Closed loop control system and method based on deep sea riser slurry concentration monitoring

Abstract

The application discloses a closed-loop control system and method based on deep-sea ore-lifting pipe slurry concentration monitoring, and the system comprises: a water surface support platform control center for outputting a control strategy; and a hydraulic ore injector for receiving the control strategy and performing adaptive adjustment according to the control strategy; wherein the water surface support platform control center comprises an intelligent control unit and a controller, the intelligent control unit is used for collecting operation parameter data of the hydraulic ore injector, processing the operation parameter data, and obtaining a solid phase volume concentration prediction value through real-time prediction; the controller is used for receiving the solid phase volume concentration prediction value, and determining the control strategy according to the solid phase volume concentration prediction value; and the intelligent control unit comprises a data acquisition module, a data processing module and a prediction module; through the system, the concentration of the slurry in the pipe can be more accurately controlled, and the system can better adapt to complex working conditions such as seabed topography and ore supply fluctuation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of ore pulp concentration control, in particular to a closed-loop control system and method based on deep-sea ore-lifting pipe ore pulp concentration monitoring. BACKGROUND

[0002] The deep-sea ore-lifting pipe is the core conveying device of the deep-sea mining system, and its function is to transport the mined ore (in the form of ore pulp, a mixture of solid mineral particles and liquid) from the seabed to the surface mining ship or platform through a vertical pipe. Deep-sea ore-lifting pipe ore pulp concentration monitoring refers to the technical process of real-time or periodic detection of the solid content of the ore pulp in the conveying pipe of the deep-sea mining system.

[0003] Currently, there are two main ore lifting systems in deep-sea mining: one is the traditional hydraulic lifting system, which uses a submersible pump (multi-stage centrifugal pump) to generate upward flow of slurry in the riser to lift the ore from the seabed to the surface support ship. This system relies on a submersible pump as the core power source, but in actual application, the rapid wear of the pump impeller (about 3 mm in 1.5 hours) and the high risk of blockage cause frequent maintenance and low production efficiency.

[0004] The other is a closed-loop circulation lifting system: seawater is driven by a ship-borne pump to circulate in a vertical closed pipe system connecting the seabed mining point and the surface operation ship. Ore is injected into the high-pressure closed pipe system through a special hydraulic ore injector, which usually consists of a flow control device, multiple ore bins, an ore feeder, and a mixer. Through sequential control of the valves, ore is quantitatively fed to the mixer through the ore feeder, and is fully mixed with circulating seawater under the action of high-pressure water flow to form stable ore pulp before being injected into the lifting pipe. The closed pipe (water supply pipe connected to the lifting pipe) forms a closed circulation loop, so that the pump body only drives seawater circulation without direct contact with the ore pulp containing coarse particles, thereby avoiding the risks of pump body wear and particle blockage in the traditional hydraulic lifting system.

[0005] However, during the use of the deep-sea closed-loop system for ore lifting, the ore bin switching moment is prone to cause pulsation of the ore pulp concentration in the pipe. The existing control logic is difficult to cope with the instantaneous flow rate fluctuations during multi-tank switching, resulting in unstable ore pulp concentration in the ore-lifting pipe, which is prone to pipe blockage at too high a concentration, and reduces commercial mining efficiency at too low a concentration, making it difficult to adapt to complex working conditions such as seabed topography and fluctuating ore supply. SUMMARY

[0006] To solve the above technical problems, the present application provides a closed-loop control system and method based on deep-sea ore-lifting pipe ore pulp concentration monitoring.

[0007] To achieve the above purpose, the technical solution of the embodiments of the present application is as follows:

[0008] In a first aspect, the present application provides a closed-loop control system based on deep-sea slurry concentration monitoring of the ore-transporting pipe, comprising:

[0009] a water surface support platform control center for outputting a control strategy;

[0010] a hydraulic ore injector for receiving the control strategy and performing adaptive adjustment according to the control strategy;

[0011] wherein the water surface support platform control center comprises an intelligent control unit and a controller, the intelligent control unit is configured to collect operation parameter data of the hydraulic ore injector, process the operation parameter data, and obtain a solid phase volume concentration prediction value through real-time prediction; and the controller is configured to receive the solid phase volume concentration prediction value, and determine a control strategy according to the solid phase volume concentration prediction value;

[0012] the intelligent control unit comprises:

[0013] a data collection module configured to collect operation parameter data of the hydraulic ore injector;

[0014] a data processing module configured to process the operation parameter data, filter the processed operation parameter data, and obtain key features;

[0015] a prediction module configured to input the key features into a constructed XGB-PSO concentration prediction model for training, obtain an optimized XGB-PSO concentration prediction model, input real-time operation parameter data into the optimized XGB-PSO concentration prediction model, and output a solid phase volume concentration prediction value of slurry in the ore-transporting pipe;

[0016] the construction process of the optimized XGB-PSO concentration prediction model comprises:

[0017] constructing an XGBoost model;

[0018] defining a hyperparameter combination of the XGBoost model as a particle P in a PSO algorithm i , each particle comprising four dimension vectors: ;

[0019] wherein, is a learning rate; is a maximum tree depth; is a split threshold; is a number of iterations;

[0020] randomly generating N particles to form an initial population, assigning random positions and velocities to each particle, and assigning a hyperparameter represented by each particle to the XGBoost model;

[0021] The velocity and position of the particle are updated to make the particle approach the individual historical optimum and the global optimum of the group, and the formula for updating the velocity and position of the particle is as follows:

[0022]

[0023]

[0024] wherein V new is the updated velocity; w is the inertia weight; V is the current velocity; c1 is the individual learning factor; r1 is random number 1; P best is the individual historical optimum position, X is the current position; c2 is the group learning factor; r2 is random number 2; G best is the global optimum position of the group; X new is the updated position; V new is the updated velocity;

[0025] The root mean square error value is calculated, and the root mean square error value is taken as the fitness value of the particle; when the iteration reaches a preset number of times or the root mean square error value no longer decreases, the parameter combination corresponding to the global optimum particle is output, and an optimized XGB-PSO concentration prediction model is obtained.

[0026] Optionally, the hydraulic ore injector comprises an ore feeding bin, a water supply pump, a flow direction control device, an ore tank, a mixer, a ore hoisting pipe, a separator and a storage device.

[0027] The output end of the ore feeding bin is connected with the input end of the flow direction control device through a feeding pipe, a feeding valve is arranged on the feeding pipe, the output end of the water supply pump is connected with the input end of the flow direction control device through a water supply pipe, a water supply valve and a flow meter are arranged on the water supply pipe, the output end of the flow direction control device is connected with a plurality of ore tanks through conveying pipes, control valves are arranged on the conveying pipes, the output end of the ore tank is connected with the input end of the mixer, a torque sensor is arranged in the mixer, the mixer is connected with the separator through the ore hoisting pipe, pressure sensors are arranged at the inlet end and the outlet end of the ore hoisting pipe, the separator is provided with a first output port and a second output port, the first output port is connected with the input end of the water supply pump through a recovery pipe, and the second output port is connected with the storage device through a slurry conveying pipe.

[0028] Optionally, the operation parameter data comprises feeding machine rotating speed data, water supply pipe flow data, injector valve timing state data, ore hoisting pipe pressure difference data and mixer power data.

[0029] Optionally, the operation parameter data is processed, and the processed operation parameter data is screened to obtain key features, which comprise:

[0030] Step S1, using leverage value analysis method to remove outliers in the collected data;

[0031] Step S2, using LASSO regression algorithm to screen out key feature variables from the data after removing outliers.

[0032] Optionally, the step S1 specifically comprises:

[0033] Step S11, normalizing the operating parameters;

[0034] Step S12, constructing a feature matrix based on the normalized data, and calculating a hat matrix according to the feature matrix;

[0035] Step S13, calculating the leverage value of each normalized operating parameter data point using the diagonal elements of the hat matrix;

[0036] Step S14, presetting an abnormality determination threshold, comparing the abnormality determination threshold with the leverage value, removing the leverage value exceeding the abnormality determination threshold, and obtaining data after outlier filtering.

[0037] Optionally, the step S2 specifically comprises:

[0038] Step S21, performing data standardization processing on the data after removing outliers;

[0039] Step S22, constructing a penalty objective function containing an L1 regularization term, and the expression of the penalty objective function is:

[0040]

[0041] wherein, is the target variable of the pulp concentration, is the input feature, is the penalty coefficient, is the regression coefficient.

[0042] Optionally, the control strategy is determined according to the solid volume concentration prediction value, comprising:

[0043] Presetting a concentration safety threshold, comparing the solid volume concentration prediction value with the concentration safety threshold, and determining the control strategy according to the comparison result;

[0044] If the solid volume concentration prediction value is greater than the concentration safety threshold, the rotational speed of the ore feeding bin is reduced or the water supply valve is opened, and the ore pulp in the mixer is diluted.

[0045] If the solid phase volume concentration prediction value is less than the concentration safety threshold, it is determined whether the solid phase volume concentration prediction value is in the optimal concentration interval, and the hydraulic ore injector is adaptively adjusted according to the determination result.

[0046] Optionally, the determination whether the solid phase volume concentration prediction value is in the optimal concentration interval and the adaptive adjustment of the hydraulic ore injector according to the determination result comprises:

[0047] If the solid phase volume concentration prediction value is in the optimal concentration interval, steady-state control is performed to control the speed fine adjustment of the ore feeding bin to maintain the pulp concentration in the optimal concentration interval.

[0048] If the solid phase volume concentration prediction value is not in the optimal concentration interval, transient compensation is performed to control the speed increase of the ore feeding bin to compensate the concentration fluctuation caused by the ore tank switching gap.

[0049] In a second aspect, the present application further provides a closed-loop control method based on deep-sea ore-lifting pipe pulp concentration monitoring, comprising the following steps:

[0050] The intelligent control unit collects real-time operation parameter data of the hydraulic ore injector, inputs the real-time operation parameter data into the optimized XGB-PSO concentration prediction model, predicts the pulp concentration in the ore-lifting pipe, and outputs the solid phase volume concentration prediction value at the current time;

[0051] The current solid phase volume concentration prediction value is sent to the controller, which compares the current solid phase volume concentration prediction value with the preset concentration safety threshold, and determines the control strategy according to the comparison result;

[0052] The controller sends the control strategy to the hydraulic ore injector, and the hydraulic ore injector responds to adjust the operation parameters of the hydraulic ore injector, and further adjusts the pulp concentration entering the ore-lifting pipe at the next time, so that the pulp concentration entering the storage device meets the requirements;

[0053] The construction process of the optimized XGB-PSO concentration prediction model comprises:

[0054] An XGBoost model is constructed.

[0055] The hyperparameter combination of the XGBoost model is defined as a particle P in the PSO algorithm i Each particle contains four dimensional vectors: ;

[0056] wherein, is the learning rate; is the maximum tree depth; is a split threshold value; is the number of iterations;

[0057] N particles are randomly generated to form an initial population, random positions and velocities are assigned, and the hyperparameters represented by each particle are assigned to the XGBoost model;

[0058] The velocity and position of the particle are updated so that the particle approaches the individual historical optimum and the global optimum of the group, and the formula for updating the velocity and position of the particle is:

[0059]

[0060]

[0061] wherein V new is the updated velocity; w is the inertia weight; V is the current velocity; c1 is the individual learning factor; r1 is random number 1; P best is the individual historical optimal position, X is the current position; c2 is the group learning factor; r2 is random number 2; G best is the global optimal position of the group; X new is the updated position; V new is the updated velocity;

[0062] The root mean square error value is calculated, and the root mean square error value is taken as the fitness value of the particle, when the iteration reaches the preset number of times or the root mean square error value no longer decreases, the parameter combination corresponding to the global optimal particle is output, and the optimized XGB-PSO concentration prediction model is obtained.

[0063] The beneficial effects of the present application are that the closed-loop control system based on deep-sea slurry concentration monitoring of the slurry in the slurry pipeline provided by the present application collects a plurality of operating parameter data through the data acquisition module, then screens the operating parameter data to obtain key features, then uses the optimized XGB-PSO concentration prediction model to predict the slurry concentration in the slurry pipeline, obtains the solid volume concentration prediction value of the slurry in the pipeline, and then determines the corresponding control strategy according to the solid volume concentration prediction value, and adjusts it accordingly, so as to more accurately control the concentration of the slurry in the slurry pipeline, and better adapt to complex working conditions such as seabed topography and ore supply fluctuations.

[0064] The solid volume concentration value in the slurry pipeline is predicted by the optimized XGB-PSO concentration prediction model in the intelligent control unit to obtain the solid volume concentration prediction value at the current time, so as to adjust the feeding / water supply operating parameters of the hydraulic ore injector at the next time according to the predicted solid volume concentration prediction value at the current time, so as to adjust the slurry concentration entering the slurry pipeline at the next time, and finally obtain slurry with the required concentration. According to this mode, the pulsation caused by the ore tank switching gap can be avoided, and the influence on the subsequent slurry concentration.

[0065] The closed-loop control system based on deep-sea slurry concentration monitoring of the slurry pipe can also avoid directly setting a density meter in the slurry pipe, and the density meter is used for detecting the slurry concentration in the pipe of the slurry pipe, so that the problem of wear of the density meter caused by direct contact with the slurry is avoided, and meanwhile, the system is easier to maintain, and the maintenance cost of the deep-sea equipment is reduced. BRIEF DESCRIPTION OF DRAWINGS

[0066] Figure 1 The structure schematic diagram of the closed-loop control system based on deep-sea slurry concentration monitoring of the slurry pipe is provided in the embodiments of the present application;

[0067] Figure 2 The adaptive optimization principle schematic diagram of the optimized XGB-PSO model is provided in the embodiments of the present application;

[0068] Figure 3 The process schematic diagram of the seawater circulation in the closed-loop control system based on deep-sea slurry concentration monitoring of the slurry pipe is provided in the embodiments of the present application;

[0069] Figure 4 The process schematic diagram of the slurry formation in the closed-loop control system based on deep-sea slurry concentration monitoring of the slurry pipe is provided in the embodiments of the present application;

[0070] Figure 5 The working process schematic diagram of the water surface support platform control center is provided in the embodiments of the present application;

[0071] Figure 6 The process schematic diagram of the closed-loop control method based on deep-sea slurry concentration monitoring of the slurry pipe is provided in the embodiments of the present application. DETAILED DESCRIPTION

[0072] The technical solutions of the present application are further described in detail below in combination with the drawings and specific embodiments. Unless otherwise defined, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs. The optional embodiments of the present application are described in detail as follows, however, the present application can have other implementation manners in addition to these detailed descriptions.

[0073] Embodiment 1

[0074] Please refer to the drawings Figure 1 and the drawings Figure 2 The present application provides a closed-loop control system based on deep-sea slurry concentration monitoring of the slurry pipe, which comprises:

[0075] A water surface support platform control center is used for outputting a control strategy.

[0076] A hydraulic ore injector for receiving a control strategy, the hydraulic ore injector being adapted to adjust itself according to the control strategy;

[0077] The water surface support platform control center comprises an intelligent control unit and a controller, the intelligent control unit is configured to collect operation parameter data of the hydraulic ore injector, process the operation parameter data, and obtain a solid phase volume concentration prediction value through real-time prediction; and the controller is configured to receive the solid phase volume concentration prediction value, and determine a control strategy according to the solid phase volume concentration prediction value.

[0078] The intelligent control unit comprises:

[0079] A data collection module configured to collect operation parameter data of the hydraulic ore injector;

[0080] A data processing module configured to process the operation parameter data, filter the processed operation parameter data, and obtain key features;

[0081] A prediction module configured to input the key features into an XGB-PSO concentration prediction model to obtain an optimized XGB-PSO concentration prediction model, input real-time operation parameter data into the optimized XGB-PSO concentration prediction model, and output a solid phase volume concentration prediction value of slurry in a pipeline of the ore drawing pipe.

[0082] The construction process of the optimized XGB-PSO concentration prediction model comprises:

[0083] Constructing an XGBoost model;

[0084] Defining a hyperparameter combination of the XGBoost model as a particle P in a PSO algorithm i Each particle comprises four dimension vectors: ;

[0085] wherein, is a learning rate; is a maximum tree depth; is a split threshold; is a number of iterations;

[0086] Randomly generating N particles to form an initial population, assigning random positions and velocities to the particles, and assigning a hyperparameter represented by each particle to the XGBoost model;

[0087] Updating the velocity and position of the particle to make the particle approach the individual historical optimum and the global optimum of the population, and the formula for updating the velocity and position of the particle is:

[0088]

[0089]

[0090] wherein, V new is the updated velocity; w is the inertia weight; V is the current velocity; c1 is the individual learning factor; r1 is random number 1; P best is the individual historical optimal position, X is the current position; c2 is the group learning factor; r2 is random number 2; G best is the group global optimal position; X new is the updated position; V new is the updated velocity;

[0091] The root mean square error value is calculated, the root mean square error value is taken as the fitness value of the particle, when the iteration reaches a preset number of times or the root mean square error value no longer decreases, the parameter combination corresponding to the global optimal particle is output, and an optimized XGB-PSO concentration prediction model is obtained.

[0092] Specifically, in the prior art, a physical flowmeter and a densimeter are used to monitor the concentration inside the mine hoist pipe. The physical flowmeter is a contact sensor, which needs to be built into the mine hoist pipe and must directly contact the ore pulp in the mine hoist pipe to obtain flow data. Since the physical flowmeter directly contacts the ore pulp, it is prone to cause local blockage and wear. Although the densimeter is a non-contact sensor and does not need to directly contact the ore pulp, the densimeter calculates the solid volume concentration through an indirect method. This method may not be able to capture the fluctuation in real time at the moment of ore bin switching due to sudden increase of the ore pulp concentration, and is prone to cause dynamic concentration pulsation response lag.

[0093] In the technical solution, the current ore pulp concentration is predicted by the intelligent control unit of the water surface support platform control center, specifically: the operation parameter data of the hydraulic ore injector is collected by the data acquisition module, the operation parameter data can reflect the resistance, water quantity and other operation states in the slurry pipeline, so as to reflect the blockage of the slurry pipeline, the conveying efficiency and the like, thereby providing a basis for subsequent prediction of the solid volume concentration in the slurry pipeline; then the operation parameter data is processed by the data processing module, the key features which have greater influence on the prediction of the solid volume concentration in the slurry pipeline are screened, and the optimized XGB-PSO concentration prediction model in the prediction module is used for prediction, so as to obtain the predicted value of the solid volume concentration of the ore pulp in the pipeline, the optimized XGB-PSO concentration prediction model is an XGBoost model based on the XGBoost algorithm, and the PSO algorithm is used to optimize the hyperparameters of the XGBoost model, by which the complex patterns in the operation parameter data can be better captured, the robustness and generalization ability of the model can be enhanced, and the accuracy of the solid volume concentration prediction can be improved, so as to obtain the predicted value of the solid volume concentration at the current time; then the predicted value of the solid volume concentration is sent to the controller, the controller determines the control strategy according to the predicted value of the solid volume concentration, so as to adjust the operation parameters of the hydraulic ore injector at the next time, and then adjust the ore pulp concentration at the new next time, so as to obtain the ore pulp with the required concentration; that is, in the technical solution, the ore pulp concentration at the next time can be adjusted according to the ore pulp concentration at the previous time, so as to realize dynamic closed-loop control, thereby avoiding the pulsation of the ore pulp concentration in the pipeline caused by the ore bin switching moment, and further avoiding the problem of unstable ore pulp concentration in the slurry pipeline, and ensuring the stability of the application in the deep sea complex working condition.

[0094] And through the control system, the problem of easy failure of deep sea sensors (coarse ore pulp can cause wear of pumps and sensors) can also be overcome, the water surface pump power and pipeline pressure difference data which are easy to maintain are used for subsequent prediction of the solid volume concentration, the problem of high wear of the densimeter caused by direct contact of the densimeter with the ore pulp is avoided, and the maintenance cost of the deep sea equipment is also significantly reduced.

[0095] As an optional embodiment, please refer to the accompanying drawings Figure 3 And 4 The hydraulic ore injector comprises an ore feeding bin, a water supply pump, a flow direction control device, an ore tank, a mixer, a slurry pipeline, a separator and a storage device.

[0096] The output end of the ore feeding bin is connected with the input end of the flow direction control device through a feeding pipe, a feeding valve is arranged on the feeding pipe, the output end of the water supply pump is connected with the input end of the flow direction control device through a water supply pipe, a water supply valve and a flow meter are arranged on the water supply pipe, the output end of the flow direction control device is connected with a plurality of ore tanks through conveying pipes, control valves are arranged on the conveying pipes, the output end of the ore tank is connected with the input end of the mixer, a torque sensor is arranged in the mixer, the mixer is connected with the separator through the ore lifting pipe, pressure sensors are arranged at the inlet end and the outlet end of the ore lifting pipe, the separator is provided with a first output port and a second output port, the first output port is connected with the input end of the water supply pump through a recovery pipe, and the second output port is connected with the storage device through a slurry conveying pipe.

[0097] Specifically, when deep-sea ore lifting pipe slurry lifting needs to be carried out, a control instruction is output to the hydraulic ore injector by the water surface support platform control center, the hydraulic ore injector functions to "inject" low-pressure ore into a high-pressure water channel, the control instruction controls the opening of the feeding valve of the ore feeding bin, the flow direction control device determines the corresponding ore tank, the corresponding control valve is opened, the ore in the ore feeding bin enters the corresponding ore tank through the flow direction control device and the control valve, if it is detected that the ore tank is full, the valve of another ore tank is controlled to be opened, the flow direction control device is switched to another ore tank, so that the ore coming out of the ore feeding bin enters another ore tank, at the same time, the water supply valve of the water supply pump is controlled to be opened, the water in the water supply pump enters the water supply pipe, the seawater in the water supply pipe enters the corresponding ore tank through the flow direction control device, if the flow direction control device is switched to another ore tank, the seawater also enters another ore tank and enters the mixer through the ore tank, in the mixer, the ore and the seawater are mixed to obtain slurry, and the slurry in the mixer is conveyed to the ore lifting pipe; in the above process, the water supply pipe flow, the feeding machine rotating speed, the ore lifting pipe pressure difference, the injector valve timing state, the ore lifting pipe pressure difference data and the mixer power data at the current time are collected by the data acquisition module of the intelligent control unit and are sent to the data processing module and the prediction module, and the solid phase volume concentration in the ore lifting pipe is predicted, a solid phase volume concentration prediction value is output, the solid phase volume concentration prediction value is sent to the controller, the controller determines the control strategy, the control strategy is sent to the hydraulic ore injector, the running parameters of the hydraulic ore injector at the next moment are controlled, the concentration at the next moment is adjusted through the concentration prediction at the previous moment, so that the solid phase volume concentration of the slurry in the ore lifting pipe finally entering the storage device meets the requirements, then under the action of the hydraulic pressure difference at both ends of the ore lifting pipe, the slurry in the ore lifting pipe is conveyed upward until entering the separator, the ore and the seawater are separated in the separator, the separated ore is stored in the storage device, and the separated seawater is treated and then enters the water supply pump again, so that a closed cycle is realized.

[0098] Meanwhile, since the ore feeding bin does not fill all the ore tanks at the same time and then carries out subsequent ore and seawater mixing, but is provided with multiple ore tanks, the number of ore tanks in the embodiment is two, and in actual application, the number of ore tanks can be set according to actual conditions, and the number of ore tanks is not limited herein; the ore feeding bin injects ore one by one, that is, when an ore tank is filled with ore, the flow direction control device switches to the next ore tank, and the ore in the ore feeding bin is injected into the next ore tank; and during the switching of the flow direction control device from one ore tank to the next ore tank, there is a switching gap, which makes the switching process take time, and the ore supply is temporarily interrupted during the time, resulting in pulsation of the slurry concentration in the pipeline, and in the technical solution, the solid volume concentration prediction value is obtained by predicting the solid volume concentration value in the ore lifting pipe through the optimized XGB-PSO concentration prediction model in the intelligent control unit, so as to determine the adjustment strategy according to the solid volume concentration prediction value, thereby avoiding the pulsation caused by the ore tank switching gap and affecting the subsequent ore slurry concentration.

[0099] It should be noted that the specific steps of detecting that the ore tank is filled with ore and switching another ore tank are as follows: through volume integration (soft measurement) based on the speed of the feeder: the speed of the feeder and the single-circle displacement of the feeder are known, when the calculated volume reaches the volume of the ore tank, the water surface support platform control center determines that it is "full" and triggers switching (i.e., from one ore tank to another ore tank).

[0100] The solid volume concentration prediction value predicted by the intelligent control unit is the concentration value of the ore slurry at the outlet of the mixer, i.e., the ore slurry about to enter the bottom of the ore lifting pipe, and the optimized XGB-PSO concentration prediction model predicts the ore slurry concentration at the current T time according to the feed-forward parameters (feeder speed data, water supply flow data, and injector valve timing state data) on the feeding side and the feedback parameters (ore lifting pipe pressure difference data and mixer power data) on the conveying side, and adjusts the feeding and water supply parameters at the next T+1 time according to the same, to realize dynamic closed-loop control; for example, if the concentration predicted at T time is high, the controller immediately reduces the speed of the feeder or increases the water injection, and the concentration of the new ore slurry entering the pipe at T+1 time will be reduced.

[0101] As an optional embodiment, the operating parameter data includes feeder speed data, water supply pipe flow data, injector valve timing state data, ore lifting pipe pressure difference data, and mixer power data.

[0102] Specifically, the feeder speed data is the real-time rotating speed of the rotating feeder, which is a source parameter for controlling the pulp concentration, and is detected by a rotating speed sensor installed on the motor drive shaft of the rotating feeder at the bottom of the ore feeding bin; the water supply pipe flow data is the real-time flow rate of the water supply pipe, which can affect the pulp concentration and dilution efficiency inside the ore pipe, and is detected by a flow meter on the water supply pipe; the injector valve timing state data is a switch timing signal directly collected by a valve limit switch, which is installed on each key valve actuator of the hydraulic ore injector, specifically on the drive shaft of the flow control device, and on the valve stems of the feeding valve and the water supply valve; the ore pipe pressure difference is the difference between the inlet pressure and the outlet pressure of the ore pipe, which can reflect the overall density of the solid-liquid two-phase flow and the flow resistance of the ore pipe, and is a key indicator for judging pipe blockage, wear or deposition; the power data of the mixer is the real-time load data in the mixer, which is detected by a torque sensor in the mixer; by cooperatively acquiring the above parameters, a basis can be provided for subsequent prediction of the solid phase volume concentration prediction value.

[0103] In another embodiment, in addition to acquiring the operating parameter data of the hydraulic ore injector, the data acquisition module can also acquire other physical parameter data, including obtaining the sound frequency of particle impact on the inner wall of the ore pipe by using an acoustic sensor, obtaining the vibration frequency inside the ore pipe by using a vibration sensor, and obtaining the solid particle concentration in the ore pipe by using an optical turbidimeter; the operating parameter data and the physical parameter data are processed, and then input into the optimized XGB-PSO concentration prediction model for prediction of the solid phase volume concentration, so as to enhance the feature capture of the solid phase content and obtain more accurate prediction results.

[0104] As an optional embodiment, the processing of the operating parameter data includes:

[0105] Step S1, using a leverage value analysis method to eliminate outliers in the collected data;

[0106] Step S2, using a LASSO regression algorithm to screen out key feature variables from the data after eliminating outliers.

[0107] As an optional embodiment, the step S1 specifically includes:

[0108] Step S11, normalizing the operating parameters;

[0109] Step S12, constructing a feature matrix based on the normalized data, and calculating a hat matrix according to the feature matrix;

[0110] Step S13, calculating the leverage value of each normalized operating parameter data point by using the diagonal elements of the hat matrix.

[0111] Step S14: Preset an anomaly detection threshold, compare the anomaly detection threshold with the lever value, and remove the sampling points corresponding to the lever values ​​that exceed the anomaly detection threshold to obtain the dataset filtered by anomalies.

[0112] Specifically, the leverage value analysis method is used to remove outliers from the collected data. The aim is to address the impact of the unique "particle impact noise" in deep-sea solid-liquid two-phase flow monitoring on the accuracy of subsequent concentration prediction models. The specific steps are as follows: The collected raw operating parameter data—including the ore lifting pipe pressure difference data P, water supply pipe flow rate data Q, mixer power data W, feeder speed data R, and discrete variable injector valve timing status S—are used to construct the raw dataset. Continuous variables (P, Q, W, R) are extracted from the raw dataset to construct an n×p dimensional feature matrix X, where n is the number of sampling points and p is the feature dimension. The formula is then used to... The hat matrix H is calculated, reflecting the geometric relationship between observed and predicted values. The leverage value of each normalized operating parameter data point is calculated using the diagonal elements of the hat matrix H (the main diagonal elements of the hat matrix H are extracted as the leverage value of the i-th sampled operating parameter data point; the leverage value quantifies the deviation of the data point from the data center in the feature space). Then, an anomaly detection threshold is set. When the leverage value calculated from the sensor data combination exceeds this threshold at a certain moment, the data is automatically identified as an outlier and removed from the dataset to prevent it from participating in subsequent calculations. This process needs to be iteratively executed, by refitting the model, updating the leverage value calculation, and repeating the threshold detection and removal operations until no new outliers appear in continuous iterations, ultimately obtaining the dataset filtered for outliers. High leverage values ​​(leverage values ​​exceeding the anomaly detection threshold) typically correspond to non-logical spikes generated by the instantaneous impact of large deep-sea ore particles on the pressure sensor probe; that is, although the value is high, it has no correlation with fluid pressure. Removing these spikes prevents the model from misjudging "impact noise" as "precursor to pipe blockage," improving the accuracy of the model's concentration prediction.

[0113] As an optional implementation, step S2 specifically includes:

[0114] Step S21: Perform data standardization on the data after outlier removal;

[0115] Step S22: Construct a penalty objective function that includes an L1 regularization term. The expression of the penalty objective function is:

[0116]

[0117] in, The target variable is pulp concentration. is an input feature, is a penalty coefficient, is a regression coefficient;

[0118] Specifically, the data set (containing P, Q, W, R, S) after removing outliers is subjected to data standardization processing to eliminate the difference between different physical dimensions, so as to prepare for LASSO regression; then the coefficient is selected by using a penalty objective function The corresponding physical quantity is taken as a key feature variable, and the screened key feature variable specifically includes: a differential pressure of the ore drawing pipe P: a physical representation of the gravity density of the solid-liquid mixed fluid; W: a physical representation of the viscous resistance of the fluid under high concentration; Q: a physical representation of the dilution effect and the conveying carrier speed; RPM: a physical representation of the volume conveying rate of the ore source (feedforward signal); an injector valve timing state: a physical representation of the flow field pulsation state caused by the switching of the hydraulic ore injector (transient feature).

[0119] In the technical solution, the sensor abnormal values are automatically removed by introducing the leverage value analysis method, and the LASSO regression algorithm is used for screening, so as to solve the problem of "multiple collinearity" (parameters interfere with each other), obtain the key feature variable that has the greatest impact on the concentration, and eliminate the interference of multiple collinearity; the global optimization of the XGBoost model hyperparameters is combined with the particle swarm optimization algorithm, so that the XGB-PSO concentration prediction model can still maintain high-precision prediction ability when facing changes in seabed topography or fluctuations in ore supply.

[0120] Specifically, the PSO algorithm is used to globally optimize the hyperparameters of the XGBoost model, the four core hyperparameters of the XGBoost model are mapped to the position coordinates of the particles in the PSO algorithm, the position of each particle is a four-dimensional vector, a plurality of particles are randomly generated in the hyperparameter value range when initializing the particle swarm, and an initial speed is randomly assigned; wherein, is a learning rate, used to control the response step to the concentration change, and the parameter determines the correction amplitude of the model to the residual error; in the present application, The optimization target of is to enable the model to quickly track the second-level concentration pulsation caused by the switching of the ore tank by the flow control device, and prevent control instability caused by response lag; is the maximum tree depth, used to control the fitting complexity of the nonlinear rheological properties, This parameter determines the nonlinear expression ability of the model; since the rheological properties of the ore slurry in the deep sea environment are significantly different from those in the shallow water area, and the relationship between the differential pressure of the ore drawing pipe and the concentration is seriously affected by the flow coupling, the PSO is used to find the best , aiming to establish a model structure that can accurately map this high-dimensional nonlinear fluid dynamics relationship and avoid underfitting; The splitting threshold is used to control the filtering ability of noise. This parameter determines the conservatism of the model's splitting nodes. For the instantaneous spike noise generated by the pressure sensor caused by the random collision of coarse ore particles in the ore lifting pipe, optimizing this parameter can force the model to ignore such pseudo-signals that are not related to concentration changes, thereby improving the system's anti-interference robustness.

[0121] Next, the fitness of each particle is evaluated. The hyperparameters corresponding to its position are substituted into the XGBoost model for pre-training, and the root mean square error (RMSE) is calculated as the fitness value. RMSE can quantify the model's prediction error and directly reflect the quality of the hyperparameter combination. The smaller the RMSE, the higher the fitness, indicating that the deviation between the model's predicted value and the true value is smaller, and the hyperparameter combination is better. The historical best position of each particle (individual best) and the historical best position of the entire swarm (global best) are recorded. The individual historical best P best This represents the model parameter configuration found by the particle in previous iterations that best suits the current seabed topography and slurry concentration prediction; the swarm global optimal G best This represents the globally optimal parameter combination found across the entire domain that minimizes the prediction residual of solid volume concentration within the ore-lifting pipe; then, position updates are performed according to the PSO update rule; the optimization is iterated continuously until the termination condition is met, finding the globally optimal hyperparameter combination that drives all particles toward G. best The optimized model output is able to overcome the shortcomings of traditional empirical formulas in terms of accuracy degradation under varying deep-sea conditions (such as water depth changes and backwater back pressure fluctuations). This hyperparameter combination represents the optimal balance between the model's ability to quickly capture early signs of pipe blockage and the random noise of the filtered fluid, enabling it to adapt to the dynamic changes in the deep-sea operating environment. Then, the hyperparameters corresponding to this globally optimal hyperparameter combination are assigned to the XGBoost model to obtain the optimized XGB-PSO concentration prediction model.

[0122] The real-time key feature variables selected using the LASSO regression algorithm ( The parameters (p, W, Q, RPM, Svalve) are input into the XGBoost model loaded with the optimal parameters. That is, the real-time key feature variables are input into the optimized XGB-PSO concentration prediction model, and the model outputs the current solid volume concentration prediction value (SC%).

[0123] It should be noted that this technical solution is not a simple application of existing general algorithms, but rather an improvement on the physical characteristics of deep-sea closed-loop ore hoisting systems. In this invention, particles represent the hyperparameters of the prediction model. This invention creatively establishes a mapping relationship between these abstract mathematical parameters and specific physical phenomena of deep-sea ore hoisting (valve pulsation, rheological nonlinearity, particle impact noise), providing a basis for subsequent concentration prediction based on this mapping relationship.

[0124] Targeted noise processing: By introducing physical correlation constraints through leverage value analysis, the problem of false signal interference in deep-sea large particle transport is specifically solved, so as to distinguish between'real concentration increase' and 'ore impact noise', and avoid the influence of noise on solid volume concentration prediction.

[0125] Unique feature engineering (variable / feature selection): The XGB-PSO model first introduces the discrete feature of 'water injector valve timing state' in this technical solution, which enables the model to perceive 'working condition switching' and accurately predict the transient concentration pulsation caused by tank switching that traditional models cannot capture.

[0126] Adaptive parameter optimization: In view of the problem of back pressure change caused by water depth change in deep-sea operation, the best tree depth and learning rate are dynamically found through PSO, solving the technical problem of accuracy decline of fixed parameter model under changing working conditions.

[0127] As an optional implementation, the control strategy is determined according to the solid volume concentration prediction value, comprising:

[0128] A preset concentration safety threshold is compared with the solid volume concentration prediction value, and a control strategy is determined according to the comparison result;

[0129] If the solid volume concentration prediction value is greater than the concentration safety threshold, the speed of the ore feeding bin is reduced or the water supply valve is opened to dilute the ore slurry in the mixer;

[0130] If the solid volume concentration prediction value is less than the concentration safety threshold, it is judged whether the solid volume concentration prediction value is in the optimal concentration interval, and the hydraulic ore injector is adaptively adjusted according to the judgment result.

[0131] The judgment of whether the solid volume concentration prediction value is in the optimal concentration interval and the adaptive adjustment according to the judgment result comprise:

[0132] If the solid volume concentration prediction value is in the optimal concentration interval, steady-state control is performed to fine-tune the speed of the ore feeding bin to maintain the ore slurry concentration in the optimal concentration interval;

[0133] If the solid volume concentration prediction value is not in the optimal concentration interval, transient compensation is performed to increase the speed of the ore feeding bin to compensate for the concentration fluctuation caused by the ore tank switching gap.

[0134] Specifically, the controller sends a control strategy to the hydraulic ore injector for response, wherein the control strategy contains control instructions on how specific devices in the hydraulic ore injector adjust their parameters, so that the specific devices in the hydraulic ore injector can respond and execute according to the control instructions.

[0135] The solid phase volume concentration prediction value is predicted by the optimized XGB-PSO concentration prediction model, and the concentration safety threshold is set according to the actual situation; if the solid phase volume concentration prediction value is greater than the concentration safety threshold, it indicates that the concentration of the ore pulp in the slurry pipeline is high, and the controller sends a control instruction to the ore feeding bin to reduce the speed of the ore feeding bin to reduce the transportation of the ore, or opens the water injection valve to dilute the ore pulp in the mixer by adding water to reduce the concentration of the ore pulp;

[0136] If the solid phase volume concentration prediction value is less than the concentration safety threshold, it is further needed to judge whether the solid phase volume concentration prediction value is in the optimal concentration interval; if it is in the optimal concentration interval, the concentration of the ore pulp needs to be maintained in the optimal concentration interval, so the speed of the ore feeding bin is not adjusted or is adjusted slightly; if it is not in the optimal concentration interval and the switching signal of the flow control device is monitored, the concentration may drop, and the controller sends a control instruction to the ore feeding bin to increase the speed in advance to compensate for the concentration fluctuation caused by the switching gap;

[0137] It should be noted that when the switching signal of the flow control device is sent, there is an inertial delay in the liquid flow during the switching process, which may cause a short-term flow instability, or the residual ore pulp in the slurry pipeline before switching is mixed with the newly flowed material unevenly, which may cause a local concentration fluctuation and further cause the concentration to drop; in the technical solution, the data acquisition module acquires the running parameter data in real time, the real-time solid phase volume concentration prediction value is obtained by real-time prediction according to the real-time acquired running parameter data, and then the real-time solid phase volume concentration prediction value is compared with the concentration safety threshold and the optimal concentration interval, and the concentration of the ore pulp is adjusted according to the comparison result; in this way, the concentration fluctuation at the switching time can be perceived and fed forwardly compensated, so that the closed-loop control of the concentration of the ore pulp is realized; this self-adaptive adjustment method not only effectively prevents the risk of pipeline blockage, but also ensures that the hoisting system always operates in the optimal efficiency interval, reduces unnecessary water drainage and energy consumption, and can also adapt to complex working conditions such as seabed topography and ore supply fluctuation.

[0138] Embodiment 2

[0139] Please refer to the accompanying drawings Figure 5 and 6 The application also provides a closed-loop control method based on deep-sea slurry pipeline ore pulp concentration monitoring, which comprises the following steps:

[0140] The real-time operation parameter data of the hydraulic ore injector is collected by the intelligent control unit, the real-time operation parameter data is input into the optimized XGB-PSO concentration prediction model, the pulp concentration in the pipe of the ore drawing pipe is predicted, and the prediction value of the solid volume concentration at the current time is output;

[0141] The current solid volume concentration prediction value is sent to the controller, the controller compares the current solid volume concentration prediction value with the preset concentration safety threshold, and determines the control strategy according to the comparison result;

[0142] The controller sends the control strategy to the hydraulic ore injector, and the hydraulic ore injector responds to adjust the operation parameters of the hydraulic ore injector, so as to adjust the pulp concentration entering the ore drawing pipe at the next time, so that the pulp concentration entering the storage device finally meets the requirements;

[0143] The construction process of the optimized XGB-PSO concentration prediction model comprises:

[0144] Constructing an XGBoost model;

[0145] The hyperparameter combination of the XGBoost model is defined as a particle P in the PSO algorithm i Each particle contains four dimension vectors: ;

[0146] Among them, is the learning rate; is the maximum tree depth; is the split threshold; is the number of iterations;

[0147] N particles are randomly generated to form an initial population, and random positions and speeds are assigned, and the hyperparameters represented by each particle are assigned to the XGBoost model;

[0148] The speed and position of the particle are updated, so that the particle approaches the individual historical optimum and the global optimum of the group, and the formula for updating the speed and position of the particle is:

[0149]

[0150]

[0151] Among them, V new is the updated speed; w is the inertia weight; V is the current speed; c1 is the individual learning factor; r1 is random number 1; P best is the individual historical optimal position, X is the current position; c2 is the group learning factor; r2 is random number 2; G best is the global optimal position of the group; X newV is the updated position; new V is the updated velocity;

[0152] The root mean square error value is calculated, and the root mean square error value is used as the fitness value of the particle; when the iteration reaches a preset number of times or the root mean square error value no longer decreases, the parameter combination corresponding to the global optimal particle is output, and an optimized XGB-PSO concentration prediction model is obtained.

[0153] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. The protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A closed-loop control system based on deep-sea ore lift pipe slurry concentration monitoring, characterized in that, include: The surface support platform control center is used to output control strategies; A hydraulic ore injector is used to receive a control strategy and perform adaptive adjustments according to the control strategy. The water surface support platform control center includes an intelligent control unit and a controller. The intelligent control unit is used to collect the operating parameter data of the hydraulic ore injector, process the operating parameter data, and predict the solid volume concentration in real time. The controller is used to receive the solid volume concentration prediction value and determine the control strategy based on the solid volume concentration prediction value. The intelligent control unit includes: The data acquisition module is used to collect the operating parameter data of the hydraulic ore injector; The data processing module is used to process the operating parameter data, filter the processed operating parameter data, and obtain key features; The prediction module is used to train the XGB-PSO concentration prediction model by inputting the key features into it, to obtain an optimized XGB-PSO concentration prediction model, to input real-time operating parameter data into the optimized XGB-PSO concentration prediction model, and to output the predicted value of the solid volume concentration of the slurry in the ore lifting pipe. The process of processing the operating parameter data and filtering the processed operating parameter data to obtain key features includes: Step S11: Normalize the operating parameters; Step S12: Construct a feature matrix based on the normalized data, and calculate the hat matrix based on the feature matrix; Step S13: Calculate the lever value for each normalized running parameter data point using the diagonal elements of the hat matrix; Step S14: Preset an anomaly detection threshold, compare the anomaly detection threshold with the leverage value, and remove leverage values ​​that exceed the anomaly detection threshold to obtain data filtered for anomalies. Step S21: Perform data standardization on the data after outlier removal; Step S22: Construct a penalty objective function that includes an L1 regularization term. The expression of the penalty objective function is: in, The target variable is slurry concentration. As input features, The penalty coefficient, These are the regression coefficients; The process of constructing the optimized XGB-PSO concentration prediction model includes: Construct the XGBoost model; The hyperparameter combination of the XGBoost model is defined as a particle P in the PSO algorithm. i Each particle contains a four-dimensional vector: ; in, The learning rate; Maximum tree depth; This is the splitting threshold; This represents the number of iterations. N particles are randomly generated to form an initial population, and their positions and velocities are assigned randomly. The hyperparameters represented by each particle are then assigned to the XGBoost model. The velocity and position of particles are updated to move them closer to their individual historical optimum and the global optimum. The formula for updating the velocity and position of particles is: Among them, V new The updated velocity is represented by w; the inertia weight is w; the current velocity is v; c1 is the individual learning factor; r1 is a random number 1; P best Let X be the individual's historical best position, and C be the current position; c2 be the group learning factor; r2 be the random number 2; G be the random number 2. best X is the globally optimal position for the group; new The updated position; V new For the updated speed; Calculate the root mean square error (RMSE) value and use it as the fitness value of the particle. When the preset number of iterations is reached or the RMS error value no longer decreases, output the parameter combination corresponding to the globally optimal particle to obtain the optimized XGB-PSO concentration prediction model.

2. The closed-loop control system based on deep-sea ore lifting pipe slurry concentration monitoring according to claim 1, characterized in that, The hydraulic ore injector includes: an ore feed silo, a water pump, a flow control device, an ore tank, a mixer, a lifting pipe, a separator, and a storage device; The output end of the ore feed hopper is connected to the input end of the flow direction control device via a feed pipe. A feed valve is installed on the feed pipe. The output end of the water supply pump is connected to the input end of the flow direction control device via a water supply pipe. A water supply valve and a flow meter are installed on the water supply pipe. The output end of the flow direction control device is connected to several ore tanks via conveying pipes. A control valve is installed on the conveying pipes. The output end of each ore tank is connected to the input end of the mixer. A torque sensor is installed inside the mixer. The mixer is connected to the separator via a lifting pipe. Pressure sensors are installed at both the inlet and outlet ends of the lifting pipe. The separator has a first output port and a second output port. The first output port is connected to the input end of the water supply pump via a recovery pipe. The second output port is connected to the storage device via a slurry conveying pipe.

3. The closed-loop control system based on deep-sea ore lifting pipe slurry concentration monitoring according to claim 2, characterized in that, The operating parameter data includes feeder speed data, water supply pipe flow data, injector valve timing status data, ore lifting pipe pressure difference data, and mixer power data.

4. The closed-loop control system based on deep-sea ore lifting pipe slurry concentration monitoring according to claim 2, characterized in that, The step of determining the control strategy based on the predicted solid volume concentration includes: A preset concentration safety threshold is set, the predicted solid phase volume concentration is compared with the concentration safety threshold, and a control strategy is determined based on the comparison result. If the predicted solid volume concentration is greater than the concentration safety threshold, the rotation speed of the ore feed hopper is reduced or the water supply valve is opened to dilute the slurry in the mixer. If the predicted solid volume concentration is less than the concentration safety threshold, it is determined whether the predicted solid volume concentration is within the optimal concentration range, and the hydraulic ore injector is controlled to make adaptive adjustments based on the determination result.

5. The closed-loop control system based on deep-sea ore lifting pipe slurry concentration monitoring according to claim 4, characterized in that, The step of determining whether the predicted solid volume concentration is within the optimal concentration range, and controlling the hydraulic ore injector to adaptively adjust based on the determination result, includes: If the predicted solid volume concentration is within the optimal concentration range, steady-state control is performed, and the rotation speed of the ore feed hopper is finely adjusted to maintain the slurry concentration within the optimal concentration range. If the predicted solid volume concentration is not in the optimal concentration range, transient compensation is performed to increase the rotation speed of the ore feed hopper and compensate for the concentration fluctuations caused by the switching gap of the ore tank.

6. A closed-loop control method based on deep-sea ore lifter slurry concentration monitoring, the method being used to implement the closed-loop control system based on deep-sea ore lifter slurry concentration monitoring as described in any one of claims 1-5, characterized in that, Includes the following steps: The real-time operating parameter data of the hydraulic ore injector is collected by the intelligent control unit. The real-time operating parameter data is input into the optimized XGB-PSO concentration prediction model to predict the slurry concentration in the ore lifting pipe and output the predicted value of the solid phase volume concentration at the current moment. The current solid volume concentration prediction value is sent to the controller, which compares the current solid volume concentration prediction value with a preset concentration safety threshold and determines the control strategy based on the comparison result. The controller sends the control strategy to the hydraulic ore injector, which responds to adjust its operating parameters, thereby adjusting the slurry concentration entering the ore lifting pipe at the next moment, so that the slurry concentration entering the storage device meets the requirements. The process of constructing the optimized XGB-PSO concentration prediction model includes: Construct the XGBoost model; The hyperparameter combination of the XGBoost model is defined as a particle P in the PSO algorithm. i Each particle contains a four-dimensional vector: ; in, The learning rate; Maximum tree depth; This is the splitting threshold; This represents the number of iterations. N particles are randomly generated to form an initial population, and their positions and velocities are assigned randomly. The hyperparameters represented by each particle are then assigned to the XGBoost model. The velocity and position of particles are updated to move them closer to their individual historical optimum and the global optimum. The formula for updating the velocity and position of particles is: Among them, V new The updated velocity is represented by w; the inertia weight is w; the current velocity is v; c1 is the individual learning factor; r1 is a random number 1; P best Let X be the individual's historical best position, and C be the current position; c2 be the group learning factor; r2 be the random number 2; G be the random number 2. best X is the globally optimal position for the group; new The updated position; V new For the updated speed; Calculate the root mean square error (RMSE) value and use it as the fitness value of the particle. When the preset number of iterations is reached or the RMS error value no longer decreases, output the parameter combination corresponding to the globally optimal particle to obtain the optimized XGB-PSO concentration prediction model.

Images