A density adjustment adaptive control system for dense medium separation

Abstract

The present invention proposes an adaptive control system for density adjustment of heavy medium separation, which relates to the field of heavy medium coal preparation, including a monitoring device for monitoring the operating parameter information of each device during the separation process, an analysis system for statistically analyzing the historical operating data of the equipment, establishing a separation model, a control unit cooperating with the analysis system to realize adaptive PMC adjustment, and an execution unit executing the instructions issued by the control unit. The monitoring device is connected to the analysis system through a network, the analysis system is connected to the control unit through a network, and the control unit is connected to the execution unit through a wire. The analysis system performs feedforward test adjustment on the entire system based on the production process parameters obtained by the monitoring device and the established separation model, and continuously optimizes the training model, thereby realizing adaptive adjustment control of the density of heavy medium separation, improving the stability and timeliness of the heavy medium separation system, and having strong adaptability.

Description

Technical Field

[0001] The invention relates to the field of heavy medium coal separation, in particular to a heavy medium separation density adjustment adaptive control system. Background Art

[0002] With the continuous advancement of intelligent coal washing plants, from the perspective of research fields and directions at home and abroad in recent years, technologies such as the Internet of Things, big data, cloud computing, data mining, and artificial intelligence have been widely studied and applied in the fields of logistics warehousing, smart cities, etc., and related concepts such as digital mines, sensory mines, and smart mines in the field of mining energy have also been proposed one after another. As an important part of the mining industry, the coal preparation industry has also entered a new stage where automation, digitization, and intelligence coexist. Although the coal industry has achieved rapid development in the past 10 years, the processing and utilization of coal resources is still relatively extensive. Although most coal preparation plants have basically achieved automation, mainly reflected in the centralized control of equipment operation, process parameter monitoring and adjustment, and equipment operation status monitoring, they still require manual participation in the entire process in production management, are highly dependent on manual experience, and are relatively backward in information and intelligent construction. Therefore, it is particularly important for coal preparation enterprises to further improve management levels, reduce production costs, and use new technologies to improve the economic benefits of enterprises.

[0003] As an indispensable part of coal washing, heavy medium separation has made great progress in automatic control of heavy medium separation with automatic water replenishment and automatic diversion as the main control objects. However, in actual production, due to the change of raw coal quality and the development of ash back control, higher requirements are put forward for the control range and speed of density adjustment, and the current heavy medium separation cannot meet this demand well. Summary of the invention

[0004] In view of the above problems, the present invention proposes a dense medium separation density adjustment adaptive control system.

[0005] The embodiment of the present invention provides a heavy medium separation density adjustment adaptive control system, the heavy medium separation density adjustment adaptive control system comprises: a monitoring device, an analysis system, a control unit, and an execution unit;

[0006] The monitoring device is used to obtain real-time data of various types of equipment and transmit it to the analysis system and the control unit;

[0007] The analysis system is used to screen, classify, count and store the real-time data, and realize real-time data display and historical data storage;

[0008] The analysis system is further used to establish a relationship model between coal quality and sorting density based on a big data analysis algorithm model, establish a sorting model according to the relationship model, and feed back a sorting density adjustment amount to the control unit according to the sorting model;

[0009] The control unit is used to issue instructions to the execution unit according to the sorting density adjustment amount to adjust the actual sorting density;

[0010] The execution unit is used to execute corresponding actions according to the instructions issued by the control unit, adjust the actual sorting density, and feed back the adjusted actual sorting density to the analysis system;

[0011] Among them, the analysis system receives the adjusted actual sorting density, and feeds back the new sorting density adjustment amount to the control unit in combination with the new real-time data collected by the monitoring device, and the control unit issues a new instruction to the execution unit according to the new sorting density adjustment amount, and the execution unit performs corresponding actions according to the new instruction, and continues to adjust the actual sorting density. The three processes are repeated until the current value of the product coal ash content adjusted by the execution unit reaches within the preset range.

[0012] Optionally, the monitoring device includes: an online raw coal ash meter, an online clean coal ash meter, a differential pressure density meter, a magnetic material monitor, a liquid level sensor for a mixed medium barrel, a liquid level sensor for a diluted medium barrel, and a medium addition scale;

[0013] The raw coal online ash meter and the clean coal online ash meter are used to obtain the ash content, moisture content and calorific value of the raw coal and each product coal in real time respectively;

[0014] The liquid level sensor for the combined medium barrel and the liquid level sensor for the diluted medium barrel are used to obtain the liquid level information of the combined medium barrel and the diluted medium barrel respectively in real time;

[0015] The differential pressure densitometer is used to obtain the density of the suspension in real time;

[0016] The magnetic substance monitor is used to obtain the magnetic substance content in the suspension in real time;

[0017] The medium adding scale is used to obtain the weight of the medium added to the concentrated medium barrel each time;

[0018] The ash content, the moisture content, the calorific value, the liquid level information, the density, the magnetic content, and the medium weight constitute the real-time data.

[0019] Optionally, the analysis system includes: a data middle platform and a big data analysis algorithm model;

[0020] The data center is used to screen, classify, count and store the real-time data, filter out invalid data, and realize the real-time data display and historical data storage;

[0021] The big data analysis algorithm model is used to analyze and compare the real-time data and historical data, and to establish a relationship model between coal quality and sorting density by combining the selectivity curve, distribution curve, and mismatch curve of the comprehensive floating and sinking test of raw coal;

[0022] The big data analysis algorithm model is also used to establish the sorting model based on the relationship model, combined with the real-time data of sorting density, magnetic content, mixed medium barrel position, dilute medium barrel position, water replenishment valve opening, and diverter valve opening, and to feed back the sorting density adjustment amount to the control unit based on the sorting model.

[0023] Optionally, the control unit includes: a control console, a PLC control unit;

[0024] The control console is used to issue instructions to the execution unit in cooperation with the PLC control unit according to the sorting density adjustment amount, so as to adjust the actual sorting density.

[0025] Optionally, the execution unit includes: a medium adding pump, a medium combining pump, a medium diluting pump, a water adding valve, a water replenishing valve, a diverter valve, and a shallow tank separator;

[0026] When the dense medium separation density adjustment adaptive control system is in a steady-state mode, controlling the openings of the water supply valve and the diverter valve to perform MPC adjustment;

[0027] When the heavy medium separation density adjustment adaptive control system is in the medium adding mode, the concentrated medium pump and the water supply valve are opened to supplement the medium to the mixed medium barrel, and the weight of the added medium is measured by the medium adding metering scale of the monitoring device, while the suspension density and the liquid level of the mixed medium barrel are monitored, and the water supply valve is opened to stabilize the separation density. When the liquid level of the mixed medium barrel reaches the predicted high liquid level setting value, the medium addition is stopped, so that the heavy medium separation density adjustment adaptive control system enters the steady-state mode as soon as possible;

[0028] When the dense medium separation density adjustment adaptive control system is in a disordered mode and the separation density is high, the density of the suspension is lowered by increasing the opening of the water replenishment valve and opening the water addition valve, and the liquid level of the combined medium barrel is monitored at the same time to prevent the barrel from overflowing due to excessive water addition. When the predicted high liquid level setting value of the combined medium barrel is reached, the water addition valve is closed, and the opening of the water replenishment valve is adjusted to make the dense medium separation density adjustment adaptive control system enter the steady state mode as soon as possible;

[0029] When the heavy medium separation density adjustment adaptive control system is in a disordered mode and the separation density is low, the suspension density is increased by reducing the water supply valve opening and opening the diverter valve, and the liquid levels of the dilute medium barrel and the combined medium barrel are monitored at the same time to prevent the dilute medium barrel from overflowing. If the combined medium barrel is lower than the low liquid level preset value due to diversion, the medium addition mode is triggered, and combined with the suspension density in the monitoring device, the heavy medium separation density adjustment adaptive control system can enter the steady-state mode as soon as possible.

[0030] Optionally, the big data analysis algorithm model analyzes and compares the real-time data and historical data, and combines the selectivity curve, distribution curve, and mismatch curve of the raw coal comprehensive floating and sinking test to establish a relationship model between coal quality and sorting density, including:

[0031] By using the data of the quantity and quality of products at several density levels in the floating and sinking test, the data are plotted into a curve using a graphical method, so as to determine the relationship between the quantity and quality of the material within a certain density range. Through the raw coal comprehensive floating and sinking test selectivity curve, the coal quality characteristics of the raw coal can be timely understood, the raw coal selectivity can be evaluated, the clean coal yield and ash content can be predicted, and the theoretical separation density can be determined;

[0032] The raw coal comprehensive floating and sinking test selectivity curve is drawn by fitting, and the floating matter accumulation curve β and the density curve δ are fitted. The remaining three curves: ash characteristic curve λ, sinking matter accumulation curve θ, and density ±0.1 curve ε are calculated by the previous two curves;

[0033] Among them, three models are used for fitting according to the shapes of curve β and curve δ:

[0034] ① Inverse tangent model

[0035] ②Hyperbolic tangent model Y = {a + c × th [k (x-x0)]} × 100, where

[0036] ③Compound hyperbolic tangent model Y={a+bx+c×th[k(x-x0)]}×100

[0037] In the above three equations, x is the accumulated ash content or density of floating coal, and Y is the accumulated output of floating coal; the other parameters are the parameters of their respective models, which are obtained by nonlinear fitting. The purpose of nonlinear fitting is to find the parameters of their respective models. For the same model, the difference in curves is reflected in the difference in parameters. The fitting process is the process of finding the optimal parameters.

[0038] In the process of heavy medium separation, only two products can be obtained by density separation: heavy product and light product. The ideal state of raw coal in heavy medium separation is: all high-density materials with density greater than the separation density should be 100% distributed to the heavy product, and all low-density materials with density less than the separation density should be 100% distributed to the light product. In practice, the better separation result is that most of the high-density materials enter the heavy product, and a small part enters the light product; most of the low-density materials enter the light product, and a small part enters the heavy product. Therefore, the actual distribution curve can be fitted by seven models;

[0039] The product forecasting idea is expressed as follows:

[0040] r=∑Y i (X)×D i (X)

[0041] Where Y(X) is the raw coal data, and D(X) is the distribution curve;

[0042] When raw coal is separated by heavy medium, materials that are not specified in each product are called mismatches. The total amount of mismatches is equal to the sum of the percentages of materials that should not be mixed in each product. The mismatch curve is used as an indicator to evaluate the effect of the separation process. The calculation formula is:

[0043] M o =M l -M h

[0044] Where M o is the total mismatch content, %; M1 is the mismatch amount of materials with density less than the sorting density in the heavy product, %; M h is the mismatch amount of materials with density greater than the sorting density in the light product, %;

[0045] The sorting density for calculating the total amount of mismatches is the distribution density [δρ] or the equal error density [δe], where the equal error density refers to the density when the mismatches are equal in the double selection products;

[0046] The mismatch curve is expressed as:

[0047]

[0048] Where Mc is the total amount of mismatches when the cut point is at, %; ρmin and ρmax are the densities of the lowest and highest density materials in the material, kg / L; C is the cut point, kg / L; Y h (ρ) is the density distribution function of heavy products, %; Y l (ρ) is the density distribution function of light products, %; ρ is the density, kg / L; the first term on the right side of the formula is the loss curve, and the second term is the pollution curve;

[0049] According to the monitored suspension density value and the magnetic content value in the suspension, the coal slime content in the mixed medium barrel is calculated, and the coal slime content in the mixed medium barrel is adjusted by controlling the opening of the diverter valve to stabilize the coal slime content in the suspension, ensure that the coal slime content is within a reasonable range, and ensure the stability and fluidity of the suspension;

[0050] The calculation method of coal slime content is as follows:

[0051] G=A(ρ-1000)-BF

[0052] Where G is the coal slime content, kg / m 3 ; F is the magnetic content, kg / m 3 ; ρ is the suspension density, kg / m 3 ; A is the coefficient related to coal slime; B is the coefficient related to coal slime and magnetic material;

[0053] in:

[0054] In the formula, δ 煤泥 is the density of coal slime, kg / m 3 ; δ 磁 is the density of magnetite powder, kg / m 3 ;therefore,

[0055] When δ 煤泥 ,δ 磁 After the respective values ​​are determined, the value of coefficient A related to coal slime and the value of coefficient B related to coal slime and magnetic material are calculated. The magnetic material content is read by a magnetic material monitor, and the suspension density ρ is read by a differential pressure density meter. The percentage of coal slime content can be obtained in real time through the above formula;

[0056] The relationship model between coal quality and sorting density is obtained by combining the above formulas.

[0057] Optionally, after the execution unit executes the corresponding action for the first time, the actual sorting density after adjustment is fed back to the analysis system;

[0058] The analysis system determines the target sorting density adjustment amount required for controlling the density of the suspension according to the difference between the adjusted actual sorting density and the preset sorting density, in combination with the liquid level of the combined medium barrel obtained by the monitoring device and the predicted liquid level of the combined medium barrel, and through the support amount obtained in advance by the sorting model training, and sends the target sorting density adjustment amount to the control unit;

[0059] The control unit issues a new instruction to the execution unit again according to the target sorting density adjustment amount;

[0060] The execution unit executes corresponding actions according to the new instructions and continues to adjust the current actual sorting density;

[0061] The above process is circulated in sequence until the ash content of the product coal is close to the set ash content value and is within the allowable error range, and the execution unit adjusts the current value of the ash content of the product coal to be within the preset range.

[0062] Optionally, the sorting model is divided into three strategy modes according to the actual production process, namely, the medium addition mode, the steady state mode, and the disorder mode;

[0063] The medium adding mode refers to a working mode in which the medium needs to be supplemented to the combined medium barrel when the medium is continuously consumed after the heavy medium separation density adjustment adaptive control system has been in production and operation for a period of time and the separation density cannot be effectively improved by splitting the flow;

[0064] Steady-state mode refers to a working mode in which the heavy medium separation density adjustment adaptive control system makes the separation density in dynamic balance by controlling the diverter valve, the water replenishment valve, the liquid level of the mixed medium barrel and the liquid level of the dilute medium barrel MPC during the production process;

[0065] The disorder mode refers to a working mode in which the sorting density needs to be quickly adjusted when the sorting density is too high or too low due to production failure or human intervention during the sorting process.

[0066] Optionally, the monitoring device is connected to the analysis system via a network, the analysis system is connected to the control unit via a network, and the control unit is connected to the execution unit via a wire.

[0067] Optionally, when the execution unit adjusts the current value of the product coal ash content to within a preset range, the dense medium separation density adjustment adaptive control system enters the steady-state mode.

[0068] The present invention provides a heavy medium separation density adjustment adaptive control system, wherein the monitoring device is used to obtain real-time data of various types of equipment and transmit the data to the analysis system; the analysis system is used to screen, classify, statistically store the real-time data, and realize real-time data display and historical data storage, and transmit the real-time data to the control unit at the same time.

[0069] The analysis system is also used to establish a relationship model between coal quality and sorting density based on the big data analysis algorithm model, establish a sorting model according to the relationship model, and feedback the sorting density adjustment amount to the control unit according to the sorting model.

[0070] The control unit is used to issue instructions to the execution unit according to the sorting density adjustment amount to adjust the actual sorting density; the execution unit is used to perform corresponding actions according to the instructions issued by the control unit, adjust the actual sorting density, and feed back the adjusted actual sorting density to the analysis system.

[0071] Among them, the analysis system receives the adjusted actual sorting density, and feeds back the new sorting density adjustment amount to the control unit in combination with the new real-time data collected by the monitoring device, and the control unit issues new instructions to the execution unit according to the new sorting density adjustment amount, and the execution unit performs corresponding actions according to the new instructions, and continues to adjust the actual sorting density. The three processes are repeated until the execution unit adjusts the current value of the product coal ash content to within the preset range.

[0072] The heavy medium separation density adjustment adaptive control system proposed in the present invention can monitor various parameters affecting the sorting effect and system stability in the whole sorting process in real time by connecting various sensors of the monitoring device, thereby ensuring the stability of the liquid levels of the heavy medium barrel and the dilute medium barrel during the sorting process, and can effectively monitor leakage in the production process.

[0073] By monitoring the density and magnetic content in the suspension, the coal slime content in the suspension can be calculated, and the diversion can be controlled by the diverter valve to adjust the sorting density and stabilize the coal slime content in the suspension to ensure that the coal slime content is within a reasonable range. This prevents shallow trough agglomeration caused by excessive coal slime content and instability of the suspension when the coal slime content is too low, thereby ensuring the stability and fluidity of the suspension.

[0074] The intelligent adjustment of the sorting density of the heavy medium shallow trough is realized through the sorting model. The sorting density can be adjusted online in real time during the production process. It has higher control accuracy and more timely response speed, can quickly adapt to the density requirements of the washing process, and has good adaptability.

[0075] By monitoring the coal quality information of raw coal and product coal, and then establishing a sorting model, the adaptability is wider. If the properties of raw coal change due to the replacement of underground comprehensive mining working faces, changes in geological conditions and other reasons, it is only necessary to collect coal quality data through the system proposed by the present invention. After being processed by the big data analysis model, it can respond quickly and match the new sorting density to guide production.

[0076] The adaptive control system for density adjustment of heavy medium separation proposed in the present invention solves the problem that the density control of the heavy medium separation system in the existing coal preparation plant needs manual participation in adjustment, solves the problems of stability and timeliness of separation density, realizes adaptive adjustment control of heavy medium separation density, improves the stability and timeliness of the heavy medium separation system, and has strong adaptability and practicality. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0078] Figure 1 This is a schematic diagram of an optimal adaptive control system for density adjustment of heavy medium separation according to an embodiment of the present invention;

[0079] Figure 2 It is a schematic diagram of the system configuration screen and heavy medium intelligent control in an embodiment of the present invention;

[0080] Figure 3 It is a process of nonlinear fitting in an embodiment of the present invention;

[0081] Figure 4 It is an algorithm architecture diagram for model training and updating and online prediction of a sorting model in an embodiment of the present invention;

[0082] Figure 5 In the embodiment of the present invention, Figure 4 An algorithm model diagram of the algorithm architecture diagram;

[0083] Figure 6 It is an architecture diagram of the embodiment of the present invention, in which the coupling model based on reinforcement learning continuously optimizes all the models used;

[0084] Figure 7 This is an architecture diagram for further optimizing and improving each model through machine learning during operation in an embodiment of the present invention. DETAILED DESCRIPTION

[0085] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, are only part of the embodiments of the present invention, not all of the embodiments, and are not used to limit the present invention.

[0086] An adaptive control system for density adjustment of heavy medium separation according to an embodiment of the present invention includes: a monitoring device, an analysis system, a control unit, and an execution unit. The monitoring device is used to obtain real-time data of various devices and transmit them to the analysis system and the control unit; the analysis system is used to screen, classify, and statistically store real-time data, and realize real-time data display and historical data storage; the analysis system is also used to establish a relationship model between coal quality and separation density based on a big data analysis algorithm model, establish a separation model based on the relationship model, and feedback the separation density adjustment amount to the control unit based on the separation model.

[0087] The control unit is used to issue instructions to the execution unit according to the sorting density adjustment amount to adjust the actual sorting density; the execution unit is used to perform corresponding actions according to the instructions issued by the control unit, adjust the actual sorting density, and feed back the adjusted actual sorting density to the analysis system.

[0088] Among them, the analysis system receives the adjusted actual sorting density, and feeds back the new sorting density adjustment amount to the control unit in combination with the real-time data of clean coal ash content collected by the monitoring device, and the control unit issues new instructions to the execution unit according to the new sorting density adjustment amount, and the execution unit performs corresponding actions according to the new instructions, and continues to adjust the actual sorting density. The three processes are repeated until the execution unit adjusts the current value of the product coal ash content to within the preset range.

[0089] For the above monitoring devices, analysis systems, control units, and execution units, the preferred choice is that the monitoring devices include: online ash meter for raw coal, online ash meter for clean coal, differential pressure density meter, magnetic material monitor, liquid level sensor for mixed medium barrel, liquid level sensor for dilute medium barrel, and medium adding scale; the analysis system includes: data center, big data analysis algorithm model; the control unit includes: control console, PLC control unit, and the execution unit includes: medium adding pump, mixed medium pump, dilute medium pump, water adding valve, water replenishment valve, diverter valve, shallow tank separator, etc. In order to better and more intuitively understand the entire system, refer to Figure 1 The following is a diagram showing an optimal architecture of a density adjustment adaptive control system for heavy medium separation. Figure 1 The intelligent sorting module refers to both the relationship model between coal quality and sorting density and the sorting model; the PLC controller refers to the PLC control unit; the shallow trough water supply valve refers to the water supply valve; the combined medium water supply valve and the concentrated medium barrel water supply valve refer to the water supply valve; the concentrated medium pump refers to the medium supply pump, the combined medium pump, and the dilute medium pump.

[0090] The preferred choice between the monitoring device and the analysis system is to be connected via a network, the preferred choice between the analysis system and the control unit is to be connected via a network, and the preferred choice between the control unit and the execution unit is to be connected via a wire.

[0091] The raw coal online ash meter and clean coal online ash meter in the monitoring device are used to obtain the ash content, moisture and calorific value of the raw coal and each product coal in real time; the combined medium barrel level sensor and the dilute medium barrel level sensor are used to obtain the liquid level information of the combined medium barrel and the dilute medium barrel in real time; the differential pressure density meter is used to obtain the density in the suspension in real time; the magnetic material monitor is used to obtain the magnetic material content in the suspension in real time; the medium adding scale is used to obtain the weight of the medium added to the concentrated medium barrel each time. These real-time information constitutes the real-time data obtained by the monitoring device.

[0092] The data center in the analysis system is used to filter, classify, and store real-time data, filter out invalid data, display real-time data and store historical data, and transmit real-time data to the control unit.

[0093] The big data analysis algorithm model in the analysis system is used to analyze and compare based on real-time data and historical data, and to establish a relationship model between coal quality and sorting density in combination with the selectivity curve, distribution curve and mismatch curve of the comprehensive floating and sinking test of raw coal. It is also used to establish a sorting model based on the relationship model, combined with the real-time data of sorting density, magnetic content, combined medium barrel position, dilute medium barrel position, water supply valve opening and diverter valve opening, and to feed back the sorting density adjustment amount to the control unit based on the sorting model.

[0094] The control console in the control unit cooperates with the PLC control unit to issue instructions to the execution unit according to the sorting density adjustment amount to adjust the actual sorting density.

[0095] For the execution unit, when the heavy medium separation density adjustment adaptive control system is in the steady-state mode, the opening of the water supply valve and the diverter valve is controlled to perform MPC adjustment; when the heavy medium separation density adjustment adaptive control system is in the medium adding mode, the concentrated medium pump and the water supply valve are opened to supplement the medium to the combined medium barrel, and the weight of the added medium is measured by the medium adding metering scale of the monitoring device, while the suspension density and the liquid level of the combined medium barrel are monitored, and the water supply valve is opened at the same time to stabilize the sorting density. When the liquid level of the combined medium barrel reaches the predicted high liquid level setting value, the medium addition is stopped, so that the heavy medium separation density adjustment adaptive control system enters the steady-state mode as soon as possible.

[0096] When the adaptive control system for density regulation of heavy medium separation is in disordered mode and the separation density is too high, the density of the suspension is lowered by increasing the opening of the water-making valve and opening the water-adding valve. At the same time, the liquid level in the combined medium barrel is monitored to prevent overflowing caused by excessive water addition. When the predicted high liquid level setting value of the combined medium barrel is reached, the water-adding valve is closed and the opening of the water-making valve is adjusted to allow the adaptive control system for density regulation of heavy medium separation to enter steady-state mode as soon as possible.

[0097] When the adaptive control system for density regulation of heavy medium separation is in disordered mode and the separation density is low, the suspension density is increased by reducing the opening of the water supply valve and opening the diverter valve. At the same time, the liquid levels of the thin medium barrel and the combined medium barrel are monitored to prevent the thin medium barrel from overflowing. If the combined medium barrel is lower than the preset low liquid level due to diversion, the medium addition mode is triggered, and combined with the suspension density in the monitoring device, the adaptive control system for density regulation of heavy medium separation can enter the steady-state mode as soon as possible.

[0098] The above whole working process can refer to Figure 2The system configuration screen and the heavy medium intelligent control schematic diagram shown in the figure can be better understood. The monitoring device obtains the real-time data of raw coal ash content, clean coal ash content, suspension density, magnetic content, combined medium tank level, dilute medium tank level, and medium metering scale, and inputs them into the analysis system.

[0099] In addition to screening, classifying, statistically storing real-time data, screening out invalid data, and realizing real-time data display and historical data storage, the analysis system also uses big data analysis algorithm models to establish relationship models and sorting models between coal quality and sorting density. Figure 2 The prediction value calculated by the sorting model (i.e., the sorting density adjustment amount) is shown in the analysis system box. After comparing and analyzing the prediction value calculated by the BP neural network, the prediction value is output. The prediction value calculated by the sorting model is exemplified by LSTM. For the real-time data obtained by data collection, core element extraction, number of hidden layers, hidden layer nodes, time synchronization, LSTM network configuration, variable input LSTM network training and learning are performed, and the termination condition is 500 iterations. Finally, a prediction value is obtained, and another prediction value is obtained by directly calculating the real-time data using the BP neural network. After comparing and analyzing the two, the prediction value is output to the control unit.

[0100] After receiving the predicted value, the control unit coordinates with the PLC control unit to send the prediction value to the execution unit ( Figure 2 In the figure, the medium adding pump, medium combining pump, medium diluting pump, water adding valve, water replenishing valve, diverter valve, shallow trough separator) are exemplarily shown to issue instructions to adjust the actual sorting density.

[0101] After the adjustment, the monitoring device continues to obtain real-time data again. The analysis system determines the target sorting density adjustment amount required for controlling the density of the suspension based on the difference between the adjusted actual sorting density and the preset sorting density, combined with the liquid level of the combined medium barrel obtained by the monitoring device and the predicted liquid level of the combined medium barrel, and the support amount obtained in advance through the sorting model training, and sends the target sorting density adjustment amount to the control unit.

[0102] The control unit issues a new instruction to the execution unit again according to the target sorting density adjustment amount; the execution unit performs corresponding actions according to the new instruction and continues to adjust the current actual sorting density. The whole process is circulated in sequence until the product coal ash value and the set ash value are close to the allowable error range, and the execution unit adjusts the current value of the product coal ash content to within the preset range. When the execution unit adjusts the current value of the product coal ash content to within the preset range, the heavy medium separation density adjustment adaptive control system enters the steady-state mode.

[0103] The above big data analysis algorithm model analyzes and compares real-time data and historical data, and combines the selectivity curve, distribution curve, and mismatch curve of the raw coal comprehensive floating and sinking test to establish a relationship model between coal quality and sorting density, including:

[0104] By using the data on the quantity and quality of products at several density levels in the floating and sinking tests, the data are plotted into curves using a graphical method, thereby determining the relationship between the quantity and quality of the material within a certain density range. Through the comprehensive floating and sinking test selectivity curve of raw coal, the quality characteristics of the raw coal can be understood in a timely manner, the selectivity of the raw coal can be evaluated, the clean coal yield and ash content can be predicted, and the theoretical sorting density can be determined.

[0105] The raw coal comprehensive floating and sinking test selectivity curve is drawn by fitting, and the floating matter accumulation curve β (i.e., β curve) and density curve δ (i.e., δ curve) are fitted. The remaining three curves: ash characteristic curve λ (i.e., λ curve), sediment accumulation curve θ (i.e., θ curve), and density ±0.1 curve ε (i.e., ε curve) are calculated through the first two curves.

[0106] Among them, three models are used for fitting according to the shapes of the β curve and the δ curve:

[0107] ① Inverse tangent model

[0108] ②Hyperbolic tangent model Y = {a + c × th [k (x-x0)]} × 100, where

[0109] ③Compound hyperbolic tangent model Y={a+bx+c×th[k(x-x0)]}×100

[0110] In the above three equations, x is the cumulative ash content or density of floating coal, and Y is the cumulative output of floating coal; the other parameters are the parameters of their respective models, and the parameters of the model are obtained by nonlinear fitting method.

[0111] The purpose of nonlinear fitting is to find the parameters of each model. For the same model, the difference in curves is reflected in the difference in parameters. The fitting process is the process of finding the optimal parameters. Figure 3 The figure shows a process of nonlinear fitting. Select any model, input the initial value of the model parameters, calculate the fitting error, and determine whether the error meets the requirements.

[0112] If the conditions are met, the result is printed directly and the process ends. If the conditions are not met, the model parameters are changed using the optimization method and the calculation of the fitting error and subsequent steps are returned.

[0113] In the process of heavy medium separation, only two products can be obtained by density separation: heavy product and light product. The ideal state of raw coal in heavy medium separation is: all high-density materials with density greater than the separation density should be 100% distributed to the heavy product, and all low-density materials with density less than the separation density should be 100% distributed to the light product. In practice, the better separation result is that most of the high-density materials enter the heavy product, and a small part enters the light product; most of the low-density materials enter the light product, and a small part enters the heavy product. Therefore, the actual distribution curve can be fitted by seven models;

[0114] The product forecasting idea is expressed as follows:

[0115] r=∑Y i (X)×D i (X)

[0116] Among them, Y(X) is the raw coal data and D(X) is the distribution curve.

[0117] When raw coal is separated by heavy medium, materials that are not specified in each product are called mismatches. The total amount of mismatches is equal to the sum of the percentages of materials that should not be mixed in each product. The mismatch curve is used as an indicator to evaluate the effect of the separation process. The calculation formula is:

[0118] M o =M l -M h

[0119] Where M o is the total mismatch content (input), %; M l M is the mismatch amount of materials with density less than the sorting density in the heavy product (accounting for the input material), %; h It is the mismatch amount of materials with density greater than the sorting density in the light product (accounting for the input material), %.

[0120] The sorting density for calculating the total amount of mismatches is calculated using the partition density [δ ρ ] or equal error density [δ e ], equal error density refers to the density when the mismatches are equal in the two-fold selection product;

[0121] The mismatch curve is expressed as:

[0122]

[0123] Where M c is the total amount of mismatches (input) when the cutting point is at, %, ρ min , max are the density of the lowest and highest density materials in the material, kg / L; C is the cutting point, kg / L; Y h (ρ) is the density distribution function of heavy products (accounting for input materials), %; Yl (ρ) is the density distribution function of light products (accounting for input material), %; ρ is the density, kg / L; the first term on the right side of the formula is the loss curve, and the second term is the contamination curve.

[0124] According to the monitored suspension density value and the magnetic content value in the suspension, the coal slime content in the mixed medium barrel is calculated, and the coal slime content in the mixed medium barrel is adjusted by controlling the opening of the diverter valve to stabilize the coal slime content in the suspension, ensure that the coal slime content is within a reasonable range, and ensure the stability and fluidity of the suspension;

[0125] The calculation method of coal slime content is as follows:

[0126] G=A(ρ-1000)-BF

[0127] Where G is the coal slime content (non-magnetic matter), kg / m 3 ; F is the magnetic content, kg / m 3 ; ρ is the suspension density, kg / m 3 ; A is the coefficient related to coal slime; B is the coefficient related to coal slime and magnetic material;

[0128] in:

[0129] In the formula, δ 煤泥 is the density of coal slime, kg / m 3 ; δ 磁 is the density of magnetite powder, kg / m 3 ;therefore,

[0130] When δ 煤泥 ,δ 磁 After the respective values ​​are determined, the value of coefficient A related to coal slime and the value of coefficient B related to coal slime and magnetic material are calculated. For example: 煤泥 Take 1.5g / cm 3 , δ 磁 Take 4.5g / cm 3 , the coefficient A=3 related to coal slime, the coefficient B=2.33 related to coal slime and magnetic material, the magnetic material content is read by the magnetic material monitor, and the suspension density ρ is read by the differential pressure density meter. The percentage of coal slime content can be obtained in real time through the above formula;

[0131] Combining the above formulas, we can get the relationship model between coal quality and sorting density.

[0132] For the sorting model, it is based on the monitored sorting density, magnetic content, mixed medium barrel position, dilute medium barrel position, water supply valve opening, diverter valve opening and other data to establish an algorithm model between the sorting density and the magnetic content, liquid level, and each valve opening. The model can be compared and optimized through the feedback mechanism. Figure 4 The diagram shows the architecture of model training, updating and online prediction of a sorting model. During model training and updating, the historical clean coal ash content and qualified medium density data are read, the data is cleaned, the ash content data is corrected, and the data is normalized. The XGBOOST model is used to train the sorting model, which can be updated once a minute.

[0133] In the online prediction, the clean coal ash content and qualified medium density data at the current moment are read, the output sample data is re-input and the data is normalized, the current XGBOOST model is used to train the sorting model, and finally the density setting value (i.e., the sorting density adjustment amount) is output.

[0134] Figure 5 Shows the corresponding Figure 4 The algorithm model of the algorithm framework, in which the input quantities U1, U2, and U3 are respectively the ash content of the clean coal, the density of the suspension, and the content of magnetic matter, G 11 (s), G 21 (s), G 31 (s) are the set range values, G 12 (s), G 22 (s), G 32 (s) is a value that exceeds the set range, Y1 is the output of the control unit, and Y2 is the value that needs to be adjusted when fed back to the analysis system. In this model, when the input quantity, i.e., the real-time monitoring data, is within the system setting range, it is directly fed back to the control unit to enter the steady-state mode. When the input quantity, i.e., the monitoring data, is not within the set range, the data is fed back to the analysis system for data comparison and analysis again, and the system parameters are adjusted to cycle again.

[0135] It should be noted that the above Figure 4 , 5 This is only an example for better understanding of the optimization of the sorting model, and does not mean that the optimization of the sorting model can only be carried out in this way. Those skilled in the art can optimize the sorting model in other ways, as long as the sorting model can achieve its function.

[0136] The sorting model is divided into three strategy modes according to the actual production process: medium addition mode, steady-state mode, and disordered mode; the sorting model can adjust the adaptive control system according to the current situation of the heavy medium sorting density, and automatically select the corresponding mode so that the execution unit performs the corresponding action according to the corresponding strategy.

[0137] The medium adding mode refers to a working mode in which the medium is continuously consumed after the heavy medium sorting density regulating adaptive control system has been in production and operation for a period of time, and when the sorting density cannot be effectively improved by diverting, it is necessary to supplement the medium to the combined medium barrel; the steady-state mode refers to a working mode in which the heavy medium sorting density regulating adaptive control system controls the diverter valve, water replenishment valve, combined medium barrel liquid level, and dilute medium barrel liquid level MPC control during the production process to keep the sorting density in dynamic balance; the disorder mode refers to a working mode in which the sorting density needs to be quickly adjusted when the sorting density is too high or too low due to production failure or human intervention during the sorting process.

[0138] In addition, the present invention also proposes to establish a coupling model based on reinforcement learning to continuously optimize all the models used, so that the current value of the ash content of the actually monitored product coal continuously approaches the set ash content value, thereby achieving a steady-state mode that can guide production.

[0139] Reference Figure 6 The schematic diagram shows the architecture of the continuous optimization of all models used by the coupling model based on reinforcement learning. Taking the optimization of the sorting model as an example, the coupling model of reinforcement learning optimizes the sorting model. Figure 6 It is defined as a liquid level density coupling model, and the control parameters are determined based on the new concept density function. The updated concept density function is conducive to reinforcement learning. At the same time, the evaluation function is introduced, which has a reward and punishment mechanism. Specifically:

[0140] Set the interval of each parameter, calculate the initial probability density distribution of the coupling model parameters, calculate the coupling model parameters, execute the coupling model, calculate the objective function, calculate the reward and punishment values, update the probability density distribution of each parameter, and finally determine whether to end the learning. If not, return to the calculation of the coupling model parameters and the steps after that.

[0141] After all models are optimized, they are connected to the control unit to test and run the entire heavy medium separation density adjustment adaptive control system, and each model is further optimized and improved through machine learning during the operation.

[0142] Reference Figure 7 The schematic diagram shows the architecture diagram for further optimization and improvement of each model through machine learning during operation. The raw coal ash content and clean coal set ash content are used as input conditions, the liquid level of the mixed medium barrel and the dilute medium barrel is used as the set range value (set liquid level), and the historical operation data of the water replenishment valve, water addition valve, diverter valve, mixed medium pump, dilute medium pump, addition medium pump, suspension density, and magnetic content are used as reference trajectories. Through model training (density-liquid level coupling model), combined with the selected product clean coal ash feedback value (feedback prediction period), the entire system model is feedback-adjusted, and training and optimization are continuously carried out until the error between the clean coal monitoring ash content and the set ash content is within the allowable range.

[0143] It should be noted that the above Figure 6 , 7 This is only an example for better understanding of the optimization of each model, and does not mean that the optimization of each model can only be carried out in this way. Those skilled in the art can optimize each model in other ways, as long as each model can achieve its function.

[0144] Generally, when the adaptive control system for density regulation of heavy medium separation is in steady-state mode, the control unit controls the opening of the water supply valve and the diverter valve for MPC regulation. In steady-state mode, the main factors affecting the quality of product coal are sorting density, magnetic content, and coal slime content. Among them, the sorting density can be adjusted by the opening of the water supply valve, and the size of the water supply valve opening can timely reflect the size of the suspension density; by adjusting the opening of the diverter valve, the size of the magnetic content and coal slime content of the suspension can be adjusted to ensure the viscosity and stability of the suspension. There is a negative correlation between the water supply valve and the diverter valve. When the water supply valve is opened large, the diverter valve is reduced; when the water supply valve is small, the diverter valve is opened large.

[0145] When the adaptive control system for density adjustment of heavy medium separation increases density by diverting, and the liquid level of the combined medium barrel reaches the predicted low liquid level setting value, it indicates that there is insufficient medium in the system and medium needs to be added. The separation model will trigger the medium addition mode, and the control unit will issue instructions to the concentrated medium pump and water supply valve in the execution unit to open, add medium to the combined medium barrel, and use the monitoring device to add medium to measure the weight of the added medium, while monitoring the density of the suspension and the liquid level of the combined medium barrel. During the process of adding medium, the density of the combined medium barrel will increase, and the water supply valve needs to be opened at the same time to stabilize the sorting density. When the liquid level of the combined medium barrel reaches the predicted high liquid level setting value, stop adding medium, so that the adaptive control system for density adjustment of heavy medium separation can enter the steady-state mode as soon as possible, and feed the monitoring data back to the analysis system.

[0146] When the density adjustment adaptive control system of heavy medium separation is in disorder mode and the separation density is high, the suspension density is adjusted to be lowered by increasing the opening of the water supply valve and opening the combined medium water supply valve, and the combined medium barrel liquid level is monitored at the same time to prevent the barrel from overflowing due to excessive water addition. Combined with the suspension density monitoring device, when the predicted high liquid level setting value of the combined medium barrel is reached, the water supply valve is closed, and the opening of the water supply valve is adjusted, so that the density adjustment adaptive control system of heavy medium separation enters the steady-state mode as soon as possible, and the monitoring data is fed back to the analysis system.

[0147] When the density adjustment adaptive control system of heavy medium separation is in disorder mode and the separation density is low, the suspension density is increased by reducing the opening of the water supply valve and opening the diverter valve, and the liquid level of the dilute medium barrel and the combined medium barrel is monitored at the same time to prevent the dilute medium barrel from overflowing. If the combined medium barrel is lower than the low liquid level prediction setting value through diverter, the medium addition mode is triggered. Combined with the suspension density monitoring device, the density adjustment adaptive control system of heavy medium separation enters the steady-state mode as soon as possible, and the monitoring data is fed back to the analysis system.

[0148] In summary, the heavy medium sorting density adjustment adaptive control system proposed in the present invention, by connecting various sensors of the monitoring device, conducts real-time monitoring of various parameters affecting the sorting effect and system stability in the entire sorting process, ensures the stability of the liquid levels of the heavy medium barrel and the dilute medium barrel during the sorting process, and can effectively monitor leakage in the production process.

[0149] By monitoring the density and magnetic content in the suspension, the coal slime content in the suspension can be calculated, and the diversion can be controlled by the diverter valve to adjust the sorting density and stabilize the coal slime content in the suspension to ensure that the coal slime content is within a reasonable range. This prevents shallow trough agglomeration caused by excessive coal slime content and instability of the suspension when the coal slime content is too low, thereby ensuring the stability and fluidity of the suspension.

[0150] The intelligent adjustment of the sorting density of the heavy medium shallow trough is realized through the sorting model. The sorting density can be adjusted online in real time during the production process. It has higher control accuracy and more timely response speed, can quickly adapt to the density requirements of the washing process, and has good adaptability.

[0151] By monitoring the coal quality information of raw coal and product coal, and then establishing a sorting model, the adaptability is wider. If the properties of raw coal change due to the replacement of underground comprehensive mining working faces, changes in geological conditions and other reasons, it is only necessary to collect coal quality data through the system proposed by the present invention. After being processed by the big data analysis model, it can respond quickly and match the new sorting density to guide production.

[0152] The adaptive control system for density adjustment of heavy medium separation proposed in the present invention solves the problem that the density control of the heavy medium separation system in the existing coal preparation plant needs manual participation in adjustment, solves the problems of stability and timeliness of separation density, realizes adaptive adjustment control of heavy medium separation density, improves the stability and timeliness of the heavy medium separation system, and has strong adaptability and practicality.

[0153] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0154] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0155] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are within the protection of the present invention.

Claims

1. A dense medium separation density adjustment adaptive control system, characterized in that: The dense medium separation density adjustment adaptive control system comprises: a monitoring device, an analysis system, a control unit, and an execution unit; The monitoring device is used to obtain real-time data of various types of equipment and transmit it to the analysis system and the control unit; The analysis system is used to screen, classify, count and store the real-time data, and realize real-time data display and historical data storage; The analysis system is further used to establish a relationship model between coal quality and sorting density based on a big data analysis algorithm model, establish a sorting model according to the relationship model, and feed back a sorting density adjustment amount to the control unit according to the sorting model; The control unit is used to issue instructions to the execution unit according to the sorting density adjustment amount to adjust the actual sorting density; The execution unit is used to execute corresponding actions according to the instructions issued by the control unit, adjust the actual sorting density, and feed back the adjusted actual sorting density to the analysis system; The analysis system receives the adjusted actual sorting density, and feeds back a new sorting density adjustment amount to the control unit in combination with the new real-time data collected by the monitoring device, and the control unit issues a new instruction to the execution unit according to the new sorting density adjustment amount, and the execution unit performs corresponding actions according to the new instruction, and the three processes of continuing to adjust the actual sorting density are repeated until the current value of the product coal ash value adjusted by the execution unit reaches within the preset range; The analysis system includes: a big data analysis algorithm model; The big data analysis algorithm model is used to analyze and compare the real-time data and historical data, and to establish a relationship model between coal quality and sorting density in combination with the selectivity curve, distribution curve, and mismatch curve of the raw coal comprehensive floating and sinking test; wherein the method for establishing the relationship model between coal quality and sorting density includes: The data of the quantity and quality of products at several density levels in the floating and sinking test are used to curve them by a graphical method, so as to determine the relationship between the quantity and quality of materials within a certain density range. Through the raw coal comprehensive floating and sinking test selectivity curve, the coal quality characteristics of the raw coal can be timely understood, the raw coal selectivity can be evaluated, the clean coal yield and ash content can be predicted, and the theoretical sorting density can be determined; the raw coal comprehensive floating and sinking test selectivity curve is drawn by fitting, and the floating cumulative curve β and the density curve δ are fitted. The remaining three curves: the ash characteristic curve λ, the sinking cumulative curve θ, and the density ±0.1 curve ε are calculated through the previous two curves; Among them, three models are used for fitting according to the shapes of curve β and curve δ: ① Inverse tangent model ②Hyperbolic tangent model , where ③Compound hyperbolic tangent model In the above three equations, x is the accumulated ash content or density of floating coal, and Y is the accumulated output of floating coal; the other parameters are the parameters of their respective models, which are obtained by nonlinear fitting. The purpose of nonlinear fitting is to find the parameters of their respective models. For the same model, the difference in curves is reflected in the difference in parameters. The fitting process is the process of finding the optimal parameters. In the process of heavy medium separation, only two products can be obtained by density separation: heavy product and light product. The ideal state of raw coal in heavy medium separation is: all high-density materials with density greater than the separation density should be 100% distributed to the heavy product, and all low-density materials with density less than the separation density should be 100% distributed to the light product. In practice, the better separation result is that most of the high-density materials enter the heavy product, and a small part enters the light product; most of the low-density materials enter the light product, and a small part enters the heavy product. Therefore, the actual distribution curve can be fitted by seven models; The product forecasting idea is expressed as follows: Among them, Y(X) is the raw coal data, and D(X) is the distribution curve; When raw coal is separated by heavy medium, materials that are not specified in each product are called mismatches. The total amount of mismatches is equal to the sum of the percentages of materials that should not be mixed in each product. The mismatch curve is used as an indicator to evaluate the effect of the separation process. The calculation formula is: Where M o is the total mismatch content, %; M1 is the mismatch content of materials with density less than the sorting density in the heavy product, %; M h is the mismatch amount of materials with density greater than the sorting density in the light product, %; The sorting density for calculating the total amount of mismatches is the distribution density [δρ] or the equal error density [δe], where the equal error density refers to the density when the mismatches are equal in the double selection products; the mismatch curve is expressed as: Where M c is the total amount of mismatches when the cutting point is at , %; ρ min , max are the density of the lowest and highest density materials in the material, kg / L; C is the cutting point, kg / L; Y h (ρ) is the density distribution function of heavy products; Y l (ρ) is the density distribution function of light products, %; ρ is the density, kg / L; the first term on the right side of the formula is the loss curve, and the second term is the pollution curve; According to the monitored suspension density value and the magnetic content value in the suspension, the coal slime content in the mixed medium barrel is calculated, and the coal slime content in the mixed medium barrel is adjusted by controlling the opening of the diverter valve to stabilize the coal slime content in the suspension, ensure that the coal slime content is within a reasonable range, and ensure the stability and fluidity of the suspension; the coal slime content calculation method reference formula is: G = A (ρ - 1000) - BF Where, G is the coal slime content, kg / m³; F is the magnetic content, kg / m³; ρ is the suspension density, kg / m³; A is the coefficient related to coal slime; B is the coefficient related to coal slime and magnetic matter; in: , In the formula, δ 煤泥 is the density of coal slime, kg / m 3 ; δ 磁 is the density of magnetite powder, kg / m 3 ;therefore, ; When δ 煤泥 , δ 磁 After the respective values ​​are determined, the value of coefficient A related to coal slime and the value of coefficient B related to coal slime and magnetic material are calculated. The magnetic material content is read by a magnetic material monitor, and the suspension density ρ is read by a differential pressure density meter. The percentage of coal slime content can be obtained in real time through the above formula; The relationship model between coal quality and sorting density is obtained by combining the above formulas.

2. The dense medium separation density adjustment adaptive control system according to claim 1 is characterized in that: The monitoring device includes: an online raw coal ash meter, an online clean coal ash meter, a differential pressure density meter, a magnetic material monitor, a liquid level sensor for a mixed medium barrel, a liquid level sensor for a diluted medium barrel, and a medium adding scale; The raw coal online ash meter and the clean coal online ash meter are used to obtain the ash content, moisture content and calorific value of the raw coal and each product coal in real time respectively; The liquid level sensor for the combined medium barrel and the liquid level sensor for the diluted medium barrel are used to obtain the liquid level information of the combined medium barrel and the diluted medium barrel in real time, respectively; The differential pressure densitometer is used to obtain the density of the suspension in real time; The magnetic substance monitor is used to obtain the magnetic substance content in the suspension in real time; The medium adding scale is used to obtain the weight of the medium added to the concentrated medium barrel each time; The ash content, the moisture content, the calorific value, the liquid level information, the density, the magnetic content, and the medium weight constitute the real-time data.

3. The dense medium separation density adjustment adaptive control system according to claim 1, characterized in that: The analysis system also includes: a data center; The data center is used to screen, classify, count and store the real-time data, filter out invalid data, and realize the real-time data display and historical data storage; The big data analysis algorithm model is also used to establish the sorting model based on the relationship model, combined with the real-time data of sorting density, magnetic content, mixed medium barrel position, dilute medium barrel position, water replenishment valve opening, and diverter valve opening, and to feed back the sorting density adjustment amount to the control unit based on the sorting model.

4. The dense medium separation density adjustment adaptive control system according to claim 1, characterized in that: The control unit comprises: a control console and a PLC control unit; The control console is used to issue instructions to the execution unit in cooperation with the PLC control unit according to the sorting density adjustment amount, so as to adjust the actual sorting density.

5. The dense medium separation density adjustment adaptive control system according to claim 2, characterized in that: The execution unit includes: a medium adding pump, a medium combining pump, a medium diluting pump, a water adding valve, a water replenishing valve, a diverter valve, and a shallow tank separator; the medium diluting pump includes: the medium adding pump, the medium combining pump, and the medium diluting pump; When the dense medium separation density adjustment adaptive control system is in a steady-state mode, controlling the openings of the water supply valve and the diverter valve to perform MPC adjustment; When the heavy medium separation density adjustment adaptive control system is in the medium adding mode, the concentrated medium pump and the water supply valve are opened to supplement the medium to the mixed medium barrel, and the weight of the added medium is measured by the medium adding metering scale of the monitoring device, while the suspension density and the liquid level of the mixed medium barrel are monitored, and the water supply valve is opened to stabilize the separation density. When the liquid level of the mixed medium barrel reaches the predicted high liquid level setting value, the medium addition is stopped, so that the heavy medium separation density adjustment adaptive control system enters the steady-state mode as soon as possible; When the dense medium separation density adjustment adaptive control system is in a disordered mode and the separation density is high, the density of the suspension is lowered by increasing the opening of the water replenishment valve and opening the water addition valve, and the liquid level of the combined medium barrel is monitored at the same time to prevent the barrel from overflowing due to excessive water addition. When the predicted high liquid level setting value of the combined medium barrel is reached, the water addition valve is closed, and the opening of the water replenishment valve is adjusted to make the dense medium separation density adjustment adaptive control system enter the steady state mode as soon as possible; When the heavy medium separation density adjustment adaptive control system is in a disordered mode and the separation density is low, the suspension density is increased by reducing the water supply valve opening and opening the diverter valve, and the liquid levels of the dilute medium barrel and the combined medium barrel are monitored at the same time to prevent the dilute medium barrel from overflowing. If the combined medium barrel is lower than the low liquid level preset value due to diversion, the medium addition mode is triggered, and combined with the suspension density in the monitoring device, the heavy medium separation density adjustment adaptive control system can enter the steady-state mode as soon as possible.

6. The dense medium separation density adjustment adaptive control system according to claim 1, characterized in that: After the execution unit performs the corresponding action for the first time, the actual sorting density after adjustment is fed back to the analysis system; The analysis system determines the target sorting density adjustment amount required for controlling the density of the suspension according to the difference between the adjusted actual sorting density and the preset sorting density, in combination with the liquid level of the combined medium barrel obtained by the monitoring device and the predicted liquid level of the combined medium barrel, and through the support amount obtained in advance by the sorting model training, and sends the target sorting density adjustment amount to the control unit; The control unit issues a new instruction to the execution unit again according to the target sorting density adjustment amount; The execution unit executes corresponding actions according to the new instructions and continues to adjust the current actual sorting density; The above process is circulated in sequence until the product coal ash value and the set ash value are close to within the allowable error range, and the execution unit adjusts the current value of the product coal ash value to be within the preset range.

7. The dense medium separation density adjustment adaptive control system according to claim 5, characterized in that: The separation model is divided into three strategic modes according to the actual production process: the medium addition mode, the steady state mode, and the disorder mode; The medium adding mode refers to a working mode in which the medium needs to be supplemented to the combined medium barrel when the medium is continuously consumed after the heavy medium separation density adjustment adaptive control system has been in production and operation for a period of time and the separation density cannot be effectively improved by splitting the flow; Steady-state mode refers to a working mode in which the heavy medium separation density adjustment adaptive control system makes the separation density in dynamic balance by controlling the diverter valve, the water replenishment valve, the liquid level of the combined medium barrel and the liquid level of the dilute medium barrel through MPC control during the production process; The disorder mode refers to a working mode in which the sorting density needs to be adjusted quickly when the sorting density is too high or too low due to production failure or human intervention during the sorting process.

8. The dense medium separation density adjustment adaptive control system according to claim 1, characterized in that: The monitoring device is connected to the analysis system via a network, the analysis system is connected to the control unit via a network, and the control unit is connected to the execution unit via a wire.

9. The dense medium separation density adjustment adaptive control system according to claim 7, characterized in that: When the current value of the product coal ash content adjusted by the execution unit reaches within a preset range, the dense medium separation density adjustment adaptive control system enters the steady-state mode.

Images