CN117193014A - QP-DMC-based rapid dense medium coal density control system - Google Patents
QP-DMC-based rapid dense medium coal density control system Download PDFInfo
- Publication number
- CN117193014A CN117193014A CN202311297510.5A CN202311297510A CN117193014A CN 117193014 A CN117193014 A CN 117193014A CN 202311297510 A CN202311297510 A CN 202311297510A CN 117193014 A CN117193014 A CN 117193014A
- Authority
- CN
- China
- Prior art keywords
- suspension
- valve
- control
- opening
- density
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000003245 coal Substances 0.000 title claims abstract description 31
- 230000004044 response Effects 0.000 claims abstract description 36
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 14
- 239000000725 suspension Substances 0.000 claims description 198
- 239000007788 liquid Substances 0.000 claims description 69
- 239000011159 matrix material Substances 0.000 claims description 46
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 45
- 230000001502 supplementing effect Effects 0.000 claims description 44
- 238000000034 method Methods 0.000 claims description 37
- 230000001105 regulatory effect Effects 0.000 claims description 33
- 238000012546 transfer Methods 0.000 claims description 27
- 230000008569 process Effects 0.000 claims description 19
- 238000013095 identification testing Methods 0.000 claims description 17
- 230000008859 change Effects 0.000 claims description 11
- 238000000926 separation method Methods 0.000 claims description 10
- 230000001276 controlling effect Effects 0.000 claims description 7
- 238000002474 experimental method Methods 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 239000000463 material Substances 0.000 claims 1
- 238000004886 process control Methods 0.000 claims 1
- 230000035945 sensitivity Effects 0.000 abstract 1
- 238000005457 optimization Methods 0.000 description 10
- 238000004088 simulation Methods 0.000 description 8
- 238000004364 calculation method Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 238000004519 manufacturing process Methods 0.000 description 4
- 238000002156 mixing Methods 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 238000002360 preparation method Methods 0.000 description 3
- 241000073677 Changea Species 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000005530 etching Methods 0.000 description 1
- 239000012530 fluid Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Abstract
The invention provides a rapid dense medium coal dressing density control system based on QP-DMC, which is used in dense medium coal dressing equipment and comprises a relay control module, a QP-DMC control model and a control executor; the feedforward decoupling module is used for decomposing the multivariable prediction task into a plurality of univariate prediction tasks by reserving important factors and performing feedforward compensation on influencing factors under different branches divided by the branch control module, combining all univariate prediction results to obtain an overall prediction result, combining a QP-DMC control model to obtain a control instruction, and sending a control command to a control actuator, so that quick prediction and control are realized, and sensitivity of a controller to disturbance and model mismatch is reduced. The system can simplify the calculated amount of the algorithm, improve the real-time performance of the control system and accelerate the control response speed, thereby realizing the real-time online adjustment of the dense medium coal density.
Description
Technical Field
The invention relates to the technical field of dense medium coal dressing density control, and particularly provides a rapid dense medium coal dressing density control system based on QP-DMC.
Background
The intelligent dense-medium coal separation process can be realized, the yield of clean coal can be improved, the quantity of coal carried in gangue can be reduced, and therefore, the production efficiency is improved, and the stability and the accuracy of a dense-medium coal separation system are obviously improved. The key of the intellectualization of dense medium coal dressing is to realize the control of the density of suspension liquid. At present, a simple PID control method is generally adopted in an industrial field, but the control effect is not ideal. This is because the dense medium density control involves the mutual coupling of the suspension density and the liquid level, and the adjustment of the opening of the diverter valve, the water replenishment valve and the medium replenishment valve all affect the suspension density and the liquid level and cannot be considered alone. Furthermore, control objects for suspension density and liquid level are typical of large inertia, large hysteresis processes, and simple PID algorithms do not control such objects well. Therefore, a multivariable predictive control method is required to uniformly analyze and control to adapt to the actual characteristics of the dense medium sorting process. However, the conventional multivariable predictive control method applied to dense medium coal density control has the problems of large calculated amount, long calculated time, difficulty in realizing online real-time control and the like. Therefore, it becomes necessary to study a rapid dense media coal density control system.
Disclosure of Invention
Aiming at the problems of large online calculated amount and long calculation time of a common prediction control algorithm in the heavy medium coal preparation process in the prior art, the invention provides a QP-DMC-based rapid heavy medium coal preparation density control system, which is used for acquiring real-time suspension density data and real-time suspension liquid level data respectively through a densimeter and a liquid level meter arranged at a preset position in heavy medium coal preparation equipment at sampling moments of a preset time interval and correspondingly adjusting the opening of a diverter valve, the opening of a medium adding valve and the opening of a water supplementing valve according to the real-time suspension density data and the real-time suspension liquid level data; including a pass control module, a QP-DMC control model and a control actuator.
The control executor is a component in the control system responsible for executing control commands, and the control commands are received and converted into corresponding control signals, so that the opening degrees of the diverter valve, the water supplementing valve and the medium supplementing valve in the process of dense medium coal dressing are adjusted to control the density and the liquid level of the suspension to fluctuate within a preset range.
Specifically, the control actuator receives a control command from the branch controller and converts the control command into a corresponding control signal, and then the opening degrees of the branch valve, the water supplementing valve and the medium supplementing valve are adjusted by driving the corresponding actuating mechanism through the actuator, so that the suspension density and the liquid level are controlled.
In the dense medium coal separation process, the opening degrees of the flow dividing valve, the water supplementing valve and the medium supplementing valve are adjusted by controlling the actuator, so that the flow, the density and the liquid level of the suspension liquid can be controlled, and the stability and the efficiency of the dense medium coal separation process are ensured. Meanwhile, the quick response capability of the control actuator can also improve the response speed and the accuracy of the control system, so that the control effect of the dense medium coal dressing process is further improved.
The split control module divides different split strokes based on a preset suspension liquid density set value, and respectively adjusts the opening of the split valve, the opening of the medium adding valve and the opening of the water supplementing valve in real time by applying a QP-DMC control model according to the real-time suspension liquid density data and the real-time suspension liquid level data according to each split stroke, and sends out a control command.
Wherein, divide different branches based on the suspension density setting value and specifically do:
real-time suspension density data y 2 When the value is smaller than the set value, a first part is entered, and the deviation between the real-time suspension liquid level data y1 and the preset suspension liquid level value and the real-time suspension liquid density data y are used for 2 Deviation from a suspension fluid tightness setting, regulating a diverter valveOpening u 1 And the opening degree u of the medium adding valve 2 Size of the product.
(2) Real-time suspension density data y 2 When the value is larger than the set value, the second part is entered, and the real-time suspension liquid level data y is used for 1 Deviation from a preset value of the suspension level and real-time suspension density data y 2 Deviation from the suspension liquid tightness set point, the diverter valve u is adjusted 1 And the opening degree u of the water supplementing valve 3 Size of the product.
The QP-DMC control model is combined with the split control module, and the multivariable QP-DMC control model is obtained through a system identification experiment, and the method specifically comprises the following steps of:
selecting the opening u of the diverter valve 1 As a system identification input signal, the suspension liquid level y is recorded 1 Suspension density y 2 Is a change in conditions of (2); using the input and output data obtained by the system identification test to establish the suspension liquid level y 1 And a split valve opening u 1 Transfer function G of (2) 11 Suspension density y 2 And a split valve opening u 1 Transfer function G of (2) 21 The method comprises the steps of carrying out a first treatment on the surface of the Selecting the opening degree u of the medium adding valve 2 Respectively acting on the suspension liquid level y 1 And suspension density y 2 Repeating the system identification test to obtain a transfer function model of the control relation corresponding to the first branch, wherein the transfer function model is as follows:
selecting control factors in other dense medium sorting processes, and controlling the opening degree u of the water supplementing valve 3 Respectively acting on the suspension liquid level y 1 Object and suspension Density y 2 The object, repeating the system identification test, and obtaining a transfer function model of the control relation corresponding to the second branch is as follows:
according to the transfer function model respectively corresponding to the established first and second branches, designing a multi-variable branch QP-DMC controlA controller including a first and a second multi-variable branch QP-DMC predictive model for controlling the density y of the dense medium suspension 2 And suspension level y 1 Fluctuating within a predetermined range.
Discretizing the first branch transfer function model to obtain a multivariable first branch QP-DMC prediction model for predicting the opening u of the shunt valve 1 And the opening degree u of the medium adding valve 2 Under the regulating action of (a) the liquid level y of the dense medium suspension 1 And suspension density y 2 Is a trend of change in (c).
Discretizing the second branch transfer function model to obtain a multivariable second branch QP-DMC prediction model for predicting the opening u of the shunt valve 1 And the opening degree u of the water supplementing valve 3 Under the regulating action of (a) the liquid level y of the dense medium suspension 1 And suspension density y 2 Is a trend of change in (c).
The multi-variable QP-DMC control model is combined with a feedforward decoupling module, the feedforward decoupling module is used for reserving important factors in the branches and performing feedforward compensation on influence factors, decomposing a multi-variable prediction task into a plurality of single-variable prediction tasks, combining all single-variable prediction results to obtain an overall prediction result, combining the QP-DMC control model to obtain a control instruction, and sending the control instruction to a control executor to realize quick prediction and control.
In order to obtain a plurality of single variable part-pass QP-DMC fast controllers, the method specifically comprises the following steps:
for the first branch, the opening u of the flow dividing valve 1 As means for regulating the level y of the suspension 1 Is the main factor of the opening degree u of the medium adding valve 2 As means for regulating the level y of the suspension 1 Is to add the influence factor of the valve opening u 2 As means for regulating the density y of the suspension 2 Is the main factor of the opening degree u of the flow dividing valve 1 As means for regulating the density y of the suspension 2 Is used for obtaining the overall prediction model as
In which y 1 (k) For suspension level y at time k 1 Is a column vectorSuspension level y at time k+1, …, k+M, …, k+P 1 Is a predictive value sequence of (1) column vector->For the suspension level y before time k 1 P, M represent the prediction time domain and the control time domain, y, respectively, of the predictive control algorithm 2 (k) For the measurement of the suspension density at time k, column vector +.>Suspension density y at time k+1, …, k+M, …, k+P 2 Is a predictive value sequence, column vector of (a)Suspension density y at time k 2 Historical numerical sequence of (u) 1 (k) For the opening u of the flow divider at time k 1 Numerical value, deltau 1 (k)＝u 1 (k)-u 1 (k-1) the split valve opening degree u before the k time 1 Increment, column vector->The opening degree u of the flow dividing valve at the moment k, … and k+M 1 Control increment sequence of u) 2 (k) For k time of opening u of medium valve 2 Numerical value, deltau 2 (k)＝u 2 (k)-u 2 (k-1) is the opening degree u of the medium valve at time k 2 Increment, column vector->The opening degree u of the medium valve at the moment k, … and k+M 2 Is controlled by a control increment sequence of (2); definition matrix A 11 For the opening u of the shunt valve 1 Variable vs. suspension level y 1 A dynamic matrix of the control is provided,
wherein [ a ] 11 (1)，…，a 11 (P)]Is the suspension level y 1 Responsive to the opening u of the shunt valve 1 A varying step response sequence; definition matrix A 22 For opening u of medium valve 2 Variable vs. suspension density y 2 A dynamic matrix of the control is provided,
wherein [ a ] 22 (1)，…，a 22 (P)]Is the suspension density y 2 Responsive to opening u of the charge valve 2 A varying step response sequence; in predictive model type [ a ] 12 (1)，…，a 12 (P)]Is the suspension level y 1 Responsive to opening u of the charge valve 2 A sequence of varying step responses, [ a ] 21 (1)，…，a 21 (P)]Is the suspension density y 2 Responsive to the opening u of the shunt valve 1 A sequence of varying step responses.
For the second branch, the opening u of the branch valve is set 1 As means for regulating the level y of the suspension 1 Main factor of (2) the opening u of the water replenishing valve 3 As means for regulating the level y of the suspension 1 Is to control the opening degree u of the water supplementing valve 3 As means for regulating the density y of the suspension 2 Is the main factor of the opening degree u of the flow dividing valve 1 As means for regulating the density y of the suspension 2 The overall prediction model is obtained by the following influence factors:
u in the formula 3 (k) For k time the opening u of the water supplementing valve 3 Numerical value, deltau 3 (k)＝u 3 (k)-u 3 (k-1) is the opening degree u of the water supplementing valve at the moment k 3 Increment, column vectorThe opening degree u of the water supplementing valve at the moment k, … and k+M 3 Is controlled by a control increment sequence of (2); definition matrix A 23 For the opening u of the water supplementing valve 3 Variable vs. suspension density y 2 A dynamic matrix of the control is provided,
wherein [ a ] 23 (1)，…，a 23 (P)]Is the suspension density y 2 Responsive to opening u of the water-replenishing valve 3 A varying step response sequence; in predictive model type [ a ] 13 (1)，…，a 13 (P)]Is the suspension level y 1 Responsive to opening u of the water-replenishing valve 3 A sequence of varying step responses.
QP-DMC problem includes performance metrics as follows: multi-objective optimization problem: the QP-DMC problem is typically a multi-objective optimization problem that requires simultaneous optimization of multiple performance metrics, such as control speed, control accuracy, power consumption, etc.
Nonlinear problem: the non-linearity problem in QP-DMC problems is typically derived from the non-linearity of the controlled object or the control algorithm.
Robustness problem: the QP-DMC problem needs to be robust in the presence of uncertainty or disturbances, yet maintain good control performance.
Real-time problem: the QP-DMC problem needs to be solved in situations where real-time requirements are high, such as real-time control problems in industrial process control.
Large scale problem: QP-DMC problems typically involve large-scale state space and control input space, requiring the use of efficient optimization algorithms and computation methods.
The invention aims at performance indexes of a plurality of single-variable part-way QP-DMC (program-controlled quick control) controllers, in order to determine the opening degree u of a control variable part-way valve 1 Opening u of medium valve 2 Opening u of water supplementing valve 3 Thereby ensuring that the controlled variable suspension liquid level and suspension density fluctuate within a predetermined range, and respectively constructing a plurality of single variable part-way QP-DMC quick controller performance indexes for each part-way process.
For the first part-pass, a plurality of single variable part-pass QP-DMC quick controller performance indexes are constructed as follows:
wherein: matrix A' 11 Is a dynamic matrix A 11 Transpose of matrix A' 22 Is a dynamic matrix A 22 Is the transposed, column vector of (2)Is the suspension level y 1 Responsive to opening u of the charge valve 2 Variable step response sequence, column vectorIs the suspension density y 2 Responsive to the opening u of the shunt valve 1 Variable step response sequence, column vectorFor suspension level y at time k 1 Target value sequence, column vector of (2)When k isDensity of etching suspension y 2 Is a controlled variable error weighting matrix in the multivariable predictive control, and the matrix R is an operation variable increment control weighting matrix in the multivariable predictive control.
For the second pass, a plurality of univariate pass QP-DMC fast controller indexes are constructed as follows:
wherein: matrix A' 23 Is a dynamic matrix A 23 Is the transposed, column vector of (2)Is the suspension level y 1 Responsive to opening u of the water-replenishing valve 3 A sequence of varying step responses.
In one embodiment, the system identification test may be a step response test or an impulse response test.
In the system identification test, in the process of establishing the transfer function model, the acquired data needs to be processed, such as filtering, fitting and other operations, so as to obtain a more accurate model.
In the system identification test, other multivariable controllers such as DMC, GPC, MAC can be adopted in the process of designing the proper multivariable predictive controller so as to control the change of the resuspension liquid level and the density simultaneously.
The branch prediction model is trained and optimized based on a system identification method.
The rapid dense medium coal dressing density control system based on QP-DMC provided by the invention has the following beneficial effects:
the QP-DMC-based rapid dense medium coal density control system can bring the following beneficial effects:
1. the efficiency of dense medium coal separation equipment is improved: the opening degrees of the flow dividing valve, the medium adding valve and the water supplementing valve are adjusted in real time, so that the operation parameters of the equipment can be adjusted in time according to the suspension density data and the real-time suspension liquid level data, and the sorting efficiency and the stability of the equipment are improved.
2. By monitoring the suspension density and the liquid level data in real time, the operation parameters of the equipment can be timely adjusted, so that the quality and the stability of the product are improved. By introducing the control actuator, the automatic control of the equipment can be realized, the manual intervention is reduced, and the production efficiency and the safety are improved.
3. By the design of the feedforward decoupling module, uncertainty and interference in a multivariable prediction task can be reduced, so that reliability and stability of the system are improved.
Drawings
FIG. 1 shows the main steps of the present invention;
FIG. 2 is a flow chart of a fast re-media suspension density control system based on QP-DMC algorithm according to the present invention;
FIG. 3 is a first split Matlab simulation;
FIG. 4 is a second pass Matlab simulation;
FIG. 5 is a simulation of a suspension density control loop;
fig. 6 is a simulation of a suspension level control loop.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
1-2, in the dense medium coal dressing equipment, the opening of the diverter valve, the opening of the medium adding valve and the opening of the water supplementing valve are correspondingly adjusted according to the real-time suspension density data and the real-time suspension liquid level data at sampling moments of preset time intervals; the system is characterized by comprising a path control module, a QP-DMC control model and a control actuator.
The split control module divides different split strokes based on a preset suspension liquid density set value, and respectively adjusts the opening of the split valve, the opening of the medium adding valve and the opening of the water supplementing valve in real time by applying a QP-DMC control model according to the real-time suspension liquid density data and the real-time suspension liquid level data according to each split stroke, and sends out a control command.
The control executor receives the control command and converts the control command into corresponding control signals, and the opening of the diverter valve, the opening of the medium adding valve and the opening of the water supplementing valve are adjusted in real time.
Wherein, divide different branches based on the suspension density setting value and specifically do:
real-time suspension density data y 2 When the value is smaller than the set value, the first part is entered, and the real-time suspension liquid level data y is used for 1 Deviation from a preset value of the suspension level and real-time suspension density data y 2 Deviation from the suspension liquid tightness set value, the opening u of the shunt valve is regulated 1 And the opening degree u of the medium adding valve 2 Size of the product.
Real-time suspension density data y 2 When the value is larger than the set value, the second part is entered, and the real-time suspension liquid level data y is used for 1 Deviation from a preset value of the suspension level and real-time suspension density data y 2 Deviation from the suspension liquid tightness set point, the diverter valve u is adjusted 1 And the opening degree u of the water supplementing valve 3 Size of the product.
The QP-DMC control model is combined with the split control module, and the multivariable QP-DMC control model is obtained through a system identification experiment, and the method specifically comprises the following steps of:
selecting the opening u of the diverter valve 1 As a system identification input signal, the suspension liquid level y is recorded 1 Suspension density y 2 Is a change in conditions of (2); using the input and output data obtained by the system identification test to establish the suspension liquid level y 1 And a split valve opening u 1 Transfer function G of (2) 11 Suspension density y 2 And a split valve opening u 1 Transfer function G of (2) 21 The method comprises the steps of carrying out a first treatment on the surface of the Selecting opening of medium valveu 2 Respectively acting on the suspension liquid level y 1 And suspension density y 2 Repeating the system identification test to obtain a transfer function model of the control relation corresponding to the first branch, wherein the transfer function model is as follows:
selecting control factors in other dense medium sorting processes, and controlling the opening degree u of the water supplementing valve 3 Respectively acting on the suspension liquid level y 1 Object and suspension Density y 2 The object, repeating the system identification test, and obtaining a transfer function model of the control relation corresponding to the second branch is as follows:
according to the transfer function models respectively corresponding to the established first and second branches, a multi-variable branch QP-DMC controller is designed, and comprises a multi-variable first branch QP-DMC prediction model and a multi-variable second branch QP-DMC prediction model for controlling the density y of the re-medium suspension 2 And suspension level y 1 Fluctuating within a predetermined range.
Discretizing the first branch transfer function model to obtain a multivariable first branch QP-DMC prediction model for predicting the opening u of the shunt valve 1 And the opening degree u of the medium adding valve 2 Under the regulating action of (a) the liquid level y of the dense medium suspension 1 And suspension density y 2 Is a trend of change in (c).
Discretizing the second branch transfer function model to obtain a multivariable second branch QP-DMC prediction model for predicting the opening u of the shunt valve 1 And the opening degree u of the water supplementing valve 3 Under the regulating action of (a) the liquid level y of the dense medium suspension 1 And suspension density y 2 Is a trend of change in (c).
The multi-variable QP-DMC control model is combined with a feedforward decoupling module, the feedforward decoupling module is used for reserving important factors in the branches and performing feedforward compensation on influence factors, decomposing a multi-variable prediction task into a plurality of single-variable prediction tasks, combining all single-variable prediction results to obtain an overall prediction result, combining the QP-DMC control model to obtain a control instruction, and sending the control instruction to a control executor to realize quick prediction and control.
Obtaining a plurality of univariate part-pass QP-DMC fast controllers, comprising the following steps:
for the first branch, the opening u of the flow dividing valve 1 As means for regulating the level y of the suspension 1 Is the main factor of the opening degree u of the medium adding valve 2 As means for regulating the level y of the suspension 1 Is to add the influence factor of the valve opening u 2 As means for regulating the density y of the suspension 2 Is the main factor of the opening degree u of the flow dividing valve 1 As means for regulating the density y of the suspension 2 Is used for obtaining the overall prediction model as
In which y 1 (k) For suspension level y at time k 1 Is a column vectorSuspension level y at time k+1, …, k+M, …, k+P 1 Is a predictive value sequence of (1) column vector->For the suspension level y before time k 1 P, M represent the prediction time domain and the control time domain, y, respectively, of the predictive control algorithm 2 (k) Column vector for measurement of suspension density at time kSuspension density y at time k+1, …, k+M, …, k+P 2 Is a predictive value sequence of (1) column vector->Suspension density y at time k 2 Historical numerical sequence of (u) 1 (k) For the opening u of the flow divider at time k 1 Numerical value, deltau 1 (k)＝u 1 (k)-u 1 (k-1) the split valve opening degree u before the k time 1 Increment, column vector->The opening degree u of the flow dividing valve at the moment k, … and k+M 1 Control increment sequence of u) 2 (k) For k time of opening u of medium valve 2 Numerical value, deltau 2 (k)＝u 2 (k)-u 2 (k-1) is the opening degree u of the medium valve at time k 2 Increment, column vector->The opening degree u of the medium valve at the moment k, … and k+M 2 Is controlled by a control increment sequence of (2); definition matrix A 11 For the opening u of the shunt valve 1 Variable vs. suspension level y 1 A dynamic matrix of the control is provided,
wherein [ a ] 11 (1)，…，a 11 (P)]Is the suspension level y 1 Responsive to the opening u of the shunt valve 1 A varying step response sequence; definition matrix A 22 For opening u of medium valve 2 Variable vs. suspension density y 2 A dynamic matrix of the control is provided,
wherein [ a ] 22 (1)，…，a 22 (P)]Is the suspension density y 2 Responsive to opening u of the charge valve 2 A varying step response sequence; in predictive model type [ a ] 12 (1)，…，a 12 (P)]Is the suspension level y 1 Responsive to opening u of the charge valve 2 A sequence of varying step responses, [ a ] 21 (1)，…，a 21 (P)]Is the suspension density y 2 Responsive to the opening u of the shunt valve 1 A sequence of varying step responses.
For the second branch, the opening u of the branch valve is set 1 As means for regulating the level y of the suspension 1 Main factor of (2) the opening u of the water replenishing valve 3 As means for regulating the level y of the suspension 1 Is to control the opening degree u of the water supplementing valve 3 As means for regulating the density y of the suspension 2 Is the main factor of the opening degree u of the flow dividing valve 1 As means for regulating the density y of the suspension 2 The overall prediction model is obtained by the following influence factors:
u in the formula 3 (k) For k time the opening u of the water supplementing valve 3 Numerical value, deltau 3 (k)＝u 3 (k)-u 3 (k-1) is the opening degree u of the water supplementing valve at the moment k 3 Increment, column vectorThe opening degree u of the water supplementing valve at the moment k, … and k+M 3 Is controlled by a control increment sequence of (2); definition matrix A 23 For the opening u of the water supplementing valve 3 Variable vs. suspension density y 2 A dynamic matrix of the control is provided,
wherein [ a ] 23 (1)，…，a 23 (P)]Is the suspension density y 2 Responsive to opening u of the water-replenishing valve 3 Step of changeA response sequence; in predictive model type [ a ] 13 (1)，…，a 13 (P)]Is the suspension level y 1 Responsive to opening u of the water-replenishing valve 3 A sequence of varying step responses.
QP-DMC problem includes performance metrics as follows: multi-objective optimization problem: the QP-DMC problem is typically a multi-objective optimization problem that requires simultaneous optimization of multiple performance metrics, such as control speed, control accuracy, power consumption, etc.
Nonlinear problem: the non-linearity problem in QP-DMC problems is typically derived from the non-linearity of the controlled object or the control algorithm.
Robustness problem: the QP-DMC problem needs to be robust in the presence of uncertainty or disturbances, yet maintain good control performance.
Real-time problem: the QP-DMC problem needs to be solved in situations where real-time requirements are high, such as real-time control problems in industrial process control.
Large scale problem: QP-DMC problems typically involve large-scale state space and control input space, requiring the use of efficient optimization algorithms and computation methods.
The invention aims at performance indexes of a plurality of single-variable part-way QP-DMC (program-controlled quick control) controllers, in order to determine the opening degree u of a control variable part-way valve 1 Opening u of medium valve 2 Opening u of water supplementing valve 3 Thereby ensuring that the controlled variable suspension liquid level and suspension density fluctuate within a predetermined range, and respectively constructing a plurality of single variable part-way QP-DMC quick controller performance indexes for each part-way process.
For the first part-pass, a plurality of single variable part-pass QP-DMC quick controller performance indexes are constructed as follows:
wherein: matrix A' 11 Is a dynamic matrix A 11 Transpose of matrix A' 22 Is a dynamic matrix A 22 Is the transposed, column vector of (2)Is the suspension level y 1 Responsive to opening u of the charge valve 2 Variable step response sequence, column vectorIs the suspension density y 2 Responsive to the opening u of the shunt valve 1 Variable step response sequence, column vectorFor suspension level y at time k 1 Target value sequence, column vector of (2)Suspension density y at time k 2 Is a controlled variable error weighting matrix in the multivariable predictive control, and the matrix R is an operation variable increment control weighting matrix in the multivariable predictive control.
For the second pass, a plurality of univariate pass QP-DMC fast controller indexes are constructed as follows:
wherein: matrix A' 23 Is a dynamic matrix A 23 Is the transposed, column vector of (2)Is the suspension level y 1 Responsive to opening u of the water-replenishing valve 3 Step of changeA response sequence.
In the system identification test, the system identification test can be a step response test or an impulse response test.
In the system identification test, in the process of establishing the transfer function model, the acquired data needs to be processed, such as filtering, fitting and other operations, so as to obtain a more accurate model.
In the system identification test, other multivariable controllers such as DMC, GPC, MAC can be adopted in the process of designing the proper multivariable predictive controller so as to control the change of the resuspension liquid level and the density simultaneously.
The branch prediction model is trained and optimized based on a system identification method.
In one embodiment, through a system identification test, the opening u of the diverter valve in the process of dense medium coal dressing is respectively used 1 Opening u of medium valve 2 Opening u of water supplementing valve 3 For step quantity, acquiring real-time liquid level data y of suspension 1 And real-time density data y of suspension 2 The step response values of (2) to obtain the transfer function of the dense medium sorting process are as follows:
(1) When the density is less than the set value, the first part-pass is entered. In this case, assuming that the initial density of the heavy medium is 1.2kg/L, the required set density is 1.5kg/L, the initial liquid level of the mixing drum is 1.6m, and the Matlab simulation result when the liquid level of the mixing drum is stabilized at 2.5m is required, as shown in FIG. 3.
(2) And when the density is greater than the set value, entering a second branch. In this case, assuming that the initial density of the heavy medium is 1.8kg/L, the required set density is 1.5kg/L, the initial liquid level of the mixing drum is 1.6m, and the Matlab simulation result when the liquid level of the mixing drum is required to be stabilized at 2.5m is shown in FIG. 4.
To illustrate the control effect of the rapid dense media coal density control, compared with the conventional PID control algorithm, FIG. 5 is Matlab simulation of the conventional PID algorithm in the density control loop and the liquid level control loop.
It can be seen from fig. 5 and 6 that the system has a large overshoot of density and liquid level under the conventional PID control, the adjustment time is about 300s, and the adjustment time is up to 200s under the control of the present invention. Therefore, the control effect of the control method of the invention on the density and liquid level object of the heavy medium suspension is better than the simple PID control effect.
According to the method, rolling optimization in common multivariable predictive control is split into 2 times of dispersion optimization, so that the computational complexity is changed from O (2, M) to 2*O (1, M). In the method, the one-time calculation time of Matlab simulation is 0.253 seconds, and the one-time calculation time of a common QP-DMC algorithm is 1.147 seconds. The method can simplify the calculation process, shorten the calculation time, meet the real-time requirement of a control system and realize the purpose of quick control.
Claims (6)
1. The quick dense medium coal dressing density control system based on QP-DMC is used for correspondingly adjusting the opening of a diverter valve, the opening of a medium adding valve and the opening of a water supplementing valve according to the real-time suspension density data and the real-time suspension liquid level data at the sampling time of a preset time interval in dense medium coal dressing equipment; the system is characterized by comprising a path control module, a QP-DMC control model and a control executor;
the split control module divides different split strokes based on a preset suspension liquid density set value, and respectively adjusts the opening of the split valve, the opening of the medium adding valve and the opening of the water supplementing valve in real time by applying a QP-DMC control model according to the real-time suspension liquid density data and the real-time suspension liquid level data aiming at each split stroke, and sends out a control command;
the control executor receives the control command and converts the control command into corresponding control signals, and the opening of the diverter valve, the opening of the medium adding valve and the opening of the water supplementing valve are adjusted in real time.
2. The rapid dense coal dense control system of claim 1, further comprising a feedforward decoupling module for retaining important factors in the partial process and performing feedforward compensation on influencing factors for each partial process divided by the partial process control module, decomposing the multivariable predictive task into a plurality of univariate predictive tasks, combining all univariate predictive results to obtain an overall predictive result, combining a QP-DMC control model to obtain a control instruction, and sending the control instruction to a control executor to realize rapid prediction and control.
3. The rapid dense media separation (rapid dense media separation) control system based on QP-DMC according to claim 1, wherein the different divisions based on the preset suspension density set value are specifically:
(1) Real-time suspension density data y 2 When the value is smaller than the set value, the first part is entered, and the real-time suspension liquid level data y is used for 1 Deviation from a preset value of the suspension level and real-time suspension density data y 2 Deviation from the suspension liquid tightness set value, the opening u of the shunt valve is regulated 1 And the opening degree u of the medium adding valve 2 Size of the material;
(2) Real-time suspension density data y 2 When the value is larger than the set value, the second part is entered, and the real-time suspension liquid level data y is used for 1 Deviation from a preset value of the suspension level and real-time suspension density data y 2 Deviation from the suspension liquid tightness set point, the diverter valve u is adjusted 1 And the opening degree u of the water supplementing valve 3 Size of the product.
4. The rapid dense media separation coal density control system based on QP-DMC according to claim 3, wherein the QP-DMC control model is combined with a separate control model, and a multivariable QP-DMC control model is obtained through a system identification experiment, and the method specifically comprises the following steps:
selecting the opening u of the diverter valve 1 As a system identification input signal, the suspension liquid level y is recorded 1 Suspension density y 2 Is a change in conditions of (2); using the input and output data obtained by the system identification test to establish the suspension liquid level y 1 And a split valve opening u 1 Transfer function G of (2) 11 Suspension liquidDensity y 2 And a split valve opening u 1 Transfer function G of (2) 21 The method comprises the steps of carrying out a first treatment on the surface of the Selecting the opening degree u of the medium adding valve 2 Respectively acting on the suspension liquid level y 1 And suspension density y 2 Repeating the system identification test to obtain a transfer function model of the control relation corresponding to the first branch, wherein the transfer function model is as follows:
selecting control factors in other dense medium sorting processes, and controlling the opening degree u of the water supplementing valve 3 Respectively acting on the suspension liquid level y 1 Object and suspension Density y 2 The object, repeating the system identification test, and obtaining a transfer function model of the control relation corresponding to the second branch is as follows:
according to the transfer function models respectively corresponding to the established first and second branches, a multi-variable branch QP-DMC controller is designed, and comprises a multi-variable first branch QP-DMC prediction model and a multi-variable second branch QP-DMC prediction model for controlling the density y of the re-medium suspension 2 And suspension level y 1 Fluctuation within a predetermined range;
discretizing the first branch transfer function model to obtain a multivariable first branch QP-DMC prediction model for predicting the opening u of the shunt valve 1 And the opening degree u of the medium adding valve 2 Under the regulating action of (a) the liquid level y of the dense medium suspension 1 And suspension density y 2 Is a trend of change in (2);
discretizing the second branch transfer function model to obtain a multivariable second branch QP-DMC prediction model for predicting the opening u of the shunt valve 1 And the opening degree u of the water supplementing valve 3 Under the regulating action of (a) the liquid level y of the dense medium suspension 1 And suspension density y 2 Is a trend of change in (c).
5. The rapid dense media separation system of claim 4, wherein the multi-variable QP-DMC control model is combined with a feedforward decoupling module to obtain a plurality of single-variable, split QP-DMC rapid controllers, comprising the steps of:
for the first branch, the opening u of the flow dividing valve 1 As means for regulating the level y of the suspension 1 Is the main factor of the opening degree u of the medium adding valve 2 As means for regulating the level y of the suspension 1 Is to add the influence factor of the valve opening u 2 As means for regulating the density y of the suspension 2 Is the main factor of the opening degree u of the flow dividing valve 1 As means for regulating the density y of the suspension 2 Is used for obtaining the overall prediction model as
In which y 1 (k) For suspension level y at time k 1 Is a column vectorSuspension level y at time k+1, …, k+M, …, k+P 1 Is a predictive value sequence of (1) column vector->For the suspension level y before time k 1 P, M represent the prediction time domain and the control time domain, y, respectively, of the predictive control algorithm 2 (k) For the measurement of the suspension density at time k, column vector +.>Suspension liquid density at k+1, …, k+M, …, k+Py 2 Is a predictive value sequence, column vector of (a)Suspension density y at time k 2 Historical numerical sequence of (u) 1 (k) For the opening u of the flow divider at time k 1 Numerical value, deltau 1 (k)＝u 1 (k)-u 1 (k-1) the split valve opening degree u before the k time 1 Increment, column vector->The opening degree u of the flow dividing valve at the moment k, … and k+M 1 Control increment sequence of u) 2 (k) For k time of opening u of medium valve 2 Numerical value, deltau 2 (k)＝u 2 (k)-u 2 (k-1) is the opening degree u of the medium valve at time k 2 Increment, column vector->The opening degree u of the medium valve at the moment k, … and k+M 2 Is controlled by a control increment sequence of (2); definition matrix A 11 For the opening u of the shunt valve 1 Variable vs. suspension level y 1 A dynamic matrix of the control is provided,
wherein [ a ] 11 (1)，…，a 11 (P)]Is the suspension level y 1 Responsive to the opening u of the shunt valve 1 A varying step response sequence; definition matrix A 22 For opening u of medium valve 2 Variable vs. suspension density y 2 A dynamic matrix of the control is provided,
wherein [ a ] 22 (1)，…，a 22 (P)]Is the suspension density y 2 Responsive to opening u of the charge valve 2 Varied byA step response sequence; in predictive model type [ a ] 12 (1)，…，a 12 (P)]Is the suspension level y 1 Responsive to opening u of the charge valve 2 A sequence of varying step responses, [ a ] 21 (1)，…，a 21 (P)]Is the suspension density y 2 Responsive to the opening u of the shunt valve 1 A varying step response sequence;
for the second branch, the opening u of the branch valve is set 1 As means for regulating the level y of the suspension 1 Main factor of (2) the opening u of the water replenishing valve 3 As means for regulating the level y of the suspension 1 Is to control the opening degree u of the water supplementing valve 3 As means for regulating the density y of the suspension 2 Is the main factor of the opening degree u of the flow dividing valve 1 As means for regulating the density y of the suspension 2 The overall prediction model is obtained by the following influence factors:
u in the formula 3 (k) For k time the opening u of the water supplementing valve 3 Numerical value, deltau 3 (k)＝u 3 (k)-u 3 (k-1) is the opening degree u of the water supplementing valve at the moment k 3 Increment, column vectorThe opening degree u of the water supplementing valve at the moment k, … and k+M 3 Is controlled by a control increment sequence of (2); definition matrix A 23 For the opening u of the water supplementing valve 3 Variable vs. suspension density y 2 A dynamic matrix of the control is provided,
wherein [ a ] 23 (1)，…，a 23 (P)]Is the suspension density y 2 Response toOpening u of water supplementing valve 3 A varying step response sequence; in predictive model type [ a ] 13 (1)，…，a 13 (P)]Is the suspension level y 1 Responsive to opening u of the water-replenishing valve 3 A sequence of varying step responses.
6. The QP-DMC based rapid dense media separation control system according to claim 5, wherein the QP-DMC based problem includes a plurality of single variable step QP-DMC rapid controller performance metrics for determining the control variable diverter valve opening u 1 Opening u of medium valve 2 And the opening degree u of the water supplementing valve 3 Ensuring that the suspension liquid level and suspension density of the controlled variable fluctuate within a preset range, and respectively constructing a plurality of single variable part-way QP-DMC (quick controller) performance indexes for each part-way process;
for the first part-pass, a plurality of single variable part-pass QP-DMC quick controller performance indexes are constructed as follows:
wherein: matrix A' 11 Is a dynamic matrix A 11 Transpose of matrix A' 22 Is a dynamic matrix A 22 Is the transposed, column vector of (2)Is the suspension level y 1 Responsive to opening u of the charge valve 2 Variable step response sequence, column vectorIs the suspension density y 2 Responsive to the opening u of the shunt valve 1 Variable step response sequence, column vectorFor suspension level y at time k 1 Target value sequence, column vector of (2)Suspension density y at time k 2 A matrix g is a controlled variable error weighting matrix in the multivariable predictive control, and a matrix R is an operation variable increment control weighting matrix in the multivariable predictive control;
for the second pass, a plurality of univariate pass QP-DMC fast controller indexes are constructed as follows:
wherein: matrix A' 23 Is a dynamic matrix A 23 Is the transposed, column vector of (2)Is the suspension level y 1 Responsive to opening u of the water-replenishing valve 3 A sequence of varying step responses.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311297510.5A CN117193014A (en) | 2023-10-09 | 2023-10-09 | QP-DMC-based rapid dense medium coal density control system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311297510.5A CN117193014A (en) | 2023-10-09 | 2023-10-09 | QP-DMC-based rapid dense medium coal density control system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117193014A true CN117193014A (en) | 2023-12-08 |
Family
ID=88997997
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311297510.5A Pending CN117193014A (en) | 2023-10-09 | 2023-10-09 | QP-DMC-based rapid dense medium coal density control system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117193014A (en) |
-
2023
- 2023-10-09 CN CN202311297510.5A patent/CN117193014A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US6826521B1 (en) | System and methodology and adaptive, linear model predictive control based on rigorous, nonlinear process model | |
| US7877154B2 (en) | Method and system for controlling a batch process | |
| CN1940780B (en) | On-line adaptive model predictive control in a process control system | |
| US20060111858A1 (en) | Computer method and apparatus for online process identification | |
| US8756039B2 (en) | Rapid process model identification and generation | |
| Zhu et al. | Toward a low cost and high performance MPC: The role of system identification | |
| US20090132064A1 (en) | Apparatuses, Systems, and Methods Utilizing Adaptive Control | |
| CN110764417B (en) | Linear quadratic optimal dynamic feedforward-feedback PID control system based on closed-loop identification model and control method thereof | |
| CN105589448A (en) | Model predictive controller with tunable integral component to compensate for model mismatch | |
| CA2620315A1 (en) | Adaptive multivariable mpc controller | |
| CN103293953A (en) | Robust adaptive model predictive controller with tuning to compensate for model mismatch | |
| US9448546B2 (en) | Deterministic optimization based control system and method for linear and non-linear systems | |
| Efheij et al. | Comparison of model predictive control and PID controller in real time process control system | |
| CN117193014A (en) | QP-DMC-based rapid dense medium coal density control system | |
| Chen et al. | Neural network-based predictive control for multivariable processes | |
| CN112130542B (en) | Control loop performance evaluation method based on normal operation data and system identification | |
| CN112379601A (en) | MFA control system design method based on industrial process | |
| Pawlowski et al. | Event-based generalized predictive control | |
| Tohidi et al. | Self-tuning adaptive multiple model predictive control with application to ph control process | |
| CN111695300A (en) | Steam turbine valve flow fitting method and system | |
| Yu et al. | Predictive control based on neural networks of the chemical process | |
| Jumaev et al. | Microcontroller device for increasing the accuracy of PID regulators | |
| Keeler | Vision of neural networks and fuzzy logic for prediction and optimization of manufacturing processes | |
| Vu et al. | The use of the pilot plant facility in teaching process control engineering at Murdoch University | |
| Joseph | Single-Loop Reactor Control Strategies |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |