CN-121983181-A - Mine water treatment agent proportioning intelligent recommendation method based on multi-objective optimization

CN121983181ACN 121983181 ACN121983181 ACN 121983181ACN-121983181-A

Abstract

The invention relates to the technical field of mine water treatment, in particular to an intelligent recommendation method for proportioning of mine water treatment agents based on multi-objective optimization, which comprises the steps of collecting multidimensional dynamic data of mine water treatment and preprocessing to obtain a standardized engineering reference data sequence; the method comprises the steps of constructing a heavy metal and suspended matter competitive adsorption index and a flocculating constituent hydraulic shearing weakness index according to a standardized engineering reference data sequence, fusing the standardized engineering reference data sequence with the heavy metal and suspended matter competitive adsorption index and the flocculating constituent hydraulic shearing weakness index, predicting the water turbidity at the future moment by using a classification enhancement gradient lifting tree prediction algorithm, constructing an adaptability objective function according to the water turbidity at the future moment by combining engineering punishment constraint, and optimizing by using a multi-objective particle swarm optimization algorithm. Through the technical scheme of the invention, the firmness of the microstructure of the flocculating constituent can be ensured under severe working conditions, and the invalid consumption of the medicament is avoided.

Inventors

SUN YANG
LI PENG
YANG ZHENXUE
SUN MINGYUAN
HUO XIAOXIA
WANG SHENGJIN
LV ZIWEI
FENG XUZHE
LIU SHAOJIE

Assignees

山东黄金集团有限公司
山东省地矿工程集团检测技术有限公司
山东黄金集团蓬莱矿业有限公司

Dates

Publication Date: 20260505
Application Date: 20260408

Claims (10)

1. The intelligent recommendation method for the proportioning of the mine water treatment medicament based on multi-objective optimization is characterized by comprising the steps of collecting mine water treatment multidimensional dynamic data and preprocessing to obtain a standardized engineering reference data sequence; Constructing a heavy metal and suspended matter competitive adsorption index and a flocculating constituent hydraulic shear weakness index according to a standardized engineering reference data sequence; fusing a standardized engineering reference data sequence with the heavy metal and suspended matter competitive adsorption index and the flocculating constituent hydraulic shearing weakness index, and predicting the effluent turbidity at the future moment by using a classification enhancement gradient lifting tree prediction algorithm; according to the turbidity of the water outlet at the future moment, constructing an adaptability objective function by combining engineering punishment constraint, optimizing by adopting a multi-objective particle swarm optimization algorithm, and outputting the optimal medicament addition amount combination as a recommended proportioning instruction so as to finish intelligent recommendation of the medicament proportioning of the mine water treatment.
2. The intelligent recommendation method for the proportioning of the mine water treatment medicament based on the multi-objective optimization of claim 1 is characterized in that the mine water treatment multidimensional dynamic data are collected and preprocessed to obtain a standardized engineering reference data sequence, wherein the standardized engineering reference data sequence comprises the steps of obtaining the instantaneous flow of water, the pH value of the water, the concentration of suspended matters in the water, the total equivalent concentration of heavy metal ions in the water, the instantaneous addition of polyaluminium chloride and the instantaneous addition of polyacrylamide in real time to form high-frequency continuous time sequence data, and periodically obtaining the total equivalent concentration of heavy metal ions in the water to form low-frequency discrete time sequence data; And smoothing the acquired high-frequency continuous time sequence data by adopting a fixed-length sliding window mean value filtering algorithm, and performing time axis interpolation filling on the acquired low-frequency discrete time sequence data by adopting a zero-order retainer to enable the low-frequency discrete time sequence data to be asynchronously aligned with the sampling rate of the smoothed high-frequency continuous time sequence data, so that a standardized engineering reference data sequence is obtained.
3. The intelligent recommendation method for the proportioning of the mine water treatment medicament based on the multi-objective optimization according to claim 2, wherein the competition adsorption index of the heavy metal and suspended matters satisfies the relation: ; In the formula, The competitive adsorption index of heavy metals and suspended substances, In order to obtain the total equivalent concentration of heavy metal ions in the water, As an exponential function with natural constants as a base, For the value of the pH value of the inlet water, Is an optimal engineering pH value constant for characteristic hydroxide precipitation of heavy metal ions, For the concentration of the suspended matter in the water, The absolute value sign is used for calculating the difference value between the real-time PH value and the optimal engineering PH value constant.
4. The intelligent recommendation method for the proportioning of the mine water treatment medicament based on multi-objective optimization according to claim 3, wherein the index of the hydraulic shear weakness of the floccules satisfies the relation: ; In the formula, For the index of the hydraulic shear weakness of the floccule, As a natural logarithmic function, For the instantaneous flow of incoming water for the current acquisition cycle, For the instantaneous flow of water inlet in the previous acquisition period, For the nominal reference volume flow rate, The instantaneous addition amount of the polyacrylamide is adopted, For the instant adding amount of the polyaluminum chloride, For the reference constant of the lowest addition quantity, the interior of the logarithmic function in the relation is defined by the steady-state absolute flow term And dynamic impact flow term And (5) overlapping.
5. The intelligent recommendation method for the proportioning of the mine water treatment medicament based on the multi-objective optimization according to claim 1, wherein the predicting the future-time effluent turbidity by using the classification-enhanced gradient-lifted tree prediction algorithm comprises the following steps: tensor splicing the standardized engineering reference data sequence and the heavy metal and suspended matter competitive adsorption index and the flocculating constituent hydraulic shearing weakness index to form an input characteristic vector; And inputting the input feature vector into a pre-constructed classification enhancement gradient lifting tree prediction algorithm model to perform forward reasoning calculation so as to output the turbidity of the water at the future moment.
6. The intelligent recommendation method for the proportioning of the mine water treatment agent based on the multi-objective optimization of claim 5, wherein a symmetrical decision tree structure is adopted in the classification enhancement gradient lifting tree prediction algorithm model, and the heavy metal and suspended matter competitive adsorption index and the flocculating body hydraulic shearing weakness index are used as node division criterion constraint conditions of the classification enhancement gradient lifting tree prediction algorithm model.
7. The intelligent recommendation method for the proportioning of the mine water treatment medicament based on the multi-objective optimization according to claim 1, wherein the constructing the fitness objective function according to the future effluent turbidity and combining engineering punishment constraint comprises the following steps: Constructing an operation cost minimization relation and a water outlet comprehensive risk minimization relation as the fitness objective function; the operation cost minimization relation is calculated by the product of the unit purchase cost constant of the polyaluminum chloride and the instantaneous addition amount of the polyaluminum chloride and the product of the unit purchase cost constant of the polyacrylamide and the instantaneous addition amount of the polyacrylamide.
8. The intelligent recommendation method for the proportioning of the mine water treatment medicament based on the multi-objective optimization of claim 7, wherein the comprehensive risk minimization relation of water outlet is: ; In the formula, For the calculation result of the water outlet comprehensive risk minimization relation, For the turbidity of the effluent at a future time, For the upper limit constant of the discharge turbidity law, For the index of the hydraulic shear weakness of the floccule, Competing adsorption index for heavy metals and suspensions.
9. The intelligent recommendation method for the proportioning of the mine water treatment medicament based on the multi-objective optimization according to claim 1, wherein the optimizing is performed by adopting a multi-objective particle swarm optimization algorithm, and the optimal medicament addition amount combination is output as a recommended proportioning instruction, and the method comprises the following steps: Initializing a group of particle swarms within a set allowable upper and lower limit range of the drug addition amount, mapping the space position coordinates of each particle in the group of particle swarms into a group of candidate drug addition proportioning schemes, updating the speed and the position of each particle in the group of particle swarms in a solution space according to social experience of the group and individual cognitive experience, calculating the value of the fitness objective function, and outputting a group of pareto optimal solutions which are mutually non-dominant after the maximum iteration step number is set.
10. The intelligent recommendation method for the proportioning of the mine water treatment medicament based on the multi-objective optimization of claim 9, wherein the optimal medicament addition amount combination is output as a recommended proportioning instruction, and further comprising: And calculating Euclidean distances between all solution vectors in the pareto optimal solution set and an ideal origin, extracting a compromise solution with the shortest Euclidean distance as an optimal recommended proportioning instruction, converting the recommended proportioning instruction into a continuous control electric signal, and sending the continuous control electric signal to an execution end to finish closed-loop control of the mine water treatment agent dosing system.

Description

Mine water treatment agent proportioning intelligent recommendation method based on multi-objective optimization Technical Field The invention relates to the technical field of mine water treatment, in particular to an intelligent recommendation method for a mine water treatment medicament proportion based on multi-objective optimization. Background In the gold exploitation and dressing industry, produced mine water contains not only high-concentration suspended matters but also heavy metal ions such as copper, lead, zinc and the like, in order to treat water with complex components, a combined water treatment process of adding polyaluminium chloride as a coagulant and adding polyacrylamide as a polymeric flocculant in the processing industry is generally adopted at present, and coprecipitation separation is carried out on the suspended matters and heavy metals by adjusting the pH value of fluid, however, the requirement on medicament addition and proportioning control is extremely high when the mine water with complex components is treated. In the existing water treatment process, when the data processing system is used for carrying out the dosing proportion of mine water, historical water quality monitoring data and historical dosing data are mainly input into an implicit nonlinear mapping model for training calculation, because heavy metal ions and suspended particles in the mine water compete with adsorption complexing sites of polyacrylamide polymer chains on objective layers together, when the fluctuation of pH value of inflow water deviates from an optimal precipitation interval, the dissolution morphology of the heavy metal ions changes, so that nonlinear deviation is generated in the competition relationship between the heavy metal ions and suspended matters, the existing pure data algorithm cannot characterize the chemical competition relationship, so that recommended medicaments are not consumed by the heavy metal ions and flocculation failure engineering problem is caused, and the method for fitting by the historical data only ignores the real physical-chemical reaction mechanism of a water treatment bottom layer, so that the data processing system can not only waste high-price medicament resources, but also lose the effectiveness of the whole chemical precipitation process and seriously threaten the water quality standard rate when the working conditions such as the pH value of water is changed. In addition, the existing method breaks away from a physical shear damage mechanism in a hydrodynamic environment, dynamic fluctuation is presented in the water inflow flow in the actual engineering, huge hydraulic shear stress is generated in a coagulation reaction tank by flow surge, if the adding proportion of polyaluminium chloride and polyacrylamide is unbalanced, irreversible physical disruption is very easy to occur in the generated floccule, the hydrodynamic environment is not taken as constraint punishment condition by the existing optimizing algorithm, the output optimal proportion has extremely poor impact resistance under the dynamic flow working condition, once the sudden water inflow is encountered, the instantaneous surge of the water inflow is caused, the physical weakness degree of the floccule cannot be early warned by a data processing system which simply pursues theoretical indexes, the firmness of a microstructure cannot be guaranteed by an output proportion instruction, finally, the condensed impurities are scattered again, and the problem of water quality exceeding is caused. Disclosure of Invention The invention provides an intelligent recommendation method for the proportioning of the mine water treatment medicament based on multi-objective optimization, aiming at solving the technical problems that chemical competitive adsorption and physical shearing damage are not considered in the existing intelligent recommendation technology, so that the medicament is invalid to consume and has poor impact resistance. The invention provides an intelligent recommendation method for proportioning of mine water treatment agents based on multi-objective optimization, which comprises the steps of collecting multidimensional dynamic data of mine water treatment and preprocessing to obtain a standardized engineering reference data sequence, constructing heavy metal and suspended matter competitive adsorption indexes and flocculating constituent hydraulic shearing weakness indexes according to the standardized engineering reference data sequence, fusing the standardized engineering reference data sequence with the heavy metal and suspended matter competitive adsorption indexes and the flocculating constituent hydraulic shearing weakness indexes, predicting water turbidity at the future moment by utilizing a classification enhancement gradient lifting tree prediction algorithm, constructing an adaptability objective function according to the water turbidity at the future moment, optimizing by adopt