operation according to the optimal grabbing parameters, and analyzes the grabbing stability of the PCB in real time, and dynamically adjusts the grabbing force and angle of the mechanical arm according to the grabbing stability of the PCB; Based on the adjusted grabbing force and angle of the mechanical arm, an adaptive control algorithm is used to initialize the board lowering speed; According to the initialization of the board lowering speed, the board lowering operation is performed, the actual motion trajectory of the PCB is monitored in real time, and the board lowering risk assessment is performed, and the board lowering speed is dynamically adjusted according to the evaluation result; Wherein, the optimal grabbing parameters of the mechanical arm are predicted based on the geometric features using a time sequence fusion mechanism, and the specific steps are as follows: Organize the geometric features according to the time sequence to construct the time sequence data; Based on the time sequence data, a bidirectional long short memory network is constructed through two LSTM units, and the time sequence features of the PCB are extracted through the bidirectional long short memory network; The geometric features and the time sequence features are fused through a dynamic interaction mechanism based on a graph neural network to generate geometric time sequence features; Based on the geometric timing characteristics, through hyperbolic orthogonal function and exponential function respectively nonlinear transformation and enhancement, and using Sigmoid function mapping, the optimal grabbing parameters of the mechanical arm are generated ; And, the grabbing force and angle of the mechanical arm are dynamically adjusted according to the grabbing stability of the PCB, and the specific steps are as follows, Based on the stability index of grabbing Deviation from the stability standard S1, according to the principle of energy balance dynamic adjustment of the strength of the grab Frictional losses between the robot arm and the PCB and the vibration velocity of the PCB The grasp angle is dynamically adjusted as a tuning factor.

2. The intelligent control method for full-automatic plate lowering as claimed in claim 1, characterized in that: The specific steps of extracting geometric features are as follows, The visual data of the PCB is preprocessed through denoising, contrast enhancement, edge sharpening and image segmentation; Based on the preprocessed visual data of the PCB, the contour feature, angle feature, surface texture feature and edge feature of the PCB are extracted using contour detection algorithm, Hough transform, GLCM and edge detection algorithm.

3. The intelligent control method for full-automatic plate lowering as claimed in claim 1, characterized in that: The specific steps of the mechanical arm performing grabbing operation according to the optimal grabbing parameters and analyzing the grabbing stability of the PCB in real time are as follows, During the process of the mechanical arm grabbing the PCB according to the optimal grabbing parameters, the state data of the PCB is collected; Based on the stress data of the PCB board, the dynamic energy interaction in the grabbing process is analyzed, an energy balance equation is established, and the grabbing stability index of the PCB board is calculated ; Based on the historical grabbing data, the stability standard S1 is defined through statistical analysis; When If S1, then the robotic arm grasping process is considered stable, otherwise it is considered unstable.

4. The intelligent control method for full-automatic plate lowering as claimed in claim 1, characterized in that: The specific steps of initializing the board lowering speed using an adaptive control algorithm based on the adjusted grabbing force and angle of the mechanical arm are as follows, Based on MRAC and RL learning mechanism, an adaptive control algorithm is formulated; The adjusted grabbing force and angle of the mechanical arm are used as the adjustment factor of the adaptive control algorithm; The adaptive control algorithm uses MRAC to preliminarily calculate the wafer speed, and optimizes the wafer speed through an RL learning mechanism to generate the initialized wafer speed .

5. The intelligent control method for full-automatic plate lowering as claimed in claim 4, characterized in that: The specific steps of performing board lowering operation according to the initialization of the board lowering speed, monitoring the motion trajectory of the PCB in real time, and performing board lowering risk assessment according to the evaluation result, and dynamically adjusting the board lowering speed according to the evaluation result are as follows, In performing the underplate process, actual motion trajectories of the PCB are collected and deviations of the actual motion trajectories from expected motion trajectories are calculated ; Based on the statistical analysis of the deviation of the motion trajectory of the historical PCB, the deviation threshold H1 is defined; When If H1, the under-plate process is at risk, otherwise, the under-plate process is not at risk; According to the risk assessment result, the board lowering speed is dynamically adjusted by using nonlinear mapping and dynamic adjustment mechanism.

6. An intelligent control system for full-automatic plate unloading, based on the intelligent control method for full-automatic plate unloading according to any one of claims 1-5, characterized in that: comprising, grabbing parameter prediction module, mechanical arm adjustment module, board lowering speed initialization module and board lowering speed adjustment module; The grabbing parameter prediction module is used for collecting the visual data of the PCB and extracting the geometric features, using a time sequence fusion mechanism based on the geometric features to predict the optimal grabbing parameters of the mechanical arm; The mechanical arm adjusting module is used for the mechanical arm to perform a grabbing operation according to the optimal grabbing parameter and to analyze the grabbing stability of the PCB in real time, and the grabbing force and angle of the mechanical arm are dynamically adjusted according to the grabbing stability of the PCB. The lower plate speed initialization module is used for initializing the lower plate speed by using an adaptive control algorithm based on the adjusted grabbing force and angle of the mechanical arm. The lower plate speed adjusting module is used for performing a lower plate operation according to the initialized lower plate speed, monitoring the actual motion track of the PCB in real time, performing a lower plate risk assessment, and dynamically adjusting the lower plate speed according to the assessment result. 7.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is characterized in that: The processor executes the computer program to implement the steps of the intelligent control method for full-automatic lower plate as claimed in any one of claims 1-5.

8. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to implement the steps of the intelligent control method for full-automatic lower plate as claimed in any one of claims 1-5.

Images