An adaptive tension control device and method for winch take-up

Abstract

The application provides an adaptive tension control device and method for winch winding and unwinding. In the application, a rope passes through a guide ring in the device. When the rope has tension, the middle guide ring is subjected to pressure and moves radially on a linear guide. The displacement of the linear guide is collected by a pull rope displacement sensor, the tension is indirectly measured by the displacement, and the tension size is fed back to a winch controller. In combination with physical modeling and feedback information, a parameter identification algorithm based on recursive least squares is used to identify system parameters, a complete system mathematical model is established, and a model predictive control algorithm is used to control the winch speed in real time, that is, the parameter identification and predictive control algorithm, which dynamically adjusts the winding and unwinding of the cable and controls the tension on the rope.

Description

Technical Field

[0001] This invention belongs to the field of electromechanical control and relates to an adaptive tension control device and method for winch operation. Background Technology

[0002] When a ship-mounted winch tows and recovers a remotely operated underwater vehicle (ROV), the ship's heave or the impact of water currents on the towed body can cause severe fluctuations in cable tension. In complex sea conditions, due to communication and execution lags in the system, traditional feedback control methods struggle to adjust cable tension quickly and accurately. This can lead to mechanical malfunctions such as winch disc misalignment or cable seizing, and in severe cases, even cable breakage, resulting in the loss or damage of valuable ROV equipment and threatening the safety and reliability of the entire operation.

[0003] Furthermore, tension control technology has wide applications in winding mechanisms, offshore hoisting equipment, mine hoisting, and submarine cable laying. In these systems, the stability of tension affects the accuracy and safety of the entire system. Excessive tension can lead to rope breakage or motor overload; insufficient tension can cause slippage and slack, affecting motion accuracy and response speed. Currently, PID control and its improved methods are commonly used in engineering practice. This type of control strategy has a relatively simple structure, low implementation cost, and can achieve relatively ideal control results when system parameters do not change significantly and external disturbances are small. However, in practical applications, the following shortcomings still exist:

[0004] (1) The measurement process is relatively complex and costly: If a tension sensor is used for direct measurement, it will increase the complexity of the system structure and the cost of hardware; while the indirect measurement method based on motor current or output torque depends on an accurate system model. Once there is a deviation in the model parameters, additional errors will inevitably be introduced, affecting the control accuracy.

[0005] (2) Lack of adaptive adjustment capability: As the system is used for a longer period of time, parameters such as the friction coefficient of the motor and the elastic modulus of the rope will change, causing the dynamic characteristics of the system to deviate from the original design parameters. Therefore, using a fixed model and parameters in situations with large load changes or environmental disturbances will lead to a decrease in controller performance and make it easy to generate overshoot, oscillation or phase lag.

[0006] Precise control of cable tension is crucial for the safety of ROV towing and retrieval operations. Parameter identification algorithms and model predictive control (MPC) algorithms offer a new technical approach to tension control due to their advantages of adaptive adjustment of system model parameters and forward predictive control. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides an adaptive tension control device and method for winch operation.

[0008] An adaptive tension control device for winch winding and unwinding, as disclosed in this invention, is installed on the winch rope exit side and includes:

[0009] case;

[0010] The first fixed guide ring and the second fixed guide ring are respectively disposed at the left and right ends of the housing;

[0011] A linear guide rail is vertically installed inside the housing;

[0012] The slider slides in conjunction with the linear guide rail and can move in the vertical direction;

[0013] A movable guide ring is disposed on the slider and located between the first fixed guide ring and the second fixed guide ring. The first fixed guide ring, the movable guide ring and the second fixed guide ring together form an S-shaped rope path. The movable guide ring moves up and down along the linear guide rail as the rope tension changes.

[0014] An elastic reset mechanism, disposed between the slider and the housing, is used to apply an elastic force to the slider toward an equilibrium position; and

[0015] A displacement sensor is installed on the housing, and its detection end is connected to the slider. It is used to detect the displacement of the slider relative to the housing in real time and output a displacement signal. The displacement signal is used to feed back to the winch controller to adjust the winch linear speed so that the slider returns to the equilibrium position, thereby realizing constant tension control of the rope.

[0016] Furthermore, the elastic reset mechanism includes a compression spring and a guide rod. One end of the guide rod is fixed to the rear plate of the housing, and the other end passes through the central hole of the slider. The compression spring is fitted on the outside of the guide rod, with one end abutting against the rear plate of the housing and the other end abutting against the back of the slider.

[0017] Furthermore, the displacement sensor is a wire rope displacement sensor, with its steel wire rope rigidly connected to the slider via a connecting bracket.

[0018] Furthermore, it also includes a pulley guide pulley, which is located at the outlet of the wire rope of the pull rope displacement sensor, and is used to change the pull rope lead-out angle and prevent wear.

[0019] Furthermore, the displacement signal is fed back to the winch controller, which judges the change in external load based on the change in slider displacement and adjusts the speed of the winch motor accordingly to match the winch linear speed with the external speed, so as to pull the slider back to the equilibrium position.

[0020] An adaptive tension control method for winch operation according to the present invention, employing the above-mentioned device, includes the following steps:

[0021] The displacement signal of the slider relative to the shell is collected in real time by a displacement sensor, and the displacement signal serves as an indirect reflection of the change in rope tension.

[0022] The displacement signal is fed back to the winch controller, which determines the trend of external load change based on the change in the displacement signal.

[0023] Based on the displacement signal and winch control input, the system dynamic model parameters are estimated and updated online using a recursive least squares algorithm.

[0024] Based on the updated system dynamic model, model predictive control algorithms are used to predict the system state within a preset time domain in the future, and the optimal control input sequence is obtained by solving a constrained quadratic programming problem; and

[0025] The current control quantity in the optimal control input sequence is applied to the winch motor to dynamically adjust the winch line speed so that the winch line speed matches the external speed, thereby pulling the slider back to the equilibrium position and achieving constant tension control during the rope winding and unwinding process.

[0026] Furthermore, in the recursive least squares algorithm, the system parameters to be identified are parameter matrices containing motor time constants and system gain-related elements, and the regression vector includes the motor linear velocity, rope displacement difference, external velocity, and control input at the previous moment.

[0027] Furthermore, the performance metrics of the model predictive control algorithm include terminal cost, operating cost, and control input cost, which are obtained by weighting the terminal state weight matrix, the operating state weight matrix, and the control input weight matrix, respectively, and accumulating them in the prediction time domain.

[0028] Furthermore, the performance index is transformed into a standard quadratic programming problem with the control input sequence as the optimization variable. The quadratic term coefficient matrix is ​​composed of the transpose of the control prediction matrix, the state weight matrix, and the control prediction matrix, and is superimposed with the control weight matrix. The linear term coefficient matrix is ​​composed of the transpose of the control prediction matrix, the state weight matrix, and the state prediction matrix.

[0029] Furthermore, the constraints of the constrained quadratic programming problem include state variable constraints and control input constraints. These constraints are constructed by the state variable constraint coefficient matrix, the control input constraint coefficient matrix, and the boundary conditions of the state and control inputs, forming a linear inequality constraint on the control input sequence.

[0030] The beneficial effects of this invention are:

[0031] This invention integrates rope path guidance and tension detection into one unit. Through an S-shaped rope path structure of "three guide rings – intermediate movement," changes in rope tension are converted into linear displacement of a slider, which is then collected using a rope displacement sensor. This solution eliminates the need for additional, expensive tension sensors, avoiding the structural complexity and increased hardware costs associated with direct measurement. It also eliminates the reliance on precise system models for indirect measurement based on motor current.

[0032] To address the issue of parameter drift in the motor friction coefficient and rope elastic modulus after long-term use of winches, this invention introduces a recursive least squares (RLS) online parameter identification algorithm to estimate and update the system's dynamic model parameters in real time. The controller can dynamically adjust the control strategy based on the identification results, overcoming the performance degradation problem caused by fixed models and parameters in traditional PID control, and significantly enhancing the system's control robustness throughout its entire lifecycle.

[0033] This invention employs a model predictive control (MPC) algorithm, which uses the identified system model to predict future states and obtains the optimal control input by solving a constrained quadratic programming problem. This strategy not only achieves forward-looking optimization of motor speed but also explicitly handles physical constraints such as motor speed, torque, and rope tension, achieving a balance between tension stability, response speed, and energy consumption, effectively suppressing overshoot and oscillation.

[0034] This invention uses slider displacement as an indirect feedback quantity for tension changes and dynamically adjusts the winch line speed through a "speed matching" strategy to keep it consistent with the line speed of external equipment or operating conditions. When external load changes abruptly or the ship's heave or sway causes tension fluctuations, the system can quickly identify the slider displacement trend and adjust the rotation speed in advance to bring the slider back to the equilibrium position, thereby maintaining constant rope tension and avoiding mechanical failures such as rope misalignment, rope biting, and rope breakage. Attached Figure Description

[0035] Figure 1 Exploded view of the winch tension control device for winding and releasing;

[0036] Figure 2 Front view of the winch tension control device for winding and releasing;

[0037] Figure 3 Rear view of the winch tension control device for winding and releasing;

[0038] Figure 4 Working analysis diagram of the winch retraction and release constant tension control device;

[0039] Figure 5 Flowchart of parameter identification algorithm for winch retraction and release constant tension control system;

[0040] Figure 6 Flowchart of the MPC algorithm for the winch's constant tension control system;

[0041] Figure 7 Overall flowchart of the winch tension control system. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail with reference to the accompanying drawings. This description illustrates specific embodiments consistent with the principles of the present invention by way of example rather than limitation. These embodiments are described in sufficient detail to enable those skilled in the art to practice the invention, use other embodiments, and modify and / or substitute the structure of various elements without departing from the scope and spirit of the invention. Therefore, the following detailed description should not be construed as limiting.

[0043] like Figures 1 to 3 As shown, the present invention provides an adaptive tension control device for winch winding and unwinding. The device is an integrated rope guide and tension detection assembly installed on the rope exit side of the winch.

[0044] The device as a whole includes a cover plate 01, a housing / mounting base 02, a fixed guide ring A03, a fixed guide ring B04, a movable guide ring 05, a linear guide rail and slider assembly 06, a spring guide rod 07, a spring 08, a pull rope displacement sensor 09, a connecting bracket 10, a pull rope guide pulley 11, a bracket pad 12, and a first fastener 13 and a second fastener 14.

[0045] The housing 01 is a cuboid structure, fixed to the front of the housing / mounting base 02, which is fixed to the winch frame. A linear guide rail and a slider assembly 06 are located on the rear side inside the housing, with the slider moving vertically along the guide rail. Fixed guide rings A03 and B04 are fixedly installed on both sides of the upper end of the housing, respectively. These two guide rings are located on the same horizontal plane and are used to guide the direction of rope entry and exit; the guide rings have plastic flared openings. A movable guide ring 05 is installed at the front end of the slider assembly, forming an "up-down-up" S-shaped rope path structure together with the two fixed guide rings.

[0046] A spring guide rod 07 is mounted on the back of the slider assembly 06. One end of the guide rod is fixed to the rear plate of the housing, and the other end passes through the center hole of the slider. The spring 08 is fitted onto the outside of the guide rod 07, with one end abutting against the rear plate of the housing and the other end abutting against the back of the slider. This structure ensures that the spring is subjected to uniform force and compressed axially during operation, preventing the spring from deflecting or jamming, thereby ensuring smooth movement of the slider.

[0047] The spring is replaceable to set different initial preloads. A cable displacement sensor 09 is installed on the outside of the housing, with its wire rope end fixed to the slider assembly via a connecting bracket 10. The output signal of the cable sensor is proportional to the slider displacement, used to detect changes in rope tension and feed it back to the winch control system. A cable guide pulley 11 is provided at the outlet of the cable displacement sensor's wire rope to change the cable lead-out angle and prevent wear. A bracket pad 12 is provided at the mounting end of the cable guide pulley 11 for positioning and distributing installation stress. A first fastener 13 is used to assemble the cable guide pulley 11, and a second fastener 14 is used for the fixed connection between the sensor rope end and the connecting bracket 10.

[0048] The rope of this device enters from the left fixed guide ring A03, winds down around the moving guide ring 05, rises, and exits through the right fixed guide ring B04. During normal operation, the slider assembly 06 is held in the neutral position by the preload of the spring 08. If the external load increases, causing the rope tension to rise, the slider moves upward and compresses the spring; when the load decreases, the spring releases energy, pushing the slider downward. The displacement sensor 09 detects the slider displacement in real time and feeds the signal back to the control system. Based on this, the control system automatically adjusts the winch speed to return the slider to the equilibrium position, achieving constant tension control.

[0049] like Figure 4 As shown, the device is used as follows:

[0050] Step 101: Initial state. The horizontal distance between fixed guide ring A03 and fixed guide ring B04 is 2D, and the vertical distance between moving guide ring 05 and the line connecting them is... The device completes pre-tensioning and zero-point calibration at this geometric position, with the slider in the neutral position and the tension at the set value. Subsequent control revolves around "returning the slider to this neutral position," at which point the tension on the rope... for:

[0051]

[0052] in The elastic force of the spring, The original length of the spring. This is the current length of the spring. The spring constant is... The angle between the rope and the vertical direction.

[0053] Step 102: Tension State. When the external load or resistance increases, the rope is straightened, and the moving guide ring moves upward with the slider, reducing sag. The displacement sensor will give a "moving upward and faster" signal before the drum is subjected to force. The controller interprets this as an indication that "the external force is pulling the rope faster," and immediately increases the winch's rope release speed (or decreases the rope take-up speed) to rematch the winch's linear speed with the external speed. Once the speeds match, the tension naturally decreases, and the slider is pulled back to the neutral position. At this point, the tension on the rope is:

[0054]

[0055] in The elastic force of the spring in the stretched state. The angle between the rope in its stretched state and the vertical direction.

[0056] Step 103: Retrieval Status. As the load decreases or the external speed slows, the rope sags more, and the guide ring moves downwards. The displacement signal is displayed as "downward / slowing down." The controller determines that the external demand on the rope has decreased and immediately reduces the winch's rope release speed or switches to slight rope retraction to prevent slack. Similarly, through speed matching, the slider gradually returns to the neutral position, and the tension returns to the set value. At this point, the tension on the rope is:

[0057]

[0058] in To recover the spring force in the recovery state, The angle between the rope and the vertical direction in the recovery state.

[0059] when At that time, the tension on the rope remained essentially unchanged.

[0060] External load changes are first reflected in rope tension changes, which are then amplified and detected by the device through slider displacement. Instead of directly measuring tension, the system uses a rope displacement sensor to collect slider displacement signals in real time and uses this as an indirect measure of tension change. Through this "speed matching-tension balancing" method, the winch speed is dynamically adjusted to achieve constant tension control of the rope during the winding and unwinding process.

[0061] When the external load increases, causing the rope tension to rise, the slider assembly moves upward under the force, compressing the spring. The rope displacement sensor detects this increased slider displacement, and the control system determines that the external speed or resistance has increased. It then increases the winch motor speed to realign the winch's rope release speed with the external speed, thereby reducing the instantaneous tension. When the load decreases and the slider moves downward, the system recognizes this as a decrease in external speed or weakening tension, and the controller automatically reduces the winch speed to prevent rope slack.

[0062] The system indirectly measures tension by collecting displacement data through a rope displacement sensor. It can accurately sense the force state of the rope and feed it back to the controller without the need for an additional tension sensor. The system then corrects the winch output speed in real time to keep it consistent with the linear speed of external equipment or operating conditions, ultimately achieving stable tension control.

[0063] The rope passes through three guide rings A, B, and C, forming an isosceles triangle structure with each guide ring as its vertex. When the external rope speed and the winch line speed are mismatched, the rope tension will change; when the tension exceeds the initial preload, the rope will move upwards relative to its initial position in the device, reducing its total length within the device, raising the middle guide ring C, and thus increasing the displacement sensor reading. Decrease.

[0064] The tension of the rope is:

[0065]

[0066] in: The elastic force of the spring, The original length of the spring. The spring constant is... The angle between the rope and the vertical direction. This is the displacement measured by the current displacement sensor.

[0067] Set the rope reference height The horizontal distance between guide rings A and B is 2. Then the rope displacement difference This can be obtained from geometric relationships:

[0068]

[0069] in The total change in length of the rope within the device during a dynamic process is the control objective, which aims to maintain stable tension on the rope. When the value approaches 0, it means that the length of the rope within the device hardly changes. There was basically no change, that is, the tension on the rope remained the same. It's stable now.

[0070] To describe the dynamic mathematical relationships of the system, this invention uses state-space expressions to establish a mathematical model, starting with motor speed control.

[0071] Analyzing the speed control of the winch, the dynamic response of the winch can be approximately described as a first-order inertial element, i.e.:

[0072]

[0073] in:

[0074] : Motor linear velocity;

[0075] : Winch control input signal;

[0076] Motor time constant;

[0077] System gain.

[0078] According to motor speed rope displacement difference Speed ​​of external pull rope Based on the coupling relationship between them, the augmented state-space equations are established:

[0079]

[0080] in:

[0081]

[0082] The output is:

[0083]

[0084] This method does not directly use forces, which require complex calculations, as state variables; instead, it selects the rope displacement difference. It reflects the responsiveness of the winch's take-up and release speed to external disturbances, and is simple and intuitive.

[0085] Let the sampling time be After the system is discretized, time:

[0086]

[0087]

[0088] in:

[0089]

[0090] As the system is used for an extended period, parameters such as the motor friction coefficient and the rope elastic modulus will change, causing the system's dynamic characteristics to deviate from the original design parameters. To address this issue, this invention employs a recursive least squares (RLS) parameter identification algorithm, an online parameter identification method, to estimate the system's dynamic model parameters in real time, enabling the controller to adjust its control strategy based on the new parameters.

[0091] Through the RLS algorithm, the dynamic parameters of the system can be updated in real time, enabling the controller to dynamically adapt to external disturbances and changes in system state in practical applications.

[0092] The mathematical form of the RLS algorithm in this system is:

[0093]

[0094] in:

[0095] The system parameters at the current moment are set as follows:

[0096]

[0097] The input vector at the current moment, also called the regression vector, includes the previous system output and control input, and in this system, it is:

[0098]

[0099] The current output of the system is the rope displacement difference in this system. ;

[0100] Gain vector, defined as:

[0101]

[0102] The covariance matrix controls the accuracy of the model. It needs to be updated in real time. The update formula is:

[0103]

[0104] Forgetting factor: Used to control the degree of influence of historical data.

[0105] In the tension control system of this invention, Restricted Scale (RLS) is used to update the dynamic model of winch tension control in real time. Specifically, the RLS algorithm continuously updates the system parameters. When the calculation converges, stable estimated parameters can be obtained, which correspond to the system parameter matrix after system modeling, i.e.:

[0106]

[0107] It can be calculated Motor time constant, The real-time estimation of system gain forms an online adaptive model. This model provides accurate prediction basis for subsequent MPC control.

[0108] like Figure 5 As shown, the main steps of the parameter identification algorithm are as follows:

[0109] Step 201: Define the state space of the system. First, perform state space modeling on the system, selecting four state variables A, B, C, and D, and then establish the mathematical model of the system based on these variables.

[0110] Step 202: Determine the number of parameters and regression vector. Based on the model, determine the number of parameters to be identified and construct the corresponding regression vector to provide a basis for subsequent calculations.

[0111] Step 203: Calculate the gain vector. Calculate the system gain vector using the input and output data.

[0112] Step 204: Perform parameter estimation. Using the established model and gain vector, perform parameter estimation operations to obtain the estimated values ​​of the current parameters.

[0113] Step 205: Update the covariance matrix. Based on the new parameter estimation results, update the covariance matrix to improve the accuracy and stability of subsequent calculations.

[0114] Step 206: Recursive Loop. Repeat the gain calculation and parameter update process until the parameter estimation results converge.

[0115] Step 207: Calculate the new state space. After parameter identification is completed, recalculate the system's state space model to obtain a more accurate system description, thereby improving control performance.

[0116] Model-based control (MPC) is a control method whose core idea is to use a system model to predict the system's behavior over a future period and optimize the control input to bring the system to its optimal state in the future. In winch tension control systems, MPC is used to optimize the motor speed, thereby precisely controlling the rope's winding and unwinding speed and maintaining stable tension.

[0117] First, it is necessary to establish system performance metrics. At any time, performance indicators As shown in the following formula:

[0118]

[0119] in:

[0120]

[0121]

[0122]

[0123] The weight matrix representing the terminal cost of the system The weight matrix represents the operating cost. The weight matrix represents the cost of control. These represent the weights of the corresponding dimensions.

[0124] The next step is to transform the performance metrics into a quadratic programming problem:

[0125] In model predictive control, This represents the prediction interval. At that moment, the system will predict Within, state variables From to The changing trend, among which The predicted value of the state variable at time t is:

[0126]

[0127] pass arrive The recursive relationship at each moment can be summarized into an equation:

[0128]

[0129] in:

[0130]

[0131]

[0132]

[0133]

[0134] for Vector, containing The state within all prediction intervals at each time point Vector, representing The state vector sequence calculated at each time step, It is a sequence of control input quantities that can be obtained through calculation. This is the state prediction matrix. To control the prediction matrix.

[0135] After calculation, the performance indicators can be... use express:

[0136]

[0137] in:

[0138] ,

[0139] ,

[0140] It is a standard quadratic programming problem. By solving the quadratic programming problem, the optimal control sequence can be found. In practical applications, the first number in the control quantity sequence can be directly taken. ,in, The coefficient matrix of the quadratic term in the quadratic programming objective function, also known as the Hessian matrix, is derived from the control prediction matrix. State weight matrix and control weight matrix Together, they form a structure used to describe the control input sequence. Secondary impact on performance indicators. This is a linear coefficient matrix used to describe the current state. With future control input sequence The coupling relationship between them.

[0141] To handle various constraint problems, further mathematical processing is required:

[0142] System constraints can be expressed by the following formula:

[0143]

[0144]

[0145] This formula constrains the combination of state variables and control inputs at each time step, where:

[0146] State variable constraint coefficient matrix, used to describe state variables Its role in constraints;

[0147] : Control input constraint coefficient matrix, used to describe the control input Its role in constraints;

[0148] : Constraint upper bound vector.

[0149] The above formula can be further simplified to the following formula:

[0150]

[0151] This represents the coefficient matrix of the state variables in the constraints. This represents the coefficient matrix of the control inputs within the constraints. This represents the upper bound vector of the state variables and control input constraints. This adds constraints to the quadratic programming problem. In this invention, the calculation method is as follows:

[0152] For system state variables Input variables , The number of state variables is 3. The number of input variables is 1; the constraints are as follows:

[0153]

[0154]

[0155] but:

[0156]

[0157] in:

[0158] ,

[0159]

[0160]

[0161]

[0162]

[0163]

[0164]

[0165]

[0166]

[0167]

[0168] , and It is a stacked matrix constructed to unify all constraints in the prediction time domain into a matrix form.

[0169] By setting an appropriate cost matrix, solving the constrained quadratic programming problem, and continuously executing the algorithm and performing rolling optimization, the optimal control output can be obtained, controlling the winch to follow the rope's release and retraction, thus achieving constant tension control.

[0170] like Figure 6 As shown, the main processing steps of the MPC algorithm are as follows:

[0171] Step 301: Parameter Identification and State Space Modeling. Based on the parameter identification results, establish the system state space model and determine the corresponding state matrix.

[0172] Step 302: Transform the performance indicators into a quadratic programming problem. Based on the control objectives (such as tension stability, response speed, and energy consumption requirements), the performance indicators are transformed into a standard quadratic programming problem in order to solve for the optimal control strategy.

[0173] Step 303: Handle constraints. Process various constraints during system operation, including range limitations on input, output, and state variables, such as motor speed, torque, and wire rope tension, to ensure that the control results meet the requirements of actual operating conditions.

[0174] Step 304: Solve the optimization problem. An online optimization algorithm is used to solve the above quadratic programming problem to obtain the optimal control quantity at the current time step.

[0175] Step 305: Apply the optimal control quantity to the system and obtain feedback information through sensors to update the system status in real time.

[0176] Step 306: Iterative Execution and Rolling Optimization. Repeat the above process within each control cycle, continuously refining the prediction results through rolling optimization, enabling the system to respond promptly to external disturbances and changes in operating conditions.

[0177] like Figure 7 As shown, the main steps of the overall process of the winch retraction and release constant tension control system are as follows:

[0178] Step 401: System Modeling. First, establish an initial system model. This model comprehensively considers known and unknown parameters and reflects the basic dynamic characteristics of the winch system.

[0179] Step 402: Parameter Identification. Based on this, parameter identification is performed using the input and output data to identify unknown parameters in the system, thereby obtaining a more accurate model.

[0180] Step 403: Model Predictive Control (MPC). Based on the identified model, a model predictive control method is introduced to predict the future state of the system and obtain the optimal control input through optimization calculation, so that the tension output meets the expected target.

[0181] Step 404: Status / Output Feedback. Based on status or output feedback information, the control input is dynamically corrected to ensure system stability and control accuracy.

[0182] In this winch constant tension control system, the recursive least squares parameter identification method is combined with model predictive control. By identifying the dynamic characteristics of the system online and optimizing the control input in real time, problems such as system nonlinearity, time-varying parameters, and external disturbances are effectively addressed, thereby improving the overall control performance and tension control accuracy.

[0183] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. An adaptive tension control device for winch operation, installed on the winch rope exit side, characterized in that, include: case; The first fixed guide ring and the second fixed guide ring are respectively disposed at the left and right ends of the housing; A linear guide rail is vertically installed inside the housing; The slider slides in conjunction with the linear guide rail and can move in the vertical direction; A movable guide ring is disposed on the slider and located between the first fixed guide ring and the second fixed guide ring. The first fixed guide ring, the movable guide ring and the second fixed guide ring together form an S-shaped rope path. The movable guide ring moves up and down along the linear guide rail as the rope tension changes. An elastic reset mechanism is disposed between the slider and the housing, and is used to apply an elastic force to the slider toward the equilibrium position; as well as A displacement sensor is installed on the housing, and its detection end is connected to the slider. It is used to detect the displacement of the slider relative to the housing in real time and output a displacement signal. The displacement signal is used to feed back to the winch controller to adjust the winch linear speed so that the slider returns to the equilibrium position, thereby realizing constant tension control of the rope.

2. The apparatus according to claim 1, characterized in that, The elastic reset mechanism includes a compression spring and a guide rod. One end of the guide rod is fixed to the rear plate of the housing, and the other end passes through the center hole of the slider. The compression spring is fitted on the outside of the guide rod, with one end abutting against the rear plate of the housing and the other end abutting against the back of the slider.

3. The apparatus according to claim 1 or 2, characterized in that, The displacement sensor is a wire rope displacement sensor, and its steel wire rope is rigidly connected to the slider via a connecting bracket.

4. The apparatus according to claim 3, characterized in that, It also includes a pulley guide pulley, which is located at the outlet of the wire rope of the pull rope displacement sensor, and is used to change the pull rope lead-out angle and prevent wear.

5. The apparatus according to claim 1, characterized in that, The displacement signal is fed back to the winch controller, which judges the change in external load based on the change in slider displacement and adjusts the speed of the winch motor accordingly to match the winch linear speed with the external speed, so as to pull the slider back to the equilibrium position.

6. An adaptive tension control method for winch operation, employing the device described in any one of claims 1-5, characterized in that, Includes the following steps: The displacement signal of the slider relative to the shell is collected in real time by a displacement sensor, and the displacement signal serves as an indirect reflection of the change in rope tension. The displacement signal is fed back to the winch controller, which determines the trend of external load change based on the change in the displacement signal. Based on the displacement signal and winch control input, the system dynamic model parameters are estimated and updated online using a recursive least squares algorithm. Based on the updated system dynamic model, the model predictive control algorithm is used to predict the system state in the future preset time domain, and the optimal control input sequence is obtained by solving a constrained quadratic programming problem. as well as The current control quantity in the optimal control input sequence is applied to the winch motor to dynamically adjust the winch line speed so that the winch line speed matches the external speed, thereby pulling the slider back to the equilibrium position and achieving constant tension control during the rope winding and unwinding process.

7. The method according to claim 6, characterized in that, In the recursive least squares algorithm, the system parameters to be identified are parameter matrices containing motor time constants and system gain-related elements, and the regression vector includes the motor linear velocity, rope displacement difference, external velocity, and control input at the previous moment.

8. The method according to claim 6, characterized in that, The performance metrics of the model predictive control algorithm include terminal cost, running cost, and control input cost, which are obtained by weighting the terminal state weight matrix, running state weight matrix, and control input weight matrix, respectively, and accumulating them in the prediction time domain.

9. The method according to claim 8, characterized in that, The performance index is transformed into a standard quadratic programming problem with the control input sequence as the optimization variable. The quadratic term coefficient matrix is ​​composed of the transpose of the control prediction matrix, the state weight matrix, and the control prediction matrix, and the control weight matrix is ​​superimposed on it. The linear term coefficient matrix is ​​composed of the transpose of the control prediction matrix, the state weight matrix, and the state prediction matrix.

10. The method according to claim 8 or 9, characterized in that, The constraints of the constrained quadratic programming problem include state variable constraints and control input constraints. The constraints are constructed by the state variable constraint coefficient matrix, the control input constraint coefficient matrix, and the boundary conditions of the state and control inputs, forming a linear inequality constraint on the control input sequence.

Images