Dynamic anti-swing control method of grab ship unloader, storage medium and computer device

Abstract

This application discloses a dynamic anti-sway control method, storage medium, and computer equipment for a grab bucket ship unloader. The method includes: acquiring real-time operating status information of the grab bucket ship unloader during operation using multi-source sensors installed on the grab bucket ship unloader; determining a target feedback gain matrix corresponding to the grab bucket ship unloader based on the real-time wire rope length and real-time load in the real-time operating status information using a gain scheduling table, wherein the gain scheduling table includes optimal feedback gain matrices corresponding to various combinations of wire rope length and load; determining a target control law based on the target feedback gain matrix, and determining a target control signal for the grab bucket ship unloader based on the target control law and the real-time operating status information; and driving the movement of the grab bucket ship unloader with the target control signal. This application can achieve dynamic anti-sway, improving operational accuracy and efficiency.

Description

Technical Field

[0001] This application relates to the field of grab unloader technology, and in particular to a dynamic anti-sway control method, storage medium and computer equipment for a grab unloader. Background Technology

[0002] In the port logistics and transportation sector, grab unloaders, as a key loading and unloading equipment, undertake the important task of efficiently and accurately unloading cargo from ships to the dock. Their operational efficiency and safety directly affect the operational effectiveness and economic benefits of the entire port logistics system. However, in actual operation, grab unloaders face many complex and urgent problems, among which the grab bucket swing control challenge is particularly prominent, becoming a key factor restricting their performance improvement.

[0003] During operation, grab unloaders inevitably experience grab swaying due to the lifting and lowering of the grab and the frequent movement of the trolley and carriage. This swaying not only severely affects the loading and unloading accuracy, leading to excessive deviations in the unloading position and increasing the workload of subsequent cargo handling and transportation, thus reducing overall operational efficiency, but also poses a significant risk of collisions with surrounding equipment and facilities, damaging the equipment itself, shortening its lifespan, and threatening the personal safety of on-site personnel, potentially causing serious accidents. Therefore, effectively suppressing grab swaying and achieving dynamic anti-sway control of grab unloaders has become a pressing technical challenge in this field.

[0004] Currently, existing anti-sway control methods for grab unloaders have many limitations. Some traditional methods employ simple open-loop control strategies, operating the grab unloader solely based on preset motion commands, lacking real-time monitoring and feedback adjustments to the actual swaying state of the grab. This control method cannot respond promptly to the dynamic changes of the grab under different operating conditions, making it difficult to effectively suppress grab swaying and resulting in poor anti-sway performance.

[0005] While some methods incorporate closed-loop control concepts and acquire partial state information of the grab bucket through sensors, their control algorithm design is relatively simplistic, often considering only a single factor affecting the grab bucket's sway, such as controlling solely based on the grab bucket's sway angle. However, the operation of a grab bucket ship unloader is a complex dynamic system, influenced by multiple factors that intertwine and collectively affect the grab bucket's sway characteristics. Existing single-factor control methods cannot comprehensively and accurately reflect the actual sway of the grab bucket, making it difficult to adjust the control strategy in real time according to complex changes under different operating conditions. This results in unstable anti-sway control performance in complex and ever-changing operating environments, failing to meet the requirements for efficient and safe operation of grab bucket ship unloaders in actual production. Summary of the Invention

[0006] In view of this, the embodiments of this application provide a dynamic anti-sway control method, storage medium and computer equipment for a grab bucket ship unloader, which can realize dynamic anti-sway, improve operation accuracy and efficiency, reduce the risk of equipment damage caused by collisions, extend equipment life, and at the same time ensure the personal safety of on-site operators.

[0007] According to one aspect of this application, a dynamic anti-sway control method for a grab unloader is provided, the method comprising:

[0008] Real-time operating status information of the grab unloader is obtained by using multi-source sensors installed on the grab unloader.

[0009] Using a gain scheduling table, the target feedback gain matrix corresponding to the grab unloader is determined based on the real-time wire rope length and real-time load in the real-time operating status information. The gain scheduling table includes the optimal feedback gain matrix corresponding to various combinations of wire rope length and load.

[0010] The target control law is determined based on the target feedback gain matrix, and the target control signal of the grab unloader is determined based on the target control law and the real-time operating status information.

[0011] The movement of the grab unloader is driven by the target control signal.

[0012] Optionally, the multi-source sensors include a spreader end IMU, a wire rope length sensor, a hoisting speed encoder, a grab bucket position encoder, a grab bucket speed encoder, and a weighing sensor; the real-time operating status information includes real-time spreader swing angle, real-time spreader angular velocity, real-time wire rope length, real-time hoisting speed, real-time grab bucket position, real-time grab bucket speed, and real-time load.

[0013] Optionally, using a gain scheduling table, based on the real-time wire rope length and real-time load in the real-time operating status information, the target feedback gain matrix corresponding to the grab unloader is determined, including:

[0014] Check if the gain scheduling table contains a combination that is the same as the real-time wire rope length and the real-time load;

[0015] If included, the optimal feedback gain matrix corresponding to the combination of the same real-time wire rope length and the real-time load is obtained as the target feedback gain matrix;

[0016] If not included, bilinear interpolation is performed on each combination in the gain scheduling table, and the target feedback gain matrix matching the real-time wire rope length and the real-time load is determined based on the bilinear interpolation result.

[0017] Optionally, before determining the target feedback gain matrix corresponding to the grab unloader based on the real-time wire rope length and real-time load in the real-time operating status information using the gain scheduling table, the method further includes:

[0018] A dynamic model of the grab bucket ship unloader is constructed, and the Laplace transform is performed on the dynamic model to obtain the transfer function of the control system of the grab bucket ship unloader.

[0019] A state-space expression is established based on the transfer function of the control system, and the Riccati equation is determined based on the state-space expression.

[0020] Define the performance index function of the control system, and solve the Ricardi equation based on multiple control sample data of the grab unloader to obtain the feedback gain matrix that minimizes the value of the performance index function.

[0021] Based on the wire rope length sample data and load sample data in each control sample data, as well as the corresponding feedback gain matrix, a gain scheduling table is constructed.

[0022] Optionally, the state-space expression is: A represents the system state matrix, B represents the input matrix, C represents the output matrix, D represents the direct transfer matrix, x represents the operating state information, f represents the driving force of the grab unloader, and y represents the output vector. Represent the state equation;

[0023] The expression for the Riccati equation is: P represents a symmetric positive definite matrix, R represents the control input weight matrix, and Q represents the state weight matrix;

[0024] The feedback gain matrix is ​​determined by solving for R, Q, and P, which minimize the value of the performance index function. .

[0025] Optionally, after obtaining the feedback gain matrix that minimizes the value of the performance index function, the method further includes:

[0026] The convergence time, overshoot, spreader swing angle, and control input energy of the control system are weighted and summed based on the corresponding weights to define the fitness function;

[0027] The particle swarm optimization algorithm is used to optimize the control input weight matrix R and the state weight matrix Q in the LQR controller with the fitness function as the optimization objective, and the feedback gain matrix is ​​updated based on the optimized control input weight matrix R and state weight matrix Q.

[0028] Optionally, a target control law is determined based on the target feedback gain matrix, and a target control signal for the grab unloader is determined according to the target control law and the real-time operating status information, including:

[0029] The target control law is determined based on the target feedback gain matrix and the state vector corresponding to the real-time operating status information.

[0030] The real-time disturbance estimate obtained by the disturbance observer is obtained, and the target control law is compensated based on the product of the difference between the real-time disturbance estimate and the state vector corresponding to the real-time operating state information and the observation gain coefficient, so as to obtain the feedback compensation amount.

[0031] The target control signal for the grab unloader is determined based on the feedback compensation amount.

[0032] Optionally, the method further includes:

[0033] The system acquires real-time image data of the grab unloader captured by video surveillance, and uses visual algorithms to identify the real-time spatial position, real-time swing posture, and real-time three-dimensional shape of the target unloading area of ​​the grab bucket, thereby generating a real-time operation path prediction trajectory for the grab unloader.

[0034] Based on the current spatial location and the real-time operation path pre-aiming trajectory, calculate the feedforward compensation amount of the grab unloader.

[0035] The target control signal for the grab unloader is determined based on the feedback compensation amount, including:

[0036] The target control signal for the grab unloader is determined based on the feedforward compensation amount and the feedback compensation amount.

[0037] According to another aspect of this application, a storage medium is provided that stores a computer program thereon, which, when executed by a processor, implements the above-described dynamic anti-sway control method for a grab unloader.

[0038] According to another aspect of this application, a computer device is provided, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, wherein the processor executes the program to implement the above-described dynamic anti-sway control method for a grab unloader.

[0039] By employing the above technical solutions, this application provides a dynamic anti-sway control method, storage medium, and computer equipment for a grab bucket ship unloader. This method utilizes multiple sensors on the unloader, including a spreader-end IMU, wire rope length sensor, lifting speed encoder, grab bucket position encoder, and grab bucket speed encoder, to acquire real-time operating status information such as spreader swing angle, angular velocity, wire rope length, lifting speed, and grab bucket position and speed. A gain scheduling table is used to determine the target feedback gain matrix based on the real-time wire rope length and load, thereby determining the target control law and target control signal. This signal drives the unloader's movement to achieve dynamic anti-sway. This application embodiment uses multi-source sensors to collect data comprehensively, allowing the control system to accurately grasp the real-time operating details of the equipment, providing a reliable basis for precise control. The gain scheduling table flexibly determines the feedback gain matrix based on real-time operating conditions, enabling the control system to quickly adapt to different operating conditions, enhancing the system's adaptability and robustness. The generated target control signal can then adjust the control output promptly and accurately based on the real-time swing state of the grab bucket, effectively suppressing sway. Ultimately, dynamic anti-sway is achieved, improving operational accuracy and efficiency, reducing the risk of equipment damage caused by collisions, extending equipment life, and ensuring the personal safety of on-site personnel.

[0040] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0041] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0042] Figure 1 A flowchart illustrating a dynamic anti-sway control method for a grab unloader provided in an embodiment of this application is shown.

[0043] Figure 2 A flowchart illustrating another dynamic anti-sway control method for a grab unloader provided in an embodiment of this application is shown. Detailed Implementation

[0044] The present application will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the embodiments of the present application can be combined with each other.

[0045] This embodiment provides a dynamic anti-sway control method for a grab bucket ship unloader, such as... Figure 1 As shown, the method includes:

[0046] Step 101: Obtain real-time operating status information of the grab unloader during operation using multi-source sensors installed on the grab unloader. These multi-source sensors include a spreader end IMU, a wire rope length sensor, a hoisting speed encoder, a grab bucket position encoder, and a grab bucket speed encoder. The real-time operating status information includes real-time spreader swing angle, real-time spreader angular velocity, real-time wire rope length, real-time hoisting speed, real-time grab bucket position, and real-time grab bucket speed.

[0047] The dynamic anti-sway control method for grab unloaders provided in this application aims to accurately and efficiently control the grab bucket swaying problem during the operation of the grab unloader. First, real-time operating status information is collected by installing multi-source sensors on the grab unloader. Specifically, the multi-source sensors include: a spreader-end IMU (Inertial Measurement Unit): a device capable of measuring the three-axis attitude angles (or angular rates) and acceleration of an object. On the grab unloader, the spreader-end IMU can measure the sway angle and angular velocity of the spreader in three-dimensional space in real time. For example, when the grab bucket sways during lifting or movement, the IMU can quickly sense the angular changes and angular velocities of the spreader in the roll, pitch, and yaw directions, providing crucial data for subsequent analysis of the grab bucket's swaying state. A wire rope length sensor: this sensor is used to measure the length of the wire rope in real time. During the operation of the grab unloader, the length of the wire rope continuously changes as the grab bucket is lifted and lowered. A wire rope length sensor accurately obtains the actual length of the wire rope at any given moment. This data is crucial for understanding the vertical position of the grab bucket and analyzing its swaying characteristics. For example, when the wire rope is shorter, the grab bucket is closer to the main body of the unloader, and its swaying amplitude and frequency may differ from when the wire rope is longer. A hoisting speed encoder precisely measures the hoisting speed of the grab bucket. During the hoisting process, the hoisting speed encoder records the movement speed of the hoisting mechanism in real time and converts the speed information into an electrical signal, which is then transmitted to the control system. Understanding the real-time hoisting speed helps analyze the dynamic characteristics of the grab bucket during the hoisting process. For example, rapid hoisting may cause the grab bucket to sway more, while slow, smooth hoisting helps reduce swaying. A grab bucket position encoder determines the specific horizontal position of the grab bucket. During the operation of the trolley and crane of the grab bucket unloader, the position of the grab bucket constantly changes. The grab bucket position encoder can provide real-time feedback on the coordinate position of the grab bucket on the dock plane, which is important for accurately controlling the movement trajectory of the grab bucket and avoiding collisions with surrounding equipment. For example, when unloading cargo, the grab bucket needs to be moved accurately over the designated cargo. The position information provided by the grab bucket position encoder ensures the accuracy of this operation. The grab bucket speed encoder measures the horizontal speed of the grab bucket. Changes in speed during horizontal movement affect the grab bucket's sway. The grab bucket speed encoder monitors the grab bucket's speed in real time, allowing the control system to adjust its control strategy promptly based on speed changes to suppress grab bucket sway. For example, when the grab bucket accelerates, it may generate a large inertial force, leading to increased sway. The control system can take measures in advance to reduce sway based on the speed information fed back by the grab bucket speed encoder.

[0048] Step 102: Using the gain scheduling table, determine the target feedback gain matrix corresponding to the grab bucket unloader based on the real-time wire rope length and real-time load in the real-time operating status information. The gain scheduling table includes the optimal feedback gain matrix corresponding to various combinations of wire rope length and load.

[0049] Next, using a pre-defined gain scheduling table, the target feedback gain matrix is ​​determined based on the two key parameters: the real-time wire rope length and the real-time load. The gain scheduling table is pre-constructed through extensive experiments and data analysis, covering various combinations of wire rope length and load, and determining the optimal feedback gain matrix for each combination. For example, when the real-time wire rope length is 20 meters and the real-time load is 10 tons, the optimal feedback gain matrix can be found in the gain scheduling table. This matrix reflects the system's response characteristics to the control input under this specific operating condition, providing a crucial basis for generating appropriate control signals subsequently.

[0050] Step 103: Determine the target control law based on the target feedback gain matrix, and determine the target control signal of the grab bucket unloader according to the target control law and the real-time operating status information.

[0051] Then, based on the determined target feedback gain matrix, the target control law is further determined. The target control law is the core logic of the entire control system; it determines how to adjust the control output according to the current operating state. After obtaining the target control law, the target control signal of the grab unloader is calculated using a specific algorithm, combined with the acquired real-time operating state information. For example, if the real-time operating state information shows that the grab bucket's swing amplitude is large, the calculated target control signal, based on the target feedback gain matrix and the target control law, may increase the control force on the lifting or movement of the grab bucket to quickly reduce the swing amplitude.

[0052] Step 104: Drive the movement of the grab unloader with the target control signal.

[0053] Finally, the calculated target control signal is transmitted to the drive system of the grab unloader. This target control signal drives the movement of the grab unloader, achieving dynamic adjustment and control of the grab's swing, ensuring that the grab can operate according to the expected trajectory and state. This embodiment acquires comprehensive and accurate real-time operating status information through multi-source sensors, enabling real-time monitoring of various dynamic changes in the grab during operation. This provides precise data support for subsequent control, avoiding control errors caused by incomplete or inaccurate information. The target feedback gain matrix is ​​determined using a gain scheduling table based on the real-time wire rope length and real-time load, fully considering the impact of various key factors on system control during grab unloader operation. The optimal feedback gain matrix corresponding to different length and load combinations allows the control system to adjust its response characteristics in real-time according to actual working conditions, enhancing the adaptability and flexibility of the control system. The target control law is determined based on the target feedback gain matrix, and a target control signal is generated by combining real-time operating status information. This allows for timely and accurate adjustment of the control output according to the actual swing of the grab, effectively suppressing grab swing. Ultimately, the target control signal drives the movement of the grab unloader, achieving dynamic and precise control of the grab's swing. This greatly improves the operational accuracy of the grab unloader, reduces the deviation of the cargo unloading position, reduces the workload of subsequent cargo sorting and handling, and enhances overall operational efficiency, providing a strong guarantee for the efficient and safe operation of port logistics.

[0054] By applying the technical solution of this embodiment, multi-source sensors, including a spreader IMU, wire rope length sensor, hoisting speed encoder, grab bucket position encoder, and grab bucket speed encoder, are configured on the ship unloader to acquire real-time operating status information such as spreader swing angle, angular velocity, wire rope length, hoisting speed, and grab bucket position and speed. Using a gain scheduling table, a target feedback gain matrix is ​​determined based on the real-time wire rope length and load, thereby determining the target control law and target control signal. This signal drives the ship unloader's movement to achieve dynamic anti-sway. This embodiment uses multi-source sensors to collect data comprehensively, allowing the control system to accurately grasp the real-time operating details of the equipment, providing a reliable basis for precise control. The gain scheduling table flexibly determines the feedback gain matrix based on real-time operating conditions, enabling the control system to quickly adapt to different operating conditions, enhancing the system's adaptability and robustness. The generated target control signal can adjust the control output promptly and accurately based on the real-time swing state of the grab bucket, effectively suppressing sway. Ultimately, efficient dynamic anti-sway is achieved, improving operational accuracy and efficiency, reducing the risk of equipment damage caused by collisions, extending equipment life, and ensuring the personal safety of on-site personnel.

[0055] In this embodiment of the application, optionally, a target feedback gain matrix corresponding to the grab bucket unloader is determined using a gain scheduling table based on the real-time wire rope length and real-time load in the real-time operating status information. This includes: querying whether the gain scheduling table contains a combination that is the same as the real-time wire rope length and the real-time load; if it is, obtaining the optimal feedback gain matrix corresponding to the combination that is the same as the real-time wire rope length and the real-time load as the target feedback gain matrix; if it is not, performing bilinear interpolation on each combination in the gain scheduling table, and determining the target feedback gain matrix that matches the real-time wire rope length and the real-time load based on the bilinear interpolation result.

[0056] In this embodiment, the gain scheduling table is first queried to see if there is a combination that is exactly the same as the currently obtained real-time wire rope length and real-time load. For example, assuming the real-time measured wire rope length is 20 meters and the real-time load is 12 tons, the gain scheduling table is searched for a combination with a wire rope length of 20 meters and a load of 12 tons. If the gain scheduling table happens to contain such a combination, the optimal feedback gain matrix corresponding to that combination is directly obtained and used as the target feedback gain matrix for subsequent control calculations. However, in actual operation, there may be situations where the combination of real-time wire rope length and real-time load does not exist in the gain scheduling table. For example, the real-time wire rope length is 21 meters and the real-time load is 11 tons, but this exact combination is not found in the gain scheduling table. In this case, bilinear interpolation is required for each combination in the gain scheduling table. Bilinear interpolation is a method of data interpolation in two-dimensional space. It uses the known data values ​​of four adjacent points to estimate the data value of the target point through a certain calculation method. In this embodiment, the real-time wire rope length and real-time load are used as target points. Utilizing the four adjacent combinations in the gain scheduling table and their corresponding optimal feedback gain matrices, a target feedback gain matrix matching the real-time wire rope length and real-time load is calculated using bilinear interpolation. The bilinear interpolation method can reasonably estimate the feedback gain matrix under unknown operating conditions based on existing data through scientific calculation. This allows the control system to accurately control various non-standard operating conditions, avoiding control failures or inaccuracies caused by the inability to obtain a suitable feedback gain matrix. This ensures that the control system obtains a relatively accurate target feedback gain matrix under various operating conditions, thereby improving the stability and reliability of the entire grab unloader's dynamic anti-sway control system and further enhancing the grab unloader's operating efficiency and safety.

[0057] In this embodiment of the application, optionally, before determining the target feedback gain matrix corresponding to the grab unloader based on the real-time wire rope length and real-time load in the real-time operating status information using the gain scheduling table, the method further includes: constructing a dynamic model of the grab unloader; performing a Laplace transform on the dynamic model to obtain the control system transfer function of the grab unloader; establishing a state-space expression based on the control system transfer function; determining the Riccati equation based on the state-space expression; defining a performance index function of the control system; and solving the Riccati equation based on multiple control sample data of the grab unloader to obtain the feedback gain matrix that minimizes the value of the performance index function; and constructing a gain scheduling table based on the wire rope length sample data and load sample data in each control sample data and the corresponding feedback gain matrix.

[0058] In this embodiment, a dynamic model of the grab unloader is first constructed. This process requires in-depth analysis of the physical structure and motion characteristics of the grab unloader, considering factors such as the interaction forces between its various components, mass distribution, and inertia. For example, during the lifting and lowering of the grab, the elastic deformation of the wire rope and the interaction force between the grab and the cargo all affect the dynamic behavior of the entire system. By applying mechanical principles and mathematical methods, these factors are comprehensively considered to construct a mathematical model that can accurately describe the motion state of the grab unloader. The grab anti-sway control system can be abstracted into a linearized model of a typical pendulum lifting system: Where M represents the equivalent mass of the grab bucket and its drive mechanism, which is usually obtained through equipment design parameters or on-site static weighing calculations; m is the equivalent mass of the grab bucket and the material it carries (real-time load), which can be obtained by converting the weighing sensor or the hoisting motor current with the calibration curve. This refers to the length of the wire rope. The swing angle of the grab relative to the vertical direction is obtained through IMU detection; This refers to the driving force (or control input force) acting in the horizontal direction of the grab bucket, generated by a drive motor or servo system, used to adjust the movement of the grab bucket to achieve anti-sway control of the grab bucket; and These are the angular velocity and angular acceleration, respectively, which can be obtained by integrating the IMU angular velocity or by filtering and differentiating. and These represent the horizontal displacement and acceleration of the grab bucket, respectively, calculated by the slide rail encoder or the speed feedback of the drive motor. This refers to gravitational acceleration. These parameters, determined through a combination of structural measurements, sensor fusion, and experimental calibration, can fully describe the nonlinear motion characteristics of the grab under horizontal acceleration / deceleration and external disturbances.

[0059] When the swing angle is small (which can be approximated as) , This allows for linear simplification of the system, ignoring higher-order terms. And the nonlinear centrifugal force component, yielding the simplified dynamic equations of the control system, as follows: ,in, The overall damping coefficient of the system reflects the combined effects of rope elasticity, air damping, and friction. It can be obtained through experimental identification (fitting of free decay curves) or system identification algorithms (such as the least squares method).

[0060] Further performing a Laplace transform on the equation yields the transfer function form of the system: Where s represents a complex frequency variable, it can be seen that the input of the system is the driving force of the grab bucket. The output is the grab angle. .

[0061] Next, the constructed dynamic model is subjected to a Laplace transform. The Laplace transform is a mathematical tool that converts time-domain functions into complex frequency-domain functions, transforming differential equations into algebraic equations, thus simplifying system analysis and calculation. Through the Laplace transform, the transfer function of the grab unloader's control system can be obtained. This transfer function describes the relationship between system inputs (such as control signals) and outputs (such as the grab's position and speed), and is an important basis for analyzing and designing the control system. A state-space expression is then established based on the control system transfer function. The state-space expression is a method of describing the dynamic characteristics of a system using state variables; it represents the relationship between the system's inputs, outputs, and state variables in matrix form. Establishing a state-space expression facilitates state feedback control and optimization design of the system. The Riccati equation is then determined based on the state-space expression. The Riccati equation is an important equation in control theory, closely related to the optimal control problem of a system. In the control system of the grab unloader, the solution to the Riccati equation directly affects the system's control performance. Specifically, the state-space expression is as follows: A represents the system state matrix, B represents the input matrix, C represents the output matrix, D represents the direct transfer matrix, x represents the operating state information, f represents the driving force of the grab unloader, and y represents the output vector. This represents the state equation. , , , The expression for the Riccati equation is: P represents a symmetric positive definite matrix, R represents the control input weight matrix, and Q represents the state weight matrix; the feedback gain matrix is ​​determined by solving for R, Q, and P that minimize the value of the performance index function. The feedback gain matrix corresponding to a specific combination of wire rope length L and load m can be represented as K(L, m).

[0062] Then, define the performance index function of the control system. The performance index function is a standard used to measure the performance of the control system, and it can be set according to actual needs. For example, the performance index function can be defined as the amplitude of the grab's swing, the energy consumption of the control system, etc. By optimizing the performance index function, the system can achieve better performance while meeting control requirements.

[0063] The Riccati equation was solved based on multiple control sample data from the grab unloader. This control sample data contained operational information of the grab unloader under different working conditions, such as varying wire rope lengths, loads, and control inputs. By analyzing and processing this sample data, the Riccati equation was solved to obtain the feedback gain matrix that minimizes the performance index function. This feedback gain matrix is ​​a key parameter for achieving optimal control of the control system. Finally, a gain scheduling table was constructed based on the wire rope length and load sample data from each control sample data set, along with the corresponding feedback gain matrix. The gain scheduling table maps different combinations of wire rope lengths and loads to the corresponding feedback gain matrix, facilitating the subsequent determination of the target feedback gain matrix based on real-time operating conditions. For example, when the real-time wire rope length is 25 meters and the real-time load is 18 tons, the corresponding feedback gain matrix can be quickly found from the gain scheduling table for real-time control.

[0064] This application's embodiments construct and analyze a dynamic model of the grab unloader, enabling a deep understanding of the system's inherent characteristics and operational laws, providing a solid theoretical foundation for subsequent control design. By establishing a state-space expression and determining the Riccati equation, the complex control system problem is transformed into a mathematical problem, making the control design more scientific and precise. Defining a performance index function and solving the Riccati equation to obtain the optimal feedback gain matrix ensures that the control system achieves performance optimization while meeting actual needs, such as effectively reducing the grab bucket's swing amplitude and lowering energy consumption. Furthermore, constructing a gain scheduling table provides the control system with the ability to adapt to different operating conditions. In actual operation, the length of the grab unloader's wire rope and the load constantly change. The gain scheduling table allows for the rapid and accurate acquisition of a feedback gain matrix matching the current operating conditions, enabling the control system to adjust its control strategy in a timely manner and maintain good control performance. This improves the grab unloader's operational efficiency, stability, and safety, and reduces the risk of equipment failure and accidents.

[0065] In this embodiment of the application, optionally, after obtaining the feedback gain matrix that minimizes the value of the performance index function, the method further includes: performing a weighted summation calculation on the convergence time, overshoot, hanger swing angle, and control input energy of the control system based on corresponding weights to define a fitness function; using a particle swarm optimization algorithm with the fitness function as the optimization objective, optimizing the control input weight matrix R and the state weight matrix Q in the LQR controller, and updating the feedback gain matrix based on the optimized control input weight matrix R and the state weight matrix Q.

[0066] In this embodiment, after obtaining the feedback gain matrix that minimizes the performance index function value, the fitness function is first defined by weighting and summing the convergence time, overshoot, spreader swing angle, and control input energy of the control system based on corresponding weights. Convergence time reflects the time required for the control system to reach a steady state from its initial state; overshoot reflects the degree to which the system output exceeds the target value; spreader swing angle is a key controllable indicator in grab unloader operations; and control input energy relates to the system's energy consumption. For example, in grab unloader operations, excessively long convergence time reduces operational efficiency; excessive overshoot may cause the grab to collide with surrounding objects; excessively large spreader swing angle affects the accuracy of cargo loading and unloading; and excessively high control input energy increases operating costs. By assigning different weights to these four indicators, they are combined to form the fitness function, which comprehensively measures the overall performance of the control system. Next, the particle swarm optimization algorithm is used to optimize the control input weight matrix R and state weight matrix Q in the LQR controller, with the fitness function as the optimization objective. The LQR controller (Linear Quadratic Regulator) is a classic optimal control method widely used in automatic control, robotics, aerospace, vehicle engineering, and process control. Its core idea is to design a state feedback control law within a linear system model to make the system state quickly and smoothly approach the desired value, while achieving an optimal balance between control performance and control energy consumption. The particle swarm optimization algorithm finds the optimal solution by having a swarm of particles continuously search and update their positions in the solution space. In this process, each particle represents a set of values ​​for the control input weight matrix R and the state weight matrix Q. They continuously adjust their positions based on their own experience and the experience of the group, moving towards directions with smaller fitness function values. For example, initially, the particle swarm is randomly distributed in the solution space. After multiple iterations, the particles gradually gather in regions with smaller fitness function values, thus finding the optimal control input weight matrix R and state weight matrix Q. In one example, the fitness function is defined as... This is used to measure the overall performance of a control system. Among them, It represents the convergence time of the system, reflecting the time required for the grab to reach a stable state after a disturbance or change in command; Overshoot is the amount by which the system response exceeds the target value, and it is used to measure control stability. It is the time integral of the square of the grab bucket's swing angle, reflecting the degree of accumulation of swing energy during operation. The smaller the value, the better the anti-swing effect. The integral of the input energy is used to control the smoothness of the driving force (or motor torque) and energy consumption; while These are the weighting coefficients for each performance metric, used to balance the priorities of speed, stability, anti-sway performance, and energy consumption. The PSO algorithm uses this function as its optimization objective, continuously adjusting the weight matrix parameters of LQR to achieve a globally optimal balance between fast convergence, small overshoot, strong anti-sway capability, and low energy consumption. PSO is updated to... , ,in, Indicates the first The particle in the first The position vector at the next iteration, i.e., the current solution (such as the parameters of the LQR weight matrix); The velocity vector of the particle determines the search direction and step size; The inertia weight coefficient controls the particle's ability to maintain its original motion trend, and is used to balance global search and local convergence; , These are individual learning factors and group learning factors, which regulate the particle's trajectory towards its historical best position. and the global optimal position of the group The degree of closeness; , The random number in the interval [0,1] is used to increase the randomness of the search and the global exploration capability. Through the above iterative process, the particle swarm continuously adjusts its position and velocity in the multidimensional parameter space, eventually converging to the optimal solution. Thus, the optimal feedback gain matrix is ​​obtained. This is so that the original feedback gain matrix can be updated.

[0067] Finally, the feedback gain matrix is ​​updated based on the optimized control input weight matrix R and state weight matrix Q. Because the feedback gain matrix is ​​closely related to the control input weight matrix R and state weight matrix Q, changes in R and Q necessitate corresponding adjustments to ensure better performance of the control system under new parameters. This embodiment defines a fitness function that includes convergence time, overshoot, spreader swing angle, and control input energy, enabling a comprehensive evaluation of the control system's performance across multiple key dimensions, resulting in a more comprehensive and accurate assessment. The particle swarm optimization algorithm is used to optimize the control input weight matrix R and state weight matrix Q of the LQR controller. This algorithm possesses strong global search capabilities, quickly finding approximate optimal solutions in complex solution spaces. Compared to traditional optimization methods, it significantly improves optimization efficiency and reduces the time and computational resources required for optimization. Updating the feedback gain matrix based on the optimization results allows it to better adapt to various operating conditions in actual operations, further optimizing the control system's performance. In actual operation, this makes the control system of the grab unloader more stable and reliable, effectively reduces grab swing, shortens convergence time, reduces overshoot, and reduces control input energy consumption, thereby improving operating efficiency, reducing operating costs, and enhancing the safety and stability of the equipment.

[0068] In this embodiment of the application, optionally, determining the target control law based on the target feedback gain matrix, and determining the target control signal of the grab unloader based on the target control law and the real-time operating status information, includes: determining the target control law based on the target feedback gain matrix and the state vector corresponding to the real-time operating status information; obtaining the real-time disturbance estimate value estimated by the disturbance observer, and compensating the target control law based on the product of the difference between the real-time disturbance estimate value and the state vector corresponding to the real-time operating status information and the observation gain coefficient to obtain the feedback compensation amount; and determining the target control signal of the grab unloader based on the feedback compensation amount.

[0069] In this embodiment, the target feedback gain matrix reflects the system's response characteristics to control inputs under different wire rope lengths and loads. The state vector corresponding to the real-time operating status information contains many key state parameters of the grab unloader at the current moment, such as the grab's position, speed, and acceleration. By performing specific mathematical operations (e.g., matrix multiplication) on the target feedback gain matrix and the state vector, the target control law can be obtained. This target control law provides the basis for generating accurate control signals, determining the control actions the system should take under ideal, disturbance-free conditions. Further, the real-time disturbance estimate obtained through the disturbance observer is acquired, and the target control law is compensated accordingly. In actual operating environments, the grab unloader is subject to various disturbances, such as wind, uneven cargo distribution, and frictional changes in mechanical components. The disturbance observer's role is to estimate the magnitude and direction of these disturbances in real time, obtaining real-time disturbance estimates. Then, the difference between the real-time disturbance estimate and the state vector corresponding to the real-time operating status information is calculated, and this difference is multiplied by the observation gain coefficient to obtain the feedback compensation amount. The observation gain coefficient is preset based on the system characteristics and control requirements, determining the intensity of disturbance compensation. For example, if the wind is strong, the disturbance estimate obtained by the disturbance observer will also be large. Through calculation with the state vector difference and the observation gain coefficient, the resulting feedback compensation will increase accordingly to better offset the wind's influence on the grab bucket's sway. After further calculation and processing, the feedback compensation can generate the final target control signal. This target control signal will be sent to the actuators of the grab bucket unloader, such as motors, driving the grab bucket to move in the desired manner, achieving dynamic anti-sway control. For example, if the feedback compensation indicates a need to increase control intensity to suppress grab bucket sway, the target control signal will be adjusted accordingly, causing the motor to output greater torque, thereby stabilizing the grab bucket. This application's embodiment introduces a disturbance observer and feedback compensation mechanism, considering the existence of various disturbances in the actual operating environment. It can correct and compensate the target control law in real time, overcoming the impact of disturbances on system stability and improving the system's adaptability and robustness in complex environments. For example, when encountering sudden strong winds, the control strategy can be adjusted in time to prevent the grab bucket from swinging excessively due to the wind force.

[0070] In this embodiment of the application, optionally, as shown in the example... Figure 2 As shown, the method further includes:

[0071] Step 201: Obtain real-time image data of the grab unloader captured by video surveillance, identify the real-time spatial position, real-time swing posture and real-time three-dimensional shape of the target unloading area of ​​the grab bucket through visual algorithms, and generate the real-time operation path prediction trajectory of the grab unloader.

[0072] Step 202: Calculate the feedforward compensation amount of the grab unloader based on the current spatial location and the real-time operation path prediction trajectory;

[0073] Step 203: Determine the target control signal for the grab unloader based on the feedforward compensation amount and the feedback compensation amount.

[0074] In this embodiment, in addition to the feedback-based control method mentioned above, a feedforward control loop based on visual information is also introduced. First, real-time image data of the grab unloader is captured by video surveillance equipment. These devices are typically installed in suitable locations to comprehensively and clearly capture the operation of the grab unloader. Next, visual algorithms are used to process the acquired real-time image data. Visual algorithms possess powerful image recognition and analysis capabilities; they can accurately identify the real-time spatial position of the grab from complex images, such as determining the grab's coordinates in three-dimensional space; identifying the grab's real-time swing posture, such as its tilt angle, swing direction, and amplitude; and identifying the real-time three-dimensional shape of the target unloading area, including its size, shape, and height. Based on these identification results, a real-time operation path prediction trajectory for the grab unloader is generated. This prediction trajectory is like an ideal operation route pre-planned for the grab, taking into account the current state of the grab and the situation of the target unloading area. Then, based on the grab's spatial position at the current moment and the real-time operation path prediction trajectory, the feedforward compensation amount of the grab unloader is calculated. The feedforward compensation is calculated based on the deviation between the current position of the grab bucket and the pre-aiming trajectory. It predicts potential deviations in the grab bucket's subsequent movement and provides corresponding compensation. For example, if the pre-aiming trajectory indicates that the grab bucket needs to move to the upper left to accurately reach the unloading area, but the current grab bucket position is slightly to the lower right, then the feedforward compensation will include a control amount that adjusts to the upper left to guide the grab bucket in the correct direction. When determining the target control signal for the grab bucket unloader, it no longer relies solely on the feedback compensation, but considers both the feedforward and feedback compensation. The feedback compensation primarily corrects for disturbances and deviations that have already occurred during actual operation, while the feedforward compensation proactively prevents potential deviations. Combining the two, the final target control signal is determined. This target control signal is then sent to the grab bucket unloader's actuators, such as motors, to drive the grab bucket to move in the desired manner. Using both feedforward and feedback compensation to determine the target control signal fully leverages the advantages of both feedforward and feedback control, achieving complementary strengths. Feedforward control prevents deviations in advance, while feedback control corrects existing deviations. The synergistic effect of the two enables grab unloaders to achieve more precise and stable dynamic anti-sway control under various complex working conditions, improving the accuracy and efficiency of operations, reducing the risk of equipment failures and accidents, extending the service life of the equipment, and enhancing the safety and reliability of the entire operation process.

[0075] This application also provides a computer device, specifically a personal computer, server, network device, etc. The computer device includes a bus, processor, memory, and communication interface, and may also include input / output interfaces and a display device. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the computer device stores location information. The network interface of the computer device is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements the steps in the various method embodiments.

[0076] Those skilled in the art will understand that the structure of the computer device described above is only a partial structure related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. A specific computer device may include more or fewer components, or combine certain components, or have different component arrangements.

[0077] In one embodiment, a computer-readable storage medium is provided, which may be non-volatile or volatile, having stored thereon a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0078] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0079] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0080] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, graphics processors, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0081] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0082] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A dynamic anti-sway control method for a grab bucket ship unloader, characterized in that, The method includes: The system acquires real-time image data of the grab unloader captured by video surveillance, and uses visual algorithms to identify the real-time spatial position, real-time swing posture, and real-time three-dimensional shape of the target unloading area of ​​the grab bucket, generating a real-time operation path prediction trajectory for the grab unloader; based on the current spatial position and the real-time operation path prediction trajectory, the system calculates the feedforward compensation amount of the grab unloader. Real-time operating status information of the grab unloader is obtained by using multi-source sensors installed on the grab unloader. Using a gain scheduling table, the target feedback gain matrix corresponding to the grab unloader is determined based on the real-time wire rope length and real-time load in the real-time operating status information. The gain scheduling table includes the optimal feedback gain matrix corresponding to various combinations of wire rope length and load. The process of determining a target control law based on the target feedback gain matrix and determining a target control signal for the grab unloader based on the target control law and the real-time operating status information includes: determining the target control law based on the state vector corresponding to the target feedback gain matrix and the real-time operating status information; obtaining a real-time disturbance estimate estimated by a disturbance observer, and compensating the target control law by multiplying the difference between the real-time disturbance estimate and the state vector corresponding to the real-time operating status information by the observation gain coefficient to obtain a feedback compensation amount; and determining the target control signal for the grab unloader based on the feedforward compensation amount and the feedback compensation amount. The movement of the grab unloader is driven by the target control signal.

2. The method of claim 1, wherein, The multi-source sensors include a spreader end IMU, a wire rope length sensor, a hoisting speed encoder, a grab bucket position encoder, a grab bucket speed encoder, and a weighing sensor; the real-time operating status information includes real-time spreader swing angle, real-time spreader angular velocity, real-time wire rope length, real-time hoisting speed, real-time grab bucket position, real-time grab bucket speed, and real-time load.

3. The method of claim 1, wherein, Using the gain scheduling table, based on the real-time wire rope length and real-time load in the real-time operating status information, the target feedback gain matrix corresponding to the grab unloader is determined, including: Check if the gain scheduling table contains a combination that is the same as the real-time wire rope length and the real-time load; If included, the optimal feedback gain matrix corresponding to the combination of the same real-time wire rope length and the real-time load is obtained as the target feedback gain matrix; If not included, bilinear interpolation is performed on each combination in the gain scheduling table, and the target feedback gain matrix matching the real-time wire rope length and the real-time load is determined based on the bilinear interpolation result.

4. The method according to any one of claims 1 to 3, characterized in that, Before determining the target feedback gain matrix corresponding to the grab unloader based on the real-time wire rope length and real-time load in the real-time operating status information using the gain scheduling table, the process also includes: A dynamic model of the grab bucket ship unloader is constructed, and the Laplace transform is performed on the dynamic model to obtain the transfer function of the control system of the grab bucket ship unloader. A state-space expression is established based on the transfer function of the control system, and the Riccati equation is determined based on the state-space expression. Define the performance index function of the control system, and solve the Ricardi equation based on multiple control sample data of the grab unloader to obtain the feedback gain matrix that minimizes the value of the performance index function. Based on the wire rope length sample data and load sample data in each control sample data, as well as the corresponding feedback gain matrix, a gain scheduling table is constructed.

5. The method according to claim 4, characterized in that, The state-space expression is: A represents the system state matrix, B represents the input matrix, C represents the output matrix, D represents the direct transfer matrix, x represents the operating state information, f represents the driving force of the grab unloader, and y represents the output vector. Represent the state equation; The expression for the Riccati equation is: P represents a symmetric positive definite matrix, R represents the control input weight matrix, and Q represents the state weight matrix; determining feedback gain matrices R, Q and P by solving for R, Q and P that minimize the value of the performance index function .

6. The method of claim 5, wherein, After obtaining the feedback gain matrix that minimizes the value of the performance index function, the process further includes: The convergence time, overshoot, spreader swing angle, and control input energy of the control system are weighted and summed based on the corresponding weights to define the fitness function; The particle swarm optimization algorithm is used to optimize the control input weight matrix R and the state weight matrix Q in the LQR controller with the fitness function as the optimization objective, and the feedback gain matrix is ​​updated based on the optimized control input weight matrix R and state weight matrix Q.

7. A storage medium having stored thereon a computer program, characterized in that When the computer program is executed by a processor, it implements the method of any one of claims 1 to 6.

8. A computer device, comprising a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 6.

Images