CN-121523071-B - Mineral processing equipment process parameter self-adaptive control method, system and storage medium

CN121523071BCN 121523071 BCN121523071 BCN 121523071BCN-121523071-B

Abstract

The application relates to the technical field of mineral processing equipment control, and discloses a mineral processing equipment process parameter self-adaptive control method, a mineral processing equipment process parameter self-adaptive control system and a storage medium. The method comprises the steps of improving state estimation reliability through redundant measurement data fusion, enabling a model to track process characteristic change by adopting online parameter identification, realizing prospective prediction control by utilizing rolling optimization, suppressing process fluctuation caused by multivariable coupling through hierarchical coordination execution, and enabling controller parameters to be optimized continuously based on gradient adjustment driven by performance indexes. The application solves the problems of poor reliability of measured data, incapability of adapting to ore property change due to fixed model parameters, control response lag and insufficient multivariable coordinated control capability in the prior art.

Inventors

XU JUNQIANG
WU YUNQIAN
LIANG YI
WANG HONGTAO
LI ZHENLONG
PAN FENG
WANG MINGYANG
ZHANG YIHENG
YAO QISHENG
ZHAO CHUNXIA

Assignees

河南中矿能源有限公司

Dates

Publication Date: 20260508
Application Date: 20260119

Claims (9)

1. The self-adaptive control method for the technological parameters of the mineral processing equipment is characterized by comprising the following steps of: s1, acquiring a first rheological parameter at an inlet of a flotation tank, a second rheological parameter in the middle and a bubble characteristic parameter, a third rheological parameter at an outlet and an electrochemical parameter, and performing redundant fusion treatment on the first rheological parameter, the second rheological parameter and the third rheological parameter by a weighted least square method to obtain a fusion state estimation value; s2, constructing a state transition matrix and an input matrix according to the fusion state estimation value, and carrying out online identification on parameter vectors of the state transition matrix and the input matrix through a recursive least square algorithm to obtain updated parameter vectors; S3, substituting the updated parameter vector into a rolling optimization objective function, and performing quadratic programming solution on the control increment sequence to obtain optimal control output; s4, decomposing the optimal control output into an inflation quantity adjusting sequence, a stirring intensity adjusting sequence and a medicament quantity adjusting sequence, and performing hierarchical execution on the inflation quantity adjusting sequence, the stirring intensity adjusting sequence and the medicament quantity adjusting sequence according to a time sequence priority to obtain state response data; S5, calculating tracking deviation indexes and control change indexes according to the state response data, carrying out gradient adjustment on the output weights and the control weights in the rolling optimization objective function to obtain adjusted weight parameters, wherein the step comprises the steps of extracting output sequences from the state response data, carrying out time sequence alignment on the output sequences and a reference track, calculating point-by-point difference values, carrying out square summation on the difference values, averaging to obtain tracking deviation indexes, extracting execution tracks of all control channels from the state response data, carrying out square summation on the control quantity difference values at adjacent moments of all the control channels, averaging to obtain control change indexes, carrying out ratio calculation on the tracking deviation indexes and the target tracking deviation, carrying out ratio calculation on the control change indexes and the target control change, carrying out weighted square summation on the two ratios to obtain comprehensive performance indexes, judging the performance state according to the comprehensive performance indexes, calculating the gradient of the output weights and the control weights in the rolling optimization objective function, updating the output weights and the control weights according to the gradient direction when the comprehensive performance indexes are lower than a threshold value, and obtaining numerical parameters after adjustment.
2. The process parameter adaptive control method of beneficiation equipment according to claim 1, wherein the step S1 comprises: acquiring a first rheological parameter at an inlet of a flotation tank, a second rheological parameter in the middle of the flotation tank and a third rheological parameter at an outlet of the flotation tank, constructing a constraint equation set based on a material conservation relation, substituting the first rheological parameter, the second rheological parameter and the third rheological parameter into the constraint equation set, and carrying out weighted least square solution to obtain a primary fusion value; calculating a measurement residual vector of the constraint equation set, and carrying out standardization processing on the measurement residual vector to obtain a standardization residual of each measuring point; Judging an abnormal measuring point according to the standardized residual error, reducing a weight coefficient corresponding to the abnormal measuring point, and then carrying out weighted least square solution again to obtain a correction fusion value; And combining the corrected fusion value with the bubble characteristic parameter and the electrochemical parameter to obtain a fusion state estimated value.
3. The process parameter adaptive control method of beneficiation equipment according to claim 1, wherein the step S2 comprises: constructing a state space expression according to the fusion state estimation value, and expanding a state transition matrix and an input matrix in the state space expression into parameter vectors; collecting a control input sequence and an actual output sequence, constructing a regression vector and a covariance matrix, substituting the regression vector and the covariance matrix into a recursive least square formula to perform parameter identification, and obtaining an identification parameter vector; Calculating the parameter change rate of the identification parameter vector and the parameter vector at the previous moment, and judging the parameter convergence state according to the parameter change rate to obtain a convergence judgment result; and when the convergence judging result meets the convergence condition, outputting the identification parameter vector as an updating parameter vector, and updating the covariance matrix.
4. The process parameter adaptive control method of beneficiation equipment according to claim 1, wherein the step S3 comprises: Decomposing the updated parameter vector into a state transition matrix and an input matrix according to dimensions, and calculating an output predicted value sequence at each moment in a prediction time domain according to the state transition matrix and the input matrix; calculating point-by-point difference values of the output predicted value sequence and the reference track at the corresponding moment, and carrying out weighted square summation on the difference values at all the moments to obtain a tracking error item; The control increment of each moment in the control time domain is weighted, squared and summed to obtain a control cost item, and the tracking error item and the control cost item are added to form a rolling optimization objective function; Setting an upper limit of the aeration quantity change rate, an upper limit of the stirring intensity change rate and an upper limit of the medicament quantity change rate as constraint conditions, carrying out quadratic programming solution on the rolling optimization objective function to obtain a control increment sequence, and adding the first element of the control increment sequence with the control quantity at the previous moment to obtain the optimal control output.
5. The process parameter adaptive control method of beneficiation equipment according to claim 1, wherein the step S4 comprises: Decomposing the optimal control output into an inflation quantity control instruction, a stirring intensity control instruction and a medicament quantity control instruction according to a control channel, respectively calculating the difference value between each control instruction and the control quantity at the previous moment to obtain an inflation quantity adjustment sequence, a stirring intensity adjustment sequence and a medicament quantity adjustment sequence; After the execution of the inflation quantity adjusting sequence is finished, converting the stirring intensity adjusting sequence into a slope tracking instruction, gradually adjusting the stirring intensity according to a set slope, synchronously monitoring a state coupling index, and attenuating the adjusting amplitude when the state coupling index exceeds an upper limit; and when the stirring intensity adjusting sequence is executed to a preset progress, starting filling operation of the medicament quantity adjusting sequence, and recording the execution track of each control channel and the fusion state estimated value at the corresponding moment to obtain state response data.
6. The method for adaptively controlling technological parameters of mineral processing equipment according to claim 1, wherein the step of determining a performance state according to the comprehensive performance index, when the comprehensive performance index is lower than a threshold value, calculating a gradient of an output weight and a control weight in the rolling optimization objective function on the comprehensive performance index, and performing numerical update on the output weight and the control weight according to a gradient direction to obtain an adjusted weight parameter comprises: Comparing the comprehensive performance index with a performance threshold, judging that the performance is not up to standard when the comprehensive performance index is smaller than the performance threshold, and extracting the relative deviation of the tracking deviation index and the control change index; determining an output weight adjustment direction and a control weight adjustment direction according to the deviation relation between the tracking deviation index and the target tracking deviation and the deviation relation between the control change index and the target control change; Combining the output weight and the control weight in the rolling optimization objective function into a weight parameter vector, and calculating partial derivatives of the comprehensive performance indexes about the weight parameter vector to obtain a gradient vector; multiplying the gradient vector with the learning rate to obtain a weight adjustment amount, and superposing the weight adjustment amount to the current output weight and the current control weight to obtain an adjusted weight parameter.
7. A beneficiation equipment process parameter adaptive control system for implementing a beneficiation equipment process parameter adaptive control method according to any one of claims 1 to 6, the beneficiation equipment process parameter adaptive control system comprising: The fusion module is used for acquiring a first rheological parameter at the inlet of the flotation tank, a second rheological parameter in the middle and a bubble characteristic parameter, a third rheological parameter at the outlet and an electrochemical parameter, and performing redundant fusion treatment on the first rheological parameter, the second rheological parameter and the third rheological parameter by a weighted least square method to obtain a fusion state estimation value; The identification module is used for constructing a state transition matrix and an input matrix according to the fusion state estimation value, and carrying out online identification on the parameter vectors of the state transition matrix and the input matrix through a recursive least square algorithm to obtain updated parameter vectors; the solving module is used for substituting the updated parameter vector into a rolling optimization objective function, and carrying out quadratic programming solving on the control increment sequence to obtain optimal control output; the grading module is used for decomposing the optimal control output into an aeration quantity adjusting sequence, a stirring intensity adjusting sequence and a medicament quantity adjusting sequence, and grading the aeration quantity adjusting sequence, the stirring intensity adjusting sequence and the medicament quantity adjusting sequence according to the time sequence priority to obtain state response data; The adjustment module is used for calculating tracking deviation indexes and control change indexes according to the state response data, carrying out gradient adjustment on the output weights and the control weights in the rolling optimization objective function to obtain adjusted weight parameters, and comprises the steps of extracting output sequences from the state response data, carrying out time sequence alignment on the output sequences and a reference track, calculating point-by-point difference values, carrying out square summation on the difference values, averaging to obtain tracking deviation indexes, extracting execution tracks of all control channels from the state response data, carrying out square summation on the control quantity difference values at adjacent moments of all the control channels, averaging to obtain control change indexes, carrying out ratio calculation on the tracking deviation indexes and the target tracking deviation, carrying out weighted square summation on the control change indexes and the target control change, obtaining comprehensive performance indexes, judging the performance state according to the comprehensive performance indexes, calculating the gradient of the output weights and the control weights in the rolling optimization objective function, updating the output weights and the control weights according to the gradient direction when the comprehensive performance indexes are lower than a threshold value, and obtaining numerical parameters after adjustment.
8. A beneficiation equipment process parameter self-adaptive control device, characterized by comprising a memory and a processor, wherein the memory stores a computer program capable of running on the processor, and the processor realizes the beneficiation equipment process parameter self-adaptive control method according to any one of claims 1 to 6 when executing the computer program.
9. A computer readable storage medium having stored thereon a computer program, which when run by a processor causes the processor to perform a method of adaptive control of process parameters of a beneficiation plant as claimed in any one of claims 1 to 6.

Description

Mineral processing equipment process parameter self-adaptive control method, system and storage medium Technical Field The application relates to the technical field of mineral processing equipment control, in particular to a mineral processing equipment process parameter self-adaptive control method, a mineral processing equipment process parameter self-adaptive control system and a storage medium. Background The technological parameter control of the mineral processing equipment is a key technology for guaranteeing stable operation of the mineral processing process and improving technical and economic indexes, the traditional mineral processing technological parameter control mainly depends on indexes such as foam layer state, concentrate grade and the like observed by operators according to experience, and control parameters such as aeration quantity, stirring intensity, medicament addition quantity and the like are manually adjusted, so that obvious hysteresis and subjectivity exist in the manual control mode. In order to improve the control effect, an automatic control system is introduced in the prior art, a single-point measuring sensor is adopted to collect the state parameters of the flotation process, the process parameters are automatically adjusted according to the concentrate grade deviation through a PID feedback control algorithm, a part of advanced systems apply a machine learning method such as an LSTM-DNN neural network to establish a prediction model between the process parameters and technical economic indexes, a cloud computing platform is utilized to process massive historical data to carry out model training, parameter optimization regulation and control based on the prediction result is realized, and the prior art reduces the dependence on manual operation to a certain extent and improves the control precision. However, the prior art still has the remarkable defects that firstly, state information acquired in a single-point measurement mode is easily affected by local disturbance and measurement noise, when a sensor drifts or fails, an effective data correction mechanism is lacked, so that a control system makes an erroneous decision based on unreliable data, secondly, a control model with fixed parameters cannot adapt to dynamic changes of ore properties, when the hardness, the oxidation rate or the mineral composition of the selected ore fluctuates, parameters of a pre-trained neural network model remain unchanged, the model prediction precision is rapidly reduced, the control effect is seriously deteriorated, and secondly, the response speed of a hysteresis feedback control mode based on result indexes such as concentrate grade is slow, the feedback of a grade detection result is usually required to be adjusted from the process parameters to be 15 to 20 minutes, and the control delay causes the flotation process to be in a non-optimal state in the time of the abrupt change of the ore properties. Disclosure of Invention The application provides a self-adaptive control method, a system and a storage medium for technological parameters of mineral processing equipment, which are used for solving the problems that the reliability of measured data is poor, the parameters of a model are fixed and cannot be suitable for the property change of ore, the control response is delayed and the control capacity of the multivariable coordination is insufficient in the prior art by constructing a state estimation system based on redundant measured data fusion, a prediction control model for online parameter identification, a hierarchical coordinated multivariable execution strategy and a performance index-driven controller parameter self-adaptive adjustment mechanism, and improving the accuracy, the instantaneity and the robustness of the technological parameter control of the mineral processing equipment. In a first aspect, the present application provides a method for adaptively controlling process parameters of a mineral processing apparatus, where the method for adaptively controlling process parameters of the mineral processing apparatus includes: s1, acquiring a first rheological parameter at an inlet of a flotation tank, a second rheological parameter in the middle and a bubble characteristic parameter, a third rheological parameter at an outlet and an electrochemical parameter, and performing redundant fusion treatment on the first rheological parameter, the second rheological parameter and the third rheological parameter by a weighted least square method to obtain a fusion state estimation value; s2, constructing a state transition matrix and an input matrix according to the fusion state estimation value, and carrying out online identification on parameter vectors of the state transition matrix and the input matrix through a recursive least square algorithm to obtain updated parameter vectors; S3, substituting the updated parameter vector into a rolling optimization objective function, and