Anti-swing intelligent lifting trolley hoisting system and device thereof

Abstract

The application relates to the technical field of intelligent hoisting technology, and discloses an anti-swing intelligent hoisting and carrying system and device thereof, which is applied to a hoisting and carrying vehicle with a large vehicle and a small vehicle. The system comprises a parameter real-time estimation unit, which is used for online estimation of load mass, wire rope length and the change rate thereof through an extended Kalman filtering algorithm based on sensor measurement values, and outputs state estimation values; and an optimal feedforward trajectory planning unit, which is used for solving an acceleration trajectory with minimum swing energy through a variation method as a feedforward control reference according to a target position and a rope length change plan. Double-axis swing synchronous suppression under the conditions of continuous rope length change and unknown load mass is realized. Load mass, rope length and the change rate thereof are estimated in real time through the extended Kalman filtering, optimal feedforward trajectory planning and adaptive time-varying frequency input shaping are combined through the variation method, and the shaper is always aligned with the instantaneous natural frequency of the system.

Description

Technical Field

[0001] This invention relates to the field of intelligent hoisting technology, specifically to an anti-sway intelligent crane hoisting system and device. Background Technology

[0002] Overhead cranes are widely used in ports, workshops, warehouses, and other locations for lifting heavy objects. During the lifting process, the starting, acceleration, deceleration, and braking actions of the trolley and hoist inevitably induce load swaying on the wire rope; simultaneously, the lifting or lowering actions of the winch mechanism further exacerbate the swaying due to the Coriolis effect. This longitudinal and lateral load swaying not only prolongs the lifting cycle and reduces positioning accuracy, but may also lead to safety accidents such as disengagement and collisions. Therefore, anti-sway control has always been an important research topic in the field of overhead cranes.

[0003] While existing mechanical anti-sway devices can suppress swaying to some extent, they are complex in structure, costly to maintain, and have limited adaptability to changes in rope length and load weight. Their anti-sway effect significantly decreases when the hoisting height or load weight changes. Electronic anti-sway methods rely on precise knowledge of the system's natural frequency; however, under conditions of continuously changing rope length, the system's natural frequency drifts in real time, and a controller with fixed parameters cannot track this change, leading to residual swaying or even instability. Furthermore, existing solutions generally treat crane acceleration / deceleration anti-swaying separately from winch lifting / lowering anti-swaying, neglecting the mechanism by which the Coriolis force injects energy into the swaying system and causes sway angle divergence when the winch speed is too high, and lacking active coordination constraints on the lifting action. On the other hand, many advanced anti-sway algorithms are computationally intensive, making it difficult to simultaneously meet the requirements of high precision and real-time performance on embedded controllers, thus limiting their engineering applications. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an anti-sway intelligent crane hoisting system and device, which solves the problem of synchronously suppressing the longitudinal and lateral swaying of the load during crane hoisting under conditions where the rope length changes continuously and the load mass is unknown.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an anti-sway intelligent crane hoisting system, applicable to cranes with a main trolley and a secondary trolley, the system comprising:

[0006] The real-time parameter estimation unit is used to estimate the load mass, wire rope length and its rate of change online based on sensor measurements using an extended Kalman filter algorithm, and output the state estimate.

[0007] The optimal feedforward trajectory planning unit is used to solve the acceleration trajectory with minimum swing energy using the variational method based on the target position and rope length change plan, and serves as the feedforward control reference.

[0008] An adaptive time-varying frequency input shaping unit is used to calculate the instantaneous natural frequency based on the real-time estimated rope length, dynamically adjust the pulse time interval of the input shaper, filter the feedforward acceleration trajectory, and generate the shaped desired acceleration command.

[0009] The model reference adaptive control unit is used to establish a swing-free reference model. Based on the error between the actual state and the reference model, and combined with the swing angle and angular velocity in the state estimate, it calculates the driving force of the large vehicle and the small vehicle, and updates the load mass estimate online through an adaptive law.

[0010] The hoisting speed coordination constraint unit is used to calculate the upper limit of the hoisting speed in real time based on the current lateral swing angle, longitudinal swing angle and angular velocity, and to output the hoisting speed command with a limited amplitude.

[0011] The drive execution unit is used to control the motion according to the driving force and the hoisting speed command after the amplitude limit.

[0012] Preferably, the extended Kalman filter algorithm in the real-time parameter estimation unit adopts an eleven-dimensional augmented state vector, including the lateral displacement and velocity of the main vehicle, the longitudinal displacement and velocity of the trolley, the lateral swing angle and its angular velocity, the longitudinal swing angle and its angular velocity, and the load mass; the measurement vector includes the displacement of the main vehicle, the displacement of the trolley, the lateral swing angle, and the longitudinal swing angle; the filtering period is 10 milliseconds.

[0013] Preferably, the optimal feedforward trajectory planning unit uses a variational method to solve the two-point boundary value problem. The performance index is the weighted integral of the square of the swing angle and the square of the control acceleration. The optimal acceleration curve is obtained by the shooting method, and the B-spline coefficients for typical working conditions are stored offline and called online by interpolation.

[0014] Preferably, the adaptive time-varying frequency input shaping unit adopts a three-pulse ZVDD shaper, whose pulse amplitude is a fixed constant and whose pulse time interval is calculated in real time according to the instantaneous natural frequency, wherein the instantaneous natural frequency is equal to the square root of the gravitational acceleration divided by the real-time rope length; the shaped acceleration command is equal to the sum of the products of each pulse amplitude and the feedforward acceleration value at the corresponding delay time.

[0015] Preferably, the reference model in the model reference adaptive control unit is a second-order system with a damping ratio of 0.7 and a natural frequency of π radians per second. The driving force calculation formula of the unit includes a position error feedback term, a velocity error feedback term, a lateral swing angle feedback term, a lateral swing angle velocity feedback term, and a load mass estimate. The adaptive law of the unit is that the rate of change of the load mass estimate is equal to the negative adaptive gain multiplied by the error-swing angle coupling term, wherein the error-swing angle coupling term includes the lateral error multiplied by the lateral velocity error plus the lateral swing angle multiplied by the lateral swing angle velocity, plus a longitudinal term of the same kind.

[0016] Preferably, in the winch speed coordination constraint unit, the maximum allowable winch speed for lateral swing is equal to the rope length multiplied by the absolute value of the lateral swing angular velocity, multiplied by the maximum acceleration limit, divided by the rope length multiplied by the square of the lateral swing angular velocity, plus the gravitational acceleration multiplied by the square of the lateral swing angle, plus a decimal value to prevent zeroing; the maximum allowable winch speed for longitudinal swing is calculated similarly, only the lateral swing angle and angular velocity are replaced by the longitudinal swing angle and angular velocity; the final upper limit of the winch speed is the minimum value of the two and the mechanical speed limit.

[0017] Furthermore, the present invention also provides an anti-sway intelligent crane hoisting device, comprising:

[0018] Two parallel tracks are fixedly laid on the ground;

[0019] The trolley traveling mechanism is installed on the track and can move back and forth along the track. The trolley traveling mechanism is driven by a variable frequency motor to achieve lateral movement.

[0020] The main frame is fixedly supported on the traveling mechanism of the main vehicle.

[0021] The trolley traveling mechanism is mounted on the main frame and can reciprocate along the guide rails on the main frame. The trolley traveling mechanism is driven by a servo motor to achieve longitudinal movement.

[0022] The hoisting mechanism includes a double-folded drum fixed on the trolley traveling mechanism, a hoisting motor that drives the drum, and an encoder coaxially mounted with the double-folded drum. The surface of the double-folded drum is wound with steel wire rope. The hoisting mechanism is used to wind up and unwind the steel wire rope.

[0023] The lifting device is suspended below the trolley traveling mechanism by a steel wire rope. A dual-axis tilt sensor is fixedly installed on the lifting device to measure the lateral and longitudinal tilt angles of the lifting device.

[0024] The electrical control cabinet, mounted on the main frame, is used to house the control system components.

[0025] Preferably, a triaxial force sensor is also fixedly installed on the lifting device, the triaxial force sensor is connected in series between the wire rope and the lifting device, a laser rangefinder is installed on the trolley traveling mechanism and the trolley frame respectively, and a rotary encoder is installed on the trolley traveling mechanism.

[0026] Preferably, it also includes an industrial computer and a programmable automation controller, both of which are installed in an electrical control cabinet; the industrial computer is connected to the programmable automation controller, and the output of the programmable automation controller is respectively connected to the frequency converter of the trolley, the servo driver of the trolley and the frequency converter of the hoist motor.

[0027] Preferably, the hoisting mechanism is further provided with an electromagnetic brake, which is installed between the output shaft of the hoisting motor and the double-folded drum.

[0028] Working principle: After system startup, the real-time parameter estimation unit fuses sensor signals through extended Kalman filtering, outputting real-time state estimates such as rope length, swing angle, angular velocity, and load mass. When the operator issues a new target position, the optimal feedforward trajectory planning unit uses variational method to solve for the acceleration trajectory that minimizes swing energy based on the target displacement and rope length change plan. This trajectory enters the adaptive time-varying frequency input shaping unit, which calculates the instantaneous natural frequency of the system in real time based on the current rope length, dynamically adjusts the time interval of the shaping pulses, and filters the feedforward trajectory to eliminate frequency components that may excite swing. The shaped desired acceleration command is then sent to... The system incorporates an adaptive control unit that uses a sway-free second-order system as its reference model. Based on the error between the actual state and the reference model, as well as feedback on the sway angle and angular velocity, it calculates the driving force for the trolley and crane. Simultaneously, it uses an adaptive law to online correct the load mass estimate to compensate for parameter uncertainties. On the other hand, the hoisting speed coordination and constraint unit dynamically calculates the upper limit of the hoisting speed based on the real-time sway angle and angular velocity, limiting the hoisting command to prevent Coriolis force excitation. Finally, the drive execution unit outputs the driving force and the limited hoisting speed command to the drivers of the trolley, crane, and hoisting mechanism, achieving coordinated anti-sway control of trolley acceleration / deceleration and hoisting actions. Throughout this process, each unit executes cyclically with a period of 5 to 10 milliseconds, forming a closed-loop suppression chain of "feedforward planning—adaptive filtering—feedback compensation—hoisting constraint," ensuring that the longitudinal and lateral sway of the load is always controlled within a safe range.

[0029] This invention provides an anti-sway intelligent crane hoisting system and device. It has the following beneficial effects:

[0030] 1. This invention achieves synchronous suppression of dual-axis sway under conditions of continuous rope length variation and unknown load mass. By using extended Kalman filtering to estimate the load mass, rope length, and rate of change in real time, combined with variational optimal feedforward trajectory planning and adaptive time-varying frequency input shaping, the shaper is always aligned with the system's instantaneous natural frequency, solving the problem of failure of traditional fixed-parameter anti-sway methods when the rope length changes. Experiments show that, under conditions of rope length variation from 2 meters to 15 meters and unknown load mass, the maximum lateral and longitudinal sway angles can be controlled within 0.8 degrees, and the residual sway decays to below 0.2 degrees within 1.5 seconds after reaching the target, reducing the sway amplitude by approximately 80% compared to traditional methods.

[0031] 2. This invention actively suppresses Coriolis force excitation from an energy perspective by coordinating the hoisting speed constraint with the electromagnetic brake, preventing swing angle divergence during lifting or lowering. This constraint dynamically calculates the upper limit of the hoisting speed based on the real-time swing angle and its angular velocity. When the swing angle exceeds 3 degrees, the speed is automatically reduced and braking is triggered, achieving coordinated anti-swaying during lifting and trolley acceleration / deceleration. This significantly improves the safety of the hoisting process and avoids the risk of swing divergence caused by excessive hoisting speed.

[0032] 3. This invention adopts a hardware architecture that divides the tasks between an industrial computer and a programmable automation controller, balancing the computational power of complex algorithms with the real-time control response speed. The industrial computer undertakes computationally intensive tasks such as extended Kalman filtering, optimal trajectory planning, and adaptive shaping, while the programmable automation controller performs high-frequency feedback tasks such as model reference adaptive control and hoisting constraints. The two work together through a real-time industrial Ethernet network, enabling the system to maintain anti-sway accuracy while achieving a control cycle of up to 5 milliseconds, meeting the engineering requirements for rapid response of cranes. Attached Figure Description

[0033] Figure 1 This is a perspective view of the anti-sway intelligent crane hoisting device of the present invention;

[0034] Figure 2 This is a schematic diagram of the anti-sway intelligent crane hoisting device of the present invention;

[0035] Figure 3 This is a schematic diagram of the main frame in this invention;

[0036] Figure 4 This is a schematic diagram of the hoisting mechanism in this invention;

[0037] Figure 5 This is a diagram illustrating the hoisting mechanism in this invention;

[0038] Figure 6 For the present invention Figure 5 Enlarged view of point A in the image;

[0039] Figure 7 This is a diagram illustrating the architecture of the anti-sway intelligent crane hoisting system of the present invention.

[0040] The components include: 1. Track; 2. Trolley traveling mechanism; 3. Trolley frame; 4. Auxiliary trolley traveling mechanism; 5. Guide rail; 6. Hoisting mechanism; 7. Double-folded drum; 8. Hoisting motor; 9. Encoder 1; 10. Wire rope; 11. Lifting device; 12. Dual-axis tilt sensor; 13. Triaxial force sensor; 14. Laser rangefinder; 15. Rotary encoder; 16. Electrical control cabinet; 17. Industrial computer; 18. Programmable logic controller; 19. Electromagnetic brake. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Please see the appendix Figure 1 - Appendix Figure 7 This invention provides an anti-sway intelligent crane hoisting system, applicable to cranes with a main trolley and a trolley, comprising:

[0043] The real-time parameter estimation unit is used to estimate the load mass, wire rope length and its rate of change online based on sensor measurements using an extended Kalman filter algorithm, and output the state estimate.

[0044] The extended Kalman filter algorithm in the real-time parameter estimation unit uses an eleven-dimensional augmented state vector, including the lateral displacement and velocity of the main vehicle, the longitudinal displacement and velocity of the trolley, the lateral swing angle and its angular velocity, the longitudinal swing angle and its angular velocity, and the load mass; the measurement vector includes the displacement of the main vehicle, the displacement of the trolley, the lateral swing angle, and the longitudinal swing angle; the filtering period is 10 milliseconds.

[0045] Specifically, after system initialization, it is essential to acquire the load's swing state and the unknown load mass in real time to provide an accurate data foundation for subsequent adaptive shaping, model reference adaptive control, and hoisting constraints. Since measurement components such as tilt sensors and encoders inevitably introduce noise, and the load mass cannot be directly measured, this step employs an extended Kalman filter (EKF) to optimally estimate the system state. This estimator utilizes signals from the dual-axis tilt sensor 12 mounted on the lifting device 11, the laser rangefinder 14 or encoder 9 on the trolley and hoist, and the hoist encoder, combined with a dynamic model derived from the Lagrange equation, to recursively output estimated values ​​of rope length, swing angle, angular velocity, and load mass at each moment.

[0046] In this embodiment, the real-time parameter estimation unit runs within an industrial computer, with a filtering period of 10 milliseconds. First, an eleven-dimensional augmented state vector is defined:

[0047]

[0048] in, , The values ​​represent the lateral displacement of the trolley and the longitudinal displacement of the trolley (unit: meters). , The horizontal and vertical swing angles (in radians). This refers to the length of the wire rope (in meters). For the load mass (kg). The measurement vector is five-dimensional:

[0049]

[0050] These components are derived from a laser rangefinder, a tilt sensor, and a winch encoder, respectively. The state equations are determined by a dynamic model and include the oscillation equations:

[0051]

[0052] The system includes a longitudinally symmetrical form and the motion equations of the large and small vehicles. Discretization uses the fourth-order Runge-Kutta method with a sampling period Ts = 0.01 seconds. The prediction step calculates the prior state estimate and covariance. The process noise covariance Q is empirically set as a diagonal matrix (position variance 1e-4, velocity variance 1e-2, swing angle variance 1e-6, angular velocity variance 1e-4, rope length variance 1e-4, rope length change rate variance 1e-2, mass variance 100). The measurement noise covariance R is set based on sensor accuracy (displacement measurement variance 1e-6, swing angle measurement variance 1e-6, rope length measurement variance 2.5e-5). The measurement matrix H is a 5×11 constant matrix, used only in corresponding... , , , , The position is 1. The update step calculates the Kalman gain:

[0053]

[0054] Corrected state estimation:

[0055]

[0056] And update the covariance. In the initial state, the displacement and rope length are taken from the first reading of the sensor, the velocity is set to zero, the pendulum angle is taken from the first tilt angle reading, and the load mass estimate is set to half of the rated load; the mass uncertainty in the initial covariance is set to 1e6, and the rest are set according to the order of magnitude.

[0057] After filtering, the real-time parameter estimation unit outputs an estimated rope length. Rate of change of rope length Lateral swing angle estimation and its angular velocity Longitudinal swing angle estimation and its angular velocity Load quality estimation These data are transmitted via real-time industrial Ethernet to an adaptive time-varying frequency input shaping unit (used to calculate the instantaneous natural frequency). The model references an adaptive control unit (for swing angle feedback and quality updates in the adaptive law) and a hoisting speed coordination constraint unit (for calculating the upper limit of the hoisting speed). As an extension, outlier detection can be added to the filtering: when the innovation exceeds three times the measurement standard deviation, the R value of the corresponding dimension is temporarily amplified to suppress the effects of sensor interruptions or strong electromagnetic interference.

[0058] The optimal feedforward trajectory planning unit is used to solve the acceleration trajectory with minimum swing energy using the variational method based on the target position and rope length change plan, and serves as the feedforward control reference.

[0059] The optimal feedforward trajectory planning unit uses the variational method to solve the two-point boundary value problem. The performance index is the weighted integral of the square of the swing angle and the square of the control acceleration. The optimal acceleration curve is obtained by the shooting method, and the B-spline coefficients for typical working conditions are stored offline and called online by interpolation.

[0060] Specifically, the rope length and its rate of change information provided by the real-time parameter estimation unit lay the foundation for the generation of the feedforward trajectory. When the operator issues a new target position via a handle or host computer, the control system needs to plan an acceleration curve to minimize the load sway energy during the continuous change of rope length. If only subsequent feedback control is relied upon to suppress sway, significant residual sway may occur due to response lag. Therefore, this step uses a variational method to solve the two-point boundary value problem to obtain the theoretically optimal feedforward acceleration trajectory, which serves as the input reference for the adaptive shaper. This trajectory is not directly applied to the drive system but is fed into the adaptive time-varying frequency input shaping unit for filtering, thus forming a feedforward channel of "optimal planning + adaptive filtering".

[0061] In this embodiment, the optimal feedforward trajectory planning unit is deployed within an industrial computer and is triggered only when the target position or the hoisting plan changes. Taking the lateral direction as an example, the state variable is the swing angle. angular velocity Displacement of large vehicles and the speed of the large vehicle The control input is acceleration:

[0062]

[0063] The state equation is:

[0064]

[0065] in From the real-time parameter estimation unit. Performance metrics:

[0066]

[0067] in, Set the value to 0.001, with boundary conditions of zero initial and final velocities and zero swing angle. Construct the Hamiltonian function, derived from the optimal conditions:

[0068]

[0069] The costate equations constitute a two-point boundary value problem, solved using the target-shooting method. To reduce online computation, 320 working conditions were solved offline, including typical rope length change rates (-0.5, -0.2, 0, 0.2, 0.5 m / s), initial and final rope lengths (2, 5, 10, 15 m), and displacements (1, 5, 10, 20 m). The optimal acceleration curves were stored using cubic B-spline coefficients. During online computation, the industrial computer... Interpolation with target displacement The data is then stored in a circular buffer. The vertical planning is completely symmetrical. If an extreme condition occurs and the data is not in the library, online testing is triggered to expand the library. The output of this planning unit will be used as the input to the adaptive time-varying frequency shaping unit.

[0070] An adaptive time-varying frequency input shaping unit is used to calculate the instantaneous natural frequency based on the real-time estimated rope length, dynamically adjust the pulse time interval of the input shaper, filter the feedforward acceleration trajectory, and generate the shaped desired acceleration command.

[0071] The adaptive time-varying frequency input shaping unit uses a three-pulse ZVDD shaper, whose pulse amplitude is a fixed constant and whose pulse time interval is calculated in real time according to the instantaneous natural frequency, where the instantaneous natural frequency is equal to the square root of the gravitational acceleration divided by the real-time rope length; the shaped acceleration command is equal to the sum of the products of each pulse amplitude and the feedforward acceleration value at the corresponding delay time.

[0072] Specifically, after the acceleration curve generated by the optimal feedforward trajectory planning unit is fed into this unit, due to the drift in the system's natural frequency caused by changes in rope length, direct application may still excite residual oscillations. Therefore, adaptive filtering is required. In this embodiment, this unit runs within the industrial computer 17, and reads the rope length estimate from the real-time parameter estimation unit every control cycle (10 milliseconds). Calculate the instantaneous natural frequency Where g = 9.80665 m / s². The unit uses a three-pulse ZVDD shaper with a damping ratio ζ = 0.1, and a pre-calculated fixed amplitude. =0.167、 =0.487、 =0.346, pulse time interval , , (To prevent) Too small, set a lower limit of 0.02 seconds). Post-shaping acceleration command:

[0073]

[0074] in The data is taken from a circular buffer (length 200, storing 2 seconds of history). The same applies to the longitudinal direction. The shaped instructions are transmitted to the model reference adaptive control unit via real-time Ethernet. When the rate of change of rope length exceeds 0.3 m / s, ζ can be temporarily increased to 0.2 to enhance robustness.

[0075] The model reference adaptive control unit is used to establish a swing-free reference model. Based on the error between the actual state and the reference model, combined with the swing angle and angular velocity in the state estimate, it calculates the driving force of the large vehicle and the small vehicle, and updates the load mass estimate online through an adaptive law.

[0076] The reference model in the model reference adaptive control unit is a second-order system with a damping ratio of 0.7 and a natural frequency of π radians per second. The driving force calculation formula of this unit includes position error feedback terms, velocity error feedback terms, lateral sway angle feedback terms, lateral sway angle velocity feedback terms, and load mass estimates. The adaptive law of this unit is that the rate of change of the load mass estimate is equal to the negative adaptive gain multiplied by the error-sway angle coupling term, where the error-sway angle coupling term includes lateral error multiplied by lateral velocity error plus lateral sway angle multiplied by lateral sway angle velocity, plus a longitudinal term of the same kind.

[0077] Specifically, although the desired acceleration command output by the adaptive time-varying frequency input shaping unit has filtered out the main frequency components, the actual system still contains load mass estimation errors, unmodeled friction, and wind load disturbances. These uncertainties cannot be completely eliminated through open-loop shaping. Therefore, this step introduces Model Reference Adaptive Control (MRAC). Using the swing angle, angular velocity, and load mass estimates provided by the real-time parameter estimation unit, combined with the shaped acceleration command, the driving force of the trolley and carriage is calculated in real time, enabling the actual motion trajectory to track an ideal, sway-free reference model. The mass estimation is then corrected online using an adaptive law. This unit is deployed within the programmable automation controller 18, executes with a 5-millisecond cycle, and sends the output driving force to the trolley's frequency converter and the carriage's servo driver.

[0078] In this embodiment, the reference model adopts a second-order system:

[0079]

[0080] Damping ratio natural frequency After discretization, the recursion is performed.

[0081] Define tracking error:

[0082]

[0083] Control:

[0084]

[0085] in Given the known mass of the vehicle, Feedback gain for load quality estimation ;

[0086] Through first-order filtering:

[0087]

[0088] Update (α=10).

[0089] Adaptive law online correction quality estimation:

[0090]

[0091] Longitudinal control force Symmetrical. When the vehicle is moving at low speeds (command speed less than 0.05 m / s and sway angle less than 0.1°), the feedback gain is halved to suppress noise. Updated The real-time parameter estimation unit is sent back for verification. This unit, together with the adaptive shaper, forms a dual anti-shake structure of "feedforward filtering + feedback compensation".

[0092] The hoisting speed coordination constraint unit is used to calculate the upper limit of the hoisting speed in real time based on the current lateral swing angle, longitudinal swing angle and angular velocity, and to output the hoisting speed command with a limited amplitude.

[0093] In the winch speed coordination constraint unit, the maximum allowable winch speed for lateral swing is equal to the rope length multiplied by the absolute value of the lateral swing angular velocity, multiplied by the maximum acceleration limit, divided by the rope length multiplied by the square of the lateral swing angular velocity, plus the gravitational acceleration multiplied by the square of the lateral swing angle, and then a decimal value to prevent division by zero. The maximum allowable winch speed for longitudinal swing is calculated in the same way, only by replacing the lateral swing angle and angular velocity with the longitudinal swing angle and angular velocity. Finally, the upper limit of the winch speed is the minimum value of the two and the mechanical speed limit.

[0094] Specifically, the driving force output by the model reference adaptive control unit ensures sway suppression during vehicle acceleration and deceleration, but the lifting or lowering action of the hoisting mechanism 6 itself may also induce load sway. When the hoisting speed is too high, the Coriolis force term will inject energy into the swaying system, causing the originally stable sway angle to diverge, especially when the rope length is short or the sway angle already exists. Therefore, while the driving force command is issued, the hoisting speed also needs to be constrained in real time. This step uses a hoisting speed coordination constraint unit, which uses the rope length estimate, lateral and longitudinal sway angles and their angular velocities provided by the real-time parameter estimation unit to dynamically calculate the upper limit of the hoisting speed, and limits the speed command from the control panel or automatic lifting module. The limited hoisting speed command is sent to the hoisting frequency converter to work in conjunction with the drive execution unit;

[0095] In this embodiment, the unit is deployed within a programmable automation controller and executes with a 5-millisecond cycle. The formula for calculating the maximum permissible lateral hoisting speed is:

[0096]

[0097] in This is an estimated value for the rope length (in meters). The lateral angular velocity (radians / second) is the lateral swing velocity. The maximum acceleration limit for driving is 0.5 m / s², and g is the acceleration due to gravity (9.80665 m / s²). The horizontal swing angle (in radians). To prevent small discrepancies (1e-6), the maximum permissible longitudinal hoisting speed is [specified]. , In the formula Replace with vertical swing angle and its angular velocity That's it. The final maximum hoisting speed is the minimum of the three factors:

[0098]

[0099] in Speed ​​limit for the hoisting machinery (taken as 0.8 m / s). Original speed command. After limiting:

[0100]

[0101] The output is then sent to the hoist frequency converter drive. When the swing angle is large, the upper limit automatically decreases, achieving active safety limiting. As an extension, it can be used for... Apply a first-order low-pass filter (time constant 0.1 seconds) to avoid frequent fluctuations; when the rope length is less than the minimum safety value (2 meters), force the upper limit to be set to a slow speed of 0.1 meters per second.

[0102] The drive actuator is used to control the motion according to the driving force and the hoisting speed command after the amplitude is limited.

[0103] Furthermore, the present invention also provides an anti-sway intelligent crane hoisting device, applied to an anti-sway intelligent crane hoisting system, comprising:

[0104] Two parallel tracks 1 are fixedly laid on the ground;

[0105] The trolley traveling mechanism 2 is installed on the track 1 and can move back and forth along the track 1. The trolley traveling mechanism 2 is driven by a variable frequency motor to achieve lateral movement.

[0106] The main frame 3 is fixedly supported on the main traveling mechanism 2;

[0107] The trolley traveling mechanism 4 is mounted on the main frame 3 and can reciprocate along the guide rail 5 on the main frame 3. The trolley traveling mechanism 4 is driven by a servo motor to achieve longitudinal movement.

[0108] The hoisting mechanism 6 includes a double-folded drum 7 fixed on the trolley traveling mechanism 4, a hoisting motor 8 that drives the drum, and an encoder 9 coaxially mounted with the double-folded drum 7. The surface of the double-folded drum 7 is wound with a steel wire rope 10. The hoisting mechanism 6 is used to wind up and unwind the steel wire rope 10.

[0109] The lifting device 11 is suspended below the trolley traveling mechanism 4 by a steel wire rope 10. A dual-axis tilt sensor 12 is fixedly installed on the lifting device 11 to measure the lateral and longitudinal tilt angles of the lifting device 11.

[0110] Electrical control cabinet 16, mounted on the main frame 3, is used to house control system components.

[0111] A triaxial force sensor 13 is also fixedly installed on the lifting device 11. The triaxial force sensor 13 is connected in series between the wire rope 10 and the lifting device 11 to measure the real-time tension of the wire rope 10. The measured tension data is used to assist the extended Kalman filter algorithm in estimating the load mass. Laser rangefinders 14 are installed on the trolley traveling mechanism 2 and the trolley frame 3 respectively to measure the lateral displacement and speed of the trolley. A rotary encoder 15 is installed on the trolley traveling mechanism 4 to measure the longitudinal displacement and speed of the trolley.

[0112] It also includes an industrial computer 17 and a programmable automation controller 18, both of which are installed in the electrical control cabinet 16. The industrial computer 17 is connected to the programmable automation controller 18 via a real-time industrial Ethernet. The industrial computer 17 runs an extended Kalman filter algorithm, an optimal feedforward trajectory planning algorithm, and an adaptive time-varying frequency input shaping algorithm. The programmable automation controller 18 runs a model reference adaptive control algorithm and a hoisting speed coordination constraint algorithm. The output terminals of the programmable automation controller 18 are respectively connected to the variable frequency drive of the trolley, the servo drive of the trolley, and the variable frequency drive of the hoisting motor 8, for outputting driving force commands and speed commands.

[0113] An electromagnetic brake 19 is also provided on the hoisting mechanism 6. The electromagnetic brake 19 is installed between the output shaft of the hoisting motor 8 and the double-folded drum 7. When the programmable automation controller 18 detects that the lateral swing angle or the longitudinal swing angle exceeds the preset safety threshold, the programmable automation controller 18 sends a braking command to the electromagnetic brake 19 and automatically reduces the hoisting speed through the frequency converter driver of the hoisting motor 8. The preset safety threshold is that the absolute value of the lateral swing angle is greater than 3 degrees or the absolute value of the longitudinal swing angle is greater than 3 degrees.

[0114] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An anti-swing intelligent lifting trolley hoisting system, characterized in that, It is applied to cranes with trolleys and overhead cranes, and the system includes: The real-time parameter estimation unit is used to estimate the load mass, wire rope length and its rate of change online based on sensor measurements using an extended Kalman filter algorithm, and output the state estimate. The optimal feedforward trajectory planning unit is used to solve the feedforward acceleration trajectory with minimum swing energy using the variational method based on the target position and rope length change plan, and serves as the feedforward control reference. An adaptive time-varying frequency input shaping unit is used to calculate the instantaneous natural frequency based on the real-time estimated rope length, dynamically adjust the pulse time interval of the input shaper, filter the feedforward acceleration trajectory, and generate the shaped desired acceleration command. The model reference adaptive control unit is used to establish a swing-free reference model. Based on the error between the actual state and the reference model, and combined with the swing angle and angular velocity in the state estimate, it calculates the driving force of the large vehicle and the small vehicle, and updates the load mass estimate online through an adaptive law. The hoisting speed coordination constraint unit is used to calculate the upper limit of the hoisting speed in real time based on the current lateral swing angle, longitudinal swing angle and angular velocity, and to output the hoisting speed command with a limited amplitude. A drive execution unit is used to control the motion according to the driving force and the hoisting speed command after the amplitude limit; The extended Kalman filter algorithm in the real-time parameter estimation unit uses an eleven-dimensional augmented state vector, including the lateral displacement and velocity of the trolley, the longitudinal displacement and velocity of the trolley, the lateral swing angle and its angular velocity, the longitudinal swing angle and its angular velocity, and the load mass; the measurement vector includes the trolley displacement, the trolley displacement, the lateral swing angle, the longitudinal swing angle, the wire rope length and its rate of change; the filtering period is 10 milliseconds; The reference model in the model reference adaptive control unit is a second-order system with a damping ratio of 0.7 and an angular frequency in rad / s. The driving force calculation formula of the model reference adaptive control unit includes a position error feedback term, a velocity error feedback term, a lateral sway angle feedback term, a lateral sway angle velocity feedback term, and a load mass estimate. The adaptive law of the model reference adaptive control unit is that the rate of change of the load mass estimate is equal to the negative adaptive gain multiplied by the error-sway angle coupling term, wherein the error-sway angle coupling term includes the lateral displacement error multiplied by the lateral velocity error plus the lateral sway angle multiplied by the lateral sway angle velocity, plus a longitudinal term of the same kind.

2. The anti-swing intelligent lifting and trolleying hoisting system according to claim 1, characterized in that, The optimal feedforward trajectory planning unit uses the variational method to solve the two-point boundary value problem. The performance index is the weighted integral of the square of the swing angle and the square of the control acceleration. The optimal acceleration curve is obtained by the shooting method, and the B-spline coefficients for typical working conditions are stored offline and called online by interpolation.

3. The anti-swing intelligent lifting and trolleying hoisting system according to claim 1, characterized in that, The adaptive time-varying frequency input shaping unit uses a three-pulse ZVDD shaper, whose pulse amplitude is a fixed constant and whose pulse time interval is calculated in real time according to the instantaneous natural frequency, where the instantaneous natural frequency is equal to the square root of the gravitational acceleration divided by the real-time rope length; the shaped acceleration command is equal to the sum of the products of each pulse amplitude and the feedforward acceleration value at the corresponding delay time.

4. The anti-sway intelligent crane hoisting system according to claim 1, characterized in that, In the winch speed coordination constraint unit, the maximum allowable winch speed for lateral swing is equal to the rope length multiplied by the absolute value of the lateral swing angular velocity, multiplied by the maximum acceleration limit, divided by the rope length multiplied by the square of the lateral swing angular velocity, plus the gravitational acceleration multiplied by the square of the lateral swing angle, and then plus a preset positive small amount. This positive small amount is used to prevent the denominator from being zero. The maximum allowable winch speed for longitudinal swing is calculated in the same way, only the lateral swing angle and angular velocity are replaced by the longitudinal swing angle and angular velocity. Finally, the upper limit of the winch speed is the minimum value of the two and the mechanical speed limit.

5. An anti-sway intelligent crane hoisting device, characterized in that, An anti-sway intelligent crane hoisting system according to any one of claims 1-4, comprising: Two parallel tracks are fixedly laid on the ground (1); The trolley traveling mechanism (2) is installed on the track (1) and can move back and forth along the track (1). The trolley traveling mechanism (2) is driven by a variable frequency motor to realize lateral movement. The main frame (3) is fixedly supported on the main vehicle traveling mechanism (2); The trolley walking mechanism (4) is installed on the main frame (3) and can reciprocate along the guide rail (5) on the main frame (3). The trolley walking mechanism (4) is driven by a servo motor to realize longitudinal movement. The hoisting mechanism (6) includes a double-folded drum (7) fixed on the trolley traveling mechanism (4), a hoisting motor (8) that drives the drum, and an encoder (9) coaxially mounted with the double-folded drum (7). The surface of the double-folded drum (7) is wound with a steel wire rope (10). The hoisting mechanism (6) is used to wind up and unwind the steel wire rope (10). The lifting device (11) is suspended below the trolley traveling mechanism (4) by a steel wire rope (10). A dual-axis tilt sensor (12) is fixedly installed on the lifting device (11) to measure the lateral and longitudinal tilt angles of the lifting device (11). An electrical control cabinet (16) is mounted on the main frame (3) to house the control system components.

6. The anti-sway intelligent crane hoisting device according to claim 5, characterized in that, A triaxial force sensor (13) is also fixedly installed on the lifting device (11). The triaxial force sensor (13) is connected in series between the wire rope (10) and the lifting device (11). A laser rangefinder (14) is installed on the trolley traveling mechanism (2) and the trolley frame (3). A rotary encoder (15) is installed on the trolley traveling mechanism (4).

7. The anti-sway intelligent crane hoisting device according to claim 5, characterized in that, It also includes an industrial computer (17) and a programmable automation controller (18), both of which are installed in the electrical control cabinet (16); the industrial computer (17) is connected to the programmable automation controller (18), and the output of the programmable automation controller (18) is connected to the frequency converter of the trolley, the servo driver of the trolley and the frequency converter of the hoisting motor (8), respectively.

8. The anti-sway intelligent crane hoisting device according to claim 5, characterized in that, An electromagnetic brake (19) is also provided on the hoisting mechanism (6), which is installed between the output shaft of the hoisting motor (8) and the double-folded drum (7).

Images