Six-degree-of-freedom wave-compensated embarkation equipment suitable for complex sea conditions and control method thereof

Abstract

The application discloses a six-degree-of-freedom wave compensation boarding equipment suitable for complex sea conditions and a control method thereof. The equipment comprises a base platform fixed to a deck of a work ship; a Stewart parallel mechanism, a lower platform of which is connected with the base platform, and an upper platform of which is connected with the lower platform through six hydraulic cylinders; a boarding gangway system installed on the upper platform; a hydraulic servo driving system for driving the six hydraulic cylinders; an intelligent sensing system for collecting ship motion data and relative pose data with a target platform; an intelligent control system electrically connected with the hydraulic servo driving system and the intelligent sensing system; wherein the intelligent control system is configured to execute a composite control algorithm based on a nonlinear extended state observer and a nonlinear model predictive control to drive the hydraulic servo driving system and compensate for ship motion. The boarding equipment has the advantages of high precision, high stability and strong robustness, and can realize accurate and stable compensation under complex sea conditions.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of offshore engineering equipment, and particularly relates to a six-degree-of-freedom wave-compensated boarding equipment suitable for complex sea conditions and a control method thereof. BACKGROUND

[0002] Offshore wind power, oil and gas resource development and other marine engineering operations often need to transfer personnel and materials between a ship and a fixed platform (such as an offshore wind turbine or a drilling platform) or between ships. However, under the action of environmental loads such as wind, waves and currents, the operating ship will produce six-degree-of-freedom motion. This motion, especially in complex sea conditions with long wave periods and frequent swells, will cause the traditional boarding equipment to shake violently, which not only seriously affects the operation efficiency, but also poses a great threat to the safety of personnel.

[0003] To overcome the adverse effects of ship motion, wave compensation technology has gradually developed and been applied to boarding equipment.

[0004] Early compensation devices mainly focused on passive or active compensation in the heave direction to alleviate the impact of ship vertical motion. However, compensating for heave alone cannot eliminate the spatial and pose errors at the end of the gangway caused by other degrees of freedom such as roll and pitch, making it difficult to meet the spatial stability requirements for safe boarding. Therefore, six-degree-of-freedom compensation systems based on Stewart platforms and other parallel mechanisms have become an important development direction, as they can theoretically achieve full-degree-of-freedom motion compensation in three-dimensional space.

[0005] Existing technologies can achieve a certain degree of wave compensation, but the existing technical solutions still have the following shortcomings when facing high-standard operation requirements in harsh sea conditions:

[0006] Insufficient compensation accuracy and dynamic stability under complex motion: existing six-degree-of-freedom compensation systems, especially those relying on traditional PID control algorithms, have large residual errors when dealing with long-period and large-amplitude waves due to the nonlinear dynamics, parameter perturbations and control loop response lags of the system, making it difficult to accurately and stably maintain the boarding point.

[0007] Poor anti-interference robustness in harsh sea conditions: the system faces complex and variable external disturbances (such as wave impact, wind load) and model uncertainties in actual operations. Traditional linear control methods cannot effectively suppress such disturbances, while advanced control algorithms (such as sliding mode control) can improve robustness, but often rely on accurate system models and have problems such as chattering, which are limited in actual engineering applications.

[0008] High energy consumption and low energy utilization efficiency: The active compensation system needs to continuously provide power to counteract the ship movement, resulting in huge energy consumption and high operating cost. The existing system generally lacks efficient energy recovery and management mechanism, and fails to fully utilize the reverse kinetic energy and potential energy generated during the compensation process.

[0009] Lack of intelligent sensing and adaptive control capability: Reliance on a single type of motion sensor is prone to data loss or accuracy decline in harsh environments. Lack of multi-source information fusion and forward-looking prediction control based on sea state characteristics makes the system unable to predict wave motion and take action in advance, limiting the further improvement of compensation performance.

[0010] In summary, the existing wave compensation boarding equipment, especially in complex sea conditions, has significant shortcomings in compensation accuracy, dynamic stability, energy efficiency, and robustness to complex disturbances. SUMMARY

[0011] The first object of the application is to overcome the shortcomings and deficiencies in the prior art, and to provide a six-degree-of-freedom wave compensation boarding equipment suitable for complex sea conditions, which has the advantages of high precision, high stability and strong robustness, and can realize intelligent and accurate compensation in complex sea conditions.

[0012] The second object of the application is to provide a six-degree-of-freedom wave compensation control method, which can effectively solve the problems of insufficient compensation accuracy and response lag of existing compensation systems in long-period wave and complex coupled motion conditions.

[0013] The object of the application is achieved by the following technical scheme: a six-degree-of-freedom wave compensation boarding equipment suitable for complex sea conditions, comprising: a base platform, a Stewart parallel mechanism, a boarding gangway system, a hydraulic servo driving system, an intelligent sensing system and an intelligent control system,

[0014] The base platform is fixed to the deck of the work ship; the Stewart parallel mechanism, whose lower platform is connected to the base platform, and whose upper platform is connected to the lower platform through six hydraulic cylinders; the boarding gangway system is installed on the upper platform; the hydraulic servo driving system is used to drive the six hydraulic cylinders; the intelligent sensing system is used to collect ship motion data and relative pose data with the target platform; the intelligent control system is electrically connected with the hydraulic servo driving system and the intelligent sensing system;

[0015] The intelligent control system is configured to execute a compound control algorithm based on a nonlinear extended state observer and a nonlinear model predictive control to drive the hydraulic servo driving system, compensate for ship movement, and keep the end of the boarding gangway system stable in space.

[0016] Preferably, the hydraulic servo drive system comprises a hydraulic pump station, a servo valve group and an energy recovery device, the energy recovery device is used to recover and store the kinetic energy and potential energy of the system during the boarding gangway lowering or the Stewart parallel mechanism resetting, and the hydraulic pump station controls the movement of the six hydraulic cylinders through the servo valve group.

[0017] Preferably, the intelligent control system specifically comprises:

[0018] A multi-sensor data fusion module is configured to fuse multi-source data collected by the intelligent perception system to generate a motion state measurement value of the ship and the target platform.

[0019] A nonlinear extended state observer module is connected to the multi-sensor data fusion module and is configured to estimate the system state and the total disturbance in real time.

[0020] A nonlinear model predictive control module is connected to the nonlinear extended state observer module and is configured to perform multi-step prediction and rolling optimization based on the system state and the total disturbance, generate a preliminary control signal, and perform feedforward compensation on the total disturbance to form a final control amount.

[0021] A motion solving and servo control module is configured to perform kinematic inverse solution of the Stewart parallel mechanism and joint motion solving of the boarding gangway system on the final control amount respectively, and the solving result is sent to the hydraulic servo drive system to perform compensation actions.

[0022] Preferably, the nonlinear extended state observer is configured to estimate the system state and the total disturbance in real time based on a system model and an observer equation containing a nonlinear function.

[0023] The system model comprises an extended state equation:

[0024] x1=x2,

[0025] x2=x3+b0*u(t),

[0026] x3≈h(t),

[0027] wherein x1 is a position, x2 is a velocity, x3 is an extended state, h(t) is a rate of change of the total disturbance, b0 is a nominal value of a control gain, and u(t) is a control input.

[0028] The observer equation is:

[0029] e=z1-y,

[0030] ż1=z2-β1*fal(e,α1,δ),

[0031] ż2=z3-β2*fal(e,α2,δ)+b0*u(t),

[0032] z3 = -β3 * fal(e, a3, δ),

[0033] where e is the estimation error, y is the measurement output, z1, z2, z3 are the estimates of states x1, x2, x3 respectively, a1 is the strength of the non-linear response of the control to the position estimation error, a2 is the strength of the non-linear response of the control to the velocity estimation error, a3 is the strength of the non-linear response of the control to the augmented state (sum of disturbances) estimation error, β1, β2, β3 are the observer gains, the expression of the non-linear function fal(·) is fal(e, a, δ) = { |e|a*sign(e), |e| > δ, e / δ^(1-a), |e|≤δ}, a is the non-linear power, δ is the linear region threshold.

[0034] Preferably, the non-linear model predictive control is configured to roll out an objective function and handle system constraints in a prediction horizon using a discretized prediction model based on the system states and the sum of disturbances to generate a preliminary control signal,

[0035] where the prediction model is:

[0036] x(k+1) = f_d(x(k), u(k)),

[0037] where f_d is the discretized form of the system dynamics, x(k) is the current state variable, u(k) is the current control variable,

[0038] the objective function is:

[0039] min J = Σ [(x(k+i) - x ref (k+i))^T * Q * (x(k+i) - x ref (k+i)) + u(k+i)^T * R * u(k+i)],

[0040] where x ref is the desired reference trajectory, Q and R are weight matrices penalizing the state tracking error and the control variable size respectively, T denotes the transpose.

[0041] Preferably, the constraints include:

[0042] control input constraints: u min ≤ u(k+i) ≤ u max ,

[0043] state constraints: x min ≤ x(k+i) ≤ x max ,

[0044] control increment constraints: |Δu(k+i)| ≤ Δu max,

[0045] wherein, u represents control input, including output force and speed of the hydraulic servo driving system, x represents system state, including workspace boundary of the Stewart parallel mechanism and joint angle limit of the gangway, Δu max represents preset maximum control increment.

[0046] Preferably, the feedforward compensation is to feed forward the total disturbance estimated by the nonlinear model predictive control to the control end to obtain the final control amount u:

[0047] u=(u NMPC -z3) / b0,

[0048] wherein, u NMPC is the preliminary control signal of the nonlinear model predictive control output, z3 is the total disturbance, and b0 is the nominal value of the control gain.

[0049] Preferably, the observation bandwidth of the nonlinear extended state observer is 10-20 Hz, and the disturbance estimation update period is not more than 0.02 seconds; the prediction time domain of the nonlinear model predictive control is 1.5-3 seconds, the control time domain is 0.5-1.5 seconds, and the control period is not more than 0.02 seconds.

[0050] Preferably, the intelligent perception system comprises an inertial integrated navigation system, a laser range finder and a visual recognition system.

[0051] A six-degree-of-freedom wave compensation control method applied to the boarding equipment, comprising the steps of:

[0052] Step S1, acquiring motion and pose data of the ship and the target platform through the intelligent perception system, and fusing and processing to generate motion state measurement values of the ship and the target platform;

[0053] Step S2, based on the motion state measurement values, estimating system state and total disturbance in real time through the nonlinear extended state observer;

[0054] Step S3, based on the system state and the total disturbance, performing multi-step prediction and rolling optimization through the nonlinear model predictive control to generate a preliminary control signal, and performing feedforward compensation on the total disturbance to form a final control amount;

[0055] Step S4, performing kinematic inverse solution of the Stewart parallel mechanism and joint motion solution of the boarding gangway system on the final control amount respectively, and sending the solution results to the hydraulic servo driving system to perform compensation actions.

[0056] The present application has the following advantages and effects relative to the prior art:

[0057] (1) The intelligent control system of the application adopts a compound control algorithm based on a nonlinear extended state observer (NLESO) and a nonlinear model predictive control (NMPC), and the wave compensation accuracy is improved from the order of magnitude of ±15 cm of the traditional method to within ±3 cm, and the accuracy is improved by more than 80%;

[0058] The total disturbance of the system is estimated and predicted in real time by the nonlinear extended state observer, feedforward compensation is realized, the system response delay is reduced to within 50 ms from more than 200 ms of the traditional PID control, the system response delay is significantly reduced, and the inherent phase lag problem of the traditional PID control based on the current error for adjustment is overcome;

[0059] The nonlinear model predictive control is based on the current state and the prediction model, and the control sequence in the future multiple control periods is optimized, the wave disturbance can be foreseeably offset, the compensation accuracy can be maintained under complex sea conditions with measurement noise and model uncertainty, and the strong anti-interference robustness is achieved;

[0060] The application realizes intelligent and accurate wave compensation based on sea state adaptive boarding operation by combining the six-degree-of-freedom motion capability of the Stewart parallel mechanism and the NLESO-NMPC compound controller, and is suitable for complex sea conditions.

[0061] (2) The hydraulic servo driving system with energy recovery function is innovatively designed, and the energy recovery device is adopted to recover kinetic energy and potential energy during the lowering of the load or the resetting of the Stewart parallel mechanism, and the recovered energy is preferentially used for the operation of the system itself, thereby reducing the dependence on external power supply.

[0062] (3) The intelligent perception system of the application integrates inertial integrated navigation system, laser range finder, visual recognition system and other multi-source sensors, constructs a data fusion algorithm based on Kalman filtering, realizes accurate measurement of relative motion of the boarding point, and overcomes the measurement limitations of single sensor in harsh environment. BRIEF DESCRIPTION OF DRAWINGS

[0063] Figure 1 It is a front view of a six-degree-of-freedom wave compensation boarding equipment suitable for complex sea conditions.

[0064] Figure 2 It is a top view of a six-degree-of-freedom wave compensation boarding equipment suitable for complex sea conditions.

[0065] Figure 3 It is a structure diagram of the compound control algorithm based on the nonlinear extended state observer and the nonlinear model predictive control of the application.

[0066] Figure 4The block diagram of the logic implementation of the NLESO-NMPC composite controller (a composite control algorithm based on nonlinear extended state observer-nonlinear model predictive control) of the present invention.

[0067] Among them, 1 is the base platform, 2 is the Stewart parallel mechanism, 3 is the boarding gangway system, 4 is the hydraulic servo drive system, 5 is the intelligent sensing system, 6 is the intelligent control system, 7 is the upper platform, 8 is the hydraulic cylinder, and 9 is the lower platform. Detailed Implementation

[0068] The present invention will be further described in detail below with reference to the embodiments and accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0069] Example 1

[0070] like Figures 1-2 As shown, a six-degree-of-freedom wave-compensated boarding equipment suitable for complex sea conditions includes: a base platform, a Stewart parallel mechanism, a boarding gangway system, a hydraulic servo drive system, an intelligent sensing system, and an intelligent control system.

[0071] The base platform is fixed to the deck of the working vessel; the Stewart parallel mechanism has its lower platform connected to the base platform and its upper platform connected to the lower platform via six hydraulic cylinders; the boarding gangway system is installed on the upper platform; the hydraulic servo drive system is used to drive the six hydraulic cylinders; the intelligent sensing system is used to collect ship motion data and relative position and posture data with respect to the target platform; the intelligent control system is electrically connected to the hydraulic servo drive system and the intelligent sensing system.

[0072] The intelligent control system is configured to execute a composite control algorithm based on a nonlinear extended state observer and a nonlinear model predictive control to drive the hydraulic servo drive system, compensate for ship motion, and keep the end of the boarding gangway system in a stable spatial position.

[0073] Specifically, in this embodiment, the base platform is a circular steel structure platform, which serves as the mounting base of the entire equipment and is fixedly connected to the ship deck by welding or bolts to bear and disperse the load during operation of the equipment. The two ends of the six hydraulic cylinders of the Stewart parallel mechanism (i.e., the Stewart platform) are connected to the upper and lower platforms through spherical hinges, forming a spatial six-degree-of-freedom parallel structure, which serves to realize accurate position adjustment of the boarding gangway in three-dimensional space and compensate for the wave motion of the ship. The boarding gangway system includes a gangway body, handrails, step plates, and a docking mechanism. The gangway body is a truss steel structure in the shape of a channel with a certain slope; the handrails are arranged on both sides of the gangway body; the step plates are evenly distributed on the bottom surface of the gangway body; and the docking mechanism is installed at the end of the gangway body, which serves to provide a safe passage for personnel and realize flexible docking with the target platform. The hydraulic servo drive system includes a hydraulic pump station, a servo valve group, an accumulator, and a cooling device. The hydraulic pump station accurately controls the movement of the six hydraulic cylinders through the servo valve group, the accumulator is used to absorb impact load, and the cooling device ensures thermal balance during continuous operation of the system, which serves to provide accurate power output for the Stewart parallel mechanism. The intelligent control system includes a new generation of strapdown inertial integrated navigation system developed for ship motion attitude monitoring, an industrial computer, control algorithm software, and an operation interface. Various sensors of the inertial integrated navigation system are installed on the ship body and the target platform, the industrial computer runs the composite control algorithm, and the operation interface provides human-computer interaction. The intelligent control system serves to detect the ship motion in real time, calculate the compensation amount, and drive the hydraulic servo drive system to perform compensation actions.

[0074] The working process of the boarding equipment is as follows:

[0075] (1) Initial positioning stage: the operator sets the target boarding position through the control interface of the intelligent control system, the intelligent control system determines the relative pose through the laser range finder and visual recognition unit of the intelligent sensing system, and the Stewart parallel mechanism drives the boarding gangway system to preliminarily align the target platform.

[0076] (2) Wave compensation stage:

[0077] The inertial measurement unit installed on the ship body detects the six-degree-of-freedom motion (heave, roll, pitch, sway, surge, and yaw) of the ship in real time;

[0078] The nonlinear model predictive control algorithm in the intelligent control system calculates the stable pose that the upper platform of the Stewart parallel mechanism needs to maintain based on the ship motion data and wave prediction information;

[0079] The intelligent control system calculates the target length of the six hydraulic cylinders, and drives the hydraulic cylinders to accurately extend and retract through the hydraulic servo drive system;

[0080] The upper platform of the Stewart parallel mechanism drives the embarkation gangway to make a movement opposite to the ship movement, so that the end of the gangway remains relatively static in the inertial coordinate system.

[0081] The nonlinear extended state observer estimates the unmodeled dynamics and external disturbances of the system in real time, and performs feedforward compensation to improve the anti-interference ability of the system.

[0082] (3) Safe docking and passing stage:

[0083] After the docking mechanism at the end of the gangway contacts the target platform, the contact force is detected by a force sensor, and the intelligent control system adjusts the pose of the gangway to achieve flexible docking;

[0084] When personnel pass through the gangway, the system continuously performs fine compensation to ensure stability during the passing process;

[0085] If an emergency occurs or exceeds the compensation range, the system automatically starts the safety protection mode and slowly retracts the gangway to a safe position.

[0086] (4) Energy recovery process:

[0087] During the lowering of the embarkation gangway system or the resetting of the Stewart parallel mechanism, the hydraulic system switches to the pump working condition, converting the kinetic and potential energy of the mechanism into hydraulic energy and storing it in the accumulator. The stored energy is used preferentially in subsequent compensation actions, effectively reducing the total demand for external energy by the system.

[0088] Specifically, the embodiment selects specific experimental data to further explain and describe the control process of the embarkation equipment of the present application, which includes the following steps in chronological order:

[0089] Step S01: System initialization and self-checking:

[0090] Start the hydraulic servo drive system and preheat the hydraulic oil to a working temperature of -10-45℃;

[0091] Calibrate the zero position of the sensor, and the calibration time should not exceed 30 seconds;

[0092] Check the stroke of each hydraulic cylinder to ensure that it is within the range of ±50mm from the middle position;

[0093] After the system self-checking is completed, the system enters a standby state.

[0094] Step S02: Target identification and initial positioning:

[0095] Measure the relative distance to the target platform with a laser range finder, with a measurement accuracy of ±1cm;

[0096] The visual recognition system identifies the target docking point, and the identification time is ≤2 seconds;

[0097] Control the initial alignment of the gangway, the positioning error is controlled within ±5cm;

[0098] Set the safety operation area boundary, the boundary distance from the target platform edge is ≥1.5m.

[0099] Step S03: Motion detection and data acquisition:

[0100] The inertial measurement unit in the inertial integrated navigation system collects ship six-degree-of-freedom motion data in real time at a sampling frequency of 50Hz;

[0101] The pressure sensor monitors the hydraulic system pressure, and the working pressure is maintained at 21±2MPa;

[0102] The displacement sensor detects the extension and retraction amount of each hydraulic cylinder, with a resolution of 0.1mm;

[0103] The environmental sensor collects wind speed and wave height data with an update frequency of 10Hz.

[0104] Step S04: Nonlinear extended state observer calculation;

[0105] Real-time estimation of system total disturbance, including: model uncertainty disturbance, external wave force disturbance and hydraulic system nonlinear disturbance;

[0106] The bandwidth of the nonlinear extended state observer is set to 15Hz;

[0107] The disturbance estimation update period is 0.01 seconds.

[0108] Step S05: Nonlinear model predictive control calculation:

[0109] The prediction horizon is set to 2 seconds, and the control horizon is 1 second;

[0110] Rolling optimization solves the optimal control sequence of the future 20 control periods;

[0111] Consider the velocity constraint ±0.5m / s and the acceleration constraint ±2m / s² of the hydraulic cylinder;

[0112] The weight matrix of the optimization objective function is adjusted online, with an adjustment period of 0.1 seconds.

[0113] Step S06: Hydraulic drive execution:

[0114] The response time of the servo valve is ≤10ms;

[0115] The positioning accuracy of the hydraulic cylinder is ±1mm;

[0116] The system response delay is controlled within 50ms;

[0117] The oil temperature is maintained within the range of-10-45℃.

[0118] Step S07: safety monitoring and adaptive adjustment:

[0119] Real-time monitoring of the vibration acceleration at the end of the gangway, triggering a warning when it exceeds 0.3g;

[0120] Automatic adjustment of the docking posture when the docking force exceeds 500N;

[0121] Automatic adjustment of control parameters according to sea state level: precision priority mode for sea state below level 3, balance mode for sea state level 4-5, and safety priority mode for sea state level 5 and above. System state monitoring period is 0.1 seconds.

[0122] Step S08: energy recovery and management:

[0123] Energy recovery is started during the lowering of the embarkation gangway, with a recovery efficiency of ≥60%. The energy recovery device (accumulator) pressure is maintained within the range of 15-25MPa, and the energy reuse priority level is set as follows: hydraulic system auxiliary power, control system backup power, and ship power grid feedback.

[0124] Step S09: emergency handling and safety protection:

[0125] System instability detection response time ≤0.2 seconds; emergency recovery speed 0.2m / s; maximum allowed inclination angle 17°. The system automatically locks in the current safe position in case of failure.

[0126] The equipment of the present application is used in the cyclic execution of the above steps, realizing precise wave compensation for embarkation operations in complex sea conditions. The entire control period is 0.01 seconds, ensuring continuous stable compensation performance. The parameter settings of each step are verified by a large number of experiments, which can realize the optimal compensation effect under the premise of ensuring safety.

[0127] The hydraulic servo drive system includes a hydraulic pump station, a servo valve group, and an energy recovery device. The energy recovery device is used to recover and store the kinetic and potential energy of the system during the lowering of the embarkation gangway or the resetting of the Stewart parallel mechanism. The hydraulic pump station controls the movement of the six hydraulic cylinders through the servo valve group.

[0128] Specifically, the present application innovatively designs a hydraulic servo drive system with energy recovery function, with a recovery efficiency of more than 60%. In this embodiment, the energy recovery device adopts a general accumulator mode.

[0129] As Figure 3 The control structure diagram of the composite control algorithm based on the nonlinear extended state observer (NLESO) and the nonlinear model predictive control (NMPC) is shown. The intelligent control system specifically includes:

[0130] A multi-sensor data fusion module is configured to fuse multi-source data collected by an intelligent perception system to generate a motion state measurement of a ship and a target platform. The intelligent perception system includes an inertial integrated navigation system, a laser range finder, and a visual recognition system.

[0131] A nonlinear extended state observer module is connected to the multi-sensor data fusion module and is configured to estimate a system state and a total disturbance in real time.

[0132] A nonlinear model predictive control module is connected to the nonlinear extended state observer module and is configured to perform multi-step prediction and rolling optimization based on the system state and the total disturbance, generate a preliminary control signal, and perform feedforward compensation on the total disturbance to form a final control amount.

[0133] A motion solving and servo control module is configured to perform kinematic inverse solution of a Stewart parallel mechanism and joint motion solving of a boarding gangway system on the final control amount respectively, and send a solving result to a hydraulic servo drive system to perform a compensation action.

[0134] The nonlinear extended state observer module includes:

[0135] A system observation unit is configured to estimate an internal state of the system.

[0136] A disturbance estimation unit is configured to estimate a total disturbance of the system in real time. The total disturbance includes model uncertainty, external wave force, and nonlinear disturbance of the hydraulic servo drive system.

[0137] A parameter self-adaptive adjustment unit is configured to automatically adjust parameters of the nonlinear extended state observer according to a working condition change.

[0138] The nonlinear model predictive control module includes:

[0139] A motion prediction unit is configured to predict a system state in a future control period.

[0140] An optimization solving unit is configured to solve a future control sequence according to a preset target function.

[0141] A constraint processing unit is configured to perform real-time constraint processing on displacement, velocity, and output force of a hydraulic cylinder.

[0142] Specifically, the core of the equipment is a composite control algorithm executed by an intelligent control system of the equipment. The overall control structure of the equipment is as shown in Figure 3 The specific working principle is as follows:

[0143] 1. The multi-sensor data fusion module:

[0144] This module is the basis of intelligent control system perception, responsible for processing raw information from various sensors.

[0145] Data acquisition and preprocessing: The intelligent perception system continuously collects raw data from various sensors, including six-degree-of-freedom motion information (including position, attitude, and angular velocity) of the ship / platform through the inertial integrated navigation system (INS), relative distance to the target (such as the wharf) through the laser range finder, and specific markers or scene depth information through the visual recognition system to assist in pose estimation. The multi-sensor data fusion module needs to preprocess the raw data: including timestamp alignment, filtering (such as Kalman filtering) denoising, and outlier processing.

[0146] Feature-level fusion: First, extract key features from sensor data, such as low-frequency motion trends from INS data and edge or corner features from visual information. Then use weighted fusion or model-based (such as Kalman filter series algorithms) methods for fusion, finally output accurate and robust estimation of the ship / platform motion state as the observation value for subsequent control.

[0147] 2. Nonlinear extended state observer (NLESO):

[0148] The nonlinear extended state observer is a key to deal with system uncertainties and external disturbances, which reconstructs and extends the complex system dynamics (taking a single degree of freedom as an example).

[0149] System model and state extension:

[0150] Taking a degree of freedom of the Stewart platform as an example, the system dynamics model can be simplified as:

[0151] x1 = x2,

[0152] x2 = f(x, t) + w(t) + b0 * u(t),

[0153] where x1 is the position, x2 is the velocity, x1 is the position derivative, i.e., the velocity, x2 is the velocity derivative, i.e., the acceleration, f(x, t) represents the known nominal dynamics model, w(t) represents the unknown external disturbance, b0 is the nominal value of the control gain, and u(t) is the control input.

[0154] The core idea of NLESO is to regard the sum of uncertainties, unmodeled dynamics, and external wave forces f(x, t) + w(t) + (b - b0)u(t) as a new state variable x3, i.e., the extended state, where b is the system parameter, so as to obtain the extended system equation:

[0155] x1 = x2,

[0156] x2 = x3 + b0 * u(t),

[0157] x3 = h(t) (assuming the rate of change of the total disturbance is bounded, h(t) is the rate of change of the total disturbance, the total disturbance includes unmodeled dynamics plus external disturbances).

[0158] Observer design and disturbance estimation:

[0159] For the above extended system, a nonlinear observer is designed:

[0160] e = z1 - y (e is the estimation error, y is the measured output),

[0161] z1 = z2 - β1 * fal(e, α1, δ),

[0162] z2 = z3 - β2 * fal(e, α2, δ) + b0 * u(t),

[0163] z3 = - β3 * fal(e, α3, δ),

[0164] where z1, z2, z3 are the estimated values of states x1, x2, x3 respectively, α1 is the nonlinear response strength of the control to the position estimation error (e), α2 is the nonlinear response strength of the control to the speed estimation error (implicit in the change of e), α3 is the nonlinear response strength of the control to the extended state (total disturbance) estimation error. β1, β2, β3 are observer gains that need to be adjusted to ensure fast convergence. fal(e, α, δ) is a key nonlinear function, which is usually in the form of:

[0165] fal(e, α, δ) = { |e|^α * sign(e), |e| > δ, e / δ^(1-α), |e| ≤ δ},

[0166] where δ is the linear interval threshold, α is the nonlinear power, i.e. α1, α2, α3, this function makes the observer provide high gain to quickly track when the error is large, and provide low gain to suppress chatter when the error is small. Through this design, z3 can accurately estimate the total disturbance of the system in real time.

[0167] NLESO unifies the unmodeled dynamics, parameter perturbation and external disturbance of the system as "total disturbance" for observation and estimation, without the need for accurate system mathematical model to achieve effective compensation.

[0168] Parameter adaptive adjustment:

[0169] To make NLESO maintain optimal performance and further improve robustness under the condition that system parameters b drift due to changes in ship load, hydraulic oil temperature, etc., the application introduces a parameter adaptive mechanism, including designing an adaptive law (adjusting parameters in the direction of the negative gradient of performance indicators such as error squares, selecting a set of b0 and a set of β according to the indicators i ) to fine-tune b0 or observer gain β i online, so that it can track changes in system dynamics. NLESO can estimate time-varying parameters and unmodeled dynamics of the system in real time, so that the control system maintains stable performance under conditions such as changes in ship load, hydraulic oil temperature, etc.

[0170] 3. Nonlinear model predictive control (NMPC):

[0171] NMPC uses the state estimation provided by NLESO (including current state and disturbance prediction), predicts the future dynamics of the system based on the model, and solves a constrained optimization problem to obtain the optimal control.

[0172] (1) Motion prediction unit:

[0173] A discretized system model is used as the prediction model. For the kth control period, the prediction model is: x(k+1)=f_d(x(k), u(k)), where f_d is the discretized form of system dynamics, x(k) is the current state variable, i.e. position and velocity; u(k) is the current control variable, i.e. force or torque, and x(k+1) is the state variable at the next time. The model should be able to reflect the main nonlinear dynamic characteristics of the Stewart platform and the gangway.

[0174] The existing wave compensation system for complex sea conditions has significant nonlinear characteristics (such as hydraulic system dead zone, saturation characteristics, mechanism kinematics nonlinear), NMPC directly uses a nonlinear prediction model, avoiding the model mismatch problem of traditional linearization methods when moving in a large range.

[0175] (2) Optimization solving unit:

[0176] NMPC optimizes the following objective function within the prediction horizon N:

[0177] min J = Σ [(x(k+i)-x ref (k+i))^T*Q*(x(k+i)-x ref (k+i))+u(k+i)^T*R*u(k+i)], where x refis the desired reference trajectory (generated by the wave compensation strategy, ideally the inverse phase of ship motion), Q and R are weight matrices penalizing state tracking error and control magnitude, respectively. The objective is to find the optimal control sequence U = [u(k), u(k+1),..., u(k+N-1)] that makes the system future states as close as possible to the desired trajectory while minimizing the cost function J.

[0178] (3) Constraint handling unit:

[0179] One of the prominent advantages of NMPC is the ability to handle various physical constraints explicitly, constraints included in the optimization problem directly are:

[0180] Control input constraints: u min ≤ u(k+i) ≤ u max ,

[0181] u min and u max represent the minimum and maximum output force / speed of the hydraulic servo drive system,

[0182] State constraints: x min ≤ x(k+i) ≤ x max ,

[0183] x min and x max represent the minimum and maximum values of the Stewart platform workspace boundary and gangway joint angle limit,

[0184] Control increment constraints: |Δu(k+i)| ≤ Δu max ,

[0185] Δu max represents the preset maximum control increment, which is set to avoid frequent start-stop of the actuator, i.e. hydraulic servo drive system.

[0186] The NMPC framework naturally supports explicit handling of actuator displacement, velocity, force constraints, and stability constraints, avoiding overshoot oscillation and actuator saturation, ensuring that the system always operates within a safe range.

[0187] (4) Rolling optimization solver:

[0188] Rolling optimization and frontier solver: At each control cycle, the NMPC controller solves the above constrained optimization problem. Due to the strong nonlinearity of the system, efficient numerical optimization algorithms such as interior point method, sequential quadratic programming or real-time iterative algorithms specifically for NMPC are usually required. The solver outputs the optimal control sequence at the current time, but only the first control quantity u*(k) is applied to the system, and the optimization is re-solved based on the new state estimation at the next cycle, which is the principle of "rolling optimization".

[0189] Nonlinear model predictive control is based on the current state and prediction model, and rolls the control sequence in the future multiple control cycles, which can "predictively" offset the wave disturbance.

[0190] 4. Motion solving and servo control module:

[0191] Feedforward compensation:

[0192] The preliminary control instruction u NMPC is not directly output. As shown in Figure 4 , the intelligent control system feeds forward the total disturbance z3 estimated by the NLESO to the control end for feedforward compensation, forming the final control quantity: u=(u NMPC -z3) / b0,

[0193] This is equivalent to directly offsetting the impact of the disturbance on the basis of the optimal control quantity calculated by the NMPC based on the model, realizing forward-looking compensation, and greatly reducing the hysteresis of traditional feedback control, reducing the system response delay from more than 200ms of traditional PID to within 50ms.

[0194] Hydraulic servo control: u*(k) after kinematics solving is used as the instruction signal of the hydraulic servo driving system, controlling the displacement or force output of the hydraulic cylinder, and u*(k) represents the first control quantity, i.e. the hydraulic cylinder thrust in the prediction time domain N=1.

[0195] Kinematics solving: according to the overall compensation motion instruction, kinematics inverse solution of the Stewart platform (solving the extension and retraction amount of the six hydraulic cylinders from the target pose of the upper platform) and joint motion solving of the boarding gangway (solving the joint angles from the end target trajectory) are carried out respectively, and the solving results are sent to the corresponding hydraulic servo driving system for execution.

[0196] The application forms a compound control strategy with foresight and strong anti-interference ability through deep integration of real-time disturbance observation and feedforward compensation of NLESO and multi-step prediction and constraint optimization of NMPC, which is different from the traditional single compensation strategy, and the application realizes intelligent compensation based on sea state self-adaptation. The strategy enables the end of the embarkation gangway to overcome the influence of six-degree-of-freedom motion of the ship in complex sea conditions and maintain extremely high pose stability (compensation accuracy within ±3 cm) in the inertial coordinate system, thereby ensuring the safety and efficiency of personnel transfer operations.

[0197] The nonlinear model predictive control compensation algorithm based on the nonlinear extended state observer of the application breaks through the traditional control lag bottleneck: the wave compensation accuracy is improved from the traditional method of ±15 cm to within ±3 cm, with a precision increase of more than 80%, and the system response delay is reduced from more than 200 ms of traditional PID control to within 50 ms, effectively overcoming the hysteresis problem of traditional feedback control and being able to respond to rapid changes in complex sea conditions. The equipment of the application has the advantages of high precision, high stability and strong robustness, and is particularly suitable for complex disturbance environments such as irregular waves and broken waves, and can still maintain excellent compensation accuracy in the presence of measurement noise and model uncertainty.

[0198] The observation bandwidth of the nonlinear extended state observer is 10-20 Hz, and the disturbance estimation update period is not more than 0.02 seconds; the prediction time domain of the nonlinear model predictive control is 1.5-3 seconds, the control time domain is 0.5-1.5 seconds, and the control period is not more than 0.02 seconds.

[0199] Specifically, in a preferred embodiment, the observation bandwidth of the nonlinear extended state observer is 15 Hz, and the disturbance estimation update period is 0.01 seconds; the prediction time domain of the nonlinear model predictive control is 2 seconds, the control time domain is 1 second, and the control period is 0.01 seconds. A large number of experiments have verified that this group of parameters can achieve optimal compensation accuracy and response speed on the premise of ensuring system stability.

[0200] Embodiment 2

[0201] A six-degree-of-freedom wave compensation control method applied to the embarkation equipment described in embodiment 1, the flowchart thereof is shown in Figure 3 , comprising the steps of:

[0202] Step S1, acquiring the motion and pose data of the ship and the target platform through the intelligent sensing system, and fusing and processing to generate the motion state measurement values of the ship and the target platform;

[0203] Step S2, based on the motion state measurement, the system state and the total disturbance are estimated in real time by a nonlinear extended state observer;

[0204] Step S3, based on the system state and the total disturbance, multi-step prediction and rolling optimization are performed by a nonlinear model predictive control, a preliminary control signal is generated, and the total disturbance is fed forwardly compensated to form a final control amount;

[0205] Step S4, the final control amount is subjected to kinematic inverse solution of the Stewart parallel mechanism and joint motion solution of the embarkation ladder respectively, and the solution result is issued to the hydraulic servo driving system to perform compensation action.

[0206] As shown in Figure 4 , it is a logic implementation block diagram of the NLESO-NMPC (nonlinear extended state observer-nonlinear model predictive control) compound controller (compound control algorithm) of the application.

[0207] In the design, the output from the embarkation system (integrating the Stewart parallel mechanism and the embarkation ladder) is the measurement of the system motion state z1, z2 (such as the pose and the speed). The measurement is used by the NLESO (nonlinear extended state observer) to estimate the system state (such as the platform pose and the speed) and the total disturbance z3 (including the model uncertainty, the unmodeled dynamics and the external wave force and other disturbances), and the error between z1, z2 and the given reference value x ref (that is, the expected reference trajectory) is taken as the input of the NMPC (nonlinear model predictive control). The NMPC calculates the preliminary control signal u NMPC through rolling optimization. At the same time, the total disturbance z3 estimated by the NLESO is eliminated from u NMPC , so that the control amount for controlling the hydraulic servo driving system and the joint motion of the embarkation equipment, that is, the control law u= (u NMPC -z3) / b0 is obtained. The NLESO is a core component of the compound controller (NLESO-NMPC).

[0208] The compound controller design clearly shows that the NLESO only needs the motion state measurement (obtained by fusing multiple sensors such as the fiber-optic inertial integrated navigation system, the laser range finder and the visual recognition system) of the embarkation system and the control law u= (u NMPC -z3) / b0 as its input. The control method of the application can effectively solve the problems of insufficient compensation accuracy and response lag of the existing compensation system under long-period wave and complex coupling motion conditions.

[0209] The above embodiments are the preferred embodiments of the present application, and cannot limit the present application, any changes or other equivalent replacement manners without departing from the technical solutions of the present application are included in the protection scope of the present application.

Claims

1. A six-degree-of-freedom wave compensation boarding equipment suitable for complex sea conditions, characterized in that, include: Base platform, Stewart parallel mechanism, boarding gangway system, hydraulic servo drive system, intelligent sensing system, and intelligent control system. The base platform is fixed to the deck of the working vessel; the Stewart parallel mechanism has its lower platform connected to the base platform and its upper platform connected to the lower platform via six hydraulic cylinders; the boarding gangway system is installed on the upper platform; the hydraulic servo drive system is used to drive the six hydraulic cylinders; the intelligent sensing system is used to collect ship motion data and relative position and posture data with respect to the target platform; the intelligent control system is electrically connected to the hydraulic servo drive system and the intelligent sensing system. The intelligent control system is configured to execute a composite control algorithm based on a nonlinear extended state observer and a nonlinear model predictive control to drive the hydraulic servo drive system, compensate for ship motion, and keep the end of the boarding gangway system in a stable spatial position. The intelligent control system specifically includes: The multi-sensor data fusion module is used to fuse multi-source data collected by the intelligent sensing system to generate motion state measurement values ​​of the ship and target platform; The nonlinear extended state observer module, connected to the multi-sensor data fusion module, is used to estimate the system state and total disturbance in real time. The nonlinear model predictive control module, connected to the nonlinear extended state observer module, is used to perform multi-step prediction and rolling optimization based on the system state and the total disturbance, generate a preliminary control signal, and perform feedforward compensation on the total disturbance to form the final control quantity. The motion calculation and servo control module is used to perform inverse kinematics calculation of the Stewart parallel mechanism and joint motion calculation of the boarding gangway system respectively on the final control quantity. The calculation results are sent to the hydraulic servo drive system to perform compensation actions. The nonlinear extended state observer is configured to estimate the system state and total disturbance in real time based on the system model and observer equations containing nonlinear functions. The system model mentioned above includes extended state equations: ， ， ， Where x1 is position, x2 is velocity, x3 is expansion state, h(t) is the rate of change of the total disturbance, b0 is the nominal value of the control gain, and u(t) is the control input. And the observer equation is: e=z 1- yes, ， ， ， Where e is the estimation error, y is the measurement output, z1, z3, z3 are the estimated values ​​of states x1, x2, x3 respectively, α1 is the nonlinear response strength of the control to the position estimation error, α2 is the nonlinear response strength of the control to the velocity estimation error, α3 is the nonlinear response strength of the control to the extended state estimation error, β1, β2, β3 are the observer gains, and the expression of the nonlinear function fal(·) is fal(e,α,δ)= {|e|^α*sign(e),|e|>δ,e / δ^(1-α),|e|≤δ}, where α is the nonlinear power and δ is the threshold of the linear interval.

2. The six-degree-of-freedom wave compensation boarding equipment suitable for complex sea conditions according to claim 1, characterized in that, The hydraulic servo drive system includes a hydraulic pump station, a servo valve group, and an energy recovery device. The energy recovery device is used to recover and store the kinetic and potential energy of the system during the lowering of the boarding ladder or the resetting of the Stewart parallel mechanism. The hydraulic pump station controls the movement of six hydraulic cylinders through the servo valve group.

3. The six-degree-of-freedom wave compensation boarding equipment suitable for complex sea conditions according to claim 1, characterized in that, The nonlinear model predictive control is configured to, based on the system state and total disturbance, utilize a discretized predictive model to continuously optimize the objective function in the prediction time domain and handle system constraints to generate an initial control signal. The prediction model is as follows: x(k+1)=f_d(x(k),u(k)), Where f_d is the discretized form of the system dynamics, x(k) is the current state variable, and u(k) is the current control variable. The objective function is: minJ=Σ[(x(k+i)-x ref (k+i))^T*Q*(x(k+i)-x ref (k+i))+u(k+i)^T*R*u(k+i)], Where, x ref The expected reference trajectory is given by , Q and R are weight matrices, which penalize the state tracking error and the magnitude of the control variable, respectively, and T represents the transpose.

4. A six-degree-of-freedom wave compensation boarding equipment suitable for complex sea conditions according to claim 3, characterized in that, The constraints include: Control input constraints: u min ≤u(k+i)≤u max , State constraints: x min ≤x(k+i)≤x max , Control Increment Constraint: |Δu(k+i)|≤Δu max , Where u represents the control input, including the output force and speed of the hydraulic servo drive system, x represents the system state, including the workspace boundary of the Stewart parallel mechanism and the angular limit of the gangway joint, Δu max This indicates the preset maximum control increment.

5. A six-degree-of-freedom wave compensation boarding equipment suitable for complex sea conditions according to claim 4, characterized in that, The feedforward compensation involves feeding forward the sum of the nonlinear model predictive control estimates and disturbances to the control terminal to obtain the final control quantity u. u=(u NMPC -z3) / b0, Among them, u NMPC z3 is the initial control signal for the nonlinear model predictive control output, z0 is the total disturbance, and b0 is the nominal value of the control gain.

6. A six-degree-of-freedom wave compensation boarding equipment suitable for complex sea conditions according to claim 1, characterized in that, The observation bandwidth of the nonlinear extended state observer is 10-20Hz, and the disturbance estimation update period is no more than 0.02 seconds; the prediction time domain of the nonlinear model predictive control is 1.5-3 seconds, the control time domain is 0.5-1.5 seconds, and the control period is no more than 0.02 seconds.

7. A six-degree-of-freedom wave compensation boarding equipment suitable for complex sea conditions according to claim 1, characterized in that, The intelligent sensing system includes an inertial navigation system, a laser rangefinder, and a visual recognition system.

8. A six-degree-of-freedom wave compensation control method, applied to the boarding equipment described in claim 1, characterized in that, Including the following steps: Step S1: Collect motion and pose data of the ship and target platform through the intelligent sensing system, and fuse and process the data to generate motion state measurement values ​​of the ship and target platform. Step S2: Based on the measured motion state values, estimate the system state and total disturbance in real time using a nonlinear extended state observer; Step S3: Based on the system state and total disturbance, multi-step prediction and rolling optimization are performed through nonlinear model predictive control to generate preliminary control signals, and feedforward compensation is performed on the total disturbance to form the final control quantity. Step S4: Perform inverse kinematics calculation on the Stewart parallel mechanism and joint motion calculation on the boarding gangway system for the final control quantity. Send the calculation results to the hydraulic servo drive system to perform compensation actions.

Images