A method, system, device and medium for self-adaptive control of a rake head of a trailing suction hopper dredger based on soil identification

Abstract

The present application relates to the technical field of drag suction ship, and specifically provides a drag suction ship head self-adaptive control method, system, device and medium based on soil identification, comprising: collecting oil cylinder state, rake head posture, mud parameters and ship speed data in real time through a multi-source sensor network; decoupling effective cutting force from the oil cylinder pressure based on a dynamics model; calculating instantaneous power according to the effective cutting force and the speed, calculating dry soil production according to the mud flow rate and concentration, and then calculating a real-time specific energy index; comparing the specific energy index with a preset threshold value, and comprehensively determining the soil type in combination with the fluctuation trend of the effective cutting force; and adaptively adjusting the dredging parameters according to the soil type. The present application can accurately identify soil in real time and adaptively match the dredging process, significantly improving the operation efficiency and equipment safety under complex geological conditions.

Description

Technical Field

[0001] This invention belongs to the field of trailing suction hopper dredger technology, specifically relating to an adaptive control method, system, equipment, and medium for the trailing suction hopper dredger head based on soil identification. Background Technology

[0002] Soil information is crucial for controlling the dredging and loading processes of trailing suction hopper dredgers. The physical and mechanical properties of different soil types (such as shear strength, cohesion, and SPT blow count) directly determine the force required for the rake teeth to cut into the soil. Differences in cutting resistance not only affect the equipment load but also determine the soil-breaking mechanism of high-pressure water jetting. Conventional high-pressure water jetting works well in soft clay but is difficult to effectively penetrate the soil to assist in breaking up hard clay.

[0003] However, the skid head of a trailing suction hopper dredger cannot directly sense soil information. Currently, the mainstream method for obtaining soil type information is to take samples by an exploration vessel before dredging operations, send them to a laboratory for analysis, and then transmit the results back to the dredger for reference by construction personnel. This offline analysis method is time-consuming, costly, and difficult to cope with real-time changes in soil quality during construction. Summary of the Invention

[0004] To address the aforementioned shortcomings of existing technologies, this invention provides an adaptive control method, system, device, and medium for the hopper dredger head based on soil identification, in order to solve the aforementioned technical problems.

[0005] In a first aspect, the present invention provides an adaptive control method for the skid head of a skid suction hopper based on soil identification, comprising: Real-time data collection of hydraulic cylinder status, rake head attitude angle, mud flow rate, mud concentration, and ship speed is achieved through a multi-source sensor network. Using a pre-built dynamic model as a constraint, the effective cutting force is decoupled from the cylinder state data through state estimation. The decoupling process includes subtracting the rake head gravity component based on the rake head attitude angle and suppressing wave periodic loads. Instantaneous cutting power is calculated based on the effective cutting force and ship speed; dry soil yield is calculated based on mud flow rate and mud concentration; and real-time specific energy index is calculated based on the instantaneous cutting power and the dry soil yield. The specific energy index is compared with a preset threshold, and the soil type is determined comprehensively based on the fluctuation trend of the effective cutting force. If the effective cutting force reaches the preset threshold and the specific energy index increases simultaneously, it is determined to be hard soil. If the effective cutting force increases but the specific energy index does not reach the preset threshold, it is determined to be an interference signal. The dredging control parameters of the rake head are adjusted according to the soil type and the preset mapping relationship between soil type and dredging control parameters.

[0006] In an optional implementation, the effective cutting force is decoupled from the cylinder state data using a pre-built dynamic model as a constraint, including: The dynamic model is converted into a state-space equation, the effective cutting force and wave periodic load are listed as system state variables to be estimated, and the measured pressure in the cylinder state data is used as the observation variable. A state observer is constructed, using the state transition matrix of the state space equation as the equality constraint for state prediction, and the gravity component calculated in real time based on the rake head attitude angle as the feedforward input. The estimated values ​​of the system state variables are then corrected online using the residual between the measured pressure and the pressure predicted by the state observer. The cutting force component in the state estimate output by the state observer after iterative convergence is the decoupled effective cutting force.

[0007] In an alternative implementation, the state observer employs a Kalman filter or an extended state observer.

[0008] In an optional implementation, the online closed-loop correction is implemented using a Kalman filter framework, including: A state-space equation is established based on the dynamic model, wherein: the state transition matrix is ​​used to recursively calculate the state prediction value at the next moment from the state estimate value at the current moment; the observation matrix is ​​used to map the state prediction value to the observed prediction value of the cylinder pressure. The difference between the measured pressure and the observed predicted value is used as the residual signal; The residual signal is multiplied by a preset observer gain matrix to generate a correction term; The correction term is superimposed on the state prediction value to obtain the corrected state estimate at the current time, thus completing one closed-loop correction iteration.

[0009] In an optional implementation, the state observer is an extended state observer, and the system state variables are estimated in an integrated manner, taking the effective cutting force and wave periodic load as extended states; the online closed-loop correction is achieved through a nonlinear error feedback control law, including: The residual between the measured pressure and the predicted pressure is used as input; After undergoing a nonlinear transformation using a preset nonlinear function, it is multiplied by a gain coefficient; The generated correction term compensates for the first derivative of the expanded state in real time, thereby achieving active suppression of total system disturbances such as wave loads.

[0010] In an optional implementation, the instantaneous cutting power is calculated based on the effective cutting force and the ship's speed, the dry soil yield is calculated based on the mud flow rate and mud concentration, and the real-time specific energy index is calculated based on the instantaneous cutting power and the dry soil yield, including: The product of the effective cutting force and the ship's speed is taken as the instantaneous cutting power, which characterizes the instantaneous energy consumption rate of the rake head breaking the soil layer. The product of the mud flow rate and the mud concentration is taken as the dry soil yield, which represents the actual volume of soil solids broken and sucked up by the harrow head per unit time. The ratio obtained by dividing the instantaneous cutting power by the dry soil yield is used as the real-time specific energy index, which represents the mechanical energy consumed by breaking a unit volume of soil.

[0011] In an optional implementation, the dredging control parameters of the rake head are adjusted according to the soil type and a preset mapping relationship between soil type and dredging control parameters, including: A basic control mode is selected based on the soil type and the mapping relationship. The basic control mode includes basic parameter values ​​for multiple dredging control parameter items. The basic control mode is then sent to the execution agency. The deviation between the real-time specific energy index and the preset target specific energy index is obtained. The continuously obtained deviation value and the rate of change of the deviation value are used as input variables of the fuzzy controller. The fuzzy controller calculates and outputs the control parameter correction amount. Based on the control parameters, the current dredging control parameters of the road adjustment actuator are corrected. The dredging control parameters include the cylinder pressure setpoint, speed command, high-pressure flushing opening, and rake head pitch angle.

[0012] Secondly, the present invention provides an adaptive control system for the scraper head of a scraper suction hopper based on soil identification, comprising: The data acquisition module is used to collect real-time data on cylinder status, rake head attitude angle, mud flow rate, mud concentration, and ship speed through a multi-source sensor network. The first calculation module is used to decouple the effective cutting force from the cylinder state data by state estimation using a pre-built dynamic model as a constraint. The decoupling process includes subtracting the rake head gravity component based on the rake head attitude angle and suppressing wave periodic loads. The second calculation module is used to calculate the instantaneous cutting power based on the effective cutting force and the ship speed, calculate the dry soil yield based on the mud flow rate and mud concentration, and calculate the real-time specific energy index based on the instantaneous cutting power and the dry soil yield. The soil identification module is used to compare the specific energy index with a preset threshold and comprehensively determine the soil type based on the fluctuation trend of the effective cutting force. If the effective cutting force reaches the preset threshold and the specific energy index increases simultaneously, it is determined to be hard soil. If the effective cutting force increases but the specific energy index does not reach the preset threshold, it is determined to be an interference signal. The parameter adjustment module is used to adjust the dredging control parameters of the rake head according to the soil type and the preset mapping relationship between soil type and dredging control parameters.

[0013] Thirdly, a device is provided, comprising: Memory is used to store the adaptive control program for the hopper head of the hopper guillotine based on soil identification. A processor is configured to implement the steps of the soil-identification-based adaptive control method for the hopper hopper head as provided in the first aspect when executing the soil identification-based adaptive control program for the hopper hopper head.

[0014] Fourthly, a computer-readable medium is provided, on which an adaptive control program for the scraper head of a hopper dredger based on soil identification is stored. When the adaptive control program for the scraper head of a hopper dredger based on soil identification is executed by a processor, it implements the steps of the adaptive control method for the scraper head of a hopper dredger based on soil identification provided in the first aspect.

[0015] The beneficial effects of this invention are as follows: The adaptive control method, system, equipment, and medium for trailing suction hopper dredgers based on soil identification provided by this invention introduces the real-time calculated specific energy index as the core feature for soil identification, supplemented by the fluctuation trend of effective cutting force for comprehensive judgment, thereby realizing the real-time and accurate identification of soil type by the trailing suction hopper dredger during operation; the identification result directly drives the fuzzy controller to adaptively adjust process parameters such as cylinder pressure, speed, and high-pressure water flushing, so that the dredging strategy always maintains optimal matching with the current soil quality, thereby effectively avoiding the efficiency loss and equipment impact caused by soil misjudgment in traditional methods, and significantly improving the efficiency and safety of dredging operations under complex geological conditions. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic flowchart of a method according to an embodiment of the present invention.

[0018] Figure 2 This is a schematic flowchart illustrating the method for identifying soil types according to an embodiment of the present invention.

[0019] Figure 3 This is a schematic block diagram of a system according to an embodiment of the present invention.

[0020] Figure 4 This is a schematic diagram of the structure of a device provided in an embodiment of the present invention. Detailed Implementation

[0021] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0023] The adaptive control method for the scraper head of a scraper suction hopper based on soil identification provided in this embodiment of the invention is executed by a computer device, and correspondingly, the adaptive control system for the scraper head of a scraper suction hopper based on soil identification runs in the computer device.

[0024] Figure 1 This is a schematic flowchart illustrating a method according to an embodiment of the present invention. Wherein, Figure 1 The executing entity can be an adaptive control system for the skid head of a skid suction hopper based on soil identification. Depending on different requirements, the order of the steps in this flowchart can be changed, and some can be omitted.

[0025] like Figure 1 As shown, the method includes: S1. Real-time acquisition of hydraulic cylinder status data, rake head attitude angle, mud flow rate, mud concentration, and ship speed through a multi-source sensor network; S2. Using a pre-built dynamic model as a constraint, the effective cutting force is decoupled from the cylinder state data through state estimation. The decoupling process includes subtracting the rake head gravity component based on the rake head attitude angle and suppressing wave periodic loads. S3. Calculate the instantaneous cutting power based on the effective cutting force and the ship speed, calculate the dry soil yield based on the mud flow rate and mud concentration, and calculate the real-time specific energy index based on the instantaneous cutting power and the dry soil yield; S4. Compare the specific energy index with a preset threshold, and comprehensively determine the soil type based on the fluctuation trend of the effective cutting force; wherein, if the effective cutting force reaches the preset threshold and the specific energy index increases synchronously, it is determined to be hard soil; if the effective cutting force increases but the specific energy index does not reach the preset threshold, it is determined to be an interference signal. S5. Adjust the dredging control parameters of the rake head according to the soil type and the preset mapping relationship between soil type and dredging control parameters.

[0026] The following describes the process of real-time soil type identification and adaptive control based on the real-time soil type identification of this invention. The process of real-time soil type identification is as follows: Figure 2 As shown.

[0027] In one embodiment of the present invention, based on step S1, the following will provide a possible embodiment and describe its specific implementation in a non-limiting manner.

[0028] The multi-source sensor network described in this step is deployed in the scraper head system and mud delivery pipeline of the trailing suction hopper dredger, with the following specific configuration: Hydraulic Cylinder Status Sensors: Hydraulic cylinder pressure sensors and hydraulic cylinder stroke sensors are installed on the rake head lifting cylinder (or jacking cylinder). The hydraulic cylinder pressure sensors are HBM or WIKA high-precision pressure transmitters with a range of 0~40MPa and an accuracy of 0.1%FS. They are installed in the rodless and rod chambers of the cylinder, respectively, to collect real-time pressure data from both chambers. The hydraulic cylinder stroke sensors are MTS magnetostrictive displacement sensors with a range of 0~2000mm and a resolution of 0.01mm. They are installed axially along the cylinder body to collect the real-time extension length of the piston rod.

[0029] Rake Head Attitude Sensor: An inclinometer or inertial measurement unit (MRU) is installed on the rake head body to collect attitude data such as pitch and roll angles. A rake tube angle sensor is installed on the rake tube connected to the active rake head to detect the vertical and horizontal angles of the rake tube. The attitude sensor uses a high-precision Kongsberg or iXBlue inertial measurement unit to provide roll, pitch, yaw, and three-axis acceleration data, with a sampling frequency of not less than 100Hz.

[0030] The mud parameter sensors are installed on the straight section of the rake pipe using a mud flow meter and a mud concentration (density) meter. The mud flow meter selected is the KROHNE OPTIFLUX4000 series electromagnetic flow meter. This flow meter uses a full-bore flow tube and a smooth, wear-resistant lining material (ceramic discs, polyurethane, or soft rubber), featuring a robust structure and easy installation, suitable for two-phase flow measurement of mud with high sand content. The mud concentration meter can be a gamma-ray densitometer (such as Berthold) or an electrochromatographic density meter. The gamma-ray densitometer is a non-contact measuring instrument with high measurement accuracy, suitable for real-time detection of mud concentration in pipelines.

[0031] Ship speed sensor: Ship speed data is acquired via DGPS / RTK-GPS with an accuracy better than 0.1m / s, and is installed in an unobstructed position high up on the hull.

[0032] All sensors are connected to the data acquisition unit via signal cables. The upper layer of the data acquisition system uses an industrial Ethernet network consisting of servers and clients, while the lower layer uses a Siemens S7-1500 series programmable logic controller (PLC) to form a control network, enabling real-time data acquisition and monitoring of the underlying sensor devices. The PLC control program is written using Siemens STEP7. Analog signals from each sensor (such as pressure and stroke) are amplified and filtered by the signal conditioning module before being input to the PLC via the analog input module (AI); digital signals (such as DGPS speed data) are input to the PLC via a serial communication module or an Ethernet module. The sampling frequency of all sensor data is uniformly set to 100Hz to ensure the synchronization of various physical quantities on the time axis.

[0033] In one embodiment of the present invention, based on step S2, a possible embodiment will be given below, and its specific implementation will be described in a non-limiting manner.

[0034] S201, Force Analysis and Dynamic Modeling of the Rake Head-Hydraulic Cylinder System Taking the rake head and the piston rod of the hydraulic cylinder fixed to it as the objects of force analysis, a force balance equation is established along the cylinder axis. The cylinder pressure is measured by the sensor. The relationship between the various mechanical components is as follows:

[0035] in: The effective working area of ​​the hydraulic cylinder (calculated from the cylinder diameter and piston rod diameter, and is a known constant); The actual soil cutting force to be solved is generated by soil resistance and is the target output of this step. The component of the rake head's weight along the cylinder axis varies with the rake head's pitch angle, but this variation is entirely determined by the geometric structure and is a known deterministic quantity. The lumped disturbance term includes wave periodic loads, hydraulic cylinder friction, unmodeled dynamics, and other uncertainties, and represents the unknown disturbances that need to be suppressed in this step.

[0036] Rewrite the above equation in terms of pressure:

[0037] make For the actual pressure measured by the sensor, the following definition is given:

[0038] The equivalent pressure value of the gravitational component is used as a known feedforward input.

[0039] S202, Real-time calculation of gravity components and construction of feedforward input The pitch angle of the rake head relative to the horizontal plane is collected in real time using an inertial measurement unit installed on the rake head body. and roll angle According to the principles of multibody kinematics, the projection of the gravity vector at the center of gravity of the rake head onto the axial direction of the hydraulic cylinder is:

[0040] in The unloaded gravity constant of the rake head (obtained from factory parameters or weighing calibration). The angle between the line connecting the rake head hinge point and the center of gravity and the fixed structure relative to the cylinder axis.

[0041] The feedforward input term is:

[0042] because , , All are known constants. , For real-time sensor readings, therefore For deterministic quantities that can be calculated in real time, they are directly substituted into the recursive equation of ESO in each sampling period.

[0043] S203, Design of Extended State Observer (I) System Expansion State Modeling Define system state variables. The core idea of ​​ESO is to treat all unknown disturbances in the system (wave loads, friction, unmodeled dynamics, etc.) as a lumped disturbance and expand it into new state variables.

[0044] Let the state variables be defined as follows: state Equivalent total resultant pressure, i.e., the sum of cutting force and lumped disturbance (including residual dynamics after gravity feedforward):

[0045] Expansion state : Centralized disturbance The comprehensive characterization of its rate of change is defined as:

[0046] For disturbances dominated by wave loading, the rate of change Assume it is bounded.

[0047] Actual pressure As observed variables, based on force analysis, the system can be written in the following second-order extended state-space form:

[0048] in: This is the input gain coefficient (theoretical value is 1, but can be fine-tuned according to calibration in actual engineering). The feedforward gravity input is calculated in real time based on the attitude angle; Let be the rate of change of the lumped disturbance, which is assumed to be bounded and unknown; The actual pressure measured by the sensor.

[0049] It should be noted that the effective cutting force It is not explicitly stated in the above state equations. This is because, within the ESO framework, In fact, it includes both cutting force and lumped disturbance. The lumped disturbance will be subtracted in the subsequent decoupling process, thereby indirectly restoring the effective cutting force component.

[0050] (II) Construction of Nonlinear Extended State Observer For the aforementioned second-order system, a nonlinear extended state observer (NLESO) is constructed, whose discretized recursive equation is:

[0051] in: For the first The observation error of the step (the residual between the measured pressure and the state estimate); For state Tracking estimate of (total resultant force equivalent pressure); In an expanding state Tracking estimates of (lumped disturbance equivalent pressure); s is the sampling period (corresponding to a sampling frequency of 100 Hz); , The observer gain coefficient; Nonlinear factor (usually taken as) (Typical value 0.5) The width of the linear interval (typical value 0.01 to 0.1); For the first The gravity feedforward input is based on real-time attitude angle calculation; It is a nonlinear function, defined as:

[0052] The characteristic of this nonlinear function is that: when the observation error... When it is large, Function output The gain is relatively small to prevent the observer from oscillating due to excessive initial error; when the observation error... When smaller, Function output It has high gain, ensuring that the observer can quickly track small errors. This characteristic of "small gain for large errors and large gain for small errors" enables the ESO to have good performance in both the initial convergence phase and the steady-state phase.

[0053] (III) Tuning the observer gain Observer Gain , The selection of the pole placement method directly affects the convergence speed and disturbance rejection performance of the observer.

[0054]

[0055] in The observer bandwidth (unit: rad / s) is the only parameter in the ESO that needs to be tuned. The selection principle is as follows: The observer bandwidth should be much larger than the dominant frequency of the wave disturbance (typically 0.05–0.3 Hz, or 0.314–1.885 rad / s), generally taken as... rad / s, to ensure the observer's ability to quickly track wave disturbances; at the same time, It should not be too large to avoid amplifying high-frequency noise from the sensor. A typical value is... rad / s, corresponding to the convergence time of the observer at this bandwidth is approximately s, which is much faster than the characteristic timescale of wave changes.

[0056] (iv) Decoupling and output of effective cutting force After ESO converges recursively after several sampling periods, Approaching the actual total force (Superposition of cutting force and lumped disturbance). Approaching the actual lumped disturbance .

[0057] Then the effective cutting force Calculate using the following formula:

[0058] Its physical meaning is: from the estimated value of the total resultant force Subtracting the lumped disturbance estimate This yields the pure, effective cutting force equivalent pressure, which is then multiplied by the effective area of ​​the hydraulic cylinder. Restore it to the dimensions of force.

[0059] It is worth noting the gravitational component In the state update equation of ESO, explicit participation The prediction calculation. When the residual At that time, due to The predicted value includes the contribution of gravity synchronously through the feedforward term, while the measured pressure... The residual also includes the actual contribution of gravity, and the two cancel each other out in the residual, so the gravitational component does not enter. The estimated value also does not appear in the final value. In the expression, the gravitational component is thus accurately deducted.

[0060] Unlike the above implementation method which uses an extended state observer, this implementation method uses a Kalman filter to achieve statistically optimal estimation and suppression of wave periodic loads. The specific implementation method is as follows: I. Dynamic Modeling and Force Decomposition Taking the rake head-hydraulic cylinder system as the research object, its force balance equation in the vertical plane is established. The cylinder pressure measured by the sensor is used as the basis for this study. It is a composite manifestation of multiple forces, and its basic dynamic constraint equations are:

[0061] in: The effective working area of ​​the hydraulic cylinder; The actual cutting force to be solved is denoted as , and is the main state variable of interest. The component of the rake head's weight along the cylinder axis varies with the rake head's pitch angle and is treated as a known feedforward quantity. It is a periodic dynamic load caused by waves, and has low-frequency periodic characteristics; The frictional force of the hydraulic cylinder and hinge points is considered a constant compensation term. The noise observed follows a zero-mean Gaussian white noise distribution.

[0062] The pitch angle of the rake head relative to the horizontal plane is collected in real time using an inertial measurement unit installed on the rake head body. Calculate the real-time component of the rake head's weight along the cylinder axis:

[0063] in Let be the unloaded gravity constant of the rake head. The angle between the line connecting the rake head hinge point and the center of gravity and the fixed structure relative to the cylinder axis.

[0064] Subtracting the gravitational component from the original cylinder pressure yields the residual pressure signal after deducting gravity:

[0065] The residual pressure The main components include the effective cutting force. With wave load The superposition of factors, as well as measurement noise.

[0066] II. State-space modeling The effective cutting force and wave load and their rate of change are listed as system state variables, a discrete state-space model is established, and the dynamic equation is used as the equality constraint for state transition.

[0067] (1) Modeling of wave load shaping filter wave load This can be viewed as the output of a shaping filter driven by white noise. The wave load is modeled as a second-order Markov process:

[0068] in The dominant angular frequency of the wave (wave period measured by ship motion sensors) Sure), The damping ratio is typically taken as 0.1 to 0.3. This is process white noise.

[0069] (2) Discrete state-space equations Define a four-dimensional state vector:

[0070] in The effective cutting force (its main changes come from soil quality changes, and the time scale is relatively slow). , , Let represent the wave load and its derivatives.

[0071] The discretized state equation and observation equation are as follows:

[0072]

[0073] in: for The state transition matrix, obtained by discretizing the above continuous dynamic equations, contains the dynamic characteristics of the wave forming filter and the statistical characteristics of the cutting force change rate, serving as the equality constraints of the dynamic model. For the observation matrix, the state vector is mapped to residual pressure. (That is, the sum of the effective cutting force and the wave load); Let be the process noise vector, which follows a zero-mean Gaussian distribution, and its covariance matrix is ​​. ; The observed noise follows a zero-mean Gaussian distribution with variance . .

[0074] (3) Determination of the noise covariance matrix Process noise covariance matrix This reflects the uncertainty of the state transition model. The process noise intensity of the effective cutting force is related to the rate of soil change, while the process noise intensity of the wave component is determined by the sea state level. Initial values ​​were obtained through statistical analysis of sea trial data and adjusted online during operation using an adaptive update strategy.

[0075] Observation noise variance This reflects the measurement accuracy of pressure and attitude sensors, and is determined by the sensor's nominal accuracy (e.g., 0.1%FS for pressure sensors).

[0076] III. Kalman Filter Recursion The Kalman filter uses the residual pressure obtained in step one. As an observation, the state-space equation is used as the equality constraint for prediction and correction. Through the recursion of the two steps of prediction and update, each state variable is estimated in real time.

[0077] (1) Prediction steps (prior estimation) State prior estimation:

[0078] Prior estimation of error covariance:

[0079] (2) Update steps (posterior correction) Kalman gain calculation:

[0080] Posterior state estimation:

[0081] Posterior estimation of error covariance:

[0082] (3) Wave suppression and effective cutting force output After convergence through recursive iteration over several sampling periods, the estimated value of the first state variable in the posterior state estimation vector is... That is, the effective cutting force output in real time after decoupling. The estimated value of the second state variable This is the optimal estimate of the wave load at the current moment, which can serve as effective supplementary information for the calculation of the specific energy index in subsequent steps.

[0083] IV. Filter Initialization and Parameter Configuration (1) Initial state value Initial state vector Let it be the zero vector, indicating that there is no cutting force and no wave load at the initial moment; Initial error covariance matrix Set as a diagonal matrix, with each diagonal element set according to the physical range of the corresponding state variable, reflecting the uncertainty of the initial state estimation.

[0084] (2) Online acquisition of wave dominant frequency Wave Dominant Frequency Instead of a fixed value, the data on the ship's heave or roll motion is measured in real time by ship motion sensors. The main frequency is extracted by fast Fourier transform and then dynamically updated into the state transition matrix, so that the model constraints of the Kalman filter always match the actual sea conditions.

[0085] (3) Discretization of the state transition matrix The continuous-domain transfer function of the wave-shaping filter is discretized using a zero-order hold (ZOH) and then embedded... The matrix ensures the constraint accuracy of the dynamic model.

[0086] V. Algorithm Discretization and Real-time Execution In shipboard PLCs or industrial control computers, the above Kalman filter recursion is executed periodically at a sampling frequency of 100Hz. Within each sampling period, the system executes the following timing logic sequentially: (1) Read the current time , Calculate and subtract the gravitational component , and obtain the observed values ; (2) Read data from ship motion sensors and update the dominant wave frequency. Refresh the state transition matrix ; (3) Perform the prediction step of Kalman filtering and calculate the state prior estimate and prior covariance; (4) Perform the update step of Kalman filtering, calculate the Kalman gain, and use the residual between the observed and predicted values ​​to perform weighted correction on the state prior estimate; (5) Extract the effective cutting force from the posterior state estimation vector. As for the current moment Output to the energy observer module; (6) Save the posterior covariance matrix as the initial value for the next prediction step, so as to realize the continuous online recursive update of the state estimate.

[0087] Compared to the extended state observer, the Kalman filter-based implementation has the following engineering advantages: Estimation in the best statistical sense: Under the condition that the statistical characteristics of process noise and observation noise are known, the Kalman filter provides the best state estimation in the sense of minimum mean square error, and the estimation accuracy is higher. Time-varying gain automatic adjustment: Kalman gain It adapts adaptively to the covariance matrix, eliminating the need for manual tuning of fixed gains and providing better adaptability to varying sea conditions. Uncertainty Quantification: Covariance Matrix It reflects the current estimated confidence interval in real time, providing uncertainty information for subsequent decision-making; Engineering feasibility: The Kalman filter algorithm is mature and has good numerical stability, and there are many successful deployment cases in industrial embedded systems.

[0088] In one embodiment of the present invention, based on step S3, the following will provide a possible embodiment and describe its specific implementation in a non-limiting manner.

[0089] S301, Instantaneous Cutting Power Calculation Based on the effective cutting force output in step two (Unit: N), and ship speed obtained in real time via DGPS. (Unit: m / s) Calculate the instantaneous cutting power consumed by the rake head in breaking up the soil layer. (Unit: W):

[0090] This power rating represents the effective mechanical work done by the rake on the soil layer per unit time.

[0091] In practical engineering, ship speed The speed is directly taken from the ground speed measurement value of DGPS. When the ship is in a turning or speed change transition, the speed change rate is large. At this time, the speed signal can be filtered by moving average in the PLC (window duration 2-5 seconds) to obtain smooth and timely power calculation results.

[0092] S302, Calculation of Dry Soil Yield Dry soil yield (Unit: m) 3 The volume of soil solids (excluding free water) actually cut and sucked up by the rake per unit time is represented by the mud volumetric flow rate and the mud solids volume concentration.

[0093] in: Volumetric flow rate of the rake pipe mud (unit: m³) 3 / s, which is collected in real time by an electromagnetic flow meter installed in the straight section of the rake pipe; The volume concentration of solid particles in the mud (dimensionless, 0–1) is collected in real time by a mud concentration meter installed downstream of the flow meter.

[0094] The mud flow meter and concentration meter should be installed on a straight section of the rake pipe (the length of the straight pipe section before and after the rake pipe should not be less than 5 times the pipe diameter) to ensure uniform flow velocity distribution and thorough concentration mixing across the measurement cross-section. The concentration meter outputs either solid mass concentration or volume concentration. If the output is mass concentration... The volume concentration can then be obtained through the following conversion:

[0095] in The density of the mud mixture, The true density of soil particles (usually taken as 2650 kg / m³) 3 ).

[0096] When the concentration meter signal fluctuates abnormally due to wear, scaling, or bubble interference, the system automatically switches to the soft measurement backup model. Using the mud pump speed, suction vacuum, and rake tube pressure drop as input features, the system calculates the real-time mud concentration through a pre-trained offline multiple regression model or neural network model, and multiplies it with the flow velocity data to obtain a backup estimate of dry soil yield, ensuring the continuity of the specific energy index calculation.

[0097] S303, Real-time Specific Energy Index Calculation Instantaneous cutting power Divide by dry soil yield The real-time specific energy index is obtained. (Unit: kJ / m³ or MPa):

[0098] The physical meaning of the specific energy index is: the mechanical energy consumed in breaking a unit volume of soil, which is the core indicator for measuring the difficulty of soil cutting.

[0099] To ensure the numerical stability of the calculation results, the following protection logic is set in the PLC program: when Below the preset minimum value (e.g.) m 3 When the speed is / s), it is determined to be an empty excavation or loss condition. Force the value to zero or retain the previous valid value to prevent numerical overflow; when Exceeding the preset limit (e.g.) kJ / m 3 When the condition is determined to be extremely hard / rock-containing, an alarm signal is triggered.

[0100] The above calculations are performed in real time in a PLC or industrial computer at a sampling frequency of 100Hz. After each calculation, the current value is recorded. The value is stored in a circular buffer (duration 60 seconds).

[0101] In one embodiment of the present invention, based on step S4, the following will provide a possible embodiment and describe its specific implementation in a non-limiting manner.

[0102] Soil type is determined based on two dimensions: the numerical range of the specific energy index and the fluctuation trend of the effective cutting force, using a three-level classification logic. (a) Setting the specific energy index threshold Two specific energy index thresholds are preset: First preset threshold Used to distinguish between soft and hard soils; initial values ​​are determined by test excavation calibration (typical range 1000–3000 kJ / m). 3 ); Second preset threshold Used to distinguish between hard soil and extremely hard / rocky soil. (Typical value range: 8000~15000kJ / m) 3 ).

[0103] (II) Three-level classification judgment logic The following decision is performed in each sampling period: like It is determined to be soft soil (such as silt or loose sand). like : Enter the effective cutting force synchronization check and extract the effective cutting force within the window of the previous 10 seconds. The trend of change (the difference between the first and last means of the calculation window) is determined. If the difference is greater than the preset positive change threshold (i.e. the effective cutting force is increasing), it is determined to be hard soil (such as hard clay or dense sand). If the effective cutting force does not increase or the increase is insufficient, it is determined to be an interference signal (false pressure increase caused by wave impact or attitude change), and the soil type of the previous moment remains unchanged. like If the soil is determined to be extremely hard or contains stones, an audible and visual alarm will be triggered, prompting the operator to pay attention to the status of the rake head.

[0104] Soil type determination results are based on enumerated variables. Output in the form of (soft soil / hard soil / extremely hard soil).

[0105] In one embodiment of the present invention, based on step S5, a possible embodiment will be given below, and its specific implementation will be described in a non-limiting manner.

[0106] The system internally stores a pre-defined "soil type - control parameter" mapping table, covering four types of execution parameters: hydraulic cylinder pressure setpoint, speed command, high-pressure water jet opening, and drag head pitch angle. After soil type determination, the system uses the determination result as an index to look up the target parameter value corresponding to the current soil type. If the soil type changes, the control parameters do not undergo abrupt changes, but instead smoothly transition to the target value with a time constant of 3-10 seconds through first-order inertial filtering to avoid shock to the equipment. The final control command is output to the hydraulic system, propulsion system, and high-pressure water jet system via fieldbus for execution. The baseline values ​​in the mapping table are pre-calibrated according to the different models of trailing suction hopper dredgers and the operating sea area, and support subsequent online self-learning updates.

[0107] To further improve the matching degree between dredging control parameters and construction soil conditions, a fuzzy control algorithm is used to fine-tune the parameter values ​​based on the basic control mode: S501, Target Energy Index Setting Target energy index The optimal target value is automatically selected based on the current soil type: Soft soil: (The goal is to maximize the shallowing efficiency, so a lower value is usually chosen.) Hard soil: (Aiming for maximum output while taking energy consumption into account).

[0108] and The initial value was determined by offline experiments and continuously optimized during operation through a self-learning module.

[0109] S502, Fuzzy Controller Input Variables Define two input variables: deviation This reflects the degree to which the current specific energy deviates from the target; Deviation change rate This reflects the changing trend of specific energy.

[0110] S503, Input Fuzzification deviation The universe of discourse is divided into five fuzzy subsets: negative large (NB), negative small (NS), zero (ZO), positive small (PS), and positive large (PB), and a triangular membership function is used.

[0111] Deviation change rate The domain of discourse is divided into three fuzzy subsets: negative (N), zero (Z), and positive (P), and the triangular membership function is also used.

[0112] S504, Fuzzy Inference Rules The output variable of the fuzzy controller is the hydraulic cylinder pressure correction amount. (Unit: MPa) and speed correction (Unit: kn). The core reasoning rules are as follows: Rule 1: If deviation It is positive (PB) and the rate of change of deviation If it is positive (P), then the cylinder pressure correction amount is... For positive (PB), speed correction amount The value is negative (NB). This rule corresponds to the condition where the specific energy index is far above the target value and is still rising. In this case, it is necessary to significantly increase the cylinder pressure to enhance the cutting ability, while significantly reducing the speed to ensure sufficient cutting time so that the specific energy index can quickly fall back to the target range.

[0113] Rule 2: If deviation It is positive (PB) and the rate of change of deviation If it is zero (Z), then the cylinder pressure correction amount is zero. Positive small (PS), speed correction amount The value is negative (NS). This rule corresponds to the operating condition where the specific energy index is high but has stabilized. In this case, only minor adjustments to the parameters are needed.

[0114] Rule 3: If deviation The positive small (PS) and the rate of change of deviation If it is negative (N), then the cylinder pressure correction amount is... Zero (ZO), speed correction The value is negative (NS). This rule corresponds to the condition where the specific energy index is slightly higher than the target but has begun to decline. In this case, the current cylinder pressure is kept unchanged, and the speed is reduced only slightly until the specific energy index naturally returns to the target range.

[0115] Rule 4: If deviation The value is negative (NB) and the rate of change of deviation is large. If it is negative (N), then the cylinder pressure correction amount is... For negative large (NB), speed correction amount The value is PB. This rule corresponds to the condition where the specific energy index is far below the target value and is still declining. In this case, it is necessary to significantly reduce the cylinder pressure to avoid excessive cutting by the rake head, while significantly increasing the speed to improve the shoveling efficiency, so that the specific energy index can quickly rise back to the target range.

[0116] Rule 5: If deviation The value is negative (NB) and the rate of change of deviation is large. If it is zero (Z), then the cylinder pressure correction amount is zero. For negative small (NS), speed correction amount The value is positive (PS). This rule corresponds to the operating condition where the specific energy index is low but has tended to stabilize. In this case, only minor adjustments to the parameters are needed.

[0117] Rule 6: If deviation If it is zero (ZO), then regardless of the rate of change of deviation... What value should be used for the hydraulic cylinder pressure correction? and speed correction All values ​​are zero (ZO). This rule applies to situations where the specific energy index is near the target value. In this case, the current control parameters are kept unchanged to avoid unnecessary adjustments that could cause system oscillations.

[0118] The reasoning adopts Mamdani-type fuzzy reasoning. The activation strength of each rule is calculated by the minimum value (min) of the input membership degree, and the synthesis of the output of each rule adopts the maximum value (max) method.

[0119] S505, Defuzzification The centroid method is used to convert the blurred output into a precise correction value:

[0120] in For the first The activation strength of the rule, , For the first Each rule outputs the centroid x-coordinate of the membership function.

[0121] S506, Control Parameter Synthesis and Output The correction value output by the fuzzy controller is added to the current basic control parameters to obtain the final control command:

[0122]

[0123] in and These are the basic control parameters corresponding to the current soil type (determined by empirical table lookup).

[0124] The adjustment of the high-pressure water flushing opening and the pitch angle of the rake head adopts discrete logic: the step adjustment is only performed when the soil type changes (from soft soil to hard clay: turn on the water flushing and increase the pitch angle; from hard clay to soft soil: turn off the water flushing and decrease the pitch angle), to avoid coupling oscillation with the continuous adjustment of the cylinder pressure and speed.

[0125] S507, Smooth Transition of Control Parameters When the soil type changes, the hydraulic cylinder pressure setting and speed command do not change abruptly, but instead undergo a smooth transition through first-order inertial filtering:

[0126] Where the time constant The settings are adaptively adjusted based on the current operating conditions (3-10 seconds) to avoid the equipment being subjected to impact loads.

[0127] The final control commands are output to the hydraulic system drive, propulsion system frequency converter and high-pressure flushing valve drive via Profinet / Profibus fieldbus, completing a complete intelligent control closed loop from soil identification to parameter adjustment.

[0128] In some embodiments, the soil-identification-based adaptive control system for the trailing suction hopper head may include multiple functional modules composed of computer program segments. The computer programs for each program segment in the soil-identification-based adaptive control system may be stored in the memory of a computer device and executed by at least one processor to perform (see details). Figure 1 (Description) Function of adaptive control of the hopper hopper head based on soil identification.

[0129] In this embodiment, the soil-identification-based adaptive control system for the suction hopper head can be divided into multiple functional modules according to its functions, such as... Figure 3 As shown. The module referred to in this invention is a series of computer program segments that can be executed by at least one processor and perform a fixed function, and is stored in memory. In this embodiment, the functions of each module will be described in detail in subsequent embodiments.

[0130] The data acquisition module is used to collect real-time data on cylinder status, rake head attitude angle, mud flow rate, mud concentration, and ship speed through a multi-source sensor network. The first calculation module is used to decouple the effective cutting force from the cylinder state data using a pre-built dynamic model as a constraint. The decoupling process includes subtracting the rake head gravity component based on the rake head attitude angle and suppressing wave periodic loads. The second calculation module is used to calculate the instantaneous cutting power based on the effective cutting force and the ship speed, calculate the dry soil yield based on the mud flow rate and mud concentration, and calculate the real-time specific energy index based on the instantaneous cutting power and the dry soil yield. The soil identification module is used to compare the specific energy index with a preset threshold and comprehensively determine the soil type based on the fluctuation trend of the effective cutting force. If the effective cutting force reaches the preset threshold and the specific energy index increases simultaneously, it is determined to be hard soil. If the effective cutting force increases but the specific energy index does not reach the preset threshold, it is determined to be an interference signal. The parameter adjustment module is used to adjust the dredging control parameters of the rake head according to the soil type and the preset mapping relationship between soil type and dredging control parameters.

[0131] Figure 4The soil-identification-based adaptive control method for the scraper head of a suction hopper dredger provided in this application embodiment can be applied to equipment. Those skilled in the art will understand that the equipment structure involved in the embodiments of this invention does not constitute a limitation on the equipment. The equipment may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. Specifically, the equipment 400 may include: a processor 410, a memory 420, and a communication unit 430. These components communicate through one or more buses. Those skilled in the art will understand that the server structure shown in the figures does not constitute a limitation on the invention; it may be a bus topology or a star topology, and may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements.

[0132] The present invention also provides a computer medium, wherein the computer medium may store a program, which, when executed, may include some or all of the steps provided in the embodiments of the present invention.

[0133] The same or similar parts between the various embodiments in this specification can be referred to mutually. In particular, the device embodiments are basically similar to the method embodiments, so the description is relatively simple, and the relevant parts can be referred to the description in the method embodiments.

[0134] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems or modules may be electrical, mechanical, or other forms.

[0135] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0136] In addition, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0137] Although the present invention has been described in detail with reference to the accompanying drawings and preferred embodiments, the present invention is not limited thereto. Various equivalent modifications or substitutions can be made to the embodiments of the present invention by those skilled in the art without departing from the spirit and essence of the invention, and such modifications or substitutions should all be within the scope of the present invention. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should also be covered within the protection scope of the present invention.

Claims

1. An adaptive control method for the scraper head of a suction hopper dredger based on soil identification, characterized in that, include: Real-time data collection of hydraulic cylinder status, rake head attitude angle, mud flow rate, mud concentration, and ship speed is achieved through a multi-source sensor network. Using a pre-built dynamic model as a constraint, the effective cutting force is decoupled from the cylinder state data through state estimation. The decoupling process includes subtracting the rake head gravity component based on the rake head attitude angle and suppressing wave periodic loads. Instantaneous cutting power is calculated based on the effective cutting force and ship speed; dry soil yield is calculated based on mud flow rate and mud concentration; and real-time specific energy index is calculated based on the instantaneous cutting power and the dry soil yield. The specific energy index is compared with a preset threshold, and the soil type is determined comprehensively based on the fluctuation trend of the effective cutting force. If the effective cutting force reaches the preset threshold and the specific energy index increases simultaneously, it is determined to be hard soil. If the effective cutting force increases but the specific energy index does not reach the preset threshold, it is determined to be an interference signal. The dredging control parameters of the rake head are adjusted according to the soil type and the preset mapping relationship between soil type and dredging control parameters.

2. The method according to claim 1, characterized in that, Using a pre-built dynamic model as a constraint, the effective cutting force is decoupled from the cylinder state data, including: The dynamic model is converted into a state-space equation, the effective cutting force and wave periodic load are listed as system state variables to be estimated, and the measured pressure in the cylinder state data is used as the observation variable. A state observer is constructed, using the state transition matrix of the state space equation as the equality constraint for state prediction, and the gravity component calculated in real time based on the rake head attitude angle as the feedforward input. The estimated values ​​of the system state variables are then corrected online using the residual between the measured pressure and the pressure predicted by the state observer. The cutting force component in the state estimate output by the state observer after iterative convergence is the decoupled effective cutting force.

3. The method according to claim 2, characterized in that, The state observer employs a Kalman filter or an extended state observer.

4. The method according to claim 3, characterized in that, The online closed-loop correction is implemented using a Kalman filter framework, including: A state-space equation is established based on the dynamic model, wherein: the state transition matrix is ​​used to recursively calculate the state prediction value at the next moment from the state estimate value at the current moment; the observation matrix is ​​used to map the state prediction value to the observed prediction value of the cylinder pressure. The difference between the measured pressure and the observed predicted value is used as the residual signal; The residual signal is multiplied by a preset observer gain matrix to generate a correction term; The correction term is superimposed on the state prediction value to obtain the corrected state estimate at the current time, thus completing one closed-loop correction iteration.

5. The method according to claim 3, characterized in that, The state observer is an extended state observer, and the system state variables are estimated in an integrated manner by taking the effective cutting force and wave periodic load as extended states; the online closed-loop correction is achieved through a nonlinear error feedback control law, including: The residual between the measured pressure and the predicted pressure is used as input; After undergoing a nonlinear transformation using a preset nonlinear function, it is multiplied by a gain coefficient; The generated correction term compensates for the first derivative of the expanded state in real time, thereby achieving active suppression of total system disturbances such as wave loads.

6. The method according to claim 1, characterized in that, Instantaneous cutting power is calculated based on the effective cutting force and ship speed; dry soil yield is calculated based on mud flow rate and mud concentration; and real-time specific energy index is calculated based on the instantaneous cutting power and dry soil yield, including: The product of the effective cutting force and the ship's speed is taken as the instantaneous cutting power, which characterizes the instantaneous energy consumption rate of the rake head breaking the soil layer. The product of the mud flow rate and the mud concentration is taken as the dry soil yield, which represents the actual volume of soil solids broken and sucked up by the rake head per unit time. The ratio obtained by dividing the instantaneous cutting power by the dry soil yield is used as the real-time specific energy index, which represents the mechanical energy consumed by breaking a unit volume of soil.

7. The method according to claim 1, characterized in that, Based on the soil type and the preset mapping relationship between soil type and dredging control parameters, adjust the dredging control parameters of the rake head, including: A basic control mode is selected based on the soil type and the mapping relationship. The basic control mode includes basic parameter values ​​for multiple dredging control parameter items. The basic control mode is then sent to the execution agency. The deviation between the real-time specific energy index and the preset target specific energy index is obtained. The continuously obtained deviation value and the rate of change of the deviation value are used as input variables of the fuzzy controller. The fuzzy controller calculates and outputs the control parameter correction amount. The current dredging control parameters of the road adjustment actuator are corrected based on the control parameters mentioned above. The dredging control parameters include the hydraulic cylinder pressure setpoint, speed command, high-pressure flushing opening, and rake head pitch angle.

8. An adaptive control system for the scraper head of a suction hopper dredger based on soil identification, characterized in that, include: The data acquisition module is used to collect real-time data on cylinder status, rake head attitude angle, mud flow rate, mud concentration, and ship speed through a multi-source sensor network. The first calculation module is used to decouple the effective cutting force from the cylinder state data through state estimation, using a pre-built dynamic model as a constraint. The decoupling process includes subtracting the rake head gravity component based on the rake head attitude angle and suppressing wave periodic loads. The second calculation module is used to calculate the instantaneous cutting power based on the effective cutting force and the ship speed, calculate the dry soil yield based on the mud flow rate and mud concentration, and calculate the real-time specific energy index based on the instantaneous cutting power and the dry soil yield. The soil identification module is used to compare the specific energy index with a preset threshold and comprehensively determine the soil type based on the fluctuation trend of the effective cutting force. If the effective cutting force reaches the preset threshold and the specific energy index increases simultaneously, it is determined to be hard soil. If the effective cutting force increases but the specific energy index does not reach the preset threshold, it is determined to be an interference signal. The parameter adjustment module is used to adjust the dredging control parameters of the rake head according to the soil type and the preset mapping relationship between soil type and dredging control parameters.

9. An adaptive control device for the scraper head of a suction hopper dredger based on soil identification, characterized in that, include: Memory for storing the adaptive control program for the hopper head of the hopper vessel based on soil identification; A processor is configured to implement the steps of the adaptive control method for the hopper hopper head based on soil identification as described in any one of claims 1-7 when executing the adaptive control program for the hopper hopper head based on soil identification.

10. A computer-readable medium storing a computer program, characterized in that, The readable medium stores an adaptive control program for the hopper guillotine head based on soil identification. When the processor executes the adaptive control program for the hopper guillotine head based on soil identification, it implements the steps of the adaptive control method for the hopper guillotine head based on soil identification as described in any one of claims 1-7.

Images