Arsenic-gold dissociation optimization device and method for arsenic-containing gold ore

Abstract

This invention relates to an arsenic-gold dissociation optimization treatment device and method for arsenic-containing gold ore, belonging to the field of gold ore processing technology. By configuring multiple sensors in the reaction chamber of the arsenic-gold dissociation optimization treatment device, multi-source data information is collected through these sensors. A digital twin is then formed through a mechanistic model layer, a data-driven model layer, and a high-fidelity simulation layer. This allows for the prediction of the process state under different combinations of control parameters within a preset time period in a virtual space. Based on a genetic algorithm, the optimal range for the maximum arsenic oxidation rate and microbial activity is solved, calculating the optimal control instruction set, and then optimizing the control based on this optimal control instruction set. This invention, by constructing a virtual-real interactive digital twin system and integrating a genetic algorithm for dynamic optimization, achieves accurate prediction and real-time optimization of the biological oxidation process, improving the degree of arsenic-gold dissociation in arsenic-containing gold ore and increasing the purity of the prepared gold ore.

Description

Technical Field

[0001] This invention relates to the field of gold ore processing technology, and in particular to an arsenic-gold dissociation optimization treatment device and method for arsenic-containing gold ore. Background Technology

[0002] Arsenic is a highly toxic metalloid. Improper disposal of the "three wastes" (waste gas, wastewater, and waste residue) generated during the mining and beneficiation of arsenic-containing gold ores can easily lead to arsenic pollution incidents. While existing technologies, such as biological oxidation pretreatment, have the potential for lower cost and environmental friendliness, their long reaction cycles, poor adaptability to extreme environments (such as high-altitude and cold regions), and complex microbial cultivation and activity control pose challenges to stable industrial operation. Direct leaching with non-cyanide thiosulfate or thiourea, while avoiding the synergistic toxicity of cyanide and arsenic, consumes large quantities of reagents, is costly, and has limited effectiveness in dissociating gold encapsulated in arsenic-containing minerals. Summary of the Invention

[0003] This invention overcomes the shortcomings of the prior art and provides an arsenic-gold dissociation optimization treatment device and method for arsenic-containing gold ore.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0005] The first aspect of this invention provides an arsenic-gold dissociation optimization treatment apparatus for arsenic-containing gold ore, comprising:

[0006] The reaction chamber is equipped with a microbial community, which is used to oxidize and reduce arsenic in arsenic-containing gold ore.

[0007] An air inlet pipe and an air outlet pipe are installed on the reaction chamber. An oxygen flow control valve is installed on the air inlet pipe, and the amount of oxygen input is controlled by controlling the oxygen flow control valve.

[0008] An array of pH sensors, redox potential sensors, dissolved oxygen sensors, and ion-selective electrodes are deployed inside the reaction chamber. Data collected by the array of pH sensors, redox potential sensors, dissolved oxygen sensors, and ion-selective electrodes are used to predict the influence of nonlinear perturbations on bacterial community activity.

[0009] Based on the influence of the nonlinear perturbation on the bacterial community activity, the maximum arsenic oxidation rate and the optimal range of bacterial community activity are solved.

[0010] Furthermore, in the arsenic-gold dissociation optimization treatment device for arsenic-containing gold ore, a feed inlet is installed on the reaction chamber, and a sealing door is installed on the feed inlet.

[0011] Furthermore, in the arsenic-gold dissociation optimization treatment device for arsenic-containing gold ore, an exhaust duct is installed on one end of the exhaust pipe.

[0012] Furthermore, in the arsenic-gold dissociation optimization treatment device for arsenic-containing gold ore, a controller is installed outside the reaction chamber.

[0013] A second aspect of the present invention provides a method for arsenic-gold dissociation optimization treatment of arsenic-containing gold ore, applicable to the arsenic-gold dissociation optimization treatment apparatus for any of the arsenic-containing gold ore described in any one of the claims, comprising the following steps:

[0014] Digital twin technology is introduced, and multiple sensors are configured in the reaction chamber of the arsenic-gold dissociation optimization treatment device for the arsenic-gold ore. Multi-source data information is collected through the multiple sensors, and a physical layer is constructed based on the multi-source data information.

[0015] Construct a mechanism model layer, a data-driven model layer, and a high-fidelity simulation layer, and form a digital twin based on the mechanism model layer, the data-driven model layer, and the high-fidelity simulation layer;

[0016] Predict the process state under different combinations of control parameters within a preset time in a virtual space;

[0017] A genetic algorithm is introduced to solve for the optimal range of maximum arsenic oxidation rate and bacterial community activity, calculate the optimal control instruction set, and perform control optimization based on the optimal control instruction set.

[0018] Furthermore, in the arsenic-gold dissociation optimization treatment method for arsenic-containing gold ore, digital twin technology is introduced. Multiple sensors are configured in the reaction chamber of the arsenic-gold dissociation optimization treatment device, and multi-source data information is collected through these sensors. A physical layer is constructed based on this multi-source data information, specifically:

[0019] Digital twin technology is introduced, and an array of pH sensors, redox potential sensors, dissolved oxygen sensors, and ion-selective electrodes are configured in the reaction chamber of the arsenic-gold dissociation optimization treatment device for the arsenic-gold ore.

[0020] The array-type pH sensor is used to collect pH data, the redox potential sensor is used to monitor the reduction state of substances in arsenic-containing gold ore, the dissolved oxygen sensor is used to collect dissolved oxygen concentration data in the reaction chamber, and the ion-selective electrode is used to monitor arsenic ion concentration data.

[0021] All collected data are aggregated to form multi-source data information, and a data collection frequency is set. Data is collected according to the set data collection frequency.

[0022] Furthermore, in the optimized treatment method for arsenic-gold dissociation in arsenic-bearing gold ore, a mechanism model layer, a data-driven model layer, and a high-fidelity simulation layer are constructed. A digital twin is then formed based on these layers, specifically:

[0023] A mechanism model layer, a data-driven model layer, and a high-fidelity simulation layer are constructed. In the mechanism model layer, the Monod equation for the microbial growth and arsenic oxidation kinetics in the current reaction chamber is obtained through big data.

[0024] A mathematical model of key biochemical reaction rates is established based on the Mono equation of microbial growth and arsenic oxidation kinetics in the current reaction chamber, describing the dynamic relationship between microbial community concentration, substrate consumption and product formation. In the data-driven model layer, a long short-term memory network is introduced.

[0025] A massive amount of time-series data on the effect of nonlinear perturbations on bacterial community activity is collected. The data is then trained using a long short-term memory network to capture the long-term impact of nonlinear perturbations on bacterial community activity.

[0026] Furthermore, the method for optimizing the arsenic-gold dissociation process of arsenic-bearing gold ore also includes:

[0027] In the high-fidelity simulation layer, a three-dimensional model of the reaction chamber is established using computational fluid dynamics software, and the slurry flow field, gas distribution and mass transfer efficiency in the reaction chamber of the arsenic gold dissociation optimization treatment device for arsenic gold ore are simulated.

[0028] The three-dimensional model of the oxidation tank is fused with the mechanistic model layer to form a dynamic model, thereby creating a digital twin.

[0029] Furthermore, in the optimized treatment method for arsenic-gold dissociation in arsenic-bearing gold ore, the process state under different combinations of control parameters within a preset time period is predicted in virtual space, specifically as follows:

[0030] Different combinations of aeration and acid addition were set up, and the process conditions of arsenic-containing gold ore under different combinations of aeration and acid addition were simulated by digital twins. A process condition prediction model was built based on a deep learning network.

[0031] Set the model parameters of the process state prediction model, and input the process state of the arsenic-containing gold ore under different combinations of aeration and acid addition into the process state prediction model for training according to the model parameters of the process state prediction model.

[0032] When the training reaches the predetermined target conditions, the process state prediction model is trained and the current real-time multi-source data information and the combined parameters of aeration rate and acid addition rate are input into the process state prediction model for prediction to obtain the process state of the arsenic-containing gold ore in the reaction chamber.

[0033] Furthermore, in the optimized treatment method for arsenic-gold dissociation in arsenic-containing gold ore, a genetic algorithm is introduced. This algorithm is used to solve for the optimal range between the maximum arsenic oxidation rate and the optimal range of bacterial community activity, calculating the optimal control instruction set. Control optimization is then performed based on this optimal control instruction set, specifically as follows:

[0034] Set the process status evaluation index range for arsenic-containing gold ore in the reaction chamber. When the process status of arsenic-containing gold ore in the reaction chamber is within the process status evaluation index range, control the arsenic-gold dissociation optimization treatment device for arsenic-containing gold ore according to the current control parameter combination.

[0035] When the process state of the arsenic-containing gold ore in the reaction chamber is not within the range of the process state evaluation index of the arsenic-containing gold ore in the reaction chamber, a genetic algorithm is introduced, and the number of generations is set based on the genetic algorithm.

[0036] Adjust the parameters of the current control parameter combination, perform simulation based on the parameters of the current control parameter combination, and determine whether the process state of the arsenic-containing gold ore in the reaction chamber is within the optimal range of maximum arsenic oxidation rate and microbial activity.

[0037] If the process conditions of the arsenic-containing gold ore in the reaction chamber are within the range of the maximum arsenic oxidation rate and the optimal range of microbial activity, it is considered a high-quality solution; otherwise, it is not considered a high-quality solution.

[0038] All high-quality solutions are statistically analyzed to generate an optimal control instruction set, and a high-quality solution from the optimal control instruction set is selected for control optimization.

[0039] This invention addresses the shortcomings of the prior art and has the following beneficial effects:

[0040] This invention introduces digital twin technology, configuring multiple sensors in the reaction chamber of an arsenic-gold dissociation optimization treatment device for arsenic-containing gold ore. These sensors collect multi-source data to construct a mechanistic model layer, a data-driven model layer, and a high-fidelity simulation layer. A digital twin is formed based on these layers, allowing for the prediction of the process state under different combinations of control parameters within a preset timeframe in a virtual space. Finally, a genetic algorithm is introduced to solve for the optimal range of maximum arsenic oxidation rate and microbial activity, calculating the optimal control instruction set. Control optimization is then performed based on this optimal control instruction set. This invention, by constructing a virtual-real interactive digital twin system and integrating a genetic algorithm for dynamic optimization, achieves accurate prediction and real-time optimization of the biological oxidation process, improving the degree of arsenic-gold dissociation in arsenic-containing gold ore and increasing the purity of the prepared gold ore. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained from these drawings without creative effort.

[0042] Figure 1 A schematic diagram of an arsenic-gold dissociation optimization treatment device for arsenic-containing gold ore is shown.

[0043] Figure 2 A flowchart of the method for optimizing the arsenic-gold dissociation process of arsenic-containing gold ore is shown. Detailed Implementation

[0044] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner. Therefore, they only show the components related to the present invention. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other.

[0045] In the description of this application, it should be understood that the terms "center," "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships based on the orientation or positional relationships shown in the accompanying drawings, are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as limiting the scope of protection of this application. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, features defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0046] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art will understand the specific meaning of the above terms in this application based on the specific circumstances.

[0047] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Preferred embodiments of the invention are shown in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a thorough and complete understanding of the disclosure of the invention.

[0048] like Figure 1 As shown, the first aspect of the present invention provides an arsenic-gold dissociation optimization treatment apparatus for arsenic-containing gold ore, comprising:

[0049] Reaction chamber 1, in which a microbial community is configured, and the microbial community performs oxidation-reduction on the arsenic element in the arsenic-containing gold ore;

[0050] An inlet pipe 2 and an outlet pipe 3 are installed on the reaction chamber 1. An oxygen flow control valve 4 is installed on the inlet pipe 2, which controls the amount of oxygen input and serves as the core actuator. A gas switch control valve 7 is installed on the outlet pipe 3, which controls the discharge of gas from the reaction chamber, thereby controlling the gas in the reaction chamber to be within a predetermined oxygen concentration or pressure range.

[0051] An array of pH sensors, redox potential sensors, dissolved oxygen sensors, and ion-selective electrodes are deployed inside the reaction chamber. Data collected by the array of pH sensors, redox potential sensors, dissolved oxygen sensors, and ion-selective electrodes are used to predict the influence of nonlinear perturbations on bacterial community activity.

[0052] Based on the influence of the nonlinear perturbation on the bacterial community activity, the maximum arsenic oxidation rate and the optimal range of bacterial community activity are solved.

[0053] It should be noted that a Long Short-Term Memory (LSTM) network was used to train massive amounts of historical time-series data to learn the long-term effects of nonlinear perturbations such as temperature and pH fluctuations on microbial community activity. A three-dimensional model of the oxidation tank was built using computational fluid dynamics (CFD) software (such as ANSYS Fluent) to simulate the slurry flow field, gas distribution, and mass transfer efficiency, and this model was coupled with the mechanistic model. Based on the Monod equation for the growth of microorganisms (such as Leptospira ironophila) and the kinetics of arsenic / iron oxidation, a mathematical model of key biochemical reaction rates was established to describe the dynamic relationship between microbial community concentration, substrate consumption, and product formation.

[0054] It should be noted that a bacterial community and solution (such as Leptospira ironophila) are prepared inside the reaction chamber. The bacterial community then performs an oxidation-reduction reaction on the arsenic element in the arsenic-containing gold ore, thereby purifying the gold ore.

[0055] Among them, array-type pH sensors, oxidation-reduction potential (ORP) sensors, dissolved oxygen (DO) sensors, and ion-selective electrodes (for As) are deployed at key locations inside the reaction chamber (such as different height areas of the reaction chamber). 3+ Fe 2+ The sensor transmits encrypted data to the edge server at a frequency of 1Hz via an Industrial Internet of Things (IIoT) gateway.

[0056] Furthermore, in the arsenic-gold dissociation optimization treatment device for arsenic-containing gold ore, a feed inlet is installed on the reaction chamber, and a sealing door 5 is installed on the feed inlet.

[0057] Furthermore, in the arsenic-gold dissociation optimization treatment device for arsenic-containing gold ore, an exhaust duct 6 is installed on one end of the exhaust pipe 3.

[0058] Furthermore, in the arsenic-gold dissociation optimization treatment device for arsenic-containing gold ore, a controller is installed outside the reaction chamber 1.

[0059] The reaction chamber can also be equipped with humidifiers, heaters, and other environmental control devices to regulate the temperature and humidity of the redox reaction. These are not shown in the diagram, but they should include such devices.

[0060] like Figure 2 As shown, the second aspect of the present invention provides a method for arsenic-gold dissociation optimization treatment of arsenic-containing gold ore, applicable to the arsenic-gold dissociation optimization treatment apparatus for any of the arsenic-containing gold ore described in any one of the claims, comprising the following steps:

[0061] Digital twin technology is introduced, and multiple sensors are configured in the reaction chamber of the arsenic-gold dissociation optimization treatment device for the arsenic-gold ore. Multi-source data information is collected through the multiple sensors, and a physical layer is constructed based on the multi-source data information.

[0062] Construct a mechanism model layer, a data-driven model layer, and a high-fidelity simulation layer, and form a digital twin based on the mechanism model layer, the data-driven model layer, and the high-fidelity simulation layer;

[0063] Predict the process state under different combinations of control parameters within a preset time in a virtual space;

[0064] A genetic algorithm is introduced to solve for the optimal range of maximum arsenic oxidation rate and bacterial community activity, calculate the optimal control instruction set, and perform control optimization based on the optimal control instruction set.

[0065] It should be noted that this method constructs a virtual-real interactive digital twin system, thereby integrating genetic algorithms for dynamic optimization, achieving accurate prediction and real-time optimization of the biological oxidation process, improving the degree of arsenic-gold dissociation in arsenic-containing gold ore, and increasing the purity of gold ore preparation.

[0066] Furthermore, in the arsenic-gold dissociation optimization treatment method for arsenic-containing gold ore, digital twin technology is introduced. Multiple sensors are configured in the reaction chamber of the arsenic-gold dissociation optimization treatment device, and multi-source data information is collected through these sensors. A physical layer is constructed based on this multi-source data information, specifically:

[0067] Digital twin technology is introduced, and an array of pH sensors, redox potential sensors, dissolved oxygen sensors, and ion-selective electrodes are configured in the reaction chamber of the arsenic-gold dissociation optimization treatment device for the arsenic-gold ore.

[0068] The array-type pH sensor is used to collect pH data, the redox potential sensor is used to monitor the reduction state of substances in arsenic-containing gold ore, the dissolved oxygen sensor is used to collect dissolved oxygen concentration data in the reaction chamber, and the ion-selective electrode is used to monitor arsenic ion concentration data.

[0069] All collected data are aggregated to form multi-source data information, and a data collection frequency is set. Data is collected according to the set data collection frequency.

[0070] Furthermore, in the optimized treatment method for arsenic-gold dissociation in arsenic-bearing gold ore, a mechanism model layer, a data-driven model layer, and a high-fidelity simulation layer are constructed. A digital twin is then formed based on these layers, specifically:

[0071] A mechanism model layer, a data-driven model layer, and a high-fidelity simulation layer are constructed. In the mechanism model layer, the Monod equation for the microbial growth and arsenic oxidation kinetics in the current reaction chamber is obtained through big data.

[0072] It should be noted that in the mechanistic model layer, the Monod equation is used to describe the degradation rate of compounds when they are the sole carbon source. By utilizing the Monod equation for microbial growth and arsenic oxidation kinetics in the current reaction chamber to describe the degradation rate of arsenic compounds, a degradation mechanism for arsenic-containing gold ore can be provided for digital twin technology.

[0073] A mathematical model of key biochemical reaction rates is established based on the Mono equation of microbial growth and arsenic oxidation kinetics in the current reaction chamber, describing the dynamic relationship between microbial community concentration, substrate consumption and product formation. In the data-driven model layer, a long short-term memory network is introduced.

[0074] It should be noted that substrate consumption and product generation include dissolved oxygen concentration data, arsenic ion concentration data, and reduction state data of substances in arsenic-containing gold ore within the reaction chamber. By introducing a Long Short-Term Memory (LSTM) network into the data-driven model layer, the LSTM network is used to train on the time-series data of the massive nonlinear perturbations (such as ion concentration, reaction chamber temperature, pH data, etc.) on bacterial community activity, capturing the long-term impact of nonlinear perturbations on bacterial community activity. The LSTM network includes processes such as forgetting gates, input gates, candidate cell states, and output gates. LSTM training is an iterative process aimed at adjusting all network parameters (weights and biases) by optimizing the loss function, including forward propagation, loss calculation, backpropagation time-through-time (BPTT), parameter updates, and repeated iterations.

[0075] The training process proceeds sequentially as follows: Input sequences are processed in chronological order, and the forget gate, input gate, candidate states, updated cell states, output gate, and hidden states are calculated sequentially. The network's final output is compared with the true value, and the difference (loss) is calculated. Starting from the last time step, the gradient of the loss function with respect to all weights (gating weights and output weights) is calculated in reverse. The key point is that the error propagates backward along the "highway" of cell states, effectively mitigating gradient vanishing. An optimizer (such as Adam) and the calculated gradients are used to update all weights and biases of the network. These steps are repeated until the model performance no longer significantly improves. Thus, a Long Short-Term Memory (LSTM) network is trained on time-series data of bacterial community activity under massive nonlinear perturbation factors (such as ion concentration, reaction chamber temperature, pH data, etc.), capturing the long-term impact of nonlinear perturbations on bacterial community activity.

[0076] A massive amount of time-series data on the effect of nonlinear perturbations on bacterial community activity is collected. The data is then trained using a long short-term memory network to capture the long-term impact of nonlinear perturbations on bacterial community activity.

[0077] Furthermore, the method for optimizing the arsenic-gold dissociation process of arsenic-bearing gold ore also includes:

[0078] In the high-fidelity simulation layer, a three-dimensional model of the reaction chamber is established using computational fluid dynamics software, and the slurry flow field, gas distribution and mass transfer efficiency in the reaction chamber of the arsenic gold dissociation optimization treatment device for arsenic gold ore are simulated.

[0079] The three-dimensional model of the oxidation tank is fused with the mechanistic model layer to form a dynamic model, thereby creating a digital twin.

[0080] It should be noted that by fusing the three-dimensional model of the oxidation tank with the mechanism model layer to form a dynamic model, dynamic simulation is performed to form a dynamic simulation process of the oxidation-reduction process.

[0081] Furthermore, in the optimized treatment method for arsenic-gold dissociation in arsenic-bearing gold ore, the process state under different combinations of control parameters within a preset time period is predicted in virtual space, specifically as follows:

[0082] Different combinations of aeration and acid addition were set up, and the process conditions of arsenic-containing gold ore under different combinations of aeration and acid addition were simulated by digital twins. A process condition prediction model was built based on a deep learning network.

[0083] Set the model parameters of the process state prediction model, and input the process state of the arsenic-containing gold ore under different combinations of aeration and acid addition into the process state prediction model for training according to the model parameters of the process state prediction model.

[0084] When the training reaches the predetermined target conditions, the process state prediction model is trained and the current real-time multi-source data information and the combined parameters of aeration rate and acid addition rate are input into the process state prediction model for prediction to obtain the process state of the arsenic-containing gold ore in the reaction chamber.

[0085] It should be noted that the process conditions of arsenic-containing gold ore, such as arsenic oxidation rate and microbial activity status, can be predicted using this method, thereby enabling the use of genetic algorithms to find the optimal solution set.

[0086] Furthermore, in the optimized treatment method for arsenic-gold dissociation in arsenic-containing gold ore, a genetic algorithm is introduced. This algorithm is used to solve for the optimal range between the maximum arsenic oxidation rate and the optimal range of bacterial community activity, calculating the optimal control instruction set. Control optimization is then performed based on this optimal control instruction set, specifically as follows:

[0087] Set the process status evaluation index range for arsenic-containing gold ore in the reaction chamber. When the process status of arsenic-containing gold ore in the reaction chamber is within the process status evaluation index range, control the arsenic-gold dissociation optimization treatment device for arsenic-containing gold ore according to the current control parameter combination.

[0088] When the process state of the arsenic-containing gold ore in the reaction chamber is not within the range of the process state evaluation index of the arsenic-containing gold ore in the reaction chamber, a genetic algorithm is introduced, and the number of generations is set based on the genetic algorithm.

[0089] Within the range of the genetic generations, adjust the parameters of the current control parameter combination, perform simulation based on the parameters of the current control parameter combination, and determine whether the process state of the arsenic-containing gold ore in the reaction chamber is within the range of the maximum arsenic oxidation rate and the optimal range of microbial activity.

[0090] If the process conditions of the arsenic-containing gold ore in the reaction chamber are within the range of the maximum arsenic oxidation rate and the optimal range of microbial activity, it is considered a high-quality solution; otherwise, it is not considered a high-quality solution.

[0091] All high-quality solutions are statistically analyzed to generate an optimal control instruction set, and a high-quality solution from the optimal control instruction set is selected for control optimization.

[0092] It should be noted that this method can optimize the combination of control parameters (ventilation rate, acid addition rate, etc.), such as increasing the ventilation rate of the air inlet pipe from 15L / min to 18.5L / min, and the operation can be executed by the MFC and oxygen control valves on site through the OPCUA protocol, thereby improving the degree of arsenic-gold dissociation in arsenic-containing gold ore and improving the purity of gold ore preparation.

[0093] In addition, this method also includes:

[0094] An online analyzer is installed in the exhaust pipe to collect and analyze arsenic concentration and valence state (As). 3+ / As 5+ Phase data, and collected data on the status and market price of each resource production line, based on the status and market price of each resource production line, arsenic concentration, and valence state of As. 3+ / As 5+ Constructing a data chain from phase data;

[0095] A plant-wide static and dynamic material flow model of arsenic was established based on chemical process simulation software to show the distribution rate of arsenic in each process. A decision optimization model was built based on deep learning.

[0096] Among them, a state space, action space and reward function are constructed. The current arsenic material flow distribution, the status of each resource production line and the market price are obtained through the data chain, and the current arsenic material flow distribution, the status of each resource production line and the market price are used as the state space.

[0097] The allocation and processing paths of different arsenic-containing intermediate products are used as the action space, and the comprehensive economic benefits and environmental protection weights are used as the reward function. Deep learning is used to train the state space, action space and reward function to form a trained decision optimization model.

[0098] When the model determines that the purity of As2O3 in the current flue gas is high and the price of semiconductor-grade arsenic is strong, the operating power of the oxygen supply system is optimized, and at the same time, a raw material stocking suggestion is sent to the procurement system.

[0099] It should be noted that this solution achieves dynamic optimization of the resource utilization path of arsenic by constructing a digital flow of arsenic, thereby reducing the production cost of gold ore while maintaining a certain level of purity.

[0100] In addition, this method also includes:

[0101] The growth density and metabolic activity data of microorganisms in each reaction chamber were obtained, and the growth density and metabolic activity data of all microorganisms were input into the database and correlated with the genome sequencing data of the strains to form a multidimensional training dataset.

[0102] Using the multidimensional training dataset, a graph neural network model is trained, and the graph neural network model is used to learn the relationship between SNPs of specific functional genes and arsenic oxidation rate and acid resistance.

[0103] In the reaction chamber, samples were taken and the absolute number and relative activity of key functional bacterial groups were rapidly determined. The relationship between the SNP of the specific functional gene and the arsenic oxidation rate and acid resistance, along with the process parameters, were input into a lightweight convolutional neural network diagnostic model.

[0104] The convolutional neural network outputs the activity and health index of the current bacterial population and warnings of possible inhibitory factors. When the activity and health index of the bacterial population is lower than the preset threshold, ammonium phosphate, yeast extract powder or a backup high-activity bacterial culture tank is added according to the optimized formula, and fresh bacterial solution is pumped into the main reaction system according to the calculated amount.

[0105] It should be noted that in industrial production, flow cytometry coupled with in situ fluorescence hybridization (Flow-FISH) technology is used to sample and rapidly determine the absolute quantity and relative activity of key functional microbial communities hourly from the oxidation tank. This data, along with process parameters (such as pH and ORP), is input into a lightweight convolutional neural network (CNN) diagnostic model. This CNN model can output the current microbial community's activity and health index (0-1) and warnings of potential inhibitory factors (such as arsenic toxicity and nutrient deficiency) within 5 minutes. Once the index falls below the threshold, the system will automatically trigger remedial measures, such as adding specific nutrients (ammonium phosphate, yeast extract) according to the optimized formula or starting a backup high-activity microbial culture tank, pumping fresh bacterial solution into the main reaction system at a calculated volume. This method solves the problems of long selection cycles for functional strains (such as Leptospira ironophila) and delayed monitoring of microbial community activity in the biological oxidation process.

[0106] In summary, by using digital twins for real-time simulation and prediction, combined with intelligent optimization algorithms, the system can quickly respond to changes in operating conditions, ensuring that the biological oxidation process always operates in an optimal or near-optimal state, thus solving the problem of poor stability in traditional biological oxidation processes. Furthermore, precise control of oxygen and nutrient inputs maximizes the arsenic oxidation rate of microorganisms, potentially shortening the pretreatment cycle and increasing production capacity. Optimized control reduces the ineffective consumption of reagents and energy; and optimized arsenic resource utilization pathways turn waste into treasure, creating additional revenue. This invention achieves online, prospective diagnosis and intelligent regulation of functional microbial community activity, reducing the risk of microbial inactivation and enhancing adaptability to fluctuations in ore composition. While efficiently dissociating gold, this invention provides intelligent decision support for the stabilization treatment or high-value resource utilization of arsenic, resulting in significant environmental benefits.

[0107] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

[0108] The embodiments described above are merely specific implementations of this application, used to illustrate the technical solutions of this application, and are not intended to limit it. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application. All should be covered within the protection scope of this application. Therefore, the protection scope of this application should be determined by the protection scope of the claims.

Claims

1. A method for optimizing the arsenic-gold dissociation process of arsenic-containing gold ore, characterized in that, An arsenic-gold dissociation and optimization treatment device for arsenic-containing gold ore, wherein the arsenic-gold dissociation and optimization treatment device for arsenic-containing gold ore includes: The reaction chamber is equipped with a microbial community, which is used to oxidize and reduce arsenic in arsenic-containing gold ore. An air inlet pipe and an air outlet pipe are installed on the reaction chamber. An oxygen flow control valve is installed on the air inlet pipe. The amount of oxygen input is controlled by controlling the oxygen flow control valve. An array of pH sensors, redox potential sensors, dissolved oxygen sensors, and ion-selective electrodes are deployed inside the reaction chamber. Data collected by the array of pH sensors, redox potential sensors, dissolved oxygen sensors, and ion-selective electrodes are used to predict the influence of nonlinear perturbations on bacterial community activity. Based on the influence of the nonlinear perturbation on the bacterial community activity, the maximum arsenic oxidation rate and the optimal range of bacterial community activity were solved. It also includes the following steps: Digital twin technology is introduced, and multiple sensors are configured in the reaction chamber of the arsenic-gold dissociation optimization treatment device for the arsenic-gold ore. Multi-source data information is collected through the multiple sensors, and a physical layer is constructed based on the multi-source data information. Construct a mechanism model layer, a data-driven model layer, and a high-fidelity simulation layer, and form a digital twin based on the mechanism model layer, the data-driven model layer, and the high-fidelity simulation layer; Predict the process state under different combinations of control parameters within a preset time in a virtual space; A genetic algorithm is introduced to solve for the optimal range between the maximum arsenic oxidation rate and the bacterial community activity, calculate the optimal control instruction set, and perform control optimization based on the optimal control instruction set. A mechanistic model layer, a data-driven model layer, and a high-fidelity simulation layer are constructed. A digital twin is then formed based on these three layers. Specifically: A mechanism model layer, a data-driven model layer, and a high-fidelity simulation layer are constructed. In the mechanism model layer, the Monod equation for the microbial growth and arsenic oxidation kinetics in the current reaction chamber is obtained through big data. A mathematical model of key biochemical reaction rates is established based on the Mono equation of microbial growth and arsenic oxidation kinetics in the current reaction chamber, describing the dynamic relationship between microbial community concentration, substrate consumption and product formation. In the data-driven model layer, a long short-term memory network is introduced. Collect massive amounts of time-series data on the effect of nonlinear perturbations on bacterial community activity, and use a long short-term memory network to train the massive amounts of time-series data on the effect of nonlinear perturbations on bacterial community activity to capture the long-term influence of nonlinear perturbations on bacterial community activity. Also includes: In the high-fidelity simulation layer, a three-dimensional model of the reaction chamber is established using computational fluid dynamics software, and the slurry flow field, gas distribution and mass transfer efficiency in the reaction chamber of the arsenic gold dissociation optimization treatment device for arsenic gold ore are simulated. The three-dimensional model of the oxidation tank is fused with the mechanistic model layer to form a dynamic model, thereby creating a digital twin. Predicting the process state under different combinations of control parameters within a preset time period in a virtual space, specifically: Different combinations of aeration and acid addition were set up, and the process conditions of arsenic-containing gold ore under different combinations of aeration and acid addition were simulated by digital twins. A process condition prediction model was built based on a deep learning network. Set the model parameters of the process state prediction model, and input the process state of arsenic-containing gold ore under different combinations of aeration rate and acid addition rate into the process state prediction model for training according to the model parameters of the process state prediction model. When the training reaches the predetermined target conditions, the process state prediction model is trained and the current real-time multi-source data information and the combined parameters of aeration rate and acid addition rate are input into the process state prediction model for prediction to obtain the process state of the arsenic-containing gold ore in the reaction chamber. It also includes the following steps: An online analyzer is installed in the exhaust pipe to collect and analyze arsenic concentration and valence state As³⁺ / As. 5 ⁺, phase data, and collected data on the status and market price of each resource production line, and based on the aforementioned status and market price of each resource production line, arsenic concentration, valence state As³⁺ / As 5 ⁺, Constructing a data chain from phase data; A plant-wide static and dynamic material flow model of arsenic was established based on chemical process simulation software to show the distribution rate of arsenic in each process. A decision optimization model was built based on deep learning. Specifically, a state space, an action space, and a reward function are constructed. The current arsenic material flow distribution, the status of each resource production line, and the market price are obtained through the data chain, and the current arsenic material flow distribution, the status of each resource production line, and the market price are used as the state space. The allocation and processing paths of different arsenic-containing intermediate products are used as the action space, and the comprehensive economic benefits and environmental protection weights are used as the reward function. Deep learning is used to train the state space, action space and reward function to form a trained decision optimization model. When the model determines that the purity of As2O3 in the current flue gas is high and the price of semiconductor-grade arsenic is strong, the operating power of the oxygen supply system is optimized, and at the same time, a raw material stocking suggestion is sent to the procurement system.

2. The method for optimizing the arsenic-gold dissociation of arsenic-containing gold ore according to claim 1, characterized in that, The reaction chamber is equipped with a feed inlet, and a sealing door is installed on the feed inlet.

3. The method for optimizing the arsenic-gold dissociation of arsenic-containing gold ore according to claim 1, characterized in that, An exhaust duct is installed at one end of the exhaust pipe.

4. The method for optimizing the arsenic-gold dissociation of arsenic-containing gold ore according to claim 1, characterized in that, A controller is installed on the outside of the reaction chamber.

5. The method for optimizing the arsenic-gold dissociation of arsenic-containing gold ore according to claim 1, characterized in that, Digital twin technology is introduced, and multiple sensors are configured in the reaction chamber of the arsenic-gold dissociation optimization treatment device for the arsenic-gold ore. Multi-source data information is collected through these sensors, and a physical layer is constructed based on this multi-source data information. Specifically: Digital twin technology is introduced, and an array of pH sensors, redox potential sensors, dissolved oxygen sensors, and ion-selective electrodes are configured in the reaction chamber of the arsenic-gold dissociation optimization treatment device for the arsenic-gold ore. The array-type pH sensor is used to collect pH data, the redox potential sensor is used to monitor the reduction state of substances in arsenic-containing gold ore, the dissolved oxygen sensor is used to collect dissolved oxygen concentration data in the reaction chamber, and the ion-selective electrode is used to monitor arsenic ion concentration data. All collected data are aggregated to form multi-source data information, and a data collection frequency is set. Data is collected according to the set data collection frequency.

6. The method for optimizing the arsenic-gold dissociation of arsenic-containing gold ore according to claim 1, characterized in that, A genetic algorithm is introduced to solve for the optimal range between the maximum arsenic oxidation rate and the optimal range of bacterial community activity, thereby calculating the optimal control instruction set. Control optimization is then performed based on this optimal control instruction set, specifically as follows: Set the process status evaluation index range for arsenic-containing gold ore in the reaction chamber. When the process status of arsenic-containing gold ore in the reaction chamber is within the process status evaluation index range, control the arsenic-gold dissociation optimization treatment device for arsenic-containing gold ore according to the current control parameter combination. When the process state of the arsenic-containing gold ore in the reaction chamber is not within the range of the process state evaluation index of the arsenic-containing gold ore in the reaction chamber, a genetic algorithm is introduced, and the number of generations is set based on the genetic algorithm. Adjust the parameters of the current control parameter combination, perform simulation based on the parameters of the current control parameter combination, and determine whether the process state of the arsenic-containing gold ore in the reaction chamber is within the optimal range of maximum arsenic oxidation rate and microbial activity. If the process conditions of the arsenic-containing gold ore in the reaction chamber are within the range of the maximum arsenic oxidation rate and the optimal range of microbial activity, it is considered a high-quality solution; otherwise, it is not considered a high-quality solution. All high-quality solutions are statistically analyzed to generate an optimal control instruction set, and a high-quality solution from the optimal control instruction set is selected for control optimization.

Images